Temporal Clinical Chemistry and Microscopic Renal Effects Following Acute Uranyl Acetate Exposure

Abstract

Military use of depleted uranium (DU) has renewed interest in the toxicology of this metal. In this study, the nephrotoxicity of single exposure DU was assessed with and without pre-exposure stress. Adult male Sprague–Dawley rats (n = 288) were administered a single IM dose of 0, 0.1, 0.3 or 1.0 mg/kg DU. Corticosterone concentrations (ng/ml, mean ± SD) were 763.65 ± 130.94 and 189.80 ± 90.81 for swim stressed and unstressed rats. Serum and kidney uranium concentration, hematocrit, chemistry, and renal histology were assessed on sacrifice days 1, 3, 7 and 30 post-DU-dosing. Dose related increases in serum and kidney uranium were noted. DU concentration peaked day 1 in the kidney and days 3–7, in the serum. Dose-related elevations of Cr and BUN concentrations were seen on days 3 and 7. A decline in serum albumin coincided with Cr and BUN suggesting protein losing nephropathy. Dose related acute tubular necrosis and proliferative glomulonephritis were seen. Tubular regeneration in low dose rats was almost complete by day 30. High dose rats had extensive tubular necrosis and delayed regeneration with focal residual chronic interstitial nephritis and cortical scarring. Glomular changes were reversed in all treatment groups by day 30. Stress exposure had no impact on any measured renal parameter.

Introduction

Injury to the proximal renal tubular epithelium occurs with low level dosing of depleted uranium (DU) and renal concentrations of ≥3 μg uranium/gram of tissue (Diamond et al., 1989). Cessation of exposure led to diminished renal uranium concentrations and associated regeneration of tubular epithelium. Modulation of this toxic effect by stress has not been previously evaluated. This study presents sequential correlative evaluations of serum and renal uranium concentrations, tubular injury and clinical chemistry determinations to document progression/recovery of nephropathy in rats following a single dose exposure to DU. The study also assesses whether predosing stress modulates renal uranium distribution, lesions, and function. The research reported here is part of a larger project evaluating the neurotoxicity of depleted uranium exposure in rats and its modulation by stress (Barber et al., 2006).

Depleted uranium is a by-product of the enrichment process for that metal, during which the more radioactive isotopes U²³⁵ and U²³⁴ are removed (Pellmar et al., 1999; Squibb et al., 2005). This depleted form of uranium has about 60% of the radioactivity of the natural mineral, and its density, availability, and relatively low cost make it attractive for military purposes, specifically in armor and projectiles (McDiarmid et al., 2000). This use has resulted in exposures to the metal by aerosol, ingestion and wound contamination with metallic fragments localized in tissues for years. This work reports on the nephrotoxic effects following exposure to various dosages of soluble depleted uranium with and without predosing stress.

Stress may be defined as the imbalance between environmental demands for survival and the individual’s capacity to adapt (Lazarus and Folkman, 1984; Marshall et al., 2000). The degree of these responses relates to the intensity and duration of the stress. Milder forms of stress draw on reserve resources, but severe acute or chronic stress may negatively alter homeostatic metabolic events, diminishing ability to resist infectious agents or tolerate toxicant exposures. As regards the latter, stress may have varying effects on toxicity. For example, Ehrich et al. (1986) demonstrated both amelioration and enhancement of organophosphate-induced delayed neuropathy associated with elevated concentrations of stress-related hormones. However, few alterations of organophosphate-induced effects were noted in stressed animals after a single toxicant exposure at a dose that caused no clinical evidence of poisoning (Pung et al., 2006).

There are few studies on the effects of stress on heavy metal toxicity. Cory-Slechta et al. (2004) demonstrated that restraint stress in pregnant rats potentiated the toxic effects of lead in the offspring. Relative to modulation of experimental kidney disease, heat stress applied six (but not 48) hours prior to ischemic insult protected against acute renal injury (Kelly, 2002). This was thought to be related to the actions of induced heat shock proteins. However, these proteins were thought not to be a factor in protection against uranium-induced nephropathy (Tolson et al., 2005). In the present study, a predosing combination of restraint and swimming was employed to determine if such stress affected temporal nephrotoxic effects of depleted uranium as demonstrated by evaluation of uranium serum and renal tissue concentrations, renal lesions, and clinical chemistry data.

