Introduction

Cerium dioxide nanoparticles (CeO2 NPs), as one of the most widely used engineered nanomaterials, have been prioritized for environmental characterization and assessment by the Organization for Economic Cooperation and Development (OECD). The market volume of CeO2 NPs was estimated to be 9.1 kilotons in 2016 [1], and usage is expected to increase due to their wide applications as catalysts, abrasive agents, and polishing agents [1, 2]. The downside of broad applications of CeO2 NPs is that NPs are leaking into the environment. There is increasing evidence of CeO2 NPs’ occurrence in rivers [3,4,5] and of anthropogenic Ce in air/urban dust [6] as well as in sediment/sewage sludge [7, 8]. Furthermore, the possibility of CeO2 NPs entering the food chain has been shown [9, 10]. Therefore, humans are being exposed to CeO2 NPs through various routes. This arouses concern. Hence, a better understanding of CeO2 NP biological fate is necessary to protect the environment and human health.

After uptake by mammals, CeO2 NPs may interact with biomolecules/chemicals, leading to biotransformation (e.g., dissolution of NPs and aggregation/agglomeration) [11]. For example, CeO2 NPs may dissolve ionic Ce or change in size after their uptake. In order to investigate the biological fate of CeO2 NPs in organisms, it is critical to first ensure release of the NPs and possibly dissolved Ce from the animal tissues. This requires appropriate sample preparation and analytical methods to separate both ionic and nanoparticulate Ce species followed by characterization/quantification. Previous studies have reported that ionic Ce is more toxic than CeO2 NPs at a similar concentration level [12, 13], possibly due to the Ce3+-induced conformational changes to catalase [14]. Furthermore, CeO2 NP size, another physicochemical property of NPs, also affects the toxicity of NPs [12, 15]. In other words, the toxicity of CeO2 NPs is closely related to the Ce species and NP size. To address the biotransformation (i.e., dissolution and size transformation in this study) of CeO2 NPs in vivo, firstly, Ce species and size of CeO2 NPs are required to be well preserved after their release from animal tissues; secondly, the analytical methods should be able to differentiate the Ce species as well as characterize the NP size after releasing both Ce species from animal tissues. Finding a sample preparation method that does not alter properties of NPs or dissolved species during the extraction or separation of Ce species from animal tissues is a challenge for in vivo investigations of CeO2 NPs. Previous studies have applied different methods to extract NPs from animal tissues, such as alkaline pretreatment using tetramethylammonium hydroxide (TMAH) and enzymatic pretreatment using proteinase K [16,17,18,19]. However, little attention has been given to evaluate the effect of sample preparation methods on ionic Ce and CeO2 NPs. Moreover, most studies focused on the characterization of CeO2 NPs but neglected the possible dissolution of CeO2 NPs [17, 18].

Worse still, limitations of analytical techniques in characterizing and quantifying CeO2 NPs as well as the possibility that dissolved species, results of biotransformation, may occur in biological matrices complicate the in-depth investigation of their biological fate. It is technically especially difficult to quantify dissolved NPs at low particle concentrations [20]. For in vivo investigation of NPs, the conventional analytical techniques for NP characterization, such as dynamic light scattering (DLS) and nanoparticle tracking analysis (NTA), are not suitable due to their relatively high detection limit. Microscopy techniques, such as transmission electron microscopy (TEM) and scanning electron microscopy (SEM), are limited to the local characterization of NPs in tissue and cannot provide quantitative information of NPs, such as their size distribution and particle number concentration. Among various techniques for NP investigation, inductively coupled plasma-mass spectrometry (ICP-MS) is promising. When appropriate separation system is hyphenated to ICP-MS, it shows good performance in both qualitative and quantitative investigations of metal-containing NPs [21]. Size exclusion chromatography (SEC) — it is showing great potential for fractionating and quantifying (nano)particulate and ionic Ce species in environmental water samples [22]. Multi-method approach, including extraction, (ultra)filtration and asymmetric flow field flow fractionation (AF4)-ICP-MS, was applied to separate and quantify ionic Ce species dissolved from CeO2 NPs in surface waters [23], leaching solution [24], Luria–Bertani (LB) medium [25], algae growth media [26], and the enzymatic digestate of plant tissue [27]. However, these techniques are limited to the CeO2 NPs suspended in environmentally or botanically relevant samples, and none have been developed to fractionate the (nano)particulate and dissolved Ce species in animal samples. This lack of methodology is a major factor for our lack of understanding of CeO2 NP biotransformation in vivo.

