Introduction

Many toxic heavy metals are found in the environment, and certain levels of exposure are inevitable for the inhabiting human populations. The industrial release of some heavy metals, such as lead and chromium, to the environment is significantly larger than the natural sources of these metals, while the levels of other heavy metals, such as cadmium and mercury, from either natural or industrial sources are the same [1].

Human beings such as Homo sapiens have been exposed to various heavy metals since the Stone Age [2]. A rapid increase in exposure to levels of the heavy metals in the modern environment, when compared to those in prehistoric periods, may have caused adverse health effects. To investigate such a possibility, reconstruction of the exposure history of humans has recently been explored in a number of studies [3–5].

Teeth can serve as records of environmental exposure to heavy metals that are accumulated in the mineral phase of the dental tissues during tooth formation [6, 7]. In tooth enamel, this mineral phase is not subject to turnover, since it consists of biological mineral hydroxyapatite, where various ions may be substituted into the crystal lattice only during the development. Thus, the enamel encapsulates a permanent record of the trace element environment during the development of a tooth. Migration of ions may occur, but it is confined to the immediate surface exposed to the oral environment and burial soils.

In the past, several methodologies have been applied to analysis of heavy metals in the teeth [6–15]. Recently, XRF analysis using synchrotron radiation (SR) microbeams (SRXRF) has been applied to the analyses of tooth enamel [10–12]. This method uses microbeams and enables us to provide high spatial resolution with much higher sensitivity [10–12].

The aim of the present study is to test the applicability of SRXRF for the analysis of heavy metals in human teeth. Specifically, this study has three objectives. First, since the accurate quantification of the amounts of heavy metals is difficult due to the lack of suitable reference materials, we tested whether concentration ratios of various heavy metals are proportional to their absolute concentrations determined by separate analytical methods [11, 12]. In this way, we aimed to replace a semi-quantitative method typically used, which simply compares ratios of elements in teeth, with a quantitative method. Secondly, we also applied this quantitative method to a series of molar teeth samples from a single individual. We tested whether exposure level can be correlated in several molar teeth with different developmental ages. Finally, we applied this method to the historical reconstruction of exposures to various heavy metals of humans who lived in different times, from the prehistoric era (Jomon era, bc 1000) to present times. In the present study, the targeted heavy metals are lead (Pb), mercury (Hg), copper (Cu) and zinc (Zn). Some of the reasons for selection of these metals are the following: (1) human exposure to lead is reported to be increasingly significant due to recent industrialization in western countries, (2) the major source of mercury in the environment is thought to be from natural release due to geological activities [16] and coal-fired power stations [17], (3) copper is one of the essential metals, which is obtained through diets and is also released by several industrial activities [18], and (4) the levels of zinc, an essential metal in the human body, are known to be strongly influenced by nutrition [19]. The study of the levels of heavy metals may elucidate the source and effects of long-term environmental exposure to these metals as well as elucidate nutritional conditions of the prehistoric and modern humans.

Materials and methods

Cases and samples

Permanent teeth samples from modern humans were collected from the donors after we obtained informed consent. The teeth were donated to our study, following an extraction by the dentists. Donors were selected from candidates who had never used dental amalgams. This study protocol was approved by the ethical review board of Kyoto University Graduate School of Medicine.

Archeological permanent teeth samples were collected during excavations (Table 1). Those teeth were free of caries. The ages of the teeth samples were determined by archeological criteria except in the case of individual K (1776–1846 ad), who was a court painter. The samples of teeth from subject K were dated based on documented records from a Buddhist temple in the cemetery where he was buried. Other teeth (E1–E4) were collected from the excavated ruins of a town or a local village of the Edo era [17th–19th century (C) ad]. Subjects E1 and E2 were postulated to be farmers, subjects E3 was postulated to be a merchant’s wife, and E4 was assumed to be a merchant or a family member of the merchant.

Table 1 Summary of permanent teeth samples
Full size table

The medieval time of Japan includes the Heian period (8th–12th C ad), the Kamakura period (12th–14th C ad) and the Muromachi period (14th–16th C ad). Teeth samples (C1–C3) were excavated from the cemetery, which was continuously used for burials from the 10th to 16th C ad. People who lived in a town across from the cemetery were buried in this place. No information was available for individuals from the 10th C ad. However, the 14th C ad subjects from whom teeth samples were collected are considered to have been craftsmen engaged in the casting of a Buddhist statue as an ancestral business, as determined from many artifacts (china and white porcelain) buried in the tombs.

The Tumulus period (3rd–6th C ad) is considered to be a period, when the first centralized government was formed. Nomadic people had settled in villages and engaged in agriculture. Human exposure to heavy metals in this period is considered to be mostly attributable to natural sources. Subject (T1) is considered to have been a head of a local clan.

The Jomon period corresponds to the Stone Age and started in 16,000 bc, ending in 500 bc. Donors of teeth J1 and J2 were buried in a typical Jomon shell mound. The entire skeleton of the donor J3 was found in a ruin in a cave and showed features of a middle-aged woman. Exposure to heavy metals during this period is considered to be solely due to ecological sources, since people were primarily engaged in hunting, fishing and gathering of foods.