Materials and Methods

Male Sprague–Dawley rats (n = 288 from a larger study of 418 reported on by Barber, 2006) with a starting weight of 250 to 300 grams were used in this study. An internal Animal Care and Use Committee reviewed all in study use of animals for compliance with institutional and governmental guidelines. The experiment was conducted in a 2 × 4 randomized block design, with 2 levels of stress (yes/no) and four levels of DU exposure (0.0, 0.1, 0.3, 1.0 mg/kg) as shown in Table 1. Rats were sacrificed on days 1, 3, 7, and 30. On each sacrifice day, 1 cohort (n = 40/sacrifice day, n = 160 total) was used for renal pathology and clinical biochemistry determinations. A second cohort (n = 32/sacrifice day, n = 128 total) was used for analysis of blood and kidney uranium concentrations.

The rats were singly caged in a climate controlled facility, at 21–22°C with a daily 12 hour light-dark cycle. They were subject to daily stress for 5 days prior to dosing with depleted uranium. This included restraint on days 1–4 and swimming on day 5. For restraint stress, each animal was placed in a Plexiglas tube (6 centimeters diameter by 22 centimeters long) (Konarska et al., 1989) with adequate breathing holes for a 20-minute period and then returned to its home cage. For swim stress, each animal was placed in a tank of water at 23–25°C and allowed to swim for a 30-minute period. When swimming was complete, the animal was towel-dried for 1–2 minutes and placed under a warming lamp for an additional 2–3 minutes. Animals in the no stress groups were handled daily by removing them from the home cage, placing them in a box and immediately returning them to their home cage.

Within 10 minutes after day 5, swim stress or handling (no-stress groups), rats were anesthetized with isofluorane. Whole blood was collected from the orbital sinus into heparinized microcentrifuge tubes (Becton-Dickinson, Lincoln Park, NJ). Plasma, for corticosterone measurement, was separated from whole blood by centrifugation at 12,000 rpm at 4°C for 4 minutes. Separated plasma was frozen at −70°C until analysis using a corticosterone ¹²⁵I-radioimmunoassay kit (ICN Biomedicals, Costa Mesa, CA). Following blood collection, rats were intramuscularly dosed in the gastrocnemius with DU (uranyl acetate) prepared in saline at one of four dosage concentrations (0.0, 0.1, 0.3, or 1.0 mg/kg) dependent upon their group assignment. The day of DU exposure was considered day 0 of the study.

Animals from each group were randomly selected prior to the study onset for sacrifice by CO₂ euthanasia on postdosing days 1, 3, 7, and 30. At sacrifice, blood was collected from the inferior vena cava into serum separator and sodium heparin tubes (Becton-Dickinson, Franklin Lakes, NJ). The serum separator tube was used for serum uranium concentration and serum biochemistry value measurements. For rats given 0.1 mg/kg (no stress and stress), serum uranium concentrations were only measured on sample day one (due to low initial concentration found on day 1 as reported in results section). For all groups on all sample days, serum sodium (Na⁺), potassium (K⁺), chloride (Cl⁻), calcium (Ca⁺⁺), total carbon dioxide (TCO₂), glucose (Glu), blood urea nitrogen (BUN), creatinine (Cr), phosphorus (P), total protein (TP), albumin (Alb), globulin (Glb-calculated), total bilirubin (TBil) values, and alkaline phosphatase (ALP) and alanine aminotransferase (ALT) activities were measured. Hematocrit percentage was determined using heparinized whole blood placed in a microcentrifuge tube (Becton-Dickinson, Lincoln Park, NJ) and centrifugation (Hematastat, Separation Technology Inc., Altamonte Springs, Florida). Kidney tissue was collected and frozen at −70°C for uranium analyses. In addition, a complete transverse section was immersion-fixed in 10% neutral-buffered formalin for histological study.