For NPs suspended in an aqueous solution, they can be introduced and analyzed directly by ICP-MS operated in single particle mode (sp-ICP-MS). This technique can provide size distribution, mass concentration, and particle number concentration of NPs simultaneously at very low concentration levels. Although the simple sample preparation method (i.e., dilution of the aqueous samples) entails the risk of possible severe matrix effect on accuracy, precision, and sensitivity in sp-ICP-MS analysis, our previous work has shown that sp-ICP-MS combined with internal standardization accurately characterizes CeO2 NPs in complex biological matrices [28]. Zhang et al. [27] have utilized Amicon® Centrifugal filter to separate dissolved Ce in a CeO2 NP suspension. The success of these studies demonstrates that there is a potential to use ICP-MS-based techniques to investigate the biotransformation of CeO2 NPs, particularly their dissolution and size transformation in vivo. This information will give us a quantitative and more comprehensive understanding of their fate after their uptake by organisms.

In this work, we aimed to develop a multistep ICP-MS-based analytical approach for extraction of nanoparticulate as well as ionic Ce species from animal tissue, quantification of Ce species, and then characterization of CeO2 NPs. The fractionation of ionic and nanoparticulate species was achieved using ultra-centrifugal units. Conventional ICP-MS was used to quantify different Ce species and the characterization of CeO2 NPs was completed by sp-ICP-MS. With this integrated ICP-MS-based analytical approach, a comparison of alkaline and enzymatic pretreatment was done. This integrated analytical approach then was applied to investigate the in vivo biotransformation of CeO2 NPs.

Materials and methods

Chemicals and materials

CeO2 NPs dispersed in aqueous solution (40 wt%, 30–50 nm) were purchased from US Research Nanomaterials, Inc. (TX, USA). Standard gold nanoparticles (AuNPs) with a nominal diameter of 60 nm, which was used to determine the transport efficiency in sp-ICP-MS analysis, were from BBI Solutions (Cardiff, UK). Ce(NO3)3·6H2O (REacton®, 99.99% (REO)) for method validation was obtained from Alfa Aesar, Thermo Fisher Scientific, UK. The standards of ionic Ce, gold (Au), and rhodium (Rh) at 1000 mg/L were obtained from VHG Lab (Teddington, Middlesex, UK). TMAH (25% v/v) used for alkaline pretreatment was purchased from Alfa Aesar (Ward Hill, MA, USA). Chemicals for preparing the enzymatic solution, including proteinase K from Tritirachium album lyophilized powder (BioUltra, ≥ 99.0%), ethylenediaminetetraacetic acid (EDTA) dipotassium salt dihydrate (≥ 98%), sodium dodecyl sulfate (SDS, BioUltra, ≥ 99.0%), and tris(hydroxymethyl)aminomethane (Tris) (ACS reagent, ≥ 99.8%), were purchased from Sigma-Aldrich (St. Louis, MO, USA). Ca2+- and Mg2+-free Dulbecco’s phosphate buffered saline (DPBS, pH 7.4) was obtained from Gibco, Life Technologies (Carlsbad, CA, USA). Trace-metal grade concentrated HNO3 and concentrated HCl as well as 30% stabilized H2O2 were purchased from VWR Chemicals (Poole, UK). Ultrapure water (18.2 MΩ cm at 25 °C) was obtained from a Milli-Q® Reference Water Purification System (Millipore, USA).

