- Email: [email protected]
Uranium quantification in semen by inductively coupled plasma mass spectrometry
Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
Contents lists available at SciVerse ScienceDirect
Journal of Trace Elements in Medicine and Biology journal homepage: www.elsevier.de/jtemb
ANALYTICAL METHODOLOGY
Uranium quantification in semen by inductively coupled plasma mass spectrometry夽 Todor I. Todorov a,∗ , John W. Ejnik b , Gustavo Guandalini c , Hanna Xu c , Dennis Hoover d , Larry Anderson d , Katherine Squibb e , Melissa A. McDiarmid e , Jose A. Centeno c a
Crustal Geophysics and Geochemistry Science Center, United States Geological Survey, PO Box 25046, DFC, Bldg. 20, MS 964D, Denver, CO 80225, United States Department of Chemistry, University of Wisconsin-Whitewater, 800 West Main Street, Whitewater, WI 53190, United States c Division of Biophysical Toxicology, Depleted Uranium and Embedded Fragment Laboratory, The Joint Pathology Center, Silver Spring, MD 20910-1290, United States d University of Maryland, School of Medicine, Department of Anatomy and Neurobiology, 20 South Pine Street, Baltimore, MD 21201, United States e University of Maryland, School of Medicine, Department of Medicine, 11 South Paca Street, 2nd Floor, Baltimore, MD 21201, United States b
a r t i c l e
i n f o
Article history: Received 23 March 2012 Accepted 2 July 2012 Dedicated to the memory of Dr. Larry Anderson. Keywords: Uranium ICP-MS Semen
a b s t r a c t In this study we report uranium analysis for human semen samples. Uranium quantification was performed by inductively coupled plasma mass spectrometry. No additives, such as chymotrypsin or bovine serum albumin, were used for semen liquefaction, as they showed significant uranium content. For method validation we spiked 2 g aliquots of pooled control semen at three different levels of uranium: low at 5 pg/g, medium at 50 pg/g, and high at 1000 pg/g. The detection limit was determined to be 0.8 pg/g uranium in human semen. The data reproduced within 1.4–7% RSD and spike recoveries were 97–100%. The uranium level of the unspiked, pooled control semen was 2.9 pg/g of semen (n = 10). In addition six semen samples from a cohort of Veterans exposed to depleted uranium (DU) in the 1991 Gulf War were analyzed with no knowledge of their exposure history. Uranium levels in the Veterans’ semen samples ranged from undetectable (<0.8 pg/g) to 3350 pg/g. This wide concentration range for uranium in semen is consistent with known differences in current DU body burdens in these individuals, some of whom have retained embedded DU fragments. Published by Elsevier GmbH.
Introduction Uranium is a naturally occurring element with an average abundance in the earth’s crust of approximately 2.7 g/g. Its isotopic composition consists of three naturally occurring radioactive isotopes: 234 U (0.0055%), 235 U (0.72%), and 238 U (99.27%), decaying to 208 Pb and 206 Pb as final non-radioactive products. The main use of uranium is for nuclear fuel, but because the principal isotope in natural uranium is the isotope with the lowest radiologic activity, 238 U, an enrichment of 235 U is required to produce a nuclear, fuel-grade product. A by-product of the enrichment process is depleted uranium (DU) which has an approximate isotopic composition of: 234 U (0.00061%), 235 U (0.2%), and 238 U (99.79%) [1]. Additionally, some studies have shown that DU contains small quantities of 236 U coming from reprocessed uranium fuel [2]. DU is used commercially in a
夽 The opinions and assertions expressed herein are those of the authors and are not to be construed as official or as representing the views of the United States Geological Survey, The Joint Pathology Center, the Department of the Army, or the Department of Defense. ∗ Corresponding author. Tel.: +1 303 236 1243; fax: +1 303 236 3200. E-mail address: [email protected] (T.I. Todorov). 0946-672X/$ – see front matter. Published by Elsevier GmbH. http://dx.doi.org/10.1016/j.jtemb.2012.07.004
variety of applications, such as counterweights in aircraft, shielding for radiography cameras and manufactured chemicals containing uranium. DU exposure in civilian populations could occur in occupational settings or in unusual circumstances such as inhalation of burning aircraft materials following a plane crash [3]. DU has also been used by the military. Because of its high density, availability, and low relative cost, it has been incorporated into munitions as high energy kinetic penetrators and into armor plates for military vehicles. Soldiers in battle are therefore at risk of inhaling DU aerosols, ingesting DU particles, and/or experiencing wound contamination by DU particles and/or embedded fragments [4–6]. Because of the toxicological properties of uranium [7–9], the health effects of natural and depleted uranium exposure have long been of concern. Recent exposures to DU in military personnel have led to renewed studies of uranium’s chemical and radiologic health effects as an internal toxicant [6,10,11]. While uranium may be stored in the bone, it also accumulates in the kidney which is considered the “critical” organ for uranium toxicity [9,11]. Additionally, as with other metals, there are concerns about potential central nervous system and reproductive health effects [6]. Numerous reports have attempted to evaluate the relationship between essential and non-essential trace element composition of semen and fertility, semen quality, and/or sperm characteristics.
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
Among the essential elements that are usually and naturally elevated in semen are calcium (Ca) and zinc (Zn), for which concentrations in semen are higher than those in blood [12]. These two metals have important associations with male reproductive functions. Calcium in the semen is essential for sperm motility [13,14]. Zinc in semen may be part of the bacteriostatic action of semen that protects the male reproductive tract [15,16]. Zn may also be one component of the mechanism which maintains the proper balance of oxidants and antioxidants in semen [17,18]. It also has an influence on motility of ejaculated sperm [19]. In addition to the “intrinsic” trace elements in semen discussed above, humans with occupational, as well as certain nonoccupational, exposures have been shown to exhibit elevated semen levels of toxic trace elements. Studies in animals and humans have shown that male reproductive system and semen quality effects are associated with elevated body burdens of such toxic elements as lead (Pb) and cadmium (Cd). Higher semen concentrations of these metals have been reported for certain exposures [20] and levels in semen have been shown to be negatively correlated with sperm concentrations and motility in humans [21]. The effects of uranium at different concentrations in semen have not been reported, but there are recent reports investigating the reproductive effects of depleted uranium using laboratory animals. Arfsten et al. reported no significant changes in sperm motility or sperm count in rats with implanted DU pellets [22,23]. In a followup two-generation, reproductive toxicity study in rats, the same authors concluded that DU exposure is not a significant reproductive hazard, with a caution of possible developmental effects (increased mean relative heart weights in rat pups from adult males exposed to DU) [24]. Several analytical methods have been described for measuring uranium in environmental, geological, and biological samples, including thermal ionization mass spectrometry (TIMS), instrumental neutron activation analysis (INAA), delayed neutron counting (DNC), inductively coupled plasma mass spectrometry (ICP-MS), inductively coupled plasma optical emission spectrometry (ICP-OES), ␣-spectroscopy, spectrophotometry, fluorometry, and kinetic phosphorescence analysis (KPA) [25]. However, this metal has not been measured in semen to our knowledge. In this study we focused on ICP-MS as a technique of choice, as it provides low detection limits suitable for situations with low metal concentration and where limited sample is available, short sample preparation and instrument analysis time, and high sample throughput. Although uranium levels have been quantified in various types of biological samples including tissues and biological fluids, a number of investigations have focused mainly on urine U and to a smaller extent on blood U quantification [25–28]. The sample preparation for urine uranium analysis (ranging from digestion and wet or dry ashing to no pretreatment and simple dilution) for ICP-MS varies greatly between the different laboratories, based on the detection limits desired and the sample introduction system used for the analysis [29–34]. The primary objective of this study was to develop an assay that accurately measures the quantity of U in semen specimens. A simple robust method is presented herein using acid digestion followed by ICP-MS analysis. This method provides low detection limits and high accuracy and precision for determination of uranium in human semen at low levels.
