Document NN24vzKM77B2mpnnbQQJzyQVw

posium held in im. Eur. Cotn- md metabolic i lysis of traits* . Newton. D. itigations into well. England. US Environ017. Washing- J C. A Knel1 continuously Environmental dited by T. B. gart. I Exposures in iments). Lead >n Ave_ New on Reference ess. Oxford. (1975*. Levels opulations of II 27. C, Goldberg. ' lead in dnnk- ctive. National ilth Standards. il Register 43, Stanley. S. A id dehydratase icyte count. In ceedtngs of anisterdam. 2-6 munities. Lux- Kopple. J. O. sin in healthy eria J. Lead. i9l An investicumulalor fac ie. J. ind. Med. I I i $ | 1 J * FJC"'*rt T'txu.ti Vol pp 6)9 IO 650. 1911 Primed in Great Britain All rights reserved 001 61 0)0*)*.|0 Copyright O 19ft a--g r-itn ^ THE ROLE OF ZINC IN NITRILOTRIACETATE(NTA)ASSOCIATED RENAL TUBULAR CELL TOXICIT R. L. Andekson The Procter i Gamble Company. Miami Valley Laboratories. P.0. Box JV f 75. Cincuimui OH 45247, USA (Received 24 December I9SII) Summary--Both the extent and the seventy of renal tubular cell toxicity associated with constant systemic loads of nitnlotrtacetate (NTA) were shown to be dependent upon the availability of zinc in the circulation. NTA ingestion at doses that produced plasma ultrafiltrate (UF) NTA concentrations >20 mm caused an increase in UF zinc. The majority of the increased zinc but not the NTA was resorbed during renal clearance of NTA and the extent and severity of tubular cell lesions was propor tional to the amount of zinc resorbed. Intravenous infusion studies demonstrated that both zinc salts and ZnNaNTA produced nephrotoxicity that is very similar to heavy metal toxicity. The toxicity of iv ZnNaNTA was accompanied by renal tissue accumulation of zinc but not of NTA and the increased tissue zinc was not removed by saline infusion. These studies have led to the development of a model for NTA-assodated nephrotoxicity that demonstrates that the initiation, propagation and extent of renal tubular ceil toxicity are all dependent upon NTA-induced changes in zinc metabolism. This model negates the use of any mathematical models that estimate low dose (UFNTA < 20jim) effects from toxicity data obtained at doses that exceed the threshold <UF,,TA > 20 jiul Introduction primary site of NTA-assodated toxiaty. At dietary NTA doses of up to 18jimol/g diet (0-5% Nitrilotriacetate (NTA) is a divalent metal chelating NajNTA.HjO in the diet) the toxic response noted agent which has been shown to be functional in after 24 months of NTA ingestion has been described laundry detergents. Reviews of the mammalian tox as hydropic degeneration (Nixon et at. 1972). Higher icity and environmental effects and fate of NTA have doses of NTA are associated with renal tubular cell been published (Foley. Becking. Muller. Goyer, Falk adenoma and adenocarcinoma (National Cancer In & Chernoff, 1977; Shapiro, Chapman, Dick, Dillon, stitute. 1977). Study of the natural history of NTA- O'Melia. Spacie & Leduc 1978; Thayer A Kensler, assodated tubular cell toxiaty from acute, subchronic 1973). NTA is readily absorbed from the gut and is and chronic studies had led to the condusion that the excreted only no the urine (Michael A Wakim, 1971). process is a continuum, which begins with proximal In transit through mammals there is no alteration in convoluted tubule (PCT) vacuolization, progresses to the structure of the NTA molecule (Thayer A simple and nodular hyperplasia with vacuoles to ade Kensler. 1973) and renal clearance of NTA is ac nomatous hyperplasia, and subsequently appears as complished by filtration with no evidence for the tubular cell tumours (Alden. Kanerva. Anderson A existence of a tubular cell transport system for either Adkins, 1981). In addition to this spedfic sequence of secretion or resorption (A. Licht, N. S. Bricker, R. E. responses. NTA ingestion also exacerbates the spon Papendick A R. L. Anderson, unpublished data. taneous nephrosis noted in ageing rats (Alden A 1980). Kanerva. 1980). The PCT toxiaty, even at the nodu NTA ingestion, processing and excretion results in lar hyperplasia stage is reversible if NTA is removed changes in divalent cation (M1*) disposition (Ander from the diet (M. C. Myers. R. L Kanerva. C. L. son A Kanerva. 1978a; Michael A Wakim. 