Synchrotron-Based X-ray Fluorescence Imaging Elucidates Uranium Toxicokinetics in Daphnia magna

A combination of synchrotron-based elemental analysis and acute toxicity tests was used to investigate the biodistribution and adverse effects in Daphnia magna exposed to uranium nanoparticle (UNP, 3–5 nm) suspensions or to uranium reference (Uref) solutions. Speciation analysis revealed similar size distributions between exposures, and toxicity tests showed comparable acute effects (UNP LC50: 402 μg L–1 [336–484], Uref LC50: 268 μg L–1 [229–315]). However, the uranium body burden was 3- to 5-fold greater in UNP-exposed daphnids, and analysis of survival as a function of body burden revealed a ∼5-fold higher specific toxicity from the Uref exposure. High-resolution X-ray fluorescence elemental maps of intact, whole daphnids from sublethal, acute exposures of both treatments revealed high uranium accumulation onto the gills (epipodites) as well as within the hepatic ceca and the intestinal lumen. Uranium uptake into the hemolymph circulatory system was inferred from signals observed in organs such as the heart and the maxillary gland. The substantial uptake in the maxillary gland and the associated nephridium suggests that these organs play a role in uranium removal from the hemolymph and subsequent excretion. Uranium was also observed associated with the embryos and the remnants of the chorion, suggesting uptake in the offspring. The identification of target organs and tissues is of major importance to the understanding of uranium and UNP toxicity and exposure characterization that should ultimately contribute to reducing uncertainties in related environmental impact and risk assessments.


Summary
The following supplementary materials include description of experimental methods and additional results.

S1. Uranium Nanoparticle Synthesis and Characterization
All chemicals, unless stated otherwise, were of analytical grade (Merck, Czech Republic).
Deionized water was used for the preparation of aqueous solutions. Uranyl nitrate (Lachema, CSSR) was annealed at 1200 °C for 2 h. The purity of U3O8 was evaluated by X-ray powder diffraction (XRD).
A reaction mixture was prepared by dissolution of U3O8 (5.614 g, 6.7 mmol) in 4 mL of concentrated nitric acid (65 %), which corresponds to 10 % excess of nitric acid. The dissolution was accompanied by the release of nitric fumes. Following the complete dissolution of U3O8, the resulting solution was diluted with deionized water to prevent precipitation after addition of propan-2-ol. Finally, 200 mL of propan-2-ol (10 %vol) was added and the solution was again diluted with deionized water to a total volume of 2 L.
The resulting concentration of UO2 2+ was 10 mM. This solution was stirred with a magnetic stirrer and irradiated for 150 min in a photochemical reactor with immersed quartz-protected low-pressure mercury lamps (variable power input: 400 W (nominal value), wavelength 254 nm; Philips TUV 25WP SE) and cooled by air-ventilators. The formed dark grey product was separated from solution by centrifugation, washed in ethanol in order to remove synthesis residues, and subsequently air dried at 40 °C.
The final material was also characterized by XRD using a Rigaku MiniFlex 600 (Ni-filtered Cu-Kα1,2 radiation) equipped with a NaI:Tl scintillation detector. XRD patterns were compared to the relevant records in the ICDD PDF-2 database (version 2013). The angular range was 10° -80°, with a step of 0.02° and a scanning speed of 2°/min.

Dynamic Light Scattering Measurements
Zeta-average hydrodynamic diameters and zeta-potentials of the UNP stock suspensions were determined by dynamic light scattering (DLS) using a Malvern Zetasizer ZS (Malvern Instruments Ltd., Worcestershire, United Kingdom) equipped with a 633 nm laser. Zeta-average hydrodynamic diameter measurements were conducted in triplicate, 5 runs each, with autocorrection function of 10 s. Electrophoretic mobility (zeta potential) was determined by Smoluchowski approximations.

Electron Microscopy
Particle crystalline structure and individual sizes were confirmed by high resolution (HR) transmission electron microscopy (TEM), while scanning transmission electron microscopy (STEM) with energy dispersive X-ray spectroscopy (EDS) was used for elemental composition. Immediately following sonication, 10 µL of UNP stock suspension was added to a 400 mesh formvar-carbon film (Agar Scientific Ltd., Essex, United Kingdom) and allowed to air dry. Samples were measured at 200 kV accelerating voltage on a JEOL JEM-2100F equipped with a Gatan Porius 200D CCD camera (JOEL Ltd., Tokyo, Japan). Fluorescent X-rays were collected by an Oxford X-Max-80 SDD EDS detector at a 0.23 srad collection angle.

