Methyl mercury and selenium interaction in relation to mouse kidney γ-glutamyltranspeptidase, ultrastructure, and function

TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

80,

78-96

(1985)

Methyl Mercury and Selenium Interaction in Relation to Mouse Kidney y-Glutamyltranspeptidase, Ultrastructure, and Function P . H . FAIR,*” *National Marine Fisheries F.O. Box 12607, Charleston, Medical University

Received

W. J. DouGHERTY,t

AND S. A. BRADDON*

Service, Southeast Fisheries South Carolina 29412-0607, of South Carolina, Charleston,

December

17, 1984;

accepted

Center, Charleston Laboratory. and TDepartment of Anatomy, South Carolina 29425

March

26, 1985

Methyl Mercury and Selenium Interaction in Relation to Mouse Kidney y-Glutamyltranspeptidase, Ultrastructure, and Function. FAIR, P. H., DOUGHERTY, W. J., AND BRADDON, S. A. (1985). Toxicol. Appl. Pharmacol. 80, 78-96. The effects of methyl mercury (CH3Hg) and selenium (Se) on renal ultrastructure were investigated and correlated to changes in renal y-glutamyl transpeptidase (r-GTPase) activity, mercury (Hg) accumulation, and renal function (serum creatinine and urea nitrogen). Three experimental protocols were used to investigate CH3Hg and Se interactions of both Se-sufficient and Se-deficient mice involving ip injection of the following administered alone or in combination: CHJHg (4.0 mp/kg) and Se (0.16 mg/kg) daily for 7 days, CH3Hg (1.0 mg/kg) and Se (0.08 mg/kg) daily for 20 days, and a single acute dose of CHJHg (8.0 mg/kg). Acivicin (12 to 50 mg/kg), an antitumor glutamine antagonist, was also used as a highly effective specific inhibitor of the y-GTPase. Our results show that CH3Hg administered to Se-deficient mice for 7 or 20 days resulted in significant (p c 0.05) but only moderate inhibition (20%) of y-GTPase activity and extensive renal ultrastructural damage. Acivicin-treated mice had significant inhibition of y-GTPase activity (80%) following a single injection while ultrastructural damage was substantial only after several days of administration. These results may indicate different modes of action of acivicin and CH3Hg. Acivicin inhibited y-GTPase prior to renal damage while CH3Hg produced greater pathological effectswith only moderate y-GTPase inhibition. Renal damage from acute and chronic CH3Hg toxicity occurred after distinct neurological signs were present. Selenium administered to Sedeficient mice ameliorated both the neurotoxic effects and nephrotoxic action of CH3Hg. While Se and CH3Hg treatments caused some of the same ultrastructural pathology as the treatment with CH&g alone (cytoplasmic vacuolation, increased lysosomal profile, mitochondrial swelling, and extrusion of cellular masses into the tubular lumen), degeneration was not as extensive. Although the total doses administered during both the 7- and the 20day studies were similar, mice from the chronic 20&y study showed greater ultrastructural pathological effects from CH3Hg. The primary effects of CHJHg appeared to be on the lysosomal system, while acivicin exerted its effects on the mitochondrial and endoplasmic reticulum systems.The accumulation studies on Hg suggest that dietary Se may have only an initial protective effect against Hg accumulation in the kidney while injected Se offers longer protection. The Hg concentration in the kidneys cannot be used as an index of cellular pathology since chronic CH,Hg exposure produced greater nephrotoxic effects than acute CH,Hg intoxication while having only onethird the concentration of Hg. o 1985 Academic press. IX.

consumption and the 1970s when methyl mercury (CH3Hg)-contaminated grain led to a mass poisoning in Iraq (Bakir et al., 1973), numerous studies have focused on the toxicological effects of CH3Hg (Fehling et al.,

Since the 1950s when an epidemic of mercury (Hg) poisoning occurred in Minamata, Japan (Kurland et al., 1960), following seafood ’ To whom correspondence should be addressed. 0041408x/85

$3.00

Copyriat 0 1985 by Academic F’res, Inc. All rights of reproduction in any form reserved.

78

CH,Hg AND Se INTERACTION

1975; Chang, 1977). Protection against this toxicity is provided by the nutritionally required element selenium (Se) (Ganther ez al., 1972; Ohi et al., 1976; Potter and Matrone, 1974; Stillings et al., 1974; Stoewsand et al., 1974). The kidney is a major site of concentration and storage of Hg (Berlin, 1963; Berlin and Ullberg, 1963), particularly cells of the proximal convoluted tubule (Clarkson, 1972), irrespective of its form (Swensson and Ulfvarson, 1968). Several investigations have studied the effects of CH3Hg on renal morphology in adult (Chang et al., 1973; Chen et al., 1983; Fowler, 1972a,b; Ware et al., 1973) and neonate animals (Chang and Sprecher, 1976) and have found the proximal tubules to be the primary site of damage. Despite the numerous investigations of CH3Hg-induced renal damage, the effect of Se in ameliorating this damage has not yet been investigated nor related to a biochemical site of action. A common association between the metabolism of Se and CH3Hg is the thiolcontaining peptide glutathione (GSH). The metabolic cycling as well as the oxidationreduction of GSH are integral processes coupled to the activation and metabolism of Se (Gasiewicz and Smith, 1978; Hill and Burke, 1982) and the metabolism and detoxification of CH3Hg (Ballatori and Clarkson, 1982; Refsvik, 1978; Refsvik and Norseth, 1975; Thomas and Smith, 1982). The study of Chung et al. (1982) provides information on the biochemical changes of GSH metabolism in the kidney following HgC& treatment. They observed a significant increase in GSH concentration and timedependent decreases in the activities of yglutamylcysteine synthetase, GSSG reductase, GSH peroxidase, and y-glutamyl transpeptidase (y-GTPase). The kidney exhibits the highest activity of y-GTPase and histochemical studies have shown y-GTPase to be localized in the microvillus membranes of the proximal tubules (Tate and Meister, 198 1). Due to the diversity of acceptor donors demonstrated for y-GTPase (Tate and Meis-