1973). The Alden A R. L Anderson, unpublished data. 1980). divalent canons most sensitive to NTA are magne Extensive testing has shown that NTA is not muta sium (Mg) and zinc (Zn). NTA increases bone and genic or teratogenic (Foley et <iL 1977). This finding urinary Zn and decreases faecal Zn (Anderson A strongly implies that the tumorigenidty associated Kanerva. 1978a: Nixon. Buehler A Niewenhuis. with the ingestion of high doses of NTA is an epigene 1972) . In contrast NTA increases faecal Mg and tic phenomenon. Thus the primary concern is how a reduces plasma, bone and urinary Mg. Doses of molecule that is devoid of genotoxidty. is not metab NTA that alter Zn and Mg disposition (05*. olized to a genotoxtc moiety, is not transported across Na,NTA.H20 in diet) do not affect calcium (Ca) but PCT cells, and the only significant chemical activity higher doses of NTA O0-75*; in the diet) increase of which is its propensity to form stable complexes urinary Ca excretion (Anderson A Kanerva. 1978b: with MJ\ can be responsible for the PCT toxiaty Anderson A Kanerva. 1979; Michael A Wakim. (including tumours) noted after chronic exposure at 1973) . high doses. Acute (Mcrski. 1980L subchronic (Alden A A comparison of the dose-response curves showed Kanerva. 1979: Mahaffey A Goyer. 1972) and chronic a very dose similarity between PCT toxiaty and in (National Cancer Institute. 1977; Nixon et al. 1972) creased Zn exaetion in the urine. For example. 24 ingestion studies show that the renal tubules are a months of ingesting a diet containing IT jzmol NTA/g 921001 RowVerK ()<M 640 R. L. Anderson (0-03'`,, NajNTA.HjO) caused no alteration in renal morphology or urinary Zn excretion. Dietary NTA at 55jimolg (015% NajNTA.HjO) increased urinary Zn excretion and was associated with PCT vacuoles and hyperplasia, changes described as hydropic degeneration (Nixon et at. 1972). Both urinary Zn excretion and PCT lesions were enhanced when dietary' NTA was increased to l8jimol/g (05; Na,NTA . HjO and 05. CaNaNTA). This report describes the results of a series of ex periments in which the relationship between NTAinduced changes in Zn disposition and PCT toxicity was more extensively examined. The results of these experiments show that the initiation of PCT toxicity is dependent upon blood NTA attaining concen trations high enough to increase the plasma ultrafil trate Zn (UF*.) levels. A large proportion of the in creased UFZ. is resorbed by the PCT during renal clearance of NTA, and by some undefined mechanism this alters the endocytotic-lysosomal system to induce vacuoles and subsequently tubular cell hyperplasia. This response has' demonstrable no-effect levels which negate the validity of mathematical extrapolations of responses (including tumours) noted at high doses to possible incidences at low doses. Experimental All of the experiments reported used Sprague-Dawlcy-denved rats obtained from Charles River Breeding Laboratories. Inc. Wilmington, MA. The animals were housed individually in stainless-steel metabolism cages and kept at controlled temperature (70 5F) and humidity (40-60;) with a 12-hr light/dark cycle. In most of the experiments the diet was ground Pur ina Laboratory Chow (Ralston Purina Co, St Louis, MO) and the animals were given feed and distilled water ad lib. For one study the rats were trained to consume their daily intake of feed in a 1 -hr period (9-10 am.). Meal training was accomplished by reduc ing the time during which the diet was available over a 2-wlc period. In (he study in which the dietary Zn level was varied the animals were fed a semi-purified diet containing 8. 14. 21 or 52 ppm Zn. The diet was composed of 20% casein, 20% lard. 60% sucrose and all of the vitamins and minerals (except Zn) at greater than the recommended daily allowance (National Academy of Sbcnccs/National Research Council 1972) for the rat. In all instances NTA was added to the diets on a w/w basis. Weekly weight gains and twice-weekly feed consumption records were main tained on all animals. Urine samples for mineral and NTA analyses were total 24-hr collections obtained in cages with a separator which retained the faeces and allowed urine to be collected in tarcd bottles. The separators were not nnsed. For iv infusions. 