Materials and methods Semen sample collection and processing Collection of samples for a pool of semen used in the development of this assay was performed at the University of Maryland
3
Baltimore (UMB) School of Medicine. Collection of samples from Veterans of Gulf War I was performed at the University of Maryland Medical Center under the auspices of the Depleted Uranium Surveillance Program of the Baltimore Veterans Affairs Medical Center. All study participants gave informed consent and the protocol was approved by both the VA and University Institutional Review Board (IRB). All semen processing occurred at the UMB School of Medicine. All men participating in the collection of semen were requested to abstain from ejaculation for two days prior to sample production. Semen was produced by masturbation and collected in a sterile 118 mL polystyrene specimen container. Samples for pooled control semen were collected from two non-Veterans over a period of 7 months with reported abstinence of 1–5 days. Volumes ranged from 0.5 to 3.5 mL for each sample. Samples were obtained from Gulf War I Veterans during their 2009 biennial surveillance visit [10]. For these samples, abstinence ranged from 2 to 30 days (6.3 ± 7.7 days, mean ± s.d.) and semen mass ranged from 0.46 to 7.998 g (approximately 0.45–7.95 mL). After transport to the laboratory, all samples were incubated for 3 h at 37 ◦ C to encourage liquefaction before subsequent processing. Additives which can be employed to increase liquefaction of the samples were not added to the semen during sample processing. After incubation at 37 ◦ C, samples for pooled semen were transferred to 50 mL polystyrene centrifuge tubes to a maximum of approximately 40 mL (semen from the two donors was kept separate) which were stored at −20 ◦ C. To produce the semen pool, all aliquots were thawed at 4 ◦ C for 5 h. Aliquots were vortexed and poured into acid-washed teflon beakers, first combined by donor and then combined into a common pool. For sampling, the pooled semen was initially mixed in the covered teflon beakers with an acid-washed teflon-coated stir bar, and stirring continued during sampling (speed decreased as volume remaining decreased). Aliquots of 2 mL from the final pool were prepared in the vials used for uranium determination (screw-capped, teflon, 15 mL, numbered precleaned vials; Savillex, Eden Prairie, MN) with an air displacement, polystyrene-tipped pipettor. Aliquots were stored at −20 ◦ C. For Veterans, each ejaculate was transferred in its entirety after incubation at 37 ◦ C with a graduated glass pipet directly into a teflon vial for uranium determination. These samples were stored at −20 ◦ C and all Veteran and pooled semen samples were transferred frozen (on ice) to the Division of Biophysical Toxicology – Depleted Uranium and Embedded Fragment Laboratory (DBT-DU/EMF) for uranium analysis. All specimens, including pooled semen at the time of aliquotting and Veteran samples processed immediately after collection, were transferred into teflon vials of known individual weights and the weight of vial plus semen was recorded. Semen weight was determined as the difference between weight of vial plus semen and the weight of the empty vial. All subsequent quantification calculations for uranium in semen were based on semen weight rather than volume as the volume measured was approximate (particularly in samples with incomplete liquefaction or high viscosity). At DBT-DU/EMF laboratory, the samples were stored at −70 ◦ C until analysis. Preparation of spiked samples Two milliliter (2 mL) aliquots (weight 2.081 ± 0.252 g mean ± s.d., n = 97) from the pooled control semen were spiked at the following three levels for method validation and generation of quality control samples: (1) low level spike at 5 pg/g (added as 0.02 mL at 500 pg/mL to give a nominal concentration of 5.0 pg/g), (2) medium level spike 50 pg/g U (added as 0.02 mL at 5 ng/mL to achieve a concentration of 50 pg/g) and (3) high level spike at 1000 pg/g U (added as 0.02 mL at 100 ng/mL for a nominal
4
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
Table 1 Operating conditions for total uranium quantification in semen inductively coupled plasma mass spectrometry (ICP-MS). Instrument
Perkin Elmer Elan DRC II
ICP-MS operating parameters Plasma power Nebulizer gas flow rate Auxiliary gas flow Plasma gas flow Interface cones Rpa RPq Cell gas Monitored masses Dwell time Detector dead time Number of sweeps Number of replicates Scan mode
Setting 1400 W 1.04 L min−1 1.1 L min−1 15 L min−1 Pt 0 0.7 None m/z: 233.04, 238.05 100 ms 55 ns 25 3 Peak hopping
concentration of 1000 pg/g). The solutions used for the preparation of the spikes were prepared from Spex single element uranium ICP-MS standard by serial dilutions (Spex CertiPrep, Metuchen, NJ). The spiked pooled semen samples used for assay accuracy validation were prepared in triplicate. Quality control samples of spiked semen were prepared and used along with digestion blanks and unspiked semen. These samples were distributed throughout the digestion and analysis of all samples associated with this study. Sample preparation for analysis Before ICP-MS analysis: (1) semen samples were thawed at room temperature, (2) internal standard was added, (3) high purity nitric acid (Optima grade 70% HNO3 , ThermoFisher Scientific, Waltham, MA) was added in a volume equal to the mass of semen sample with a maximum of 6 mL of acid in order to avoid overfilling of the digestion vessel, and (4) samples were digested in a muffle furnace at 110 ◦ C until dryness. Following the digestion, each sample was reconstituted at a known volume of approximately 1:2 ratio of semen weight to volume of 10% nitric acid for ICP-MS, with the exception of low volume samples (<1.0 mL) that were reconstituted to a final volume of 1 mL. Chymotrypsin and bovine serum albumin samples Chymotrypsin is sometimes added to semen samples during sample processing in order to increase liquefaction. When this enzyme is employed, its action can be quenched by the addition of bovine serum albumin (BSA). In this study, the uranium content of chymotrypsin (Semen Viscosity Treatment System VTS-20, Conception Technologies, San Diego, CA) and BSA (BSA-embryo culture tested A3311, Sigma Chemical Co., Saint Louis, MO) was measured to evaluate the potential contamination of the semen with external uranium when these agents are used.
Table 2 Analytical performance/figures of merit. Sample weight, g Background, cps Sensitivity, cps per ppb Detection limit (3) in aqueous solution, pg/mL Detection limit (3) in semen, pg/g Precision, RSD (n = 34)
0.5–7 0.3 8.3 × 104 0.04 0.8 1.4–7%
25 pg/mL 233 U (CRM 111A, New Brunswick Laboratory, Argonne, IL, USA; certified for concentration, isotopic composition information values: 99.4911% 233 U, 0.1874% 234 U, 0.0790% 235 U, 0.0166 236 U and 0.2286% 238 U) was used in the blanks, standards, validation, quality control and unknown samples (added prior to digestion) to correct for instrument drift and sample matrix effects. Since the internal standard was also added to the blank that was used to correct all ICP-MS samples, the ∼0.05 ng/L 238 U that it contributed to the m/z 238 signal was subtracted from all the standards and samples. This procedure ensured accurate quantification of the 238 U as no systematic error was contributed by the 233 U internal standard. The quantification was achieved by monitoring the m/z at 233.04 for 233 U+ and 238.05 for 238 U+ . The instrument detection limit (IDL) for uranium by ICP-DRC-MS was calculated based on 3 times the standard deviation of blank measurements (n = 6) consisting of 10% nitric acid. The method detection limit was calculated using 3 times the standard deviation of pooled unspiked semen samples (n = 10). For each batch of digested unknown samples, a set of quality control samples were prepared. These included a digestion blank (which consisted of empty sample vials processed through the digestion process), and the quality control samples (an unspiked pooled semen sample and three spiked pooled semen samples at low, medium and high level). Blank correction for all semen samples was performed using the mean of digestion blank samples. Results Figures of merit The instrument and method parameters were optimized and Table 2 shows the analytical performance for the quantitative uranium determinations. The IDL obtained from previous studies for total uranium was determined to be 0.04 pg/mL [28,29] in aqueous solutions. The method detection limit (MDL) was 0.8 pg/g in semen. Samples of the pooled control semen were prepared using the same sample procedure as the unknown samples (i.e. digested, reconstituted in nitric acid and analyzed by ICP-MS) for determination of the MDL. The uncertainties (relative standard deviation, mean divided by the standard deviation) associated with the quantitative measurements of the semen samples were 1.4–7%, and were calculated using the pooled semen samples (unspiked and spiked, n = 34) used for the method validation and as QC materials. Method validation
Uranium analysis The quantitative analysis of uranium in semen samples was performed by ICP-MS employing a dynamic reaction collision cell to reduce polyatomic interferences (ELAN DRC II, manufactured by PerkinElmer, Waltham, MA, USA). The operational parameters of the instrument were set to optimize the detection of uranium, as summarized in Table 1. The concentration of uranium was determined by using an external calibration curve of 0.10, 1.0, 10, 50, and 250 ng/L U. The calibration standards were prepared using a 1000 mg/L uranium ICP-MS standard solution (Spex CertiPrep, Metuchen, NJ) through serial dilutions. An internal standard at
Three spike levels of uranium were created in pooled semen samples and analyzed along with samples of unspiked semen. The 233 U signal corrects for biological matrix effects and any instrument changes during the analysis. Since 233 U and 238 U have the same ionization properties, 233 U as an internal standard corrects for any matrix effects from the sample and instrumentation variations during the analysis. The 238 U isotope was used for quantification of total uranium. The small change in 238 U composition from natural uranium’s 99.274% to DU’s 99.794% was determined to not significantly alter quantification in samples containing DU derived from clinical studies. The results for unspiked and three different levels
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6 Table 3 Comparison of measured semen uranium with known uranium concentrations. Spike levela
Measured U concentration (pg/g)b
Unspiked (10) 2.9 (0.2) Low (3) 7.5 (0.1) Medium (3) 49.8 (0.3) 917 (7) High (3) a b
Calculated U concentration (pg/g)
Recovery (%)
7.7 50.1 942
98 100 97
Number of replicate samples shown in parentheses. Numbers in parentheses are standard deviations.