IN KIDNEY

79

ter, 1974) and the rapid absorption of CH3Hg-GSH complex by the kidney (Alexander and Aaseth, 1982), there is a potential for CH3Hg alone or in a thiol complex to serve as an acceptor or donor for y-GTPase in the kidney. The activity of y-GTPase is specifically inhibited by acivicin, an antitumor glutamine antagonist which serves as a highly effective affinity label for the y-glutamyl site of y-GTPase and causes a rapid, irreversible inactivation (Griffith and Meister, 1980; Allen et al., 1981). The objectives of the current study are to determine if a specific biochemical site of action of methyl mercury in the kidney is the transport enzyme y-glutamyl transpeptidase, to determine if the action of CH3Hg on this site can be modified by Se, and to investigate the morphology of the kidney following treatment with CH3Hg or CH3Hg and Se. METHODS Chemicals. Methyl mercury chloride was obtained from Alfa Chemicals (Danvers, MassJ2 Radioactive CHJHg (203Hg, sp act 122 mCi/g) was obtained from Amersham/Searle (Arlington Heights, Ill.). Glycylglycine, L-y-glutamyl-p-nitroanilide, and sodium selenite were obtained from Sigma Chemical Company (St. Louis, MO.). Acivicin (AT-125; y-amino-3-chloro-4,5-dihydro5-isoxazoleacetic acid) was provided by the Upjohn Company (Kalamazoo, Mich.). Animals. Weanhng female ICR mice obtained from Sprague-Dawley (Indianapolis, Ind.) were maintained in a controlled environment (22 + 2°C 45% relative humidity) under a 12-hr light-dark cycle. Animals were housed in polycarbonate cages with stainless-steel lids which contained shaved wood chips and polycarbonate water bottles. The mice were fed either selenium-deficient (10.05 pg Se/g) or selenium-control (0.5 pg Se/g) diets obtained from Teklad (Madison, Wise.) for 6 weeks before being used in experiments. Diets and doubly distilled deionized water were available ad libitum. The selenium status of the mice as previously determined in our laboratory (Balthrop and Braddon, 1985) indicated a 60% decrease in blood glutathione peroxidase and a 70% decrease in hepatic selenium in mice fed a selenium-

2 Reference to trade names does not imply endorsement by NMFS.

FAIR, DOUGHERTY,

80

deficient diet for 6 weeks. One criterion used to assess neurological disturbance was the “crossed hindlimb with flexion” behavior of mice held by the tail. The behavior was attributed to destruction of the nervous system by CHsHg (Suzuki, 1969; Klein et al., 1972). Experiment 1. One group of selenium-deficient (dietary Se < 0.05 pg/g) mice was given ip injections (dose volume = 0. I ml/10 g) of one of the following five treatments: saline (0.9% NaCl), Se (0.16 mg/kg), CHSHg (4.0 mp/kg), Se (0.16 mg/kg) administered 20 min before CH,Hg (4.0 mg/kg), or acivicin (50 mg/kg). Doses of 12.5 and 25 mg/kg acivicin yielded similar results to the 50 mg/kg dose and the latter was selected for experimentation. Selenium-sufficient animals (dietary Se = 0.5 pg/ g) received ip injections of saline. A listing of all experimental treatments is provided in Table 1. Animals were given single injections on 7 consecutive days and killed 24 hr after the first (Day 1) or last (Day 7) injection. Mice were anesthetized in a COz chamber and blood was obtained from the posterior vena cava. Blood was collected in a I-ml capillary blood/serum separator tube @&on-Dickinson Microtainer, Rutherford, N.J.) and centrifuged; serum was frozen for urea nitrogen and creatinine analyses. Kidneys were removed, rinsed in isotonic saline, blotted with paper, weighed, and analyzed (within 10 hr) for y-GTPase activity. A

TABLE

AND BRADDGN second group of Se-deficient and Se-sufficient mice, consisting of four mice per treatment, was treated similarly but kidneys were removed for light- and electron-microscopic examination. Se-deficient mice in the third group were injected with either radioactive methyl[203Hg]mercuric chloride (4.0 mg/kg) with or without Se (0.16 mg/kg) while Se-sufficient mice were injected with methyl[z”3Hg]mercuric chloride (4.0 mg/kg) for 7 consecutive’days. Each day, four animals were killed and the kidneys were treated as above, frozen, and stored (not more than 7 days) until analyzed for ‘03Hg. With a Beckman gamma counter (Model 8000, Beckman Instrument Co., Norcross, Ga.), the Hg concentrations in the tissues were calculated from the specific activity of the injected CH,Hg. Experiment 2. Selenium-deficient mice were given daily ip injections of CHJHg (1 .O mg/kg), CHSHg (1 .O mg/kg) and Se (0.08 mg/kg), or saline for 20 days. The same parameters (y-GTPase, serum urea nitrogen, and creatinine) as described in Experiment 1 were measured. Additional groups of mice were included for histopathological examination and determination of Hg uptake in the kidneys. Experiment 3. Selenium-deficient mice received a single ip injection of CHrHg (8.0 mg/kg) or saline. Mice were killed 24 hr following injection, and renal y-

1

EXPERIMENTAL TREATMENTS Se status” Expt 1 (l-7 days)

Expt 2 (20 days)

Expt 3 (1 day)

Se-

Treatment

Ihe

Owdkg)

Analyses

0.16 4.0 4.0 + 0.16 50.0 -

Kidney Kidney Serum Serum

Se+

Saline se CM-k CH& + Se Acivicin Saline

Se+ SeSe-

Methyl[203Hg]mercury Methyl[203Hg]mercury Methyl[203Hg]mercury + Se

4.0 4.0 4.0 + 0.16

Kidney ‘03Hg

Se-

CHJ-k CHjHg + Se Saline

1.0 1.0 + 0.08 -

se+ sew

Methyl[203Hg]mercury Methyl[203Hg]mercury Methyl[“‘Hg]mercury + Se

1.0 1.0 1.0 f 0.08

se-

CW-k Saline

u Se+ and Se- denote dietary Se-sufficient and Se-deficient mice. b Light and electron microscopy.