Serum biochemistry samples were measured within 1 hour of sample collection using an Olympus AU400 analyzer (Olympus, Center Valley, PA) by the Virginia-Maryland Regional College of Veterinary Medicine Clinical Pathology Laboratory (Blacksburg, VA). Statistical analysis of the results for each individual analyte was done using commercially available software (Minitab 14, Minitab Inc., State College, PA). A general linear model was used for univariate analysis of variance of mean analytes values. The regression model consisted of the following main effects: stress exposure (y/n), dose (0.0, 0.1, 0.3, and 1.0 mg/kg), sample day (1, 3, 7, and 30) and the following interaction effects: stress-dose, stress-day, dose-day, and stress-dose-day. Effects deemed significant were those with p-values less than or equal to an α-level of 0.05.

Serum and kidney uranium concentration analyses were performed by inductively coupled plasma-mass spectrometry (ICP-MS) on an HP 7500a ICP-MS (Agilent Technologies, Santa Clara, CA) performed by the Analytical Section of the Hazard Identification Core in the Southwest Hazardous Waste Program (Tucson, AZ). Samples were placed in 15 ml sealed glass pressure tubes with 0.5 ml of concentrated metal-free nitric acid (Optima, Fisher Scientific). Samples were heated to 140°C for 2 hours in a silicone oil bath, then 0.5 ml of 30% hydrogen peroxide (Ultrex II, J.T. Baker, Phillipsburg, NJ) was added and samples heated at 110°C for an additional 60 minutes. Samples were cooled, quantitatively transferred to acid-washed 5 ml volumetric flasks, and brought to volume with deionized water (ElgaStat Maxima, Elga, High Wycombe, UK). ICP-MS analysis on the HP 7500a was conducted using iridium as an internal standard. Uranium concentration was determined from a standard curve of uranium based on the m/z 238 signal. Five repetitions were performed per sample and the average used to calculate uranium concentration. The limit of quantitation of this method was 0.002 ppb. Recovery was determined from samples spiked with 0.1–10 ppb uranium and determined to be 96–108%. Statistical analyses for serum and tissue uranium concentration were done as previously described for serum chemistry analytes.

For histopathology, formalin fixed kidney tissue was embedded in paraffin, sectioned at 6 μm thickness and stained with hematoxylin and eosin (H&E). The renal proximal tubular lesions were qualitatively evaluated. A blinded assessment was made of changes in the outer medullary stripe and adjacent deep cortex of the 6 μm thick H&E stained sections, using the following classification system: stage 0—no lesions (Figure 1A); stage 1—focal injury of tubular epithelium manifest by cytoplasmic vacuolization and occasional necrotic/apoptotic cells (Figure 1B); stage 2—diffuse necrosis of tubular epithelium (Figure 1C); stage 3—diffuse necrosis with focal early tubular regeneration (Figure 1D); stage 4—extensive early tubular regeneration (Figure 1E); stage 5—advanced regeneration (Figure 1F). Selected sections were stained with terminal deoxynucleotidyl transferase biotin-dUTP nick end labeling (TUNEL) procedures. Control rat kidney sections treated with DNAse served as positive controls. The commercial TUNEL assay (Boehringer Mannheim, GmbH Mannheim, Germany) identifies apoptotic cells by using terminal deoxynucleotidyl transferase (TdT) to transfer fluorescein-dUTP to these strand breaks of cleaved DNA. In addition, sections from selected blocks were cut at 3 μm thickness and stained with the H&E and Masson’s trichrome procedures for detailed study of glomerular lesions.

Results

Corticosterone Concentration

Plasma corticosterone concentration was 189.8 ng/ml with a standard deviation (SD) of 90.81 in the nonstressed rats (Groups 1–4). Plasma corticosterone concentration of stressed groups 5–8 was 763.6 with a SD of 130.94 (N = 144 stressed, 144 not stressed). These values were significantly different (p< 0.05).

Uranium Concentration

Mean serum uranium concentrations are shown in Table 2 (concentration not determined on days 3, 7, and 30 for group2 [low dose DU, no stress] and group 6 [low dose DU, stress] due to low initial DU concentration measured on sacrifice day 1). There was no effect of predosing stress on serum uranium concentrations (p > 0.05, Figure 2A, Table 2). However, individual main effects due to dose, day, and a dose-day interaction effect on the mean serum uranium concentration were seen (Figures 2A, 2B, Table 2). Serum uranium concentrations were elevated in rats administered depleted uranium, in particular those given the 1.0 mg/kg dose (Table 2). This elevation was marked 1 day postdosing with further increase seen on days 3 and 7, and return to control values by day 30.