Characterization of CeO2 NPs

The CeO2 NPs without any surface modification were diluted with ultrapure water and were ultrasonicated to ensure the homogeneous distribution prior to usage. They were fully characterized by TEM (JEOL 2010F, Japan), DLS (Zetasizer Nano-S system, Malvern, UK), and X-ray photoelectron spectroscopy (XPS, SKL-12 spectrometer modified with VG CLAM 4 multichannel hemispherical analyzer) in our previous study [29]. The particle size of commercial CeO2 NPs was shown to be in the range of 30–50 nm by TEM, and most particles are round, which corresponds to the description provided by the manufacturer. The DLS measurement showed that the hydrodynamic size of CeO2 NPs in ultrapure water was 152.7 ± 1.6 nm and the average zeta potential was − 53.2 ± 1.0 mV. The XPS result showed that Ce4+ dominates on the surface of CeO2 NPs.

Animals and experimental design

Thirty-two 8-week-old female ICR mice were purchased from Laboratory Animal Service Centre, The Chinese University of Hong Kong (Hong Kong, China). The animal study was performed in compliance with The Hong Kong Code of Practice for Care and Use of Animals for Experimental Purposes [30]. Prior to the experimental treatment, they were acclimated in the animal room environment with a 12 h light/dark cycle (temperature of 22 ± 1 °C, relative humidity of 50–70%) and they had free access to food and water. After 5-day acclimation, they were randomly divided into vehicle control groups (n = 3) and experimental groups (n = 5). The experimental groups were intraperitoneally administered with 1.23 mg CeO2 NPs (as 1 mg Ce)/kg body weight dispersed in 200 μL DPBS, while the vehicle control groups were treated with DPBS with the same volume. All mice were weighed weekly to monitor their health. After the single intraperitoneal administration, mice were sacrificed on 1st, 7th, 28th, and 90th day(s) by cervical dislocation.

After the mice were terminated by cervical dislocation, livers, spleens, thymuses, and kidneys were removed after dissection. Parts of the liver tissues were frozen at − 80 °C for biochemical assays. The rest of the liver tissues, spleens, thymuses, and kidneys were lyophilized and stored at − 20 °C for CeO2 NP biodistribution and biotransformation analysis.

Development of an integrated ICP-MS-based analytical approach

A schematic diagram of the integrated ICP-MS-based analytical approach is shown in Fig. 1. Briefly, the total Ce concentration in samples was determined by conventional ICP-MS after the microwave-assisted acid digestion with enzymatic or alkaline pretreatment. The ionic Ce was separated by ultrafiltration and quantified by conventional ICP-MS. The characterization of CeO2 NPs was completed by sp-ICP-MS after simple dilution of the pretreated samples.

Fig. 1
figure 1

Schematic diagram of the integrated ICP-MS-based analytical approach

Full size image

Spiking experiments

Since no reference material containing CeO2 NPs or ionic Ce was available, the lyophilized ground liver tissue spiked with CeO2 NPs or ionic Ce was used to compare the effect of enzymatic and alkaline pretreatment on the recovery of each species. In addition, evaluation of the integrated ICP-MS-based analytical approach with suitable sample pretreatment methods was also completed with the spiked tissue.

Comparison of sample preparation methods

In order to extract the Ce species from animal tissues, both alkaline and enzymatic pretreatments were evaluated and their effect on ionic Ce as well as CeO2 NPs was compared. The alkaline pretreatment was developed based on a study where silver nanoparticles (AgNPs) were extracted from human placenta [19]. Briefly, 25% (v/v) TMAH solution was diluted to 20% with ultrapure water. Four milligrams of lyophilized tissue was weighed and digested with 2 mL of 20% TMAH solution. The samples were vortexed to mix thoroughly and left at room temperature overnight. The enzymatic solution was prepared by dissolving 0.2% (w/v) proteinase K, 0.8% EDTA (w/v), and 0.5% (w/v) SDS in 10 mM Tris–HCl buffer (pH 7.4). Four milligrams of lyophilized tissue was weighed and digested with 2 mL of enzyme solution (0.1 g of enzyme per g of lyophilized tissue). Samples were vortexed and then left in a water bath at 37 °C overnight.