8.0

-

Kidney Kidney Serum Serum

y-GTPase LM + EMb urea nitrogen creatinine

y-GTPase LM + EM urea nitrogen creatinine

Kidney 2osHg Kidney +TPa.se Serum urea nitrogen Serum creatinine

CHsHg AND Se INTERACTION GTPase activity and serum urea nitrogen and creatinine were determined. Histology. Animals were killed by cervical dislocation, and the kidneys were removed from the retroperitoneal space. Each of the left kidneys was sliced into l-mm cross sections perpendicular to the longitudinal axis and diced into l-mm cubes while immersed in fixative. The tissue was immersed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.3) for 2 to 3 hr at room temperature. The tissue samples were then rinsed in several changes of buffer, postfixed in 1% 0~0~ in 0.1 M cacodylate buffer for 1.5 hr (pH 7.3), dehydrated in an acetone series, then infiltrated, and embedded in a lowviscosity epoxy resin @purr, 1969). Ultrathin sections were stained with many1 acetate and lead citrate and examined with a Hitachi HU-112 electron microscope. Each of the right kidneys was bisected longitudinally and fixed in Formalin for light microscopy. These sections were embedded in paraffin by routine histologic methods and stained with hematoxylin and eosin (H + E) and periodic acid-Schiff (PAS) techniques. Assays. Kidneys were homogenized in a solution of 20 mM NazHPQ (pH 6.8) and 40 mM MgClz by a polytron homogenizer (Brinkman, Inc., N.Y.) and centrifuged at 49,000g for 60 min at 4°C. The pellet was resuspended and homogenized in a solution of 10 mM Tris-HCl (pH 8.8), 1% DGC, and 10 mM NarEDTA and heated for 30 min at 30°C. The suspension was centrifuged at 49,OOOgfor 30 min at 4°C. The pellet was discarded and cold 1-butanol (l/2 volume) was added to the supematant fraction and mixed immediately. After centrifugation for 30 min at 49,OOOg,the aqueous phase located under the protein interface was collected. This fraction (approx. 5 pg protein/& was analyzed for r-GTPase activity according to the method of Orlowski and Meister (1965). Final concentration of the l-ml incubation mixture contained 50 mM Tris-HCl (pH 8.0), 75 mM NaCl, 2.5 mM L-y-glutamyl-pnitroanilide, 20 mM glycylglycine (pH 8.0), and 50 to 100 pg protein (tissue extract). One unit of activity is defined as the amount of enzyme required to release 1 pmol pnitroaniline/min/mg protein. Protein was assayed by the procedure of Lowry ef al. (195 1) with crystalline bovine serum albumin as a standard. The concentration of urea nitrogen (modified ureaseBerthelot reaction) and creatinine in serum were both measured with Sigma Chemical diagnostic kits. Statistics. Data obtained on Days 1 and 7 of the first experiment were subjected to a one-way analysis of variance (ANOVA) for differences in concentrations of renal y-GTPase and serum urea nitrogen and creatinine (Sokal and Rohlf, 1969). Significant differences between treatments (p c 0.05) were subjected to the StudentNewman-Keuls (SNK) multiple range test for painvise comparisons between treatment means (Steel and Torrie, 1960). A one-way ANOVA was also computed at the p < 0.05 level of significance for renal y-GTPase and serum urea nitrogen and creatinine concentrations for

IN KIDNEY

81

the 20day (Experiment 2) study, followed by a SNK multiple range test. Student’s t test was performed on data from Experiment 3 to determine significance (p < 0.05) of the difference between means of renal y-GTPase and serum urea nitrogen and creatinine concentrations between control (saline) mice and CH,Hg (8.0 mg/kg)-treated mice. Accumulation data for concentrations of Hg in the kidney for the various treatments was subjected to a one-way ANOVA at the p c 0.05 level of significance, followed by a SNK multiple range test.

RESULTS

Experiment I Signs of CH3Hg poisoning (crossing reflex of hindlimbs, ataxic gait) were displayed by mice injected for 7 consecutive days with CH3Hg (4.0 mg/kg). In mice receiving Se (Se-treated or CH3Hg/Se-treated mice) these toxic signs were not present or were delayed. The activity of renal y-GTPase in mice receiving a single dose of CH3Hg (4.0 mg/kg) with and without Se (0.16 mg/kg) was not significantly different from that of the Sedeficient or Se-sufficient control mice 24 hr following injection (Fig. 1). However, renal y-GTPase activity was approximately 80% lower in acivicin-dosed mice compared with the control mice (Fig. 1). The mice receiving a single injection of Se (0.16 mg/kg) had significantly lower concentrations of yGTPase when compared to the control or CH3Hg-treated mice (Fig. 1). Following the 7-day injection regime, CH3Hg (4.0 mg/kg) with and without Se (0.16 mg/kg) produced a significant decrease in y-GTPase activity in comparison to the controls (Fig. 2). Although acivicin produced a significant decrease in renal y-GTPase activity when compared with the controls and other treatments, this decrease was not magnified by multiple injections (Fig. 2). Serum creatinine concentrations were not initially significantly different for the various treatment groups in the 7-day study. Following injections on 7 consecutive days, however, the mice receiving CH3Hg and Se or acivicin had significantly higher serum creatinine val-

82

d

FAIR, DOUGHERTY,

kw)

16 Y-GTPW

(E. u.‘hg

pddn)

20

22

X 10-z

FIG. 1. Activity of renal y-GTPase 24 hr following ip injection of the following treatments: control-saline; Se (0.16 mg/kg); CHSHg (4.0 mg/kg); CH,Hg (4.0 mg/kg) + Se (0.16 mglkg); and acivicin (50 mg/kg). Each bar represents the X + SE from 4 to 12 mice. Treatments with identical superscripts are not significantly different at the 5% level. Se+ and Se- denote the dietary selenium status of the animals. *EU is defined as I nmol p nitroaniline produced/min.

AND BRADDON

ing Day 1 of the 7-day study were significantly lower than those of control (Se-) and Setreated animals. After several days of injections, mice receiving the CH3Hg, CHjHg and Se, or acivicin all contained significantly less serum urea nitrogen than the controls (Seand Se+) or Se-treated mice. The accumulation of mercury in mouse kidneys over a 7-consecutive-day injection study is shown in Fig. 3. During Days 1 and 2, the kidneys from Se-deficient mice injected with CH3Hg (4.0 mg/kg) accumulated significantly more mercury than either Se-sufficient mice injected with CH3Hg (4.0 mg/kg) or Sedeficient mice injected with CH3Hg (4.0 mg/ kg) and Se (0.16 mg/kg). However, during Days 3,5,6, and 7 the kidneys of Se-deficient mice injected with both CH3Hg (4.0 mg/kg) and Se (0.16 mg/kg) accumulated significantly less (approx. 40%) Hg than either Se-deficient or Se-sufficient mice treated with CH3Hg alone.

ues than those of the controls (Se- and Se+) or those dosed with either CH3Hg or Se. Concentrations of serum urea nitrogen in animals treated with CH3Hg or acivicin dur-

2

4

6 Y-GTPaw

s

IO E.U.‘/mp

12

14 protrin

16

IS

so **

) X 10e2

FIG. 2. Activity of renal y-GTPase after 7 daily ip injections of the following treatments: control-saline; Se (0.16 me/kg); CHrHg (4.0 mg/kg); CHJHg (4.0 m&kg) + Se (0.16 mg/kg); and acivicin (50 mgfkg). Each bar represents the X +- SE from 4 to 13 mice. Treatments with identical superscripts are not significantly different at the 5% level. Se+ and Se- denote the selenium status of the animals. *EU is defined as 1 nmol pnitroaniline produced/min.