200-250 g male rats were fitted with catheters of Silastic 602-135 tubing (Dow Corn ing Corp. Medical Products. Midland. Ml) in the jugular vein as described by Upton (19751 The cath eters were connected to a Harvard 975 Infusion Pump (Harvard Apparatus. Millis. MA) set to deliver 10 id/ min. The animals were restrained throughout the in fusion period and were provided with chow pellets (Ralston Purina Co.) and distilled water ad lib. Dur ing infusion the voided faeces were collected on a nylon screen over a funnel and the urine was collected in tared vials. Before they were killed the animals were anaesthe tized with ether, and then a mid-line incision was made and blood was collected from the vena cava in heparinized syringes. The blood cells were separated by centrifugation and when required, plasma ultrafil trate (UF; MW < 10.000) was obtained by filtering a portion of the plasma through an Amicon PM-10 fil ter (Amicon Corp, Lexington. MA). Minerals were determined by atomic absorption spectrometry. Urine NTA analyses were by reverse isotope dilution when unlabelled NTA was fed and by direct count after 25:1 dilution with water when [,4C]NTA was used (Anderson & Kanerva, 1978b). Blood-cell, plasma and UF [I4C]NTA analyses were carried out after combustion of [14C]NTA to 14COj. All l4C levels were determined by scintillation counting. The kidneys were processed using routine histologi cal techniques followed at our laboratory. The tissue was placed in 10% neutral phosphate-buffered forma lin and allowed to fix for 18 hr. The tissues were embedded in paraffin blocks according to standard histological methods, sectioned on a microtome at 5 jim. stained with haetnatoxylin and eosin and evalu ated by light microscopy. [For further details see Alden et al. (1981) and Merski (1981).] Electron microscopy was carried out by Mr M. Myers of this laboratory. Alt forms of unlabelled NTA were prepared from a commercial sample of Na}NTA. HjO (Monsanto Co. Alvin. TX). HjNTA was prepared by treating sol utions of NajNTA.HjO with concentrated HG which precipitates the HjNTA. The divalent metal (M*) complexes of NTA were prepared by stirring equal molar amounts of the desired metal oxide and HjNTA in distilled water for about 1 hr and then titrating to a pH of 7-4 with NaOH (NHOH was used for the Ca complex). The complexes were diluted to the desired concentrations with water. Each complex was assayed for its content to define the NTA complex concentration for infusion. The ZnNaNTA for feeding was prepared in the same way but was precipitated by addition of three volumes of 95% (v/v) ethanol and dried in a vacuum oven at 30C. The [l4C]NTA used in all cases was labelled in the carboxyl groups of the acetate units. Periodic de terminations of [,4C]NTA stability were made by isotope dilution with HjNTA. Results The correlation between NTA-associated PCT tox icity aod increased urinary Zn excretion suggested that ingestion of the Zn complex of NTA (ZnNaNTA) could result in increased NTA renal toxicity. To test this hypothesis, mature male rats were fed diets con taining dS,.10, 15 or 20% ZnNaNTA for 3 wk. After 21 days of ad lib. feeding, the levels of Zn and NTA in a 24-hr urine collection were measured, and the kid ney: body weight ratio was ascertained as an indicator of renal toxicity (Anderson & Kanerva. 1978a). The urinary NTA recovery was low and approximately constant at the three lowest ingestion rates but was increased at the highest intake rate (Fig. la). Urinary Zn displayed a linear increase with ingestion rate k9 *21001 RowVerK Zn and NTA renal toxicity 641 t Fig. 1. Effect of ingestion of ZnNaNTA () at dietary levels of 0.05. 1-0. 15 or 2-0*. for 3 wlc on (a) urinary NTA levels (mean SEM for five rats) and (b) urinary Zn levels (single analysis on pooled urine from five rats) in adult male rats. For comparison, the effect on the same par ameters of NsjNTA . HjO given at 20T. in the diet is also shown (A: mean + SEM for five rats). (Fig. lb) but it did not approach the urinary NTA concentration. This clearly shows that the ZnNaNTA complex does not remain intact during passage through the body. The kidney:body weight ratio was not significantly influenced by any of the ZnNaNTA doses. For comparative purposes the effect of the highest dietary level of NTA (2%) as its trisodium salt (NajNTA.HjO) on the same parameters is included