Total mass concentration and ionic mass concentration

The sample solution right after the pretreatment was used for the total Ce (including CeO2 NPs and ionic Ce) quantification. Four hundred microliters of the enzyme-digested solution was mineralized by mixing with 3 mL concentrated HNO3 and 1 mL 30% H2O2 in a microwave digestion system (ETHOS-1 Advanced, Milestone S.r.l., Sorisole, Italy). Total Ce concentration (140Ce) was determined using ICP-MS (7900 Agilent Technologies, Inc., CA, USA). The instrument settings are listed in Table 1.

Table 1 Instrument settings for sp-ICP-MS and conventional ICP-MS analysis
Full size table

The ionic Ce was separated by ultrafiltration. Briefly, 400 μL of a sample solution was passed through Amicon® 3 K cellulose filters (Merck, Darmstadt, Germany) pre-rinsed with ultrapure water. The filtrate, containing mainly the ionic Ce content, was collected after centrifugation (\(\times\) 14,000 g, 22 °C, 30 min), and the ionic Ce was quantified by ICP-MS after microwave-assisted acid digestion as described above.

Characterization of CeO2 NPs using sp-ICP-MS

The Agilent 7900 ICP-MS operating in the time-resolved analysis (TRA) mode was used for the characterization of CeO2 NPs. The detailed instrumental settings and data acquisition parameters are listed in Table 1. The transport efficiency, defined as the ratio of the amount of target analyte entering the plasma to the amount of analyte aspirated, was determined daily using the particle size method developed by Pace et al. [31], in which the AuNP standard with a nominal diameter of 60 nm was used as the reference material. The flow rate was determined weekly by measuring the time spent by the peristaltic pump to take up 1 mL of water. Prior to the sp-ICP-MS analysis, the alkali-pretreated or enzyme-pretreated sample solutions were diluted with ultrapure water. Based on the instrumental settings, flow rate, and the transport efficiency of ICP-MS, the appropriate particle number concentration for NP characterization was in the range of 6,000–12,000 particles/mL.

Considering that the concentration of CeO2 NPs in tissue would be low in the real sample analysis, the dilution factors of the samples suited for sp-ICP-MS analysis may be small correspondingly. This would result in a pronounced matrix effect in sp-ICP-MS analysis. Therefore, the internal standardization developed in our previous work was applied in the sp-ICP-MS analysis for a more accurate characterization of NPs [28]. Briefly, the ionic rhodium (Rh) was used as the internal standard for the characterization of CeO2 NP by sp-ICP-MS, and dual mass mode (i.e., detection of two masses in a single measurement run) was applied to monitor 140Ce and 103Rh. Prior to the sp-ICP-MS analysis, the samples were spiked with 1 μg/L ionic Rh. 140Ce and 103Rh were monitored in sp-ICP-MS using dual mass mode. The raw data were processed by Microsoft Excel to generate the mass concentrations, size distributions, and particle number concentrations. The calculations were based on the presumption that CeO2 NPs are spherical, which corresponds to the TEM image of CeO2 NPs.

Statistical analysis

Values are presented as mean ± standard deviation (SD). Statistical differences were determined using one-way analysis of variance (ANOVA) followed by Tukey post hoc multiple comparison test. Statistical differences were considered to be significant when p < 0.05 and was indicated with *.