0

I

2

3

4

5

6

7

6

DAYS FIG. 3. Mercury concentration in mouse kidneys following daily injections of CH,“‘Hg (4.0 mg/kg) (Se’) (A); CHsm3Hg (4.0 mg/kg) (Se-) (0); CHzo3Hg (4.0 m& kg) + Se (0.16 mg/kg) (Se-) (0). Each point represents the X + SE from 4 mice. Asterisks denote statistical differences (p =G0.05) between treatments for each time period. Se’ and Se- denote the dietary selenium status of the animals.

CH3Hg AND Se INTERACTION

IN KIDNEY

dl) and Se. Significant differences in serum urea nitrogen were not observed between the two CH3Hg treatments (CH3Hg or CHsHg and Se). However, both CHsHg treatments resulted in significantly lower serum urea than the controls. No significant differences in accumulation of Hg by the kidney were observed for any of the treatments (Hg concentrations averaged 38 pg/g at 20 days).

20 DAY

Experiment

0

2

. 4

a 6

* 6

0 IO

0. I2

14

I6

* I6

‘1 20

22

lo-* FIG. 4. Activity of renal y-GTPase following a lday study with CHsHg (8.0 mg/kg) and 20day studies of the following treatments: control-saline; CHsHg (1 .O mg/ kg); and CHsHg (1.0 mg/kg) + Se (0.08 mg/kg). Each bar represents the X & SE from 8 to 10 mice. Treatments with identical superscripts are not significantly different at the 5% level. Se+ and Se- denote dietary selenium status of the animal. *EU is defined as a 1 nmol pnitroaniline produced/min. Y-GTPW

Experiment

E.U.*/mg

83

protein)

x

2

Twenty days of daily injections of a low dose of CH3Hg (1 .O mg/kg) alone or CH3Hg (1.0 mg/kg) and Se (0.08 mg/kg) to Sedeficient mice caused significant decreases in y-GTPase activity (30 and 22%, respectively) compared with control mice (Fig. 4). The administration of this same dose of CH3Hg to mice for 20 days resulted in the appearance of neurological signs. Administration of Se with CH3Hg to mice either prevented or delayed these signs. Creatinine concentrations in serum of mice receiving CH3Hg or CHjHg and Se were significantly higher than those of control mice. In addition, Se appeared to modify the response of elevated creatinine caused by CH3Hg. This effect was evident in that the mice treated with CH3Hg (1.53 f 0.042 mg/dl) had significantly higher creatinine concentrations than the mice treated with CH3Hg (1.23 f 0.105 mg/

3

A single ip injection of CH3Hg (acute dose, 8.0 mg/kg) significantly decreased y-GTPase activity in the mouse kidney (Fig. 4). This same high dose produced a significantly elevated concentration of serum creatinine and urea nitrogen in addition to a high mortality. Histology. Inspection of renal sections stained with H + E and PAS revealed no obvious morphologic changes in tubules, whereas electron-microscopic examination revealed subcellular changes. The most significant ultrastructural changes occurred in the proximal (PCT) and distal convoluted tubules (DCT) of CHsHg-treated Se-deficient animals and in acivicin-treated Se-deficient animals. None of the treatments had much effect on the structural components of the renal glomeruli and collecting ducts. Acivicin and CH3Hg seemed to exert their primary effects on different subcellular compartments, especially of the PCT cells. The PCT and DCT cells of animals which received Se just prior to CHsHg exhibited considerably fewer and less dramatic changes. Many lysosomes and vacuoles accumulated in the apical cytoplasm of PCT cells in animals of the 7-day CHsHg treatment group (Fig. 5a). Necrotizing PCT cells with slightly swollen mitochondria and shrunken nuclei were more frequently observed in Se-deficient mice treated with CH3Hg than in either Sedeficient or Se-sufficient control mice (Fig. 5b). Animals in the 20-day CHXHg treatment group exhibited increased numbers of apically located lysosomes and vacuoles (Fig. 5~). Indeed, there seemed to be more of each of these particles in the PCT cells of the 20-day

84

FAIR, DOUGHERTY,

AND BRADDON

FIG. 5. Low-magnification electron micrographs of PCT from Se-deficient mice treated with CH3Hg. (a) Mouse treated with CHpHg for 7 days. Note the accumulation of dense lysosomes (ly) and vacuoles (vat) in the apical cytoplasm of PCT cells. Mitochondria (m), nuclei (N), and microvilli (mv) appear normal. Cellular debris (cd) from necrotizing PCT cells occurs in the lumen. (X6820.) (b) Note necrotizing FCT cell with slightly swollen mitochondria (m) and shrunken nuclei (N) adjacent to a PCI cell of normal appearance in a mouse treated with CHpHg for 7 days. such necrotizing cells were of more frequent occurrence in Se-deficient control or normal mice. (X5060.) (c) A mouse treated with a low dose of CHsHg for 20 days. Note the increased numbers of dense lysosomes (ly) and electron-lucent vacuoles (vat) of varying sizes in the apical cytoplasm of these PCT cells. Mitochondria (M), nuclei (N), and microvilii (mv) appear normal. (X7040.)

CH3Hg

AND

Se INTERACTION

IN

KIDNEY

85

FIG. 5-Continued.

treatsed animals than for either the 1 or 7day (ZH3Hg-treated groups (Figs. 5a-c). Some lysos omes exhibited variable electron density (Fig. 6a), while others showed complex internal I&ucture (Figs. 6b-d). Some lysosomes seem ed to contain packets of smooth endo-

plasmic reticulum, presumably in the plrocess of degradation (Fig. 6c), while others ap] zmred to have fused with electron-lucent apical vacuoles (Fig. 6b). Most mitochondria of PCT, DCT, renal glomeruli, and collecting tubule cells v Iere of

FAIR,

DOUGHERTY,

FIG.