Uranium Reference Solution Preparation
To compare with the UNP exposure, a U reference (URef) solution was prepared from a 1.0 g L -1 uranium oxide (U3O8) assay and isotopic standard (CRM 129-A, U.S. Department of Energy, Argonne, Illinois, USA). A stock solution of 100 mg U L -1 was prepared in ddH2O and aliqouts of this stock were added directly to empty 50 mL plastic exposure cups (Graduated Polypropylene, VWR, Radnor, PA) to result in an exposure solution with a given U concentration. The solutions were evaporated to dryness and, 24 h prior to the start of the daphnia exposure, re-dissolved with 25 mL of exposure media (i.e., MHRW at pH 6.8).
Particulate and colloidal U species in the URef solutions were examined by TEM analysis.
A 5 mg U L -1 URef solution was prepared in MHRW (pH 6.8) by the same procedure described previously. After 24 h, the solution was sampled by filtering the 50 mL solution 5 through a 3 kDa filter (Amicon Millipore, Billerica, MA) in 15 mL increments at 5,000 g for 90 min each. Next the filter cartidge was washed using 10 mL of ddH2O (15 MΩ cm) and centrifuged again at 5,000 g for 90 min. Samples for TEM analysis were taken from 10 µL of retentate found in the bottom of the filter cartidge and placed onto a 400 mesh lacy carbon film (Agar Scientific Ltd., Essex, United Kingdom) and allowed to air dry. Samples were measured at 200 kV accelerating voltage on a JEOL JEM-2100F equipped with a TVIPS TemCam-XS416 (ES). Fluorescent X-rays were collected by an Oxford X-Max-80 SDD EDS detector at a 0.23 srad collection angle.

Exposure Media Size Fractionation
Size fractionation was used to determine the size distribution of U species at 0, 24, and 48 h after the start of the UNP and URef exposures. The size fractions were particulate (> 0.45 µm), colloidal (3 kDa < x < 0.45 µm), and low molecular mass species (LMM, < 3 kDa). In sampled exposures, 1 mL of media was passed through a pre-conditioned (1 mL) 0.45 µm syringe filter (VWR, Radnor, Pennsylvania, United States) and 100 µL was sampled from the filtrate. Next, 400 µL was sampled from the < 0.45 µm solution into a pre-conditioned (300 µL) 3 kDa Amicon cellulose membrane filter (Amicon Millipore, Billerica, MA) and centrifuged at 14,000 g for 30 min. From the filtrate, 100 µL was sampled prior to the QQQ-ICP-MS measurement.

Inductively Coupled Plasma Mass Spectrometry
Elemental analysis of exposure solutions, size fractionation samples, and digested daphnid was conducted by triple quadrupole inductively coupled plasma mass spectrometry (ICP-QQQ, Agilent 8900). Aliquots were diluted in 5 % HNO3 (V/V) 48 h prior to measurement of U. Measurements of digested daphnia had a limit of detection (LOD) of 0.008 µg 238 U L -1 and a limit of quantification (LOQ) of 0.026 µg 238 U L -1 .
Measurements of digested water samples for media characterization had a LOD of 0.003 µg 238 U L -1 and a LOQ of 0.009 µg 238 U L -1 while for daphnid measurements the LOD was 0.008 µg 238 U L -1 and the LOQ was 0.026 µg 238 U L -1 .

S4. Synchrotron X-ray Fluorescence Analysis and Image Processing
The hard X-ray microprobe facility (microXAS -X05LA) at the Swiss Light Source (Paul Scherrer Institute, SLS, Switzerland) was used to evaluate the whole body elemental distribution in prepared D. magna samples (Fig. S9). The incident beam size was 1 µm 2 using X-rays focused by a Kirkpatrick-Baez (KB) mirror system with an energy of 17.2 keV selected with a fixed-exit double-crystal monochromator. The resulting photon flux was 2 × 10 10 ph s -1 . Daphnid samples were raster scanned with a step size of 5 µm for the whole organism maps and 2 µm for regions of interest (ROI). X-ray fluorescence

Determination of Detection Limits
A reference standard containing 1.5 µg Cu cm -2 (1.4 × 10 7 atoms µm -2 ) was measured during the synchrotron experiments and yielded 1370 photon counts s -1 for Cu Kα and Kβ lines, integrated over 160 s. As a result, 100 cps was resolvable leading to a detection limit of 1 × 10 6 Cu atoms µm -2 , integrated over 1 s.

Attenuation Correction
The effects of self-absorption were not observed and no attenuation correction was applied in this experiment. The attenuation in denser regions of the daphnid samples, such as some hotspots, never exceeded 5 % absorption of the 17.2 keV beam.
Furthermore, the 4 SSDs surrounding the sample, which mostly contained empty space, were combined to minimize the effects of self-absorption.

Spectra Fitting and Image Processing
The resulting sum spectra encompassing the signals from all 4 SSDs for each XRF measurement was opened and fitted using PyMCA. Example sum spectra can be seen in Figure S13 for the UNP (A) and URef (B) exposed daphnids, respectively. Upon applying a fit, the resulting TIFF images were exported to ImageJ, where the maps were converted to a logarithmic scale and a Look Up Table (           distributions on a UNP exposed daphnid (left) and a URef exposed daphnid (right). The