Sample preparations

Each tooth was cut into half by longitudinal section (Fig. 1). From one piece, a longitudinal section 200 μm in thickness was cut using a diamond saw-cutter. To prevent contamination, diamond wire was immersed in distilled deionized water in a plastic container and the water was replaced after each sample was cut. Fresh water was used for each tooth when grinding and polishing the samples, and all samples were rinsed well with water prior to analysis.

Fig. 1
figure 1

Anatomy of a tooth and two-dimensional distributions of Ca and Zn by SRXRF. a Anatomy of a tooth by horizontal section. b Distribution of Ca by SRXRF. c Distribution of Zn by SRXRF. Relative signal intensities were expressed in gradient colors

Full size image

SRXRF analysis

SRXRF analyses using SR microbeams were performed at the Photon factory, KEK (Tsukuba, Japan) or at SPring-8 (Sayo, Japan) as previously described (Ide-Ektessabi et al. [12]). Briefly, SR from the storage ring (2.5 GeV, maximum current 400 mA, in the case of KEK) was monochromated using a multilayer film monochromater. The incident X-ray energy was 14.3 keV. Incident X-rays were focused using Kirkpatrick-Baez optics. The incident beam size was about 6 × 5 μm. The incident and transmitted photon flux was monitored with an ion chamber, and the fluorescent X-rays were detected by a solid-state detection (SSD). The measurements were conducted in air. SRXRF imaging was carried out as previously reported [12].

Quantification of the elements

Surface enamel portions (~200 μm) were abraded from the piece of the tooth, from which longitudinal tooth sections were cut for the analysis by SRXRF [12] in order to avoid the potential effects of diagenesis from the enveloping soil that would impact the surface of the tooth [20] (Fig. 2). Dentine was also removed from the tooth fragment to be analyzed. Semi-quantification of the concentration of each element was performed by the integration of the peak areas using software developed by Ide-Ektessabi et al. [20]. In this program, the background is estimated from the untreated spectra, and the peak is obtained using Gaussian curve fitting and the least square method. Linear scanning with high resolution was performed from the outside of enamel to pulp to obtain X-ray fluorescence for Ca, Pb Hg, Cu and Zn. The measurements were repeated by a 20-μm interval from the surface of enamel to pulp. The counts at individual points were integrated and standardized by the integrated count of Ca. The linear scanning was repeated five times for different lines per tooth. The mean of relative concentrations by five-time scanning was taken as a relative concentration for a given heavy metal for a given tooth. The relative concentrations of Pb, Hg, Cu and Zn were standardized by dividing the value of the peak areas of a given element by the peak area of Ca in the sample because of its similarity of behavior to that of heavy metals [13]. The coefficients of variations were within 20% for this analysis. From the remaining piece of the tooth, two enamel blocks were cut (~500 mg each) and washed thoroughly with doubly deionized distilled water. After cleaning, one piece of the enamel section was digested with hydrochloric acid. Digested samples were diluted to appropriate volumes with deionized water. The determinations of the concentrations of Pb, Cu and Zn were obtained using ICP-MS (Agilent 7500a, Tokyo, Japan) [21]. The lowest detection limits were 0.02 mg/l for Zn, 0.01 mg/l for Cu and 0.004 mg/l for Pb, respectively. The second piece of the enamel was subjected to the determination of the mercury concentration by reduction–aeration atomic absorption spectrometry (AAS). The detection limit was 0.002 μg.

Fig. 2
figure 2

Surface ablation and removal of dentin. The figure shows a half tooth after longitudinal section

Full size image

Calibrations of SRXRF by ICP-MS and AAS

To determine the concentrations of the heavy metals, we scanned the teeth samples from the oral side to dentin and the pulp through to the enamel (Fig. 1). One dimension and two dimension analyses gave patterns as shown in Fig. 1b and 1c. Concentrations of the elements were means of 20 × 20 pixels. We collected the fragment of the teeth for ICP-MS analyses as shown Fig. 2. From each tooth a sample was collected for the ICP-MS analysis.

Statistics

The collected data were analyzed using the SAS statistical package, version 8.2 (SAS Institute Inc., Cary, NC). P values for statistical tests were two-tailed. P value <0.05 was considered to be significant.

Results and discussion

Comparison between SRXRF and ICP-MS

The comparison between the relative concentrations of the elements obtained by SRXRF and ICP-MS are shown in Fig. 3a–d. The concentrations obtained by these two methods agreed significantly (R 2 > 0.758, P < 0.05). Such significant agreements allowed us to convert the relative concentrations (peak area for a given element divided by the area for Ca) obtained by SRXRF into concentrations (mass/mass). In Table 2, the converted values for heavy metal concentrations are presented together with concentrations measured by ICP-MS. The two values agreed well for the concentrations of Pb and Hg. However, for Cu and Zn, the values obtained from the analyses by the two methods were not in such good agreement.