of U spikes of pooled semen are shown in Table 3. The uranium concentration in the pooled, unspiked semen was 2.9 pg/g. The “calculated uranium concentration” in Table 3 is equal to the mean measured uranium quantity in the pooled semen plus the uranium quantity added by each spike divided by the mass of each pooled semen sample and the mass spike solution added. The recoveries of the spikes were between 97 and 100% for all three spike levels. Analyses of chymotrypsin and BSA for uranium content In a previous study in our laboratory (data not published), the quantification of uranium in chymotrypsin and BSA indicated that these additives could be a significant source of sample contamination. Chymotrypsin and BSA uranium levels were 90 ± 7 pg/mg (mean ± s.d.), and 10.6 ± 0.1 pg/mg, respectively. Addition of 5 mg of chymotrypsin to 2 g of semen sample would contribute 225 pg/g of uranium. Similarly, addition of 50 mg of BSA to a 2 g semen sample would increase the uranium content by 265 pg/g. When compared to the control pooled semen uranium concentration of 2.9 pg/g, this is a significant level of contamination. Therefore, liquefaction enhancement additives were not used in this study. Uranium analysis of semen samples from DU-exposed Gulf War I Veterans Veterans attending the Baltimore Veterans Administration Medical Center for the 2009 biennial Depleted Uranium Surveillance visit were requested to produce semen samples for uranium analysis. These Gulf War I Veterans were exposed to DU in friendlyfire incidents involving DU munitions and tanks with DU armor [10]. Data from a subset (n = 6) of the semen samples analyzed are presented here (Table 4). In our studies of this DU-exposed Gulf War I cohort, the on-going exposure to uranium is monitored by urine uranium excretion (assayed as 238 U) corrected for creatinine concentration and expressed as g of uranium per g creatinine [10]. Some of the men in this cohort have embedded DU fragments and, as the fragments oxidize in situ, these men continually excrete elevated uranium in their urine. The Veterans in the study group are divided into 2 populations: low and high uranium exposure based on their Table 4 Depleted uranium (DU) exposed Gulf War I participant semen samples analyzed for total uranium (U) content. Veteran
Semen quantity (g)
U concentration (pg/g)
Uranium exposure groupa
A B C D E F
4.028 2.488 3.082 5.597 2.918 1.486
<0.8 2.8 24.1 159 226 3350
Low Low High High High High
a Exposure group assignment based on high exposure group = urine uranium ≥ 0.1 g/g creatinine and low exposure group = urine uranium < 0.1 g/g creatinine.
5
urine uranium excretion. Men in the high urine uranium group have creatinine standardized urine uranium at or above 0.1 g/g creatinine. The exposure designations shown in Table 4 are based on this exposure criterion. For the 6 Veterans listed in Table 4, semen uranium concentration ranged from undetectable (<0.8) to 3350 pg/g semen. Values for the two low uranium exposure men include the undetectable sample and a semen uranium concentration (2.7 pg/g semen) very close to the pooled control semen obtained from two non-Veteran donors used in the method validation and characterization (2.9 pg/g semen). The high uranium exposure men had a mean semen uranium concentration of 905 pg/g semen (s.d. 1610), exhibiting a very wide range of values (24.0–3350 pg/g).
Discussion and conclusions In our previous studies with urine and blood samples using reaction cell quadrupole ICP-MS instruments (for removal of polyatomic interferences) [29], we found that the use of bandpass mass filtering (rpq increased to 0.7) improved both precision and accuracy of uranium quantitative and isotopic determinations. In the current investigation, we utilized only bandpass mass cutoff within the reaction cell quadrupole (without use of cell gas). The bandpass mass limit, rpq parameter, was set at 0.7 in order to decrease the amount of low ion masses entering the analyzer quadrupole. Background signals were the same as electronic noise and high m/z ion intensities, i.e. uranium, were unaffected by this instrument mode. In order to provide accurate uranium quantification, the mass bias needs to be evaluated and corrected. For the quantification measurements, we were not able to evaluate the accuracy of the ratio data, since to the best of our knowledge, there is no commercial standard with a 238 U/233 U ratio certification available. However, the introduced mass bias uncertainty is typically small, generally less than 0.5%. Using a REIMEP 18A interlaboratory comparison material [35] with the same instrumental setup, the mass bias per atomic mass unit based on the 235 U/238 U ratio was −0.35%. When compared to the relative standard deviations obtained from the quantitative analyses (1.4–7%), we can conclude that the mass bias correction for the total U measurements would not significantly affect the quantification analysis. As described in the results section, the uncertainties associated with U measurements were in the range of 1.4–7% based on measurements of the unspiked pooled semen samples and the three U spiked levels: low (5 pg/g semen), medium (50 pg/g semen) and high (1000 pg/g). When the analyses are broken down into the four different levels, we observe that the unspiked pooled semen analyses had an uncertainty of 7%. Uncertainty of: low level spikes was 2.7%; of medium level spikes was 1.7%; and of high level spikes was1.4%. The higher variability in the control pooled semen data could be caused by incomplete homogeneity of these samples or, more likely, by low counting statistics resulting in higher variance. Compared to reported seminal plasma concentrations of toxic metals such as cadmium and lead, the uranium concentrations measured in semen in this study are significantly lower [21] than those reported for Cd in men not occupationally exposed to Cd. Several studies have quantified Cd in seminal plasma and, depending on the study, levels range from 92 pg/mL [36] in US men to in excess of 40 ng/mL in men living in India [21]. Lead (Pb), a more abundant element in the environment, is present in even higher concentrations in seminal plasma compared to Cd, with levels ranging from 2 to 98 ng/mL in men from Germany, India, and Mexico [21]. In the Gulf War I Veteran group examined here, the two participants in the low urine uranium excretion group had semen uranium levels below 3 pg/g. The other four participants from the high urine
6
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
uranium excretion group had semen U levels substantially higher, ranging from 24 to 3350 pg/g. The wide range of semen uranium concentrations present in the Veterans described in this report most likely reflects differences in their current exposure and overall body burdens. The Veterans in our study group come from a variety of geographical areas where their uranium exposures could vary. Such variation could contribute to the observed difference in the range of semen U values, but this is an unlikely explanation for the wide variation in the high uranium exposure group for whom the body burdens are much higher due to their embedded DU fragments [10]. Other explanations, more directly related to the semen samples themselves, can be explored by examining relationships between urine, semen and plasma values. The results presented here are, to the best of our knowledge, the first to describe the quantification of uranium in human semen samples. The described method is simple, rapid and robust and should be easily applied by other laboratories. The limit of quantification shown in this report by the MDL is sufficiently low to allow study of levels of this metal in most samples of human semen with no or minimal exposure beyond that from natural environmental sources of uranium. The procedure can be applied to toxicological studies that focus on potential effects of natural and depleted uranium on male reproduction since the ICP-MS method coupled with acid digestion of the sample as reported here has the increased sensitivity necessary to allow reasonably successful measurement of a low abundance metal, uranium, in a sample which is limited and which may be further limited by the need to use sample for other purposes such as semen analysis or biochemical or other chemical analyses. References [1] Chazel V, Houpert P, Paquet F. Characteristics, biokinetics, and biological effects of depleted uranium used in weapons and the French nuclear industry. In: Miller AC, editor. Depleted uranium: properties, uses and health consequences. 