AND

BRADDON

5-Continued.

nom la1 appearance, except for the presence of colmplex matrical granules in mitochondria of PCZT cells (Fig. 7). Matrical granules were also encountered in mitochondria of PCT cells of control animals, but they were not

as complex in structure and usually occurred singly within the matrix, not in clusters. Methyl mercury appeared to produce swelling of the apical cytoplasm and loss of apical microvilli of DCT cells in Se-deficient

CH,Hg

AND

Se INTERACTION

animals both after 7 days of CH3Hg treatment at an elevated dose (Fig. 8) and after 20 days of chronic treatment at the lower dose. Acivicin effects on the cellular structure of Se-deficient animals were not noticeable after 1 day of treatment with this agent. After 7 days of treatment with acivicin, the most severely affected subcellular compartments of PCT cells were the mitochondria and endoplasmic reticulum which consistently exhibited moderate swelling (Fig. 9). Lysosomes and electron-lucent vacuoles occurred in the apical cytoplasm of PCT cells, but these were not unusual in their abundance or appearance. Acivicin also seemed to affect the DCT cells slightly as observed through the presence of fewer apical microvilli than those in the controls. DISCUSSION Results of this study indicated that CH3Hg had adverse effects on renal ultrastructure and moderate inhibition on y-GTPase activity. Renal morphological results following CH3Hg treatment confirm previous observations that the ultrastructure of the PCT cells is altered, especially in the lysosomal system. Our results extend previous observations by indicating that cells of the DCT exhibit swelling of the apical cytoplasm and loss of microvilli following CH3Hg treatment. Selenium exerted a moderate reversal of the CH3Hg inhibition of y-GTPase and a moderate protection against renal ultrastructural damage. Acivicin, a potent y-GTPase inhibitor, inhibited the enzyme prior to any noticeable subcellular effects. Although y-GTPase does not appear to be a major biochemical site of action of CH3Hg, the biochemical and morphological data support the possibility that some form of Hg or a thiol complex of Hg can serve as an acceptor or donor for yGTPase. Acivicin, an antitumor glutamine antagonist, is a highly effective affinity label for the y-glutamyl site of y-GTPase and causes a rapid, irreversible inactivation of y-GTPase

IN

KIDNEY

87

(Griffith and Meister, 1980; Allen et al., 198 1). Acivicin was shown in our study to cause a marked inhibition of y-GTPase activity that was not affected by increased dose or multiple injections. Acivicin inhibited yGTPase activity prior to any noticeable subcellular effects. Methyl mercury caused moderate inhibition of y-GTPase in conjunction with extensive renal damage. These results indicate that CH3Hg-induced renal damage preceded y-GTPase inhibition or may have resulted in the enzyme inhibition. It is also possible that renal damage and y-GTPase inhibition are independent events. The available data did not clarify whether the ultrastructural changes induced by acivicin and CH,Hg were direct or indirect. These agents appeared to influence different subcellular compartments. It is possible that CH3Hg acted by increasing the lysosomal population while acivicin acted through mitochondrial and rough endoplasmic reticulum systems. A single high dose of CH3Hg depressed yGTPase activity after a 24-hr exposure as did a low dose after a 20-day exposure. There was no early effect (24 hr) of the low dose of CH3Hg on y-GTPase activity or renal subcellular structures. Most changes in y-GTPase activity, following exposure to CH3Hg, were observed after distinct neurological signs were evident. Kidney ultrastructural evidence revealed several changes following CH3Hg exposure: vacuolation, edema, increased lysosomal profile, cytoplasmic degradation, and extrusion of cellular masses into the tubular lumen. Animals from the 20-day study exhibited the most severe cellular damage. The cellular changes were similar to those observed by others following CH3Hg administration (Chang et al., 1973; Chang and Sprecher, 1976; Chen et al., 1983; Fowler, 1972a; Ware et al., 1975). Our results, indicating mild DCT focal tubular degeneration and some widespread proximal tubular lesions, were consistent with the results of Klein et al. ( 1972) for rats intoxicated with CH3Hg. Mercury uptake via lysosomes has been impli-

FAIR, DOUGHERTY,

AND BRADDON

FIG. 6. Lysosomes (ly) from PCT cells of Se-deficient mice treated with CH,Hg. (a) Lysosomes ( IY) of a mouse after treatment for 7 days contain semicrystalline arrays of membranous structures. Some alwear U niformly electron dense and some appear to have fused with clear vacuoles (ly vat). (X 12,320 .I (b)

CH,Hg AND Se INTERACTION

IN KIDNEY

Portions of these lysosomes are very electron dense. (X8990.) (c) Lysosomes with variable internal structures and electron density in the apical cytoplasm of PCT cells in a Se-deficient mouse treated with a low dose of CHaHg for 20 days (X19,600.) (d) These lysosomes are large, limited by a single, delicate membrane and contain semicrystalline arrays of membranes and membranes of smooth endoplasmic reticulum apparently in the process of degrading within these lysosomes. (X30,400.)

89

90

FAIR, DOUGHERTY,

AND BRADDON

FIG. 7. Mitochondria in PCT cells of a Se-deficient mouse treated with CH3Hg for 7 days appear normal, except for the accumulation of dense matrical inclusions which frequently occur in clusters. (X66,ooo.)

cated in altering lysosomal function (Verity and Brown, 1970) and for sequestering and detoxifying Hg (Fowler et al., 1974, 1975). In addition to nonautophagic dense, homogenous lysosomes noted by Fowler et al. ( 1975), our study demonstrated the formation of autophagic lysosomes with complex internal structures. The noted lack of structural change in the glomerulus in this study was also consistent with other reports (Fowler, 1972a; Klein et al., 1972). The effect of HgC& on y-GTPase activity of the kidney in rats has been investigated by Chung et al. (1982). They observed a 30% decrease in y-GTPase activity (24 hr) that could be prevented by administration of Se. The possibility exists that Se-related increases in the activities of the enzymes involved in glutathione metabolism may contribute to the inactivation of Hg. The results obtained in our study are consistent with these observations. While administration of Se in our 7-day study did not reverse the decrease in

y-GTPase activity produced by CH3Hg, a moderate reversal was observed in the 20day study, perhaps reflecting the difference between HgC& and CH3Hg metabolism. During periods when y-GTPase was depressed, serum creatinine generally reflected the structural pathology of the kidneys. For example, when y-GTPase was depressed by acivicin on Day 1, no changes were evident in serum creatinine levels or in the ultrastructure of the kidney. It was not until Day 7 when y-GTPase activity remained depressed by acivicin that subcellular compartments of PCT were affected and creatinine concentrations became elevated. However, treatment with CH3Hg alone depressed yGTPase activity but caused no alterations in creatinine concentration in the 7-day study. Administration of CH3Hg for 20 days resulted in elevated creatinine concentrations and depressed y-GTPase activity, and the extent of cellular damage occurred in the following order of treatments: CHsHg > CH3Hg + Se