Fig. 3
figure 3

Correlations between the concentrations obtained using SRXRF (×10−6) (X) and ICP-MS or AAS (Y). The X-axis represents relative concentrations of metals by SRXRF (×10−6), while the Y-axis represents actual concentrations per gram of tooth enamel: a Y-axis is μg of Pb/g. b Y-axis is ng of Hg/g. c Y-axis is μg of Cu/g. d Y-axis is μg of Zn/g

Full size image
Table 2 Estimated concentrations of elements in enamels
Full size table

Long-term trend and data interpretation

Limited number of tooth samples made it impossible to draw any definitive conclusions on the long-term trend in the concentration of these four heavy metals in human teeth. However, data presented in Table 2 suggest some very interesting exposure profiles.

Concentrations of Pb are highest in teeth obtained from the skeletons of humans of the Edo era. However, Pb levels are widely scattered among samples, likely reflecting personal lifestyles or habits. As a matter of fact, it is reported that lead-oxide cosmetic powders were used by females of the Samurai classes or merchants in urban centers in the Edo era [22–25]. Thus, higher exposures to Pb found in teeth samples of two subjects in the merchant class in the Edo era can likely be explained by the use of the lead-containing cosmetics by the mother.

High concentrations of Cu in teeth of the people living in the medieval times seem to be associated with their ancestral occupations. Both subjects C1 and C2 were thought to be craftsmen, engaged in the production of the statue of Buddha from copper in their cottage industry. In this period, they inherited their occupations from their fathers and conducted their work at home, leading to heavy indoor exposure to copper at home. Therefore, subjects C1 and C2 might have been exposed to copper through dust or fumes.

The levels of Zn seem to be highest in modern subjects. The levels of Zn in human tissue are known to be associated with nutritional conditions [19]. When the results for the Zn content in the teeth were pooled from the Jomon to Edo periods, their mean levels (μg/g, n = 11) were 121.6 ± 27.9 as determined by ICP-MS or 113.2 ± 27.4 by SRXRF, significantly lower (P < 0.05) (n = 5) than those in modern subjects (156.0 ± 27.0 μg/g by ICP-MS or 150.0 ± 24.6 μg/g by SRXRF). Modern increases in the content of Zn in the human teeth probably are associated with an increase in the consumption of Zn-rich foods, such as meats.

A case study for K

K was one of the most famous court painters in the Edo era. He was born at the end of the 18th century and died in the middle of the 19th century. He was very active as a leader of the painting school, which was established by his ancestor. The mineralization of his first tooth started around year 2 or 3 of his life. The mineralization of his second molar tooth began at the age of 2 and was completed by the age of 7. The mineralization of his last molar tooth started at the age of 7 and ended by the age of 16. Based on the data collected in this study (Table 2), he was exposed to high levels of lead at the neonatal and early infantile periods and to moderate levels in his childhood period. It should be also pointed out that he was heavily exposed to Cu. On the other hand, his enamel contained only trace amounts of Zn.

In the present study, we have established a method using SRXRF to determine heavy metal concentrations in human tooth enamel collected from humans and human skeletons from 3,000 years ago to present times. The values of the concentrations of the analyzed metals (Pb, Hg, Zn and Cu) obtained using SRXRF were compared to the values obtained using ICP-MS or AAS. This process enabled us to translate the relative amounts of heavy metals of interest by dividing the values by the Ca concentrations into absolute concentrations. The calibration method using relative concentrations against Ca has been employed traditionally [9, 10]. In the present study we also confirmed the usefulness of this method.

The dental enamel has been considered an ideal material for reconstruction of the exposure histories, because heavy metals incorporated into the enamel are encapsulated as they are chronologically absorbed during the subject’s growth [6, 7]. Therefore, this ability of the enamel can be fully utilized only by in situ analysis of the enamel metal content with a high-resolution method. For this purpose, a laser-abraded method coupled with ICP-MS or SRXRF seems to be promising. SRXRF has some advantages, since this method enables detection of the distribution of heavy metals with high resolution. As an example, in our study this method showed that enamels in K’s two teeth, developed in an infantile period, had high levels of Pb, presumably due to the levels of Pb contained in the breast milk of his mother, who may have used Pb-containing cosmetics.

Preliminary observations in the current studies warrant further studies. Exposures to Pb are highest in the Edo era in Japan as reported by others [22–25]. No trend was observed in this study in the Hg content of teeth for 3,000 years in Japan. The copper exposures are considered to be associated with an individual’s occupation. It is of particular interest that Zn concentrations are highest in modern humans. Since meat and cereal grains are rich in Zn [19], this observed long-term trend may result from nutritional improvements in modern humans.

This study lacked solid standard reference materials that are matrix matched for calibration purposes. Alternatively, we calibrated using ICP-MS or AAS, which lost information of special distribution for each heavy metal. This is the major limitation of this study. Thus, at present, we cannot fully utilize the advantages of SRXRF. This drawback will be overcome in the future.

In conclusion, we have developed a quantitative method using SRXRF with a calibration by ICP-MS and AAS. This method allowed us spatial high sensitivity with high resolution with appropriate external standards. We have applied this method to the reconstruction of human environmental exposures to heavy metals as well as determined the nutritional conditions of humans by the analysis of the heavy metal content of their teeth.