1st ed. Boca Raton: CRC Press; 2007. [2] Parrish RR, Thirlwall MF, Pickford C, Horstwood M, Gerdes A, Anderson J, et al. Determination of 238 U/235 U, 236 U/238 U and uranium concentration in urine using SF-ICP-MS and MC-ICP-MS: an interlaboratory comparison. Health Phys 2006;90:127–38. [3] Uijt De Haag PAM, Smetsers RCGM, Witlox HWM, Krüs HW, Eisenga AHM. Evaluating the risk from depleted uranium after the Boeing 747-258F crash in Amsterdam, 1992. J Hazard Mater 2000;76:39–58. [4] AEPI (Army Policy Institute). Health and environmental consequences of depleted uranium use in the U.S. Army. Technical Report. Atlanta, GA; 1995. [5] Parkhurst M, Daxon E, Lodde G, Szrom F, Guilmette R, Roszell L, et al. Depleted uranium aerosol doses and risks. Summary of U.S. assessments. Columbus, OH: Batelle Press; 2005. [6] National Research Council (NRC). Review of toxicologic and radiologic risks to military personnel from exposure to depleted uranium during and after combat. Washington, DC: National Academy Press; 2008. [7] Voegtlin C, Hodge H. Pharmacology and toxicology of uranium compounds, vol. 1. New York: McGraw Hill; 1949. [8] Voegtlin C, Hodge H. Pharmacology and toxicology of uranium compounds, vol. 2. New York: McGraw Hill; 1949. [9] McDiarmid MA, Gaitens JM, Squibb KS. Uranium and thorium. In: Bingham E, Cohrssen B, editors. Patty’s industrial hygiene and toxicology. 6th ed. New York: John Wiley; 2012. [10] McDiarmid MA, Engelhardt SM, Dorsey CD, Oliver M, Gucer P, Gaitens JM, et al. Longitudinal health surveillance in a cohort of gulf war veterans 18 years after first exposure to depleted uranium. J Toxicol Environ Health A Curr Issues 2011;74:678–91. [11] McClain DE, Miller AC. Depleted uranium biological effects: introduction and early in vivo studies. In: Miller AC, editor. Depleted uranium: properties, uses and health consequences. 1st ed. Boca Raton: CRC Press; 2007.
[12] Salsabili N, Mehrsai A, Jalaie S. Concentration of blood and seminal plasma elements and their relationships with semen parameters in men with spinal cord injury. Andrologia 2009;41:24–8. [13] Morton B, Harrigan Lum J, Albagli L, Jooss T. The activation of motility in quiescent hamster sperm from the epididymis by calcium and cyclic nucleotides. Biochem Biophys Res Commun 1974;56:372–9. [14] Lineman C, Goltz J, Kanous K. Regulation of activation state and flagellar wave from in epididymal rat sperm: evidence for the involvement of both Ca2+ and cAMP. Cell Motil Cytoskeleton 1987;8:324–32. [15] Bedwal RS, Bahuguna A. Zinc, copper and selenium in reproduction. Experientia 1994;50:626–40. [16] Coffey DS. Androgen action and the sex accessory tissues. In: Knobil E, Neill JD, editors. The physiology of reproduction, vol. 1. NY: Raven Press; 1988. [17] Ebisch IM, Thomas CM, Peters WH, Braat DD, Steegers-Theunissen RP. The importance of folate, zinc and antioxidants in the pathogenesis and prevention of subfertility. Hum Reprod Update 2007;13(2):163. [18] Gavella M, Lipovac V. In vitro effect of zinc on oxidative changes in human semen. Andrologia 1998;30:317–28. [19] Sørensen MB, Stoltenberg M, Danscher G, Ernst E. Chelating of intracellular zinc ions affects human sperm cell motility. Mol Hum Reprod 1999;5:338–41. [20] Benoff S, Jacob A, Hurley IR. Male infertility and environmental exposure to lead and cadmium. Hum Reprod Update 2000;6:107–21. [21] Wirth JJ, Mijal RS. Adverse effects of low heavy metal exposure on the male reproductive system. Syst Biol Reprod Med 2010;56:147–67. [22] Arfsten DP, Bekkedal M, Wilfong ER, Rossi III J, Grasman KA, Healey LB, et al. Study of the reproductive effects in rats surgically implanted with depleted uranium for up to 90 days. J Toxicol Environ Health 2005;68:967–97. [23] Arfsten DP, Schaeffer DJ, Johnson EW, Cunningham RJ, Still KR, Wilfong ER. Evaluation of the effect of implanted depleted uranium on male reproductive success, sperm concentration, and sperm velocity. Environ Res 2006;100:205–15. [24] Arfsten DP, Still KR, Wilfong ER, Johnson EW, McInturf SM, Eggers JS, et al. Twogeneration reproductive toxicity study of implanted depleted uranium (DU) in CD rats. J Toxicol Environ Health A Curr Issues 2009;72:410–27. [25] Todorov TI, Ejnik JW, Mullick FG, Centeno JA. Chemical and histological assessment of depleted uranium in tissues and biological fluids. In: Miller AC, editor. Depleted uranium: properties, uses and health consequences. Boca Raton: CRC Press; 2007. p. 85–104. [26] Tolmachyov SY, Kuwabara J, Noguchi H. Flow injection extraction chromatography with ICP-MS for thorium and uranium determination in human body fluids. J Radioanal Nucl Chem 2004;261:125–31. [27] Ejnik JW, Hamilton MM, Adams PR, Carmichael AJ. Optimal sample preparation conditions for the determination of uranium in biological samples by kinetic phosphorescence analysis (KPA). J Pharm Biomed Anal 2000;24:227–35. [28] Todorov TI, Xu H, Ejnik JW, Mullick FG, Squibb KS, McDiarmid MA, et al. Depleted uranium analysis in blood by inductively coupled plasma mass spectrometry. J Anal Atom Spectrosc 2009;24:189–93. [29] Ejnik JW, Todorov TI, Mullick FG, Squibb KS, McDiarmid MA, Centeno JA. Uranium analysis in urine by inductively coupled plasma dynamic reaction cell mass spectrometry. Anal Bioanal Chem 2005;382:73–9. [30] Caddia M, Iversen BS. Determination of uranium in urine by inductively coupled plasma mass spectrometry with pneumatic nebulization. J Anal At Spectrom 1998;13:309–13. [31] Haldimann M, Baduraux M, Eastgate A, Froidevaux P, O’Donovan S, Von Gunten D, et al. Determining picogram quantities of uranium in urine by isotope dilution inductively coupled plasma mass spectrometry. Comparison with ␣spectrometry. J Anal At Spectrom 2001;16:1364–9. [32] Westphal CS, McLean JA, Hakspiel SJ, Jackson WE, McClain DE, Montaser A. Determination of depleted uranium in urine via isotope ratio measurements using large-bore direct injection high efficiency nebulizer-inductively coupled plasma mass spectrometry. Appl Spectrosc 2004;58:1044–50. [33] Hodge SJ, Ejnik JW, Squibb KS, McDiarmid MA, Morris ER, Landauer MR, et al. Detection of depleted uranium in biological samples from Gulf War Veterans. Mil Med 2001;166:69–70. [34] Ejnik JW, Carmichael AJ, Hamilton MM, McDiarmid MA, Squibb KS, Boyd P, et al. Determination of the isotopic composition of uranium in urine by inductively coupled plasma mass spectrometry. Health Phys 2000;78:143–6. [35] Richter S, Alonso A, Truyens J, Kühn H, Verbruggen A, Wellum R. REIMEP 18 inter-laboratory comparison for the measurement of uranium isotopic ratios in nitric acid solution – certification report. Geel, Belgium: Institute for Reference Materials and Measurements; 2006. [36] Benoff S, Hauser R, Marmar JL, Hurley IR, Napolitano B, Centola GM. Cadmium concentrations in blood and seminal plasma: correlations with sperm number and motility in three male populations (infertility patients, artificial insemination donors, and unselected volunteers). Mol Med 2009;15(7–8): 248–68.