Results and discussion

Integrated ICP-MS-based analytical approach for fractionation of Ce species in animal tissues

Comparison of alkaline and enzymatic pretreatment

Previous investigations of CeO2 NPs in biological bodies have neglected the possibility of dissolution of CeO2 NPs. As a result, the efficiency of extraction methods was evaluated merely in terms of their effect on NPs, while the performance of the sample preparation method on dissolved Ce species was ignored. To fill this knowledge gap, in this study, we compared the extraction efficiency of alkaline and enzymatic pretreatment towards both CeO2 NPs and ionic Ce. Alkaline solubilization, commonly with TMAH, has been used to liberate AuNPs or AgNPs from animal tissues with good recovery [19, 32]. Similarly, enzymatic pretreatment, using proteinase to degrade proteins, was shown to be effective to extract AuNPs or AgNPs from animal tissues [19, 32, 33]. To simulate the possible low concentration of ionic Ce or CeO2 NPs in animal tissue, the spiked ionic Ce concentration was 5 mg Ce/kg dry tissue and the CeO2 NPs were spiked at a concentration of 4.06 mg Ce/kg dry tissue. Since no standard CeO2 NPs were available, a suspension of CeO2 NPs in ultrapure water was also prepared for CeO2 NPs in sp-ICP-MS analysis for comparison.

In this integrated ICP-MS-based analytical approach, the ionic Ce in the medium was fractionated by ultrafiltration and quantified by conventional ICP-MS. The characterization of CeO2 NPs was completed by sp-ICP-MS, in which internal standardization was applied to increase the accuracy.

Table 2 shows the recoveries of alkaline and enzymatic pretreatment. In terms of the extraction of ionic Ce, the low recovery (1.1 ± 1.0%) obtained after alkaline pretreatment indicated that ionic Ce might form precipitates in the TMAH-pretreated matrix. The result obtained by sp-ICP-MS further confirmed the presence of precipitates in the TMAH-pretreated matrix (Fig. S1). For enzymatic pretreatment, the recovery of ionic Ce achieved 89 ± 3%.

Table 2 Recovery of ionic Ce and CeO2 NPs in biological lyophilized tissue after enzymatic or alkaline pretreatment (n = 3)
Full size table

When extracting the CeO2 NPs from animal tissues using proteinase K, less than 1% of the CeO2 NPs was dissolved during the extraction process. In addition, the mass concentration recovery and the particle number recovery were over 90%, and the size distribution of CeO2 NPs after extraction was comparable with those in ultrapure water by sp-ICP-MS (Fig. 2). These results demonstrated that enzymatic pretreatment does not alter the state of CeO2 NPs; it would not cause dissolution or size aggregation of CeO2 NPs. In terms of alkaline pretreatment, the dissolution test was comparable with the case of enzymatic pretreatment. However, the recovery of mass concentration and particle number in sp-ICP-MS analysis reached 114.8 ± 1.0% and 223 ± 23.7%, respectively. Worse still, the size distribution of CeO2 NPs showed obvious left skew after alkaline pretreatment, indicating that the frequency of CeO2 NPs with smaller diameter increased. The increased mass concentration and particle number concentration in TMAH-pretreated matrix suggested that the TMAH-pretreated matrix might have impact on the transport efficiency during the sp-ICP-MS analysis. To further address this problem, standard AuNPs were spiked in the TMAH-pretreated matrix (500-times diluted) and enzyme-pretreated matrix (500-times diluted), and the results obtained by sp-ICP-MS were compared with the case of AuNPs suspended in ultrapure water (Table 3). Internal standardization was also applied to correct for the matrix effect on signal intensity. The transport efficiency was determined to be 6.6 ± 0.1% in TMAH-pretreated matrix, which was significantly larger than that determined in ultrapure water. While for the enzyme-pretreated matrix, the determined transport efficiency was comparable with that in ultrapure water. This result demonstrated that the matrix after alkaline pretreatment would increase the transport efficiency at low dilution level. It is believed that the presence of TMAH and organic carbon species after sample preparation decreases the surface tension, resulting in smaller droplets generated in the nebulizer [34, 35]. Hence, the transport efficiency was increased in sp-ICP-MS measurement. Although more accurate transport efficiency can be obtained by matrix-matched determination, it requires tedious preparation steps and is not practicable in real sample analysis, where the dilution factor is not always consistent for each sample.