in Fig. 1. NajNTA.HjO ingestion resulted in about a three-fold greater urinary NTA load and a marked increase in kidney:body weight ratio (dau not shown) and an even higher urinary Zn level at a much lower NTA intake level. This suggests that if NTA-tnduced renal toxicity is causally related to altered Zn metab olism the effects are due to NTA-associated alter ations in systemic Zn metabolism and are not a con sequence of ingested ZnNaNTA absorption and excretion in the urine. This contention is supported by the results of a study of the effect of NTA on cation balance (Anderson & Kanerva, 1978a). which demonstrated that the increased urinary Zn excretion associated with NTA ingestion was compensated for by decreased faecal Zn excretion, so that NTA did not alter Zn balance compared to that of control rats. Since increasing dietary Zn had the net effect of reducing NTA absorption, the renal toxicity associ ated with lowered dietary Zn at a constant dietary NTA level was ascertained. Male rats were ted semipurified diets containing 12% NajNTA.HjO with either 52 (equivalent to chow). 21. 14, or 8/ig Zn/g diet. After 4 wk of ingestion of these diets ad tib., plasma-Zn and urinary Zn and NTA levels were measured and the kidneys were evaluated histologi cally with emphasis on the specific NTA-assodated histopathology (vacuoles and hyperplasia in the PCT). At the three lowest dietary Zn levels NTA ad ministration resulted in a reduced plasma-Zn concen tration which was overcome at the highest dietary Zn level (Fig. 2ak In contrast, at all dietary Zn levels, NTA ingestion resulted in an increase in urinary Zn excretion relative to control rats (Fig. 2b). This in crease was dependent on dietary Zn concentration. The urinary NTA level was approximately constant at all four dietary Zn levels (Fig. 2b). Thus the aim of establishing approximately constant urinary NTA excretion with markedly different 2Ln excretion was achieved. In addition, these results show that when dietary Zn is limited, the increased urinary Zn excre tion associated with NTA excretion results in de creased plasma-Zn levels. Table 1 summarizes the results of the histopathol ogy noted in the kidneys from the animals fed the two highest doses of Zn. expressed as the toxicity index, that is the product of incidence (kidneys having the specific lesion type/ten kidneys examined) and fre quency (average number of the specific lesion/kidney section). The toxicity indices for the kidneys from rats fed the two lowest levels of Zn (8 and I4j>pm) are not included in Table 1 since in the 14 sections examined only vacuolated tubules and two tubules with OMaryZn.jig/g Owwry Zn ,*/ Fig. 1 Effects of different levels of dietary Zn |8.14. 21 or 52 ppm) on (al plasma Zn and |bl unne Zn in rats given either 0 (------ 01 or 1-2% (----- ) NajNTA. H(0 in the diet Urinary NTA levels are also shown (A-----Al Plasma values are means SEM for group* of five rats. The urine values were obtained on a pooled 24-hr untie sample from five rats per treatment. 921001 rtowVerK M2 R. L. Andcsson Table 1. Eflei r ofJinan :mi lonienirunon on renal tubular t ell damage in male rats feJ semi-purified diets t untaimng I T. Na,NTA .H)0 for 4 wk Toxicity index* for rais given Tubular cell damage 21 ppm Zn 52 ppm Zn V acuoiizalion and simple hyperplasia Vacuolization and nodular hyperplasia 15 (0)t 3(0) 25 (01 50(0) Toxicuy index - incidence (number of kidney sections showing lesion/treatmem group) * frequency (mean number of lubules showing lesion/kidney section). Ten kidneys were examined in each NTA-irealed and control group. lA/alues in brackets are for control rats given diets that did not contain Na'jNTA.H.O. vacuoles and simple hyperplasia were noted. Thus low dietary levels of Zn resulted in minimal PCT tox icity. The zero incidence for the two types of lesions in the control groups confirms that the lesions noted are specific responses to NTA. The PCT toxicity index in the rats fed the diet containing 52 ppm Zn was greater for both stages of toxicity development than in the rats fed the diet containing 21 ppm Zn. The effect of dietary Zn on the PCT toxicity was particularly evi dent with respect to the more severe lesion (PCTs with vacuoles and nodular hyperplasia); for which the toxicity index in the