Contents lists available at SciVerse ScienceDirect
Journal of Trace Elements in Medicine and Biology journal homepage: www.elsevier.de/jtemb
ANALYTICAL METHODOLOGY
Uranium quantification in semen by inductively coupled plasma mass spectrometry夽 Todor I. Todorov a,∗ , John W. Ejnik b , Gustavo Guandalini c , Hanna Xu c , Dennis Hoover d , Larry Anderson d , Katherine Squibb e , Melissa A. McDiarmid e , Jose A. Centeno c a
Crustal Geophysics and Geochemistry Science Center, United States Geological Survey, PO Box 25046, DFC, Bldg. 20, MS 964D, Denver, CO 80225, United States Department of Chemistry, University of Wisconsin-Whitewater, 800 West Main Street, Whitewater, WI 53190, United States c Division of Biophysical Toxicology, Depleted Uranium and Embedded Fragment Laboratory, The Joint Pathology Center, Silver Spring, MD 20910-1290, United States d University of Maryland, School of Medicine, Department of Anatomy and Neurobiology, 20 South Pine Street, Baltimore, MD 21201, United States e University of Maryland, School of Medicine, Department of Medicine, 11 South Paca Street, 2nd Floor, Baltimore, MD 21201, United States b
a r t i c l e
i n f o
Article history: Received 23 March 2012 Accepted 2 July 2012 Dedicated to the memory of Dr. Larry Anderson. Keywords: Uranium ICP-MS Semen
a b s t r a c t In this study we report uranium analysis for human semen samples. Uranium quantification was performed by inductively coupled plasma mass spectrometry. No additives, such as chymotrypsin or bovine serum albumin, were used for semen liquefaction, as they showed significant uranium content. For method validation we spiked 2 g aliquots of pooled control semen at three different levels of uranium: low at 5 pg/g, medium at 50 pg/g, and high at 1000 pg/g. The detection limit was determined to be 0.8 pg/g uranium in human semen. The data reproduced within 1.4–7% RSD and spike recoveries were 97–100%. The uranium level of the unspiked, pooled control semen was 2.9 pg/g of semen (n = 10). In addition six semen samples from a cohort of Veterans exposed to depleted uranium (DU) in the 1991 Gulf War were analyzed with no knowledge of their exposure history. Uranium levels in the Veterans’ semen samples ranged from undetectable (<0.8 pg/g) to 3350 pg/g. This wide concentration range for uranium in semen is consistent with known differences in current DU body burdens in these individuals, some of whom have retained embedded DU fragments. Published by Elsevier GmbH.
Introduction Uranium is a naturally occurring element with an average abundance in the earth’s crust of approximately 2.7 g/g. Its isotopic composition consists of three naturally occurring radioactive isotopes: 234 U (0.0055%), 235 U (0.72%), and 238 U (99.27%), decaying to 208 Pb and 206 Pb as final non-radioactive products. The main use of uranium is for nuclear fuel, but because the principal isotope in natural uranium is the isotope with the lowest radiologic activity, 238 U, an enrichment of 235 U is required to produce a nuclear, fuel-grade product. A by-product of the enrichment process is depleted uranium (DU) which has an approximate isotopic composition of: 234 U (0.00061%), 235 U (0.2%), and 238 U (99.79%) [1]. Additionally, some studies have shown that DU contains small quantities of 236 U coming from reprocessed uranium fuel [2]. DU is used commercially in a
夽 The opinions and assertions expressed herein are those of the authors and are not to be construed as official or as representing the views of the United States Geological Survey, The Joint Pathology Center, the Department of the Army, or the Department of Defense. ∗ Corresponding author. Tel.: +1 303 236 1243; fax: +1 303 236 3200. E-mail address: [email protected] (T.I. Todorov). 0946-672X/$ – see front matter. Published by Elsevier GmbH. http://dx.doi.org/10.1016/j.jtemb.2012.07.004
variety of applications, such as counterweights in aircraft, shielding for radiography cameras and manufactured chemicals containing uranium. DU exposure in civilian populations could occur in occupational settings or in unusual circumstances such as inhalation of burning aircraft materials following a plane crash [3]. DU has also been used by the military. Because of its high density, availability, and low relative cost, it has been incorporated into munitions as high energy kinetic penetrators and into armor plates for military vehicles. Soldiers in battle are therefore at risk of inhaling DU aerosols, ingesting DU particles, and/or experiencing wound contamination by DU particles and/or embedded fragments [4–6]. Because of the toxicological properties of uranium [7–9], the health effects of natural and depleted uranium exposure have long been of concern. Recent exposures to DU in military personnel have led to renewed studies of uranium’s chemical and radiologic health effects as an internal toxicant [6,10,11]. While uranium may be stored in the bone, it also accumulates in the kidney which is considered the “critical” organ for uranium toxicity [9,11]. Additionally, as with other metals, there are concerns about potential central nervous system and reproductive health effects [6]. Numerous reports have attempted to evaluate the relationship between essential and non-essential trace element composition of semen and fertility, semen quality, and/or sperm characteristics.
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
Among the essential elements that are usually and naturally elevated in semen are calcium (Ca) and zinc (Zn), for which concentrations in semen are higher than those in blood [12]. These two metals have important associations with male reproductive functions. Calcium in the semen is essential for sperm motility [13,14]. Zinc in semen may be part of the bacteriostatic action of semen that protects the male reproductive tract [15,16]. Zn may also be one component of the mechanism which maintains the proper balance of oxidants and antioxidants in semen [17,18]. It also has an influence on motility of ejaculated sperm [19]. In addition to the “intrinsic” trace elements in semen discussed above, humans with occupational, as well as certain nonoccupational, exposures have been shown to exhibit elevated semen levels of toxic trace elements. Studies in animals and humans have shown that male reproductive system and semen quality effects are associated with elevated body burdens of such toxic elements as lead (Pb) and cadmium (Cd). Higher semen concentrations of these metals have been reported for certain exposures [20] and levels in semen have been shown to be negatively correlated with sperm concentrations and motility in humans [21]. The effects of uranium at different concentrations in semen have not been reported, but there are recent reports investigating the reproductive effects of depleted uranium using laboratory animals. Arfsten et al. reported no significant changes in sperm motility or sperm count in rats with implanted DU pellets [22,23]. In a followup two-generation, reproductive toxicity study in rats, the same authors concluded that DU exposure is not a significant reproductive hazard, with a caution of possible developmental effects (increased mean relative heart weights in rat pups from adult males exposed to DU) [24]. Several analytical methods have been described for measuring uranium in environmental, geological, and biological samples, including thermal ionization mass spectrometry (TIMS), instrumental neutron activation analysis (INAA), delayed neutron counting (DNC), inductively coupled plasma mass spectrometry (ICP-MS), inductively coupled plasma optical emission spectrometry (ICP-OES), ␣-spectroscopy, spectrophotometry, fluorometry, and kinetic phosphorescence analysis (KPA) [25]. However, this metal has not been measured in semen to our knowledge. In this study we focused on ICP-MS as a technique of choice, as it provides low detection limits suitable for situations with low metal concentration and where limited sample is available, short sample preparation and instrument analysis time, and high sample throughput. Although uranium levels have been quantified in various types of biological samples including tissues and biological fluids, a number of investigations have focused mainly on urine U and to a smaller extent on blood U quantification [25–28]. The sample preparation for urine uranium analysis (ranging from digestion and wet or dry ashing to no pretreatment and simple dilution) for ICP-MS varies greatly between the different laboratories, based on the detection limits desired and the sample introduction system used for the analysis [29–34]. The primary objective of this study was to develop an assay that accurately measures the quantity of U in semen specimens. A simple robust method is presented herein using acid digestion followed by ICP-MS analysis. This method provides low detection limits and high accuracy and precision for determination of uranium in human semen at low levels.