Fig. 2
figure 2

Size distributions of CeO2 NPs in ultrapure water and of CeO2 NPs after extraction with TMAH and enzyme from animal tissues, which were obtained by sp-ICP-MS

Full size image
Table 3 Transport efficiency determined by AuNPs suspended in ultrapure water and alkali-pretreated matrix (500-times diluted) using the particle size method (n = 3)
Full size table

Through the comparison of alkaline and enzymatic pretreatment, it was found that enzymatic pretreatment not only showed satisfactory recovery of both ionic Ce and CeO2 NPs when extracting Ce species from animal tissues but also preserved the size of CeO2 NPs.

Performance of the integrated ICP-MS-based analytical approach

Enzymatic pretreatment was proven to be more effective to release ionic Ce as well as CeO2 NPs from biological system without altering the species or size of NPs. Through this integrated analytical approach, the total Ce concentration, ionic Ce concentration, and characterization of CeO2 NPs were studied based on the same enzyme-pretreated solution, thereby reducing any variation that might be introduced. In order to evaluate the performance of the integrated ICP-MS-based analytical approach using enzymatic pretreatment for sample preparation, both ionic Ce and CeO2 NPs were spiked in lyophilized tissue, at concentrations of 5 μg Ce/g tissue and 43 μg Ce/g tissue, respectively. The recoveries are shown in Table 4. A comparison of CeO2 NP characterization is shown in Fig. 3 using both sp-ICP-MS and TEM techniques: the size distribution of CeO2 NPs was comparable with that in ultrapure water (Fig. 3a), and the TEM photographs of CeO2 NPs after enzymatic pretreatment compared with the original CeO2 NPs (Figs. 3b and c) further demonstrate that enzymatic pretreatment will not change the size of CeO2 NPs.

Table 4 Recovery of mixed ionic Ce and CeO2 NPs in biological lyophilized tissue extracted with enzymatic pretreatment and the LOQ of each step in the integrated ICP-MS-based analytical approach (n = 3)
Full size table
Fig. 3
figure 3

Characterization of CeO2 NPs in ultrapure water and in enzyme-pretreated matrix in the case of ionic Ce and CeO2 NPs mixture. a Size distribution obtained by sp-ICP-MS; b TEM image of CeO2 NPs in ultrapure water; c TEM image of CeO2 NPs after enzymatic pretreatment

Full size image

This is the first study to evaluate the impact of enzymatic and alkaline sample pretreatment on Ce species (both ionic and nanoparticulate Ce species). In previous studies, only the impacts of the sample pretreatment methods on NPs have been investigated [17, 19, 32], while the ionic species, which might dissolve from the NPs in the biological system, were neglected. The lack in analytical technique hinders the in-depth investigation of CeO2 NP biotransformation. However, the integrated ICP-MS-based analytical approach with enzymatic pretreatment developed in this study shows satisfactory analytical performance in fractionation and characterization of Ce species. In the spiking experiment, both Ce species and CeO2 NP morphology were well preserved during the sample preparation process with enzymatic pretreatment, and thus any changes of Ce species or NP size measured with this analytical approach would be regarded as the in vivo biotransformation of CeO2 NPs after their release from animal tissues in the following study. This analytical approach enables us to investigate the in vivo dissolution and size transformation of CeO2 NPs, and hence provides potential insights into the mechanism of biotransformation.

Bioaccumulation and biotransformation of Ce in mice over time

Figures 4 and 5 show the bioaccumulation of Ce (both nanoparticulate and ionic Ce) in thymus, spleen, liver, and kidney after intraperitoneal administration. The concentrations of Ce in the tested organs of the NP-treated groups were significantly greater than those of the vehicle control groups, with the concentrations of Ce below the limit of quantification. Figures 6 and 7 show the size distribution and particle number concentration of CeO2 NPs. CeO2 NPs spiked in enzyme-pretreated solution are regarded as pristine NPs in the following discussion.