Mean kidney tissue uranium concentrations are shown in Table 2. Uranium was concentrated in the renal tissue, being present in μg U/gram of tissue quantities in contrast to the ng U/ml seen in the serum. There was no effect of pre-dosing stress on tissue DU concentrations (p > 0.05, Figure 3A, Table 2). As with serum uranium concentration, individual main effects due to dose and day, and the dose-day interaction effect on the tissue uranium concentration were seen (Figures 3A, 3B) Renal concentration increased rapidly in all dosage groups receiving DU in a dose dependent fashion (Figures 3, 4, Table 2). Values were highest on day one then declined throughout the sample period (Figure 5).

Clinical Pathology

Cr, BUN, and Alb demonstrated significant (p < 0.05) mean analyte value differences associated with main and interaction effects as shown in Table 3. Mean and SD values for serum chemistry analytes with significant effect differences are shown in Table 4. Of the statistically different means, the following were considered to be of clinical and toxicological significance: BUN (dose, day, dose-day [Figure 6]), Cr (dose, day, dose-day [Figure 7]), and Alb (dose, day [Figure 8]). Stress had no impact on analyte response patterns across dosage groups or across sample days. Serum Cr and BUN values were elevated in comparison to control groups 1 and 5 (0.0 mg/kg DU) in a dose-day interaction effect for dosage groups three (0.3 mg/kg [2-fold increase]) and 4 (1.0 mg/kg [four fold increase]) beginning on day 3, peaking on day 7, and returning to control values by day 30 (Figures 6B, 7B). Serum albumin mean values varied significantly with separate main effects of day and dosage (Figure 8A). Again variance suggested a dose-dependent relationship with the most noteworthy decline in albumin concentration occurring in the unstressed and stressed high dosage groups (groups 4 and 8, 1.0 mg/kg) as shown in Figure 8A. Inversely and simultaneously to the increases seen with BUN and Cr, albumin declined on day 3 with a maximum decline on day 7 and a return to normal on day 30 (Figure 8A). Although not statistically significant (p = 0.167), the interaction effect of dose-day on mean serum albumin value followed an inverse pattern to that seen with BUN and Cr with the highest dosage group demonstrating the most noteworthy decline on day seven (Figure 8B). Changes in these analyte results were preceded by increases in kidney uranium concentrations (Figures 4, 5, Table 2).

Renal Histopathology

Lesions were most prominent in the epithelial cells lining the proximal tubules. Early injury was manifest by loss of the brush borders as seen with the PAS stain. Examination of the H&E stained sections revealed a sequence of change, as indicated in the lesion classification scale (see Methods). Compared to controls (Figure 1A), there was cytoplasmic swelling and vacuolization, and the presence of occasional necrotic or apoptotic cells (stage 1, Figure 1B). This was followed by expansion of necrosis to involve large numbers of tubular epithelial cells, with sloughing of cellular debris into the lumen (stage 2, Figure 1C). Subsequently, focal regions of flattened basophilic regenerating cells appeared in the walls of the tubules, displacing necrotic cells (stage 3, Figure 1D), initially noted in the outer medulla. These regenerating cells became more prominent, occupying large regions of the affected tubular surface (stage 4, Figure 1E). With time the regenerating epithelium matured to resemble control cells (stage 5, Figure 1F), although there was some residual focal scarring and interstitial inflammation in the high-dose rats. These proximal tubular changes primarily involved the outer stripe of the medulla, adjacent deep cortex and the medullary rays (Figure 9). TUNEL staining revealed only occasional positive nuclei in contracted tubular epithelial cell during periods of active injury. This was absent in other dying cells. The DNAse-treated (positive control) sections had prominent TUNEL-positive nuclei.