Materials and methods Semen sample collection and processing Collection of samples for a pool of semen used in the development of this assay was performed at the University of Maryland
3
Baltimore (UMB) School of Medicine. Collection of samples from Veterans of Gulf War I was performed at the University of Maryland Medical Center under the auspices of the Depleted Uranium Surveillance Program of the Baltimore Veterans Affairs Medical Center. All study participants gave informed consent and the protocol was approved by both the VA and University Institutional Review Board (IRB). All semen processing occurred at the UMB School of Medicine. All men participating in the collection of semen were requested to abstain from ejaculation for two days prior to sample production. Semen was produced by masturbation and collected in a sterile 118 mL polystyrene specimen container. Samples for pooled control semen were collected from two non-Veterans over a period of 7 months with reported abstinence of 1–5 days. Volumes ranged from 0.5 to 3.5 mL for each sample. Samples were obtained from Gulf War I Veterans during their 2009 biennial surveillance visit [10]. For these samples, abstinence ranged from 2 to 30 days (6.3 ± 7.7 days, mean ± s.d.) and semen mass ranged from 0.46 to 7.998 g (approximately 0.45–7.95 mL). After transport to the laboratory, all samples were incubated for 3 h at 37 ◦ C to encourage liquefaction before subsequent processing. Additives which can be employed to increase liquefaction of the samples were not added to the semen during sample processing. After incubation at 37 ◦ C, samples for pooled semen were transferred to 50 mL polystyrene centrifuge tubes to a maximum of approximately 40 mL (semen from the two donors was kept separate) which were stored at −20 ◦ C. To produce the semen pool, all aliquots were thawed at 4 ◦ C for 5 h. Aliquots were vortexed and poured into acid-washed teflon beakers, first combined by donor and then combined into a common pool. For sampling, the pooled semen was initially mixed in the covered teflon beakers with an acid-washed teflon-coated stir bar, and stirring continued during sampling (speed decreased as volume remaining decreased). Aliquots of 2 mL from the final pool were prepared in the vials used for uranium determination (screw-capped, teflon, 15 mL, numbered precleaned vials; Savillex, Eden Prairie, MN) with an air displacement, polystyrene-tipped pipettor. Aliquots were stored at −20 ◦ C. For Veterans, each ejaculate was transferred in its entirety after incubation at 37 ◦ C with a graduated glass pipet directly into a teflon vial for uranium determination. These samples were stored at −20 ◦ C and all Veteran and pooled semen samples were transferred frozen (on ice) to the Division of Biophysical Toxicology – Depleted Uranium and Embedded Fragment Laboratory (DBT-DU/EMF) for uranium analysis. All specimens, including pooled semen at the time of aliquotting and Veteran samples processed immediately after collection, were transferred into teflon vials of known individual weights and the weight of vial plus semen was recorded. Semen weight was determined as the difference between weight of vial plus semen and the weight of the empty vial. All subsequent quantification calculations for uranium in semen were based on semen weight rather than volume as the volume measured was approximate (particularly in samples with incomplete liquefaction or high viscosity). At DBT-DU/EMF laboratory, the samples were stored at −70 ◦ C until analysis. Preparation of spiked samples Two milliliter (2 mL) aliquots (weight 2.081 ± 0.252 g mean ± s.d., n = 97) from the pooled control semen were spiked at the following three levels for method validation and generation of quality control samples: (1) low level spike at 5 pg/g (added as 0.02 mL at 500 pg/mL to give a nominal concentration of 5.0 pg/g), (2) medium level spike 50 pg/g U (added as 0.02 mL at 5 ng/mL to achieve a concentration of 50 pg/g) and (3) high level spike at 1000 pg/g U (added as 0.02 mL at 100 ng/mL for a nominal
4
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
Table 1 Operating conditions for total uranium quantification in semen inductively coupled plasma mass spectrometry (ICP-MS). Instrument
Perkin Elmer Elan DRC II
ICP-MS operating parameters Plasma power Nebulizer gas flow rate Auxiliary gas flow Plasma gas flow Interface cones Rpa RPq Cell gas Monitored masses Dwell time Detector dead time Number of sweeps Number of replicates Scan mode
Setting 1400 W 1.04 L min−1 1.1 L min−1 15 L min−1 Pt 0 0.7 None m/z: 233.04, 238.05 100 ms 55 ns 25 3 Peak hopping
concentration of 1000 pg/g). The solutions used for the preparation of the spikes were prepared from Spex single element uranium ICP-MS standard by serial dilutions (Spex CertiPrep, Metuchen, NJ). The spiked pooled semen samples used for assay accuracy validation were prepared in triplicate. Quality control samples of spiked semen were prepared and used along with digestion blanks and unspiked semen. These samples were distributed throughout the digestion and analysis of all samples associated with this study. Sample preparation for analysis Before ICP-MS analysis: (1) semen samples were thawed at room temperature, (2) internal standard was added, (3) high purity nitric acid (Optima grade 70% HNO3 , ThermoFisher Scientific, Waltham, MA) was added in a volume equal to the mass of semen sample with a maximum of 6 mL of acid in order to avoid overfilling of the digestion vessel, and (4) samples were digested in a muffle furnace at 110 ◦ C until dryness. Following the digestion, each sample was reconstituted at a known volume of approximately 1:2 ratio of semen weight to volume of 10% nitric acid for ICP-MS, with the exception of low volume samples (<1.0 mL) that were reconstituted to a final volume of 1 mL. Chymotrypsin and bovine serum albumin samples Chymotrypsin is sometimes added to semen samples during sample processing in order to increase liquefaction. When this enzyme is employed, its action can be quenched by the addition of bovine serum albumin (BSA). In this study, the uranium content of chymotrypsin (Semen Viscosity Treatment System VTS-20, Conception Technologies, San Diego, CA) and BSA (BSA-embryo culture tested A3311, Sigma Chemical Co., Saint Louis, MO) was measured to evaluate the potential contamination of the semen with external uranium when these agents are used.
Table 2 Analytical performance/figures of merit. Sample weight, g Background, cps Sensitivity, cps per ppb Detection limit (3) in aqueous solution, pg/mL Detection limit (3) in semen, pg/g Precision, RSD (n = 34)
0.5–7 0.3 8.3 × 104 0.04 0.8 1.4–7%
25 pg/mL 233 U (CRM 111A, New Brunswick Laboratory, Argonne, IL, USA; certified for concentration, isotopic composition information values: 99.4911% 233 U, 0.1874% 234 U, 0.0790% 235 U, 0.0166 236 U and 0.2286% 238 U) was used in the blanks, standards, validation, quality control and unknown samples (added prior to digestion) to correct for instrument drift and sample matrix effects. Since the internal standard was also added to the blank that was used to correct all ICP-MS samples, the ∼0.05 ng/L 238 U that it contributed to the m/z 238 signal was subtracted from all the standards and samples. This procedure ensured accurate quantification of the 238 U as no systematic error was contributed by the 233 U internal standard. The quantification was achieved by monitoring the m/z at 233.04 for 233 U+ and 238.05 for 238 U+ . The instrument detection limit (IDL) for uranium by ICP-DRC-MS was calculated based on 3 times the standard deviation of blank measurements (n = 6) consisting of 10% nitric acid. The method detection limit was calculated using 3 times the standard deviation of pooled unspiked semen samples (n = 10). For each batch of digested unknown samples, a set of quality control samples were prepared. These included a digestion blank (which consisted of empty sample vials processed through the digestion process), and the quality control samples (an unspiked pooled semen sample and three spiked pooled semen samples at low, medium and high level). Blank correction for all semen samples was performed using the mean of digestion blank samples. Results Figures of merit The instrument and method parameters were optimized and Table 2 shows the analytical performance for the quantitative uranium determinations. The IDL obtained from previous studies for total uranium was determined to be 0.04 pg/mL [28,29] in aqueous solutions. The method detection limit (MDL) was 0.8 pg/g in semen. Samples of the pooled control semen were prepared using the same sample procedure as the unknown samples (i.e. digested, reconstituted in nitric acid and analyzed by ICP-MS) for determination of the MDL. The uncertainties (relative standard deviation, mean divided by the standard deviation) associated with the quantitative measurements of the semen samples were 1.4–7%, and were calculated using the pooled semen samples (unspiked and spiked, n = 34) used for the method validation and as QC materials. Method validation
Uranium analysis The quantitative analysis of uranium in semen samples was performed by ICP-MS employing a dynamic reaction collision cell to reduce polyatomic interferences (ELAN DRC II, manufactured by PerkinElmer, Waltham, MA, USA). The operational parameters of the instrument were set to optimize the detection of uranium, as summarized in Table 1. The concentration of uranium was determined by using an external calibration curve of 0.10, 1.0, 10, 50, and 250 ng/L U. The calibration standards were prepared using a 1000 mg/L uranium ICP-MS standard solution (Spex CertiPrep, Metuchen, NJ) through serial dilutions. An internal standard at
Three spike levels of uranium were created in pooled semen samples and analyzed along with samples of unspiked semen. The 233 U signal corrects for biological matrix effects and any instrument changes during the analysis. Since 233 U and 238 U have the same ionization properties, 233 U as an internal standard corrects for any matrix effects from the sample and instrumentation variations during the analysis. The 238 U isotope was used for quantification of total uranium. The small change in 238 U composition from natural uranium’s 99.274% to DU’s 99.794% was determined to not significantly alter quantification in samples containing DU derived from clinical studies. The results for unspiked and three different levels
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6 Table 3 Comparison of measured semen uranium with known uranium concentrations. Spike levela
Measured U concentration (pg/g)b
Unspiked (10) 2.9 (0.2) Low (3) 7.5 (0.1) Medium (3) 49.8 (0.3) 917 (7) High (3) a b
Calculated U concentration (pg/g)
Recovery (%)
7.7 50.1 942
98 100 97
Number of replicate samples shown in parentheses. Numbers in parentheses are standard deviations.