Fig. 4
figure 4

Concentrations of total Ce measured with conventional ICP-MS in a thymus; b liver; c spleen; d kidney (n = 5)

Full size image
Fig. 5
figure 5

Concentrations of ionic Ce measured with conventional ICP-MS in a thymus; b liver; c spleen; d kidney (n = 5)

Full size image
Fig. 6
figure 6

Size distributions of CeO2 NPs detected by sp-ICP-MS. CeO2 NPs showed minor size transformation in thymus (a, b), liver (c, d), spleen (e, f), and kidney (g, h) compared with pristine NPs and along with time

Full size image
Fig. 7
figure 7

Particle number concentrations of CeO2 NPs in a thymus; b liver; c spleen; and d kidney (n = 5)

Full size image

In our 90-day study, the highest concentration of Ce was found in the thymus (Fig. 4a). After the single intraperitoneal injection, the concentration of total Ce in the thymus was increased significantly on day 1 and showed continued increase on day 7. However, on day 28, the Ce concentration dropped significantly and remained at a similar level until day 90. It is believed that the prolonged, continuous increase of Ce concentration in the first week after exposure was caused by the way of administration — in the case of intraperitoneal administration, the NPs are taken up primarily by peritoneal macrophages in the abdomen, which then migrate to the thymus [36, 37]. During the first week post-exposure, the macrophages carrying Ce translocated slowly from the peritoneum to thymus [36], and then the macrophages entered the lymphatic system through lymph vessels, leading to the decrease of Ce in thymus after day 7. However, on day 28, no obvious increase of total Ce concentration was observed in the tested organs, especially in the spleen, which plays an important role in lymphatic circulation. We infer that the Ce translocated to the parathymic lymph nodes [37, 38]. In terms of the change of total Ce concentration, the thymus can be regarded as the “holding station” for in vivo NP translocation. Furthermore, the concentration of ionic Ce was quite low (< 0.5 μg/g) (Fig. 5a); ionic Ce accounted for only < 2% of total Ce (data not shown) in thymus. The size of CeO2 NPs did not change much compared with the pristine NPs (Fig. 6a). Gaussian fitted curves representing the CeO2 NP size distributions of the 1-day, 7-day, 28-day, and 90-day post-exposure groups overlapped, indicating the median diameter remained ~ 30 nm over time (Fig. 6b). The mass recoveries of NPs extracted from thymus in sp-ICP-MS analysis were all above 80% in all the NP-treated groups, illustrating that over 80% of the NPs can be detected by sp-ICP-MS (Table 5). These results suggest that CeO2 NPs barely changed in size after they translocated to thymus up to 90 days, further supporting the hypothesis that thymus plays the role of “holding station” in CeO2 NPs’ biological fate.

Table 5 Mass recoveries of NPs in sp-ICP-MS (n = 5)
Full size table

The concentration of total Ce was relatively stable in liver in this 90-day investigation (Fig. 4b), showing no statistical difference among the four NP-treated groups sacrificed on days 1, 7, 28, and 90. However, the dissolved Ce showed the highest concentration on day 1 and decreased to the lowest after 1 week. CeO2 NPs showed agglomeration on day 1 compared to the pristine NPs (Fig. 6c) and the obvious right skew from day 1 to day 7 suggested the median size of CeO2 NPs increased in the first week. From day 7 to day 90, the ionic Ce concentration in liver increased, while the size distribution of CeO2 NPs kept right-skewed, indicated by the rise in the median size from ~ 38 on day 7 to 45 nm on day 90 (Fig. 6d). These results indicate that ionic Ce was likely to be dissolved from the NPs with smaller diameters (25–30 nm) since the larger surface-to-volume ratio made them more susceptible to the complex environment after their translocation to liver. In the post-exposure days, CeO2 NPs agglomerated in liver.