group ingesting the diet with 2t ppm Zn was only three compared to a value of SO for the group given 52 ppm Zn. These feeding studies demonstrate that the NTAspecific toxicity in the renal PCT. assessed after com parable levels of NTA have been excreted in the urine, is proportional to the amount of Zn but independent of the amount of OTA that has been excreted in the urine. This suggests that Zn availability in the circu lation is a limiting factor in NTA-tnduced PCT tox icity. This hypothesis was further tested by ascertain ing the effect of Zn administered by iv injection (ZnSO. at 03 mmol/kg given 30 min after NTA) on the response of PCTs to an acute oral dose of OTA (7-3 mmol/kg of NajOTA.HjO titrated to pH 8 5 with HO). The number of PCTs containing vacuoles was markedly increased in the animals given iv Zn after the OTA dose compared to those only given the oral NTA (Plate ia.bk No PCT vacuoles were noted in kidneys from rats given iv Zn but no NTA. Thus both the acute (PCT vacuoles) and the subchronic nephro toxicity (PCTs with vacuoles and hyperplasia) associ ated with a constant NTA dose are proportional to available Zn in the circulation. The fact that renal clearance of NTA is tin filtra tion with no evidence of a tubular cell transport sys tem for OTA (Licht er at. 1980k coupled with the established relationship of PCT toxicity to Zn avail ability, prompted a detailed examination of the effects of NTA on Zn disposition and excretion. The effects of a nephrotoxic (73 /msol/g dietk a non-nephrotoxic (0-73 umol/g diet) and an intermediate (7-3 pmol/g diet) dose of ['*C]NTA on Zn metabolism were com pared in rats equilibrated with [l4C]NTA b> 10 days of ad lib. ingestion. [J*C]NTA and Zn levels were measured in the plasma UF and a 24-hr voided urine sample to determine if renal clearance of NTA resulted in a change in the NTA:Zn ratio in the urine relative to that in its precursor, plasma UF. A lower OTA:Zn ratio in the urine relative to that in the UF would show an OTA-induced Zn secretion, whereas a higher ratio in the urine than in the UF would show specific Zn resr-ption. For reference, in control rats the majority of the filtered Zn is resorbed by tubular tissue (Yunice. King. Kraikitpanitch. Haygood & Un deman, 1978k The two lower doses of NTA resulted in similar UF*. levels but the highest NTA dose pro duced a five-fold increase in UF*. levels (Table 2). At all of the OTA doses, the OTA:Zn ratio in the urine (product) was increased over the ratio in the UF (pre cursor) showing net renal resorption of the Zn. A con stant proportion of the filtered Zn was resorbed since in each instance the OTA:Zn ratio in the unne was 2-3 times that in the UF. The mass of Zn resorbed (estimated from the UF OTA :Zn ratio and the unne total OTA and Zn levels) was only 0-4 or 1-7 pmol day following ingestion of 073 or 7-3 *xmol NTA/g diet, respectively, but there was a twelve-fold increase in Zn resorption (11-9 pmol/day) at the nephrotoxic dose of NTA (73 fimol/g dietk This suggests that PCT toxicity is likely to be a consequence of tubular cell Zn resorption. To determine the dose of OTA required to increase UFz*. rats trained to consume their daily intake of diet in a 1-hr period were used to measure UFNTA and UF*. levels at the time of maximum blood-NTA levels (4 hr after the mealk Figure 3 shows that NTA doses that resulted in UFNTa < 21 iM did not alter the UF*. level but that the higher UFNTA level (52 jim) was accompanied by an increased UF*. level. Thus ingestion of high doses of OTA induces an increase in plasma UF*. concentration that is largely resorbed during passage through the kidney. Further, the plasma UF*tA level must exceed a threshold of about 20pM in order to induce an elevation in UF*. levels. Table 1 Effect of dietary NTA concentration on plasmi ultrafiltrate (lIF) and urinary concentrations of NT A anj zinc Concentration (nequiv ml) of Dietary NTA Ratio Uunol/g diet) NTA Zn NTA Zn 0-73 7-3 73 073 7-3 73 Pinson UF 1-41 OI0 079 + 045 119 06 077 018 119 11 3 8 070 Urine 030 004 006 + 001 2-77 039 007 + 001 27 2 >1 025 + 005 15 31 5 40 109 Each value for UF and urinary Zn and NTA is (he mean SEM for groups of 12 rats id males and 6 fcmateak Three rats of each sex were fed the indi cated dietary concentration of either