CH,Hg AND Se INTERACTION

IN KIDNEY

91

FIG. 8. DCT cells of a Se-deficient mouse treated with CHJHg for 7 days. Note swelling of apical cytoplasm and lack of microvilli on the apical plasma membrane. Nuclei (N) and basal mitochondria (m) are of normal appearance. Portion of a cilium (ci). (X 12,320.)

> control. In contrast to expected increases in serum urea nitrogen concentrations refleeting renal damage, mice treated with CH3Hg, acivicin, or CHsHg and Se had

significantly depressed urea nitrogen concentrations compared to control mice. Only in the experiment using an acute single dose of CHsHg (8.0 mg/kg) were significant increases

92

FAIR, DOUGHERTY,

AND BRADDON

FIG. 9. PCT cells of a Se-deficient mouse treated with acivicin for 7 days. Note slightly swollen mitochondria (m) and endoplasmic reticulum (er). Microvilli (mv) and nuclei (N) do not appear to be adversely affected. (X12,320.)

observed in serum urea nitrogen concentrations compared to the controls. This acute dose also caused a significant decrease in yGTPase activity and elevated creatine con-

centration. As an indicator of renal func:tion serum urea nitrogen was not as sensitit re as serum creatinine. Since urea is a by-pro duct of protein metabolism synthesized in the

CH3Hg

AND

Se INTERACTION

liver, any toxic insult that influences protein metabolism (hepatotoxicity, poor nutrition) could affect blood concentrations of urea. Thus, the depressed concentrations of urea nitrogen observed after CH3Hg and acivicin treatments in this study may indicate impaired hepatic function. This observation is in agreement with Stroo and Hook (1977) who found that blood urea nitrogen concentrations in rats were affected by HgC& but not by CH3Hg. Klein et al. (1972) found higher concentrations of urea nitrogen in rats but only 15 days after CH3Hg administration of 10 mg/kg and after numerous cellular changes were evident in the kidney. In this study, both Se-deficient and Seinjected mice were treated with CH3Hg, by acute and short-term chronic doses, to evaluate the action of Se on CH3Hg toxicity. Se was shown to exert a protective effect against neurological and nephrological damage as evidenced by behavioral indices and kidney histology. The PCT and DCT cells of animals which received Se just prior to CH3Hg exhibited considerably fewer lysosomes, vacuoles, and necrotic cells. In the 20-day study, Se modified the inhibition of y-GTPase and the change in creatinine concentration produced by CH3Hg treatment. Selenium has been shown to protect against the nephrotoxic effects of HgQ (Parizek and Ostadalova, 1967; Groth et al., 1973, 1976). The results from this study demonstrate that Se (in the 7- and 20-day studies) exerts a protective effect against morphological changes produced by CH3Hg in the kidney. Although Se modification of CH3Hg toxicity is well established, the interactions of Se and Hg are complex, and the mechanism is unknown (review, Magos and Webb, 1980). One possible protective effect of Se on CH3Hg toxicity is an in vivo alteration in the distribution of Hg. Many discrepancies are found in studies investigating the effect of Se on the accumulation of Hg by the kidneys and can be attributed to variation in the dose, exposure time, form of Hg, and experimental protocol. Despite the different forms of Hg used in

IN

KIDNEY

93

many studies (Chen et al., 1974, 1975; Groth et al., 1976; Magos and Webb, 1976; Ohi et al., 1975, Potter and Matrone, 1974), it was generally found that Se produced a lower Hg content in kidneys if treated with Hg and Se for a relatively short time. Our acute 7-day study showed that dietary Se was only able to initially effect a decrease in Hg accumulation by the kidney. After the second consecutive day of injections, only Se-deficient animals injected with Se prior to CH3Hg administration accumulated significantly less Hg compared to Se-deficient and Se-sufficient animals injected with CHsHg alone. No significant difference in Hg accumulation occurred in the 20-day study. Although mice in both the 7- and 20&y studies received equivalent total doses of CH3Hg, the overall accumulation of Hg in the kidneys of mice from the 7-day study averaged 120 &g, while the kidneys from the 20&y chronic study averaged 38 pg/g. Renal cellular injury was not as severe at the end of the 7-day study as it was at the end of the 20-day study. These data support observations that concentration of CHsHg in organs is not always correlated with expressions of CH3Hg toxicity (Ohi et al., 1975; Stoewsand et al., 1974). An important factor to consider when assessing the effects of CHjHg on Hg accumulation or toxicity is the biotransformation or organic Hg to inorganic Hg. Organomercury compounds are converted to inorganic Hg in the kidneys (Berlin et al. 1975; Norseth and Clarkson, 1970) where they are concentrated to high levels (Berlin and Ullberg, 1963; Swensson and Ulfvarson, 1968). Inorganic Hg has been shown to be a potent nephrotoxin, acting primarily on the PT (proximal tubule), inducing necrosis of tubular epithelium, and resulting in acute renal failure. A time sequence study of HgC& effects on renal morphology by Ganote et al. (1975) revealed selective necrosis of the PT with some changes similar to those induced by CHJHg (vacuolization, cellular swelling, and mitochondrial changes). In contrast,

94

FAIR, DOUGHERTY,

other changes caused by HgC12 administration, such as loss of border and clumping of nuclear chromatin in the PT, are not present following CHsHg exposure. The kidneys from mice in the 20&y study had time to reach an equilibrium state in the conversion of CHsHg to inorganic Hg, thus resulting in a higher accumulation of Hg. Renal accumulation of Hg in the 7-day study did not approach steady state and thus less conversion of CHsHg to inorganic Hg occurred. Results from our study show Se effects on Hg distribution in the kidney seem to be dependent on both dose and exposure time as well as the Se status of the animal. In acute CHsHg intoxication, renal injury occurred only after distinct neurological signs were present, while chronic CH,Hg poisoning exhibited a greater nephrotoxic effect suggesting the potential importance of the conversion of CHsHg to Hg in renal pathogenesis of CH3Hg. ACKNOWLEDGMENTS The authors gratefully acknowledge the expert assistance of Ms. M. Dougherty in electron microscopy and Mr. J. Wade in technical support. We thank Ms. S. Semple for animal care and Ms. J. Carson for typing the manuscript.