of U spikes of pooled semen are shown in Table 3. The uranium concentration in the pooled, unspiked semen was 2.9 pg/g. The “calculated uranium concentration” in Table 3 is equal to the mean measured uranium quantity in the pooled semen plus the uranium quantity added by each spike divided by the mass of each pooled semen sample and the mass spike solution added. The recoveries of the spikes were between 97 and 100% for all three spike levels. Analyses of chymotrypsin and BSA for uranium content In a previous study in our laboratory (data not published), the quantification of uranium in chymotrypsin and BSA indicated that these additives could be a significant source of sample contamination. Chymotrypsin and BSA uranium levels were 90 ± 7 pg/mg (mean ± s.d.), and 10.6 ± 0.1 pg/mg, respectively. Addition of 5 mg of chymotrypsin to 2 g of semen sample would contribute 225 pg/g of uranium. Similarly, addition of 50 mg of BSA to a 2 g semen sample would increase the uranium content by 265 pg/g. When compared to the control pooled semen uranium concentration of 2.9 pg/g, this is a significant level of contamination. Therefore, liquefaction enhancement additives were not used in this study. Uranium analysis of semen samples from DU-exposed Gulf War I Veterans Veterans attending the Baltimore Veterans Administration Medical Center for the 2009 biennial Depleted Uranium Surveillance visit were requested to produce semen samples for uranium analysis. These Gulf War I Veterans were exposed to DU in friendlyfire incidents involving DU munitions and tanks with DU armor [10]. Data from a subset (n = 6) of the semen samples analyzed are presented here (Table 4). In our studies of this DU-exposed Gulf War I cohort, the on-going exposure to uranium is monitored by urine uranium excretion (assayed as 238 U) corrected for creatinine concentration and expressed as g of uranium per g creatinine [10]. Some of the men in this cohort have embedded DU fragments and, as the fragments oxidize in situ, these men continually excrete elevated uranium in their urine. The Veterans in the study group are divided into 2 populations: low and high uranium exposure based on their Table 4 Depleted uranium (DU) exposed Gulf War I participant semen samples analyzed for total uranium (U) content. Veteran
Semen quantity (g)
U concentration (pg/g)
Uranium exposure groupa
A B C D E F
4.028 2.488 3.082 5.597 2.918 1.486
<0.8 2.8 24.1 159 226 3350
Low Low High High High High
a Exposure group assignment based on high exposure group = urine uranium ≥ 0.1 g/g creatinine and low exposure group = urine uranium < 0.1 g/g creatinine.
5
urine uranium excretion. Men in the high urine uranium group have creatinine standardized urine uranium at or above 0.1 g/g creatinine. The exposure designations shown in Table 4 are based on this exposure criterion. For the 6 Veterans listed in Table 4, semen uranium concentration ranged from undetectable (<0.8) to 3350 pg/g semen. Values for the two low uranium exposure men include the undetectable sample and a semen uranium concentration (2.7 pg/g semen) very close to the pooled control semen obtained from two non-Veteran donors used in the method validation and characterization (2.9 pg/g semen). The high uranium exposure men had a mean semen uranium concentration of 905 pg/g semen (s.d. 1610), exhibiting a very wide range of values (24.0–3350 pg/g).
Discussion and conclusions In our previous studies with urine and blood samples using reaction cell quadrupole ICP-MS instruments (for removal of polyatomic interferences) [29], we found that the use of bandpass mass filtering (rpq increased to 0.7) improved both precision and accuracy of uranium quantitative and isotopic determinations. In the current investigation, we utilized only bandpass mass cutoff within the reaction cell quadrupole (without use of cell gas). The bandpass mass limit, rpq parameter, was set at 0.7 in order to decrease the amount of low ion masses entering the analyzer quadrupole. Background signals were the same as electronic noise and high m/z ion intensities, i.e. uranium, were unaffected by this instrument mode. In order to provide accurate uranium quantification, the mass bias needs to be evaluated and corrected. For the quantification measurements, we were not able to evaluate the accuracy of the ratio data, since to the best of our knowledge, there is no commercial standard with a 238 U/233 U ratio certification available. However, the introduced mass bias uncertainty is typically small, generally less than 0.5%. Using a REIMEP 18A interlaboratory comparison material [35] with the same instrumental setup, the mass bias per atomic mass unit based on the 235 U/238 U ratio was −0.35%. When compared to the relative standard deviations obtained from the quantitative analyses (1.4–7%), we can conclude that the mass bias correction for the total U measurements would not significantly affect the quantification analysis. As described in the results section, the uncertainties associated with U measurements were in the range of 1.4–7% based on measurements of the unspiked pooled semen samples and the three U spiked levels: low (5 pg/g semen), medium (50 pg/g semen) and high (1000 pg/g). When the analyses are broken down into the four different levels, we observe that the unspiked pooled semen analyses had an uncertainty of 7%. Uncertainty of: low level spikes was 2.7%; of medium level spikes was 1.7%; and of high level spikes was1.4%. The higher variability in the control pooled semen data could be caused by incomplete homogeneity of these samples or, more likely, by low counting statistics resulting in higher variance. Compared to reported seminal plasma concentrations of toxic metals such as cadmium and lead, the uranium concentrations measured in semen in this study are significantly lower [21] than those reported for Cd in men not occupationally exposed to Cd. Several studies have quantified Cd in seminal plasma and, depending on the study, levels range from 92 pg/mL [36] in US men to in excess of 40 ng/mL in men living in India [21]. Lead (Pb), a more abundant element in the environment, is present in even higher concentrations in seminal plasma compared to Cd, with levels ranging from 2 to 98 ng/mL in men from Germany, India, and Mexico [21]. In the Gulf War I Veteran group examined here, the two participants in the low urine uranium excretion group had semen uranium levels below 3 pg/g. The other four participants from the high urine
6
T.I. Todorov et al. / Journal of Trace Elements in Medicine and Biology 27 (2013) 2–6
uranium excretion group had semen U levels substantially higher, ranging from 24 to 3350 pg/g. The wide range of semen uranium concentrations present in the Veterans described in this report most likely reflects differences in their current exposure and overall body burdens. The Veterans in our study group come from a variety of geographical areas where their uranium exposures could vary. Such variation could contribute to the observed difference in the range of semen U values, but this is an unlikely explanation for the wide variation in the high uranium exposure group for whom the body burdens are much higher due to their embedded DU fragments [10]. Other explanations, more directly related to the semen samples themselves, can be explored by examining relationships between urine, semen and plasma values. The results presented here are, to the best of our knowledge, the first to describe the quantification of uranium in human semen samples. The described method is simple, rapid and robust and should be easily applied by other laboratories. The limit of quantification shown in this report by the MDL is sufficiently low to allow study of levels of this metal in most samples of human semen with no or minimal exposure beyond that from natural environmental sources of uranium. The procedure can be applied to toxicological studies that focus on potential effects of natural and depleted uranium on male reproduction since the ICP-MS method coupled with acid digestion of the sample as reported here has the increased sensitivity necessary to allow reasonably successful measurement of a low abundance metal, uranium, in a sample which is limited and which may be further limited by the need to use sample for other purposes such as semen analysis or biochemical or other chemical analyses. References [1] Chazel V, Houpert P, Paquet F. Characteristics, biokinetics, and biological effects of depleted uranium used in weapons and the French nuclear industry. In: Miller AC, editor. Depleted uranium: properties, uses and health consequences. 