Hi[>4C]NTA or Naj[`*C)NTA.HiO ad lib. for 10 days. The urine values are based on a total 24-hr urine collection and the UF samples were obtained between 8 and 10 a.m. \321001 RowVerK 096i> Zn and NTA renal toxicity 643 Fig. 3. Effect of plasma ultrafiltrate NTA levels (UFmt*) on plasma ultrafiltrate Zn levels l(JF2.) in rats ingesting their NTA dose in a t-hr period. Blood was obtained 4 hr a/ter the tenth_ NTA meal. Each value is the mean + SEM for three samples. In the experiments described, the total-plasma and piasma-UF concentrations of Ca and Mg were also measured. NTA ingestion did not alter total-plasma Ca or UFCa levels in any study but it did reduce total-plasma Mg and UFm, levels at the time of maxi mum levels of blood NTA (4 hr after the meal). The reduced Mg levels probably reflect changes in Mg absorption since NTA ingestion increases faecal Mg (Anderson & Kanerva. 1978a). To examine further the role of Zn in the PCT tox icity associated with renal clearance of NTA, the effect of iv infusions of the forms of NTA most likely to be present in the plasma were compared. The iv dose chosen (6 mmol/kg/24 hr, except for ZnNaNTA) was equivalent to twice the systemic load of NTA in animals consuming diets containing 73 pmol/g. Five forms of NTA were examined--the Zn, Ca and Mg complexes and the Na and K salts. The infusions with the two salts and the Ca and Mg complexes of NTA at 6 mmol/kg/24 hr for 3-7 days all induced some PCT vacuoles (Plate II). These infusion studies resulted in no more PCT toxicity than had been noted after 3 days of ingestion of a diet delivering approximately one-half the systemic dose of NTA (Plate III). In contrast, infusion of the ZnNaNTA at its limiting solubility of 3 mmol/Vg/day was lethal to every animal in less than 48 hr. The animals displayed a very consistent response during ZnNaNTA in fusion: for the 0-24-hr period the animals all voided urines containing glucose, protein and haemoglobin. At about 24 hr the animals ceased to urinate and soon became comatose. Kidneys removed after 48 hr of ZnNaNTA infusion at 3 mmol/kg/24 hr had a spec kled surface and when sectioned showed a distinct white precipitate at the cortico-medullary junction. Histologically, the kidneys displayed massive renal coagulative necrosis and tubular celt sloughing. Elec tron-microscope evaluation of the material accumu lated at the cortico-medullary junction showed sloughed tubular cells and intra- and intercellular crystalline material. The recovery of [l4C]NTA in the urine following iv infusion of various "C-labelled forms of NTA showed that with all NTA species examined, includ ing ZnNaNTA. more than 90*; of the infused [l4C]NTA was recovered in the urine. The high and constant urinary recovery of NTA. regardless of NTA species infused, indicated that the nephrotoxicity induced by ZnNaNTA was a consequence of the Zn and not of increased retention of NTA. This conten tion was further investigated by comparing the renal toxicity of Zn salts and Zn ethyienediaminetetraacetic acid (ZnEDTA), a more stable complex than ZnNaNTA. When infused at 3 mmol/kg/day, the Zn salts produced the same in eico response as ZnNaNTA but in a shorter time period, and also caused PCT necrosis. In contrast, ZnEDTA infused at 3 mmol/kg/day for 3 days produced only minima) renal damage (a few PCT vacuoles). Comparison of urinary recovery of the Zn when given as an infusion of ZnNaNTA, ZnEDTA. Zn acetate or Zn gluconate (Fig. 4) showed that the renal toxicity induced was inversely proportional to the urinary recovery of the infused Zn and, therefore, proportional to the carcass retention of the infused Zn. The renal toxicity induced by iv Zn (necrosis and cell sloughing) was different from that associated with NTA ingestion (vacuoles and hyperplasia). However, when the iv dose of ZnNaNTA was reduced to c. 1 pmol/rat/day, the PCT response was similar to that induced by dietary NTA (Plate IV). In order to contrast the acute effects of ZnNaNTA infused iv with those of dietary NTA, the totalplasma, plasma UF and urinary [l4C]NTA:Zn ratios were determined after 72 hr of infusion with ZnNa[,4C]NTA. These parameters showed that at least after 48 hr of infusion the Zn and NTA were distributed differently in the blood pools than in the infusion solution (Table 3). The blood cells and plasma proteins contained mote Zn than NTA. In contrast the plasma UF contained far more [t4C]NTA than Zn. This was accompanied by a uri nary NTA:Zn ratio that was lower than the NTA:Zn ratio in the UF--a result opposite from that noted in rats fed NTA (Table 2). Thus ZnNaNTA infused iv dissociates in the circulation resulting in a marked elevation in plasma-protein and cellular Zn and Zn secretion into the urine during NTA clearance. This implies that the difference in renal lesions between ZnNaNTA infused iv and dietary NTA could be due to either or both of the following: (I) a difference in the route of Zn entry into the tubular cell--from Fig 4. Comparison of urinary recovery of Zn during iv infusion of ZnEDTA (). ZnNaNTA (Bl Zn acetate ill or Zn gluconate () at levels to give 7-3. 7-6. 7-7 or 7 4 mg Zn/24 hr. respectively. $21001 RowVerK o< 644 R. L. Anderson Table 3 Blood component and urine STA and Zn levels after 4H hr of iv infusion of ZnSaSTA Cone (|iM) Fluid or tissue NTA Zn Infusion fluid Blood cells Plasma Plasma ultrafiltrate Urine Infusion fluid Blood cells Plasma Plasma ultrafiltrate Urine Rats given 42 imol ZnNaNTA/hr 7100 7000 3 2 0-5 292 + 25 26 3 159 4 19 2 1-3 0-7 2600 300 1000 200 Rats given 124 jimol ZnNaNTA/hr 21400 21000 20 1 1159 + 103 85 7 407 20 56 5 7-4 20 10200 600 6400 + 400 Ratio NTA in IOI 0-011 0164 14 6 2 5 + 05 102 0017 0209 7-6 16 Ot . Values are means + SEM for groups of three male rats (c. 230 g) infused iv with ihe indicated dose of ZnNa[l*C]NTA for 48 hr. The infusion volume was c. 10 jtl/min the circulation into tubular cells with infused ZnNaNTA and from the tubular lumen into PCT cells with dietary NTA; (2) differences in the Zn con centration obtained in the circulation--dietary NTA does not appreciably increase Zn levels in the total plasma but does shift the distribution from the plas ma-protein pool (MW > 10,000) to the UF while iv ZnNaNTA causes a marked increase in Zn in the total-plasma pool, but most of the Zn is proteinbound. The fact that infusion of Zn salts produces a response similar to ZnNaNTA suggests that the renal tubular necrosis is associated with the increase in plasma-protein Zn levels since very little of the infused Zn from the salts enters the urine. Further, infusion of Zn gluconate results in a dose-dependent increase in renal-tissue Zn which is readily removed by short-term infusion of saline after Zn gluconate infusion (Fig. 5). The results with iv-infused Zn (both as the NTA complex and as salts) are indicative of acute heavymetal toxicity (Hammond & Belitcs, 1980). This simi larity was confirmed by showing that HgCl2 infused iv (21 jimol/rat for 4 hr) resulted in the same series of in ciro responses with iv Zn. and histological evalu ation of the kidneys from these animals showed marked tubular cell necrosis (Plate V). I0r 1 O 4 i o-i 3 5 MOOr PM ? 5 OOlI i 23 Zn tfrtuaon,dap 0 2 i Stfra inMon, a3 Zn gkjoonot*nfUMd, It mol /kg Fig. 3. Renal tissue Zn concentration as a function of the amount of Zn gluconate infused iv: with only Zn gluconate infusion !): with Zn gluconate infusion followed by a 4-hr saline infusion (SI Values are means + SEM for groups of ihree rats. Fig. 6. la) Urinary Zn excretion rate during 3 days of iv infusion !------1 of Zn gluconate (AL ZnNaNTA <) or ZnEDTA () at c. SS0 |imol/kg.'day and for two sub sequent 2-hr periods of saline infusion (-----1 Ibl Levels of Zn in kidney after 4 hr of saline infusion following Zn glu conate 1(31 ZnEDTA (Ml or ZnNaNTA () infusion. Renal zinc values arc means SEM for groups of two rats. Rats infused with ZnEDTA. ZnNaNTA and Zn gluconate excreted, respectively. 72. 17 and <1*. of the infused zinc in their urine. ` ' RowVerk OSlhV i the NTA ute heavy* This sixniJlj infused ne series of icaJ evaluIs showed $ 1 * f * f i < \ it) N ts days of iv TA 1*1 or two nib Levels of ig Zn gin* ion. Rcnai rats. Rats gluconate `used one late (. Photomicrographs of kidneys from rats 6 hr after a gavage of 7-3 mmol NTA/kg body weight: I kidney of a rat given iv 03 mmol ZnSO^kg body weight 30 mm after the NTA gavage: (bl kidney of rat given iv saiine after oral NTA. Hacmatotylin and cosin * 130. M 21001 RowVft! V (Ai l Plate II. Photomicrograph of the kidney of a rat infused for 72 hr with 6 mmol NTA. day. The photo micrograph is from an animal infused with MgNaNTA but similar effects were induced in the kidneys of animals infused with Na3NTA.H20, KjNTA.HjO or CaNaNTA at the same exposure rate. Haematoxylm and cosin x 70. Plate III. Kidney section from a rat fed a diet containing 2*. Na,NTA.H:0 for 3 days. Note the focus of tubules with vacuoles and hyperplasia. Haematoxylm and eosin x 35Q, _ 921001 RowVerV .