AND BRADDGN BALTHROP, J. E., AND BRADDON, S. A. (1985). Effect of selenium and methylmercury upon glutathione and glutathione-S-transferase in mice. Arch. Toxicol. Environ. Contam. 14, 197-202. BERLIN, M. (1963). On estimating threshold limits for mercury in biological material. Acfa Med. Stand. (SuppI.)

396,

Environ.

Health

ALEXANDER, J., AND AASETH, J. (1982). Organ distribution and cellular uptake of methylmercury in the rat as influenced by the intra- and extracellular glutathione concentration. B&hem. Pharmacol. 31, 685690.

ALLEN, L. M., CORRIGAN, M. V., AND MEINKING, T. (1981). Interaction of AT-125, (aS,SS)-amino-3chloro4,5dihydro-isoxazoleacetic acid, with bovine kidney y-glutamyl transpeptidase. Chem. Biol. Interact. 33, 36 l-365. BAKIR, F., DAMLUJI, S. F., AMIN-ZAKI, L., MURTADHA, M., KHALIDI, A., AL-RAWI, N. Y., TIKRITI, S., DHAHIR, H. I., CLARKSON, T. W., SMITH, J. C. AND DOHERTY, R. A. (1973). Methylmercury poisoning in Iraq. Science (Washington, D.C.) 181, 230-241. BALLATORI, N., AND CLARKSON, T. W. (1982). Developmental changes in the biliary excretion of methylmercury and glutathione. Science (Washington, D.C.) 216,62-63.

12, 407-4

16.

CHEN, R. W., LACY, V. L., AND WHANGER, P. D. (1975). Effect of selenium on methylmercury binding to subcellular and soluble proteins in rat tissues. Res. Commun.

REFERENCES

1-29.

BERLIN, M., CARLSON, J., AND NORSETH, T. (1975). Dose-dependence of methylmercury metabolism: A study of distribution, biotransformation, and excretion in the squirrel monkey. Arch. Environ. Health 25, 7791. BERLIN, M., AND ULLBERG, S. (1963). Accumulation and retention of mercury in the mouse. III. An autoradiographic comparison of methylmercuric dicyandiamide with inorganic mercury. Arch. Environ. Health 6, 6 10-6 16. CHANG, L. W. (1977). Neurotoxic effects of MercuryA review. Environ. Res. 14, 329-373. CHANG, L. W., AND SPRECHER,J. A. (1976). Degenerative changes in the neonatal kidney following in utero exposure to methylmercury. Environ. Res. 11, 392406. CHANG, L. W., WARE, R. A., AND DESNOYERS,P. A. (1973). A histochemical study on some enzymatic changes in the kidney, liver and brain after chronic mercury intoxication. Food Cosmet. Toxicol. 11,283286. CHEN, W., BODY, R. L., AND MOTTET, N. K. (1983). Biochemical and morphological studies of monkeys chronically exposed to methylmercury. J. Toxicol.

Chem.

Pathol.

Pharmacol.

12, 297-308.

CHEN, R. W., WHANGER, P. D., AND FANG, S. C. (1974). Diversion of mercury binding in rat tissues by selenium: A possible mechanism of protection. Pharmacol.

Res. Commun.

6, 571-579.

CHUNG, A. S., MAINE& M. D., AND REYNOLDS, W. A. (1982). Inhibition of the enzymes of glutathione metabolism by mercuric chloride in the rat kidney: Reversal by selenium. Biochem. Pharmacol. 31,30933100. CLARKSON, T. W. (1972). The pharmacology of mercury compounds. Annu. Rev. Pharmacol. 12, 375-406. FEHLING, C., ABDULLA, M., BRUN, A., DICTOR, M., SCHUTZ, A., AND SKERFVING, S. (1975). Methylmercury poisoning in the rat: A combined neurological, chemical and histopathological study. Toxicol. Appl. Pharmacol.

33,27-37.

FOWLER, B. A. (1972a). Ultrastructural evidence for nephropathy induced by long-term exposure to small amounts of methyl mercury. Science (Washington, D.C.)

175, 780-781.

FOWLER, B. A. (1972b). The morphologic

effects of

CH,Hg AND Se INTERACTION dieldrin and methyl mercuric chloride on pars recta segment of rat kidney proximal tubules. Amer. J. Pathol.

69, 163-178.

FOWLER, B. A., BROWN, H. W., LUCIER, G. W., AND BEARD, M. E. (1974). Mercury uptake by renal lysosomes of rats ingesting methyl mercury hydroxide. Arch. Pathol. 98, 297-30 1. FOWLER, B. A., BROWN, H. W., LUCIER, G. W., AND KFUGMAN, M. R. (1975). The effects of chronic oral methyl mercury exposure on the lysosomal system of rat kidney. Lab. Invest. 32, 313-322. GANOTE, D. E., REIMER, K. A., AND JENNINGS, R. B. (1975). Acute mercuric chloride nephrotoxicity: An electron microscope and metabolic study. Lab. Invest. 31,633-647.

GANTHER, H. E., GOUDIE, C., SUNDE, M. L., KOPECKY, M. J., WAGNER, P., OH, S-H., AND HOEKSTRA, W. G. (1972). Selenium: Relation to decreased toxicity of methylmercury added to diets containing tuna. Science (Washington, D.C.) 175, 1122-I 124. GASIEW~CZ,T. A., AND SMITH, J. C. (1978). Interaction between cadmium and selenium in rat plasma. Environ. Health Perspect. 25, 133- 136. GRIFL~TH, 0. W., AND MEISTER, A. (1980). Excretion of cysteine and y-glutamylcysteine moieties in human and experimental animal y-glutamyl transpeptidase deficiency. Proc. Nat/. Acad. Sci. USA 77, 3384-3387. GROTH, D. H., STETTLER, L., AND MACKAY, G. (1976). Interactions of mercury, cadmium, selenium, telhuium, arsenic and beryllium. In Effects and Dose-Response Relationships of Toxic Metals (G. F. Nordberg, ed.), pp. 527-543. Elsevier, Amsterdam. GROTH, D. H., VIGNATI, L., LOWRY, L., MACKAY, G., AND STOKINGER, H. E. (1973). Mutual antagonistic and synergistic effects of inorganic selenium and mercury salts in chronic experiments. Trace Subst. Environ. Health 4, 187-189. HILL, K. E., AND BURKE, R. F. (1982). Effect of selenium deficiency and vitamin E deficiency on glutathione metabolism in isolated rat hepatocytes. J. Biol. Chem. 257, 10668-10672.