1st ed. Boca Raton: CRC Press; 2007. [2] Parrish RR, Thirlwall MF, Pickford C, Horstwood M, Gerdes A, Anderson J, et al. Determination of 238 U/235 U, 236 U/238 U and uranium concentration in urine using SF-ICP-MS and MC-ICP-MS: an interlaboratory comparison. Health Phys 2006;90:127–38. [3] Uijt De Haag PAM, Smetsers RCGM, Witlox HWM, Krüs HW, Eisenga AHM. Evaluating the risk from depleted uranium after the Boeing 747-258F crash in Amsterdam, 1992. J Hazard Mater 2000;76:39–58. [4] AEPI (Army Policy Institute). Health and environmental consequences of depleted uranium use in the U.S. Army. Technical Report. Atlanta, GA; 1995. [5] Parkhurst M, Daxon E, Lodde G, Szrom F, Guilmette R, Roszell L, et al. Depleted uranium aerosol doses and risks. Summary of U.S. assessments. Columbus, OH: Batelle Press; 2005. [6] National Research Council (NRC). Review of toxicologic and radiologic risks to military personnel from exposure to depleted uranium during and after combat. Washington, DC: National Academy Press; 2008. [7] Voegtlin C, Hodge H. Pharmacology and toxicology of uranium compounds, vol. 1. New York: McGraw Hill; 1949. [8] Voegtlin C, Hodge H. Pharmacology and toxicology of uranium compounds, vol. 2. New York: McGraw Hill; 1949. [9] McDiarmid MA, Gaitens JM, Squibb KS. Uranium and thorium. In: Bingham E, Cohrssen B, editors. Patty’s industrial hygiene and toxicology. 6th ed. New York: John Wiley; 2012. [10] McDiarmid MA, Engelhardt SM, Dorsey CD, Oliver M, Gucer P, Gaitens JM, et al. Longitudinal health surveillance in a cohort of gulf war veterans 18 years after first exposure to depleted uranium. J Toxicol Environ Health A Curr Issues 2011;74:678–91. [11] McClain DE, Miller AC. Depleted uranium biological effects: introduction and early in vivo studies. In: Miller AC, editor. Depleted uranium: properties, uses and health consequences. 1st ed. Boca Raton: CRC Press; 2007.
[12] Salsabili N, Mehrsai A, Jalaie S. Concentration of blood and seminal plasma elements and their relationships with semen parameters in men with spinal cord injury. Andrologia 2009;41:24–8. [13] Morton B, Harrigan Lum J, Albagli L, Jooss T. The activation of motility in quiescent hamster sperm from the epididymis by calcium and cyclic nucleotides. Biochem Biophys Res Commun 1974;56:372–9. [14] Lineman C, Goltz J, Kanous K. Regulation of activation state and flagellar wave from in epididymal rat sperm: evidence for the involvement of both Ca2+ and cAMP. Cell Motil Cytoskeleton 1987;8:324–32. [15] Bedwal RS, Bahuguna A. Zinc, copper and selenium in reproduction. Experientia 1994;50:626–40. [16] Coffey DS. Androgen action and the sex accessory tissues. In: Knobil E, Neill JD, editors. The physiology of reproduction, vol. 1. NY: Raven Press; 1988. [17] Ebisch IM, Thomas CM, Peters WH, Braat DD, Steegers-Theunissen RP. The importance of folate, zinc and antioxidants in the pathogenesis and prevention of subfertility. Hum Reprod Update 2007;13(2):163. [18] Gavella M, Lipovac V. In vitro effect of zinc on oxidative changes in human semen. Andrologia 1998;30:317–28. [19] Sørensen MB, Stoltenberg M, Danscher G, Ernst E. Chelating of intracellular zinc ions affects human sperm cell motility. Mol Hum Reprod 1999;5:338–41. [20] Benoff S, Jacob A, Hurley IR. Male infertility and environmental exposure to lead and cadmium. Hum Reprod Update 2000;6:107–21. [21] Wirth JJ, Mijal RS. Adverse effects of low heavy metal exposure on the male reproductive system. Syst Biol Reprod Med 2010;56:147–67. [22] Arfsten DP, Bekkedal M, Wilfong ER, Rossi III J, Grasman KA, Healey LB, et al. Study of the reproductive effects in rats surgically implanted with depleted uranium for up to 90 days. J Toxicol Environ Health 2005;68:967–97. [23] Arfsten DP, Schaeffer DJ, Johnson EW, Cunningham RJ, Still KR, Wilfong ER. Evaluation of the effect of implanted depleted uranium on male reproductive success, sperm concentration, and sperm velocity. Environ Res 2006;100:205–15. [24] Arfsten DP, Still KR, Wilfong ER, Johnson EW, McInturf SM, Eggers JS, et al. Twogeneration reproductive toxicity study of implanted depleted uranium (DU) in CD rats. J Toxicol Environ Health A Curr Issues 2009;72:410–27. [25] Todorov TI, Ejnik JW, Mullick FG, Centeno JA. Chemical and histological assessment of depleted uranium in tissues and biological fluids. In: Miller AC, editor. Depleted uranium: properties, uses and health consequences. Boca Raton: CRC Press; 2007. p. 85–104. [26] Tolmachyov SY, Kuwabara J, Noguchi H. Flow injection extraction chromatography with ICP-MS for thorium and uranium determination in human body fluids. J Radioanal Nucl Chem 2004;261:125–31. [27] Ejnik JW, Hamilton MM, Adams PR, Carmichael AJ. Optimal sample preparation conditions for the determination of uranium in biological samples by kinetic phosphorescence analysis (KPA). J Pharm Biomed Anal 2000;24:227–35. [28] Todorov TI, Xu H, Ejnik JW, Mullick FG, Squibb KS, McDiarmid MA, et al. Depleted uranium analysis in blood by inductively coupled plasma mass spectrometry. J Anal Atom Spectrosc 2009;24:189–93. [29] Ejnik JW, Todorov TI, Mullick FG, Squibb KS, McDiarmid MA, Centeno JA. Uranium analysis in urine by inductively coupled plasma dynamic reaction cell mass spectrometry. Anal Bioanal Chem 2005;382:73–9. [30] Caddia M, Iversen BS. Determination of uranium in urine by inductively coupled plasma mass spectrometry with pneumatic nebulization. J Anal At Spectrom 1998;13:309–13. [31] Haldimann M, Baduraux M, Eastgate A, Froidevaux P, O’Donovan S, Von Gunten D, et al. Determining picogram quantities of uranium in urine by isotope dilution inductively coupled plasma mass spectrometry. Comparison with ␣spectrometry. J Anal At Spectrom 2001;16:1364–9. [32] Westphal CS, McLean JA, Hakspiel SJ, Jackson WE, McClain DE, Montaser A. Determination of depleted uranium in urine via isotope ratio measurements using large-bore direct injection high efficiency nebulizer-inductively coupled plasma mass spectrometry. Appl Spectrosc 2004;58:1044–50. [33] Hodge SJ, Ejnik JW, Squibb KS, McDiarmid MA, Morris ER, Landauer MR, et al. Detection of depleted uranium in biological samples from Gulf War Veterans. Mil Med 2001;166:69–70. [34] Ejnik JW, Carmichael AJ, Hamilton MM, McDiarmid MA, Squibb KS, Boyd P, et al. Determination of the isotopic composition of uranium in urine by inductively coupled plasma mass spectrometry. Health Phys 2000;78:143–6. [35] Richter S, Alonso A, Truyens J, Kühn H, Verbruggen A, Wellum R. REIMEP 18 inter-laboratory comparison for the measurement of uranium isotopic ratios in nitric acid solution – certification report. Geel, Belgium: Institute for Reference Materials and Measurements; 2006. [36] Benoff S, Hauser R, Marmar JL, Hurley IR, Napolitano B, Centola GM. Cadmium concentrations in blood and seminal plasma: correlations with sperm number and motility in three male populations (infertility patients, artificial insemination donors, and unselected volunteers). Mol Med 2009;15(7–8): 248–68.