()Hbi V7 64* wm*;,.* vH> . t r t p l ***> 4 Plate IV. Photomicrograph of the kidney from a rat infused with 2-25 jimol ZnNaNTA/day for 2 days. Note proximal convoluted tubule with vacuoles. Haematoxylin and eosin x 130. 4 Plate V Photomicrograph of a kidney section of a rat infused iv with 125 nmol HgCTjAg.24hr for 4 hr. i Haematoxylin and eosin x 20. r 921001 RowVtr. ()'' ii I 647 i I Zn and NTA renal loxicily Table 4 Kidney Zn and NTA concentrations after ZnNaNTA infusions Results following ZnNaNTA infusion at (jimol/kg/dayt Parameter 400 1200 Kidney excess Zn*(mmol/g) Kidney NTA (mmol/g) Kidney Zn.NTA Urine Zn:NTA 240 + 21 174 + 22 1-43 026 033 + 004 1020 65 588 110 1 81 + 022 0-62 + 004 Kidney Zn in rats infused with ZnNaNTA minus kidney Zn in rats infused with saline. Each value is the mean + SEM for samples from three raw infused with the indicated level of ZnNaNTA for 3 days. 649 To further differentiate between the renal effects of ZnNaNTA. Zn gluconate and ZnEDTA. the renal Zn levels were determined after 3 days of iv infusion of each of-these forms of Zn at c. 550 jimol/kg/day. To minimize tissue contamination with entrapped urine Zn. the animals were infused with saline for 4 hr before they were killed. Figure 6 shows that the Zn level in the kidneys from the rats infused with Zn gluconate and ZnEDTA were comparable even though the former excreted < 1% of the infused Zn in their urine and the latter excreted > 70%. In contrast, the kidney Zn level after infusion with ZnNaNTA was about 100% greater than that after infusion with Zn gluconate or ZnEDTA. This increased kidney Zn level could not be attributed to a higher urinary Zn concentration at the time of death compared with that of rats infused with ZnEDTA. Thus NTA results in a specific renal tissue Zn accumulation that is re sistant to washout and that is. not simply the result of either increased systemic Zn (as brought about by Zn gluconate) or increased urinary Zn excretion (as caused by ZnEDTA). To determine if the NTAassoctated renal accumulation of Zn was due to ZnNaNTA accumulation, the excess kidney Zn (i.e. levels in ZnNaNTA-infused kidneys - levels in salineinfused kidneys) and [,4C]NTA concentrations were ascertained after 48 hr of iv infusion of ZnNa[,4C]NTA at 400 or 1200 jimol/kg/day. At both ZnNaNTA infusion rates there was a greater excess kidney Zn than [,4C]NTA (Table 4). Further, since the urine contained more NTA than Zn the increased tissue Zn cannot be attributed entirely to urine con tamination. To place the tissue NTA levels in perspec tive the entrappment of 59 6 jil urine;g tissue could account for ail of the kidney NTA at both doses but for only 70 and 400 nmol Zn/g tissue at the low and high doses, respectively Thus not all of the renal tissue Zn that accumulates during NTA clearance ts derived from ZnNaNTA. This becomes even more significant when it is recalled that in feeding studies it was shown that renal clearance of NTA was accompanied by a net Zn resorption. Discwkw The results of the experiments reported show that the extent of damage induced in the renal PCT during clearance of constant amounts of NTA is directly pro portional to the- availability of Zn in the circulation. Both the induction of vacuoles in PCT cells and the subsequent development of hyperplasia in these cells, at a constant level of NTA excretion, are proportional to the availability of Zn in the circulation. In ad dition, the results show that Zn infused iv is a renal toxin in the absence of NTA when it is retained in the carcass (Zn salts) but its toxicity is markedly reduced when it is excreted in the urine as a complex (ZnEDTA). In this sense Zn nephrotoxicity is similar to mercury nephrotoxicity--mercury salts are potent nephrotoxins but mercurial diuretics are not (for dis cussion see Zbinden. 1971). Examination of the effects of NTA on