Assessing the proximal tubular lesions with the scoring scale demonstrated progression was related to dosage and postexposure day. Stress was not a factor. The earliest stage of injury (stage 1) was frequent on postdosing day 1 in the 0.3 and 1.0 mg/kg dose groups, but was occasionally seen in the 0.1 mg/kg group rats (Figure 1B). There was further dose-dependent differentiation in lesion progression in tissues studied the third postdosing day. The 0.1 mg/kg animals had largely progressed to stage 3, in which extensive tubular epithelial necrosis was seen, with associated early regeneration (stress > no stress) (Figure 1D). The 1.0 mg/kg rats were mostly restricted to stage 2, where diffuse necrosis was seen without regeneration (Figure 1C). The mid-dose animals were intermediate in these day 3 pathological responses. By postdosing day 7, all dosages were in the stage 4 stage, indicating extensive active regeneration, and on day 30, this process was largely completed (Figures 1E, 1F).

Glomeruli from 0.3 and 1.0 mg/kg DU dose groups had focal, segmental of varying degree, mesangial expansion and hypercellularity, consistent with mesangial proliferation (Brun and Olsen, 1981), plus the presence of red-stained (by Masson’s trichrome) droplets in this region (Figure 10). This was noted on postdosing days 3 and 7, but was absent on day 30.

Discussion

Uranium, as other heavy metals, is a well-known nephrotoxin, accumulating in the kidney and injuring proximal tubular epithelial cells, but preserving the basal lamina (Lim et al., 1987). Following absorption the metal is carried in the blood plasma as uranium-transferrin or in low molecular weight anionic complexes. Some 80% of this uranium is filtered by the glomerulus. It subsequently binds to anionic sites in the proximal tubular epithelial brush border in the form of UO₂ ⁺⁺, and enters the cell by endocytosis (Leggett, 1989). Injury to the cell is related to alterations in cell and lysosomal membranes, injured mitochondria leading to impaired energy production and altered calcium ion homeostasis (Leggett, 1989). These may be based upon uranium-related induction of cellular oxidative stress (Tanlan et al., 2004).

Rats were exposed by the IM route in the present study, as that represents the location of unremoved DU particles following human exposure. A single intramuscular exposure to 0.1, 0.3 or 1.0 mg/kg of soluble depleted uranium to male Sprague–Dawley rats, given as uranyl acetate, leads to transient prominent necrosis of proximal tubular epithelium, with associated dose-related elevations of renal uranium concentration and plasma creatinine and BUN. Serum uranium concentrations were elevated in rats administered IM depleted uranium, in particular those given the 1.0 mg/kg dose. Maximum serum uranium concentrations were delayed 3–7 day postdosage, possibly as a result of delayed absorption at the injection site.

The necrotic lesions evolve into widespread early regeneration in the seven-day post-dosing period. There is dose-related delay in the initiation of regeneration, as the latter is seen earliest in the low-dose group. However, the mid-and high-dose groups “catch up” as renal uranium levels are reduced. Regeneration is largely complete by day 30 in all exposure levels.

The significant difference in plasma corticosterone concentration between groups exposed to the stress event and those not provides convincing evidence that stress difference existed between various animals. Since dosing with depleted uranium was done shortly after blood sample collection, it is likely that the toxic exposure was achieved under conditions of markedly elevated plasma corticosterone for the stressed rats. In this study, no evidence was found that stress at the time of exposure affected renal uranium concentrations, or the nature, severity and progression of the uranium-induced tubular injury.

There have been other experimental animal studies of soluble depleted uranium nephrotoxicity. Depleted uranium salts have been administered using single or multiple exposures (Haley et al., 1982; Lim et al., 1987; Diamond et al., 1989; Kato et al., 1994). The injury in these appeared to be due to heavy metal toxicity rather than radiation effects. Pertinent to the present study are reports of single exposure studies. Haley et al. (1982) used a single subcutaneous exposure of 10 mg/kg uranyl nitrate to rats. Although their dose was much higher (10 mg/kg compared to a high of 1 mg/kg in the present study) and with subcutaneous administration rather than by an intramuscular route, they noted similar changes, such as proximal tubular injury over a 5 day post-dosing period, with subsequent regeneration. There was some residual tubular atrophy and fibrosis in affected regions. Glomeruli had widespread flattened podocytes with cytoplasmic droplets. The association of these changes with renal uranium concentrations was not provided. Lim et al. (1987) performed a single exposure study in rats, using dosages of 5, 15 or 30 mg/kg administered as intravenous uranyl nitrate. Despite these considerably higher doses, the sequences of clinical and renal pathology findings were consistent with ours. Renal uranium concentration data were not provided, so values cannot be compared with those reported here. The very high doses of uranium used by Lim et al. (1987) did not appear to result in mortality.