KLEIN, R., HERMAN, S. P., BRUBAKER, P. E., LUCIER, G. W., AND KRIGMAN, M. R. (1972). A model of acute methyl mercury intoxication in rats. Arch. Pathol. 93,408-418.

KURLAND, L. T., FARO, S. W., AND SIEDLER, H. (1960). Minamata disease: The outbreak of a neurologic disorder in Minamata, Japan and its relationship to the ingestion of seafood contaminated by mercuric compounds. World Neurol. 1, 370-39 1. LOWRY, 0. H., ROSEBROUGH,N. J., FARR, A. L., AND RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265275.

MAGOS, L., AND WEBB, M. (1976). Differences in distribution and excretion of selenium and cadmium or mercury after their simultaneous administration sub

IN KIDNEY

95

cutaneously in equimolar doses. Arch.

Toxicol.

36,

63-69.

MAGOS, L., AND WEBB, M. (1980). The interactions of selenium with cadmium and mercury. Crit. Rev. Toxicol. 8, l-42. NORSETH, T., AND CLARKSON, T. W. (1970). Studies on the biotransformation of 203Hg-labelled methylmercury chloride in rats. Arch. Environ. Health 21, 7 17-727.

OHI, G., NISHIGAKI, S., SEKI, H., TAMURA, Y., MAKI, T., KONNO, H., OCHIAI, S., YAMADA, H., SHIMAMURA, Y., MIZOGUCHI, I., AND YAGYU, H. (1976). Efficacy of selenium in tuna and selenite in modifying methylmercuty intoxication. Environ. Res. 12, 49-58. OHI, G., NISHIGAKI, S., SEKI, H., TAMURA, Y., MAKI, T., MAEDA, H., OCHIAI, S., YAMADA, H., SHIMAMURA, Y., AND YAGYU, H. (1975). Interaction of dietary methylmercury and selenium on accumulation and retention of these substances in rat organs. Toxicol. Appl. Pharmacol.

32, 527-533.

ORLOWSKI, M., AND MEISTER, A. (1965). Isolation of y-glutamyl transpeptidase from hog kidney. J. Biol. Chem.

240, 338-347.

PARIZEK, J., AND OSTADALOVA, I. (1967). The protective effect of small amounts of selenite in sublimate intoxication. Experientia 23, 142-143. POTTER, S., AND MATRONE, G. (1974). Effect of selenite on the toxicity of dietary methylmercury and mercuric chloride in the rat. J. Nutr. 104, 638-647. REFSVIK, T. (1978). Excretion of methylmercury in rat bile: The effect of diethylmaleate cyclohexene oxide and acrylamide. Acta Pharmacol. Toxicol. 42, 135141. REFSVIK, T., AND NORSETH, T. (1975). Methyl mercuric compounds in rat bile. Acta Pharmacol. Toxicol. 36, 67-78.

SOKAL, R. R., AND ROHLF, F. J. (1969). Principles and Practice search. Freeman, San

of Statistics

Biometry: in Biological

The Re-

Francisco. SPURR, A. R. (1969). A low viscosity resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26, 31-43.

STEEL, R. G. D. AND TORRIE, J. H. (1960). Principles and Procedures ofStatistics. McGraw-Hill, New York. STILLINGS, B. R., LAGALLY, H., BAUERSFELD, P., AND SOARES,J. (1974). Effect of cystine selenium, and fish protein on the toxicity and methylmercury in rats. Toxicol.

Appl.

Pharmacol.

30, 243-254.

STOEWSAND, G. S., BACHE, C. A., AND LISK, D. J. ( 1974). Dietary selenium protection of methylmercury intoxication of Japanese quail. Bull. Environ. Contam. Toxicol. 11, 152-156. STROO, W. E., AND HOOK, J. B. (1977). Renal functional correlates of methyl mercury intoxication: Interaction with acute mercuric chloride toxicity. Toxicol. Appl. Pharmacol. 42, 399-4 10. SUZUKI, T. (1969). Neurological symptoms from con-

96

FAIR,

DOUGHERTY,

centration of mercury in the brain, In Chemical (M. W. Miller and G. C. Berg, eds.), pp. 24% 257. Thomas, Springfield, III. SWENSSON,A., AND ULFVARSON, V. (1968). Distribution and excretion of mercury compounds in rats over a long period following a single injection. Acta PharFallout

macol.

Toxicol.

26, 273-283.

TATE, S. S., AND MEISTER, A. (1974). Interaction of yghttamyl transpeptidase with amino acids, peptides, and derivatives and analogs of ghrtathione. J. Biol. Chem.

249,7593-7602.

TATE, S. S., AND MEISTER, A. (198 I). y-Glutamyl transpeptidase: Catalytic, structural and functional aspects. Mol. Cell. Biochem. 39, 357-368. THOMAS, D. J., AND SMITH, J. C. (1982). Effects of

AND

BRADDON

coadministemd low-molecular weight thiol compounds on short-term distribution of methyl mercury in the rat. Toxicol. Appl. Pharmacol. 62, 104-l 10. VERITY, M. A., AND BROWN, W. J. (1970). Hg*+induced kidney necrosis: Subcellular localization and structure-linked lysosomal enzyme changes. Amer. J. Pathol. 618, 57-69. WARE, R. A., BURKHOLDER, P. M., AND CHANG, L. W. (1975). Ultrastructural changes in renal proximal tubules after chronic organic and inorganic mercury intoxication. Environ. Res. 10, 12 I-140. WARE, R. A., CHANG, L. W., AND BURKHOLDER, P. M. (1973). Ultrastructural pathology of the kidney following organic and inorganic mercury intoxication. Fed. Proc. 32, 823.