In spleen, the ionic Ce followed a similar tendency as in liver. CeO2 NPs maintained a similar size distribution after they were translocated to spleen on day 1 compared to the pristine NPs (Fig. 6e). In the post-exposure period, the Gaussian fitted curves of CeO2 NPs showed little change (Fig. 6f). However, in the 90-day post-exposure group, the mass recovery as shown by sp-ICP-MS analysis was significantly lower, implying that ca. 40% of the Ce-containing particulates were smaller than 15 nm, approaching the size LOD in sp-ICP-MS analysis (Table 5). The significant drop of particle number concentration on day 90 (Fig. 7c) also suggested that the particles that can be detected by sp-ICP-MS decreased. In other words, by day 90, CeO2 NPs were smaller in spleen. Graham et al. [39] have observed ultrafine Ce particulates (< 3 nm) in spleen after a 90-day intravenous exposure to 30 nm cubic CeO2 NPs through high-resolution TEM.

In kidney, the Ce and ionic Ce concentrations consistently decreased during the 90-day investigation. The concentrations of CeO2 NPs of the four time point groups, obtained by subtracting ionic Ce concentration from total Ce concentration, showed no significant difference among groups (data not shown), indicating that the dissolved Ce was excreted from kidney. According to Molina et al. [40], the ionic Ce in kidney was excreted in urine. In terms of CeO2 NPs, an obvious decrease in normalized frequency of NPs with 25 nm was observed on day 1 compared to pristine NPs (Fig. 6g), and the median diameters of CeO2 NPs were ~ 30 nm in the following days (Fig. 6h). In addition, the particle number concentrations of CeO2 NPs showed no statistical difference during the post-exposure (Fig. 7d). In other words, CeO2 NPs with 25 nm diameter dissolved in kidney on day 1. These results suggested that dissolution of CeO2 NPs took place once they translocated to the kidney within the first day. In the following days, dissolved Ce probably was excreted from kidney in urine, while the rest of CeO2 NPs were retained in the kidney and their biotransformation seemed to be minor.

In summary, thymus is shown to be a “holding station” in CeO2 NP translocation in mice. The biotransformation of CeO2 NPs is organ-specific. In addition, the dissolution of CeO2 NPs in spleen and liver was prominent, indicating that it cannot be neglected when conducting in vivo investigation of CeO2 NPs.

Conclusion

In this study, we firstly compared the effect of alkaline and enzymatic pretreatment on ionic Ce and CeO2 NPs during the release of Ce species from animal tissues. Our study showed that, after alkaline pretreatment, ionic Ce formed precipitates in the sample matrix. Worse still, the TMAH-pretreated matrix resulted in increased transport efficiency in sp-ICP-MS analysis, leading to the overestimation of CeO2 NP mass concentration and particle number concentration. Enzymatic pretreatment resulted in excellent recovery of ionic Ce and CeO2 NPs without altering their properties after the sample preparation. The performance of this integrated ICP-MS-based analytical approach, in which enzymatic pretreatment is used to release Ce species from animal tissues and multistep analysis is applied to fractionate, quantify, and characterize Ce species, was evaluated using lyophilized tissue spiked with a mixture of ionic and nanoparticulate Ce. The satisfactory recoveries and the comparable size distribution of CeO2 after enzymatic pretreatment and that of primary CeO2 NPs demonstrated the competence of this ICP-MS-based approach in quantifying and characterizing ionic and nanoparticulate Ce species. We then applied this integrated approach to investigate the bioaccumulation and biotransformation of CeO2 NPs in mice after the single intraperitoneal administration. It was found that CeO2 NP biotransformation is organ-specific, and thymus is likely to be a “holding station” during the in vivo translocation of CeO2 NPs. The success of the integrated ICP-MS-based analytical approach in analyzing CeO2 NP in vivo biotransformation highlights its potential in analyzing NP dissolution or size transformation in biological matrices. With this powerful analytical approach to characterize the transformed species after NP biotransformation, the mechanism of biotransformation can be better understood.