In the present study, the degree of proximal tubular necrosis and associated elevations of serum creatinine and BUN concentrations peaked at day seven post-dosing in correlation with peak serum DU concentration. Prerenal azotemia related to dehydration was excluded as a possible explanation for this transient azotemia; this interpretation is supported by the lack of significant mean differences in total protein and globulin values, and hematocrit percentage in the regression model for a dose related effect. Leggett (1983) suggested 2 mechanisms for explaining this azotemia following exposure to soluble uranium: (1) reduced effective glomerular filtration surface area, (2) feedback of solutes across damaged tubules.

Ultrastructural examination by electron microscopy (EM) by Lim et al. (1987) and Haley et al. (1982) demonstrated numerous changes in tubular epithelial cells in support of the solute feedback azotemia explanation. Lim et al. (1987) did not provide any EM evidence for the former azotemia explanation (reduced glomular filtration) reporting only that both tubular as well as glomular basement membranes remained intact. However, Haley et al. (1982) did note on EM that glomular visceral epithelium pedicles were broad and flattened in support of direct glomular damage. Intriguingly, in the present study the noteworthy dose relationship decline in serum albumin (p = 0.167) coincided with maximum microscopic evidence of tubular damage, and companying azotemia. This finding taken in consideration with the previously mentioned lack of significant main dose effect on mean group differences for hematocrit, total protein or globulin values suggests that the declining albumin concentration may be related to glomular loss of albumin and resulting albuminuria due to functional impairment of glomerular podocytes following exposure to soluble uranium. While Diamond et al. (1989) reported proteinuria following low dose exposure of uranyl fluoride in rats, noteworthy changes in serum albumin concentrations have not been previously reported following low-dose acute or chronic soluble uranium exposure. Addition evidence of this altered glomerular function is seen in the mesangial proliferation noted by light microscopy in the mid- and high-dosage groups of this study. This proliferative glomulonephritis (Brun, 1981) is also a unique finding of this study.

In contrast to the present study, in which DU was administered IM, Diamond et al. (1989) administered rats 8 intraperitoneal doses of uranyl fluoride over a 9–33 day period at 2 levels, with aggregate doses of 660 or 1329 μg/kg body weight. These led to concentrations of 3.45 or 5.5 μg uranium/g of renal tissue (wet weight) 3 days after the last exposure (day 36 of the study), which were diminished by ~90% 18 days later. There was dose-related proximal tubular necrosis and apoptosis (especially affecting the S2 and S3 segments) beginning on day 19 and peaking on day 36, with subsequent regeneration. These investigators noted that tubular injury began at renal concentrations below 1 μg uranium/g.

In addition to the short-term studies discussed here, sub-chronic exposure to uranium has been obtained by dosing in the drinking water. Gilman et al. (1998) administered uranium at levels ranging from 0.96 to 600 mg/L of drinking water to weanling rats for 91 days, and found non-lethal changes in renal tubular epithelium consisting of apical nuclear displacement and cytoplasmic vesiculation along with glomerular capsular sclerosis. Similar changes were seen by McDonald-Taylor et al. (1997) in rabbits dosed in the drinking-water at 24 or 600 mg/liter for 91 days, with subsequent 45- or 90-day recovery periods. These workers noted sublethal changes to proximal tubular epithelium consisting of loss of brush border, and increased number and size of cytoplasmic vacuoles, lysosomes, and mitochondria.

Given the concentration of uranium in the kidney, there has been much interest in determining safe levels. For many years a concentration of 3 μg uranium/gram was considered a chemical toxicity threshold for limiting occupational exposure (Leggett, 1989; Russell et al., 1996). While exposure routes, rates and forms of uranium are factors, experimental studies demonstrate severe tubular injury with concentrations of renal uranium well below this threshold (Diamond et al., 1989; Leggett, 1989). Clinical studies suggest a potential for altered proximal tubular function at uranium concentrations well below the 3 μg uranium/gram concentration (Squibb et al., 2005). In the present study, even doses as low as 0.1 mg/kg produce significant renal tubular injury consistent with that of prior studies, an injury that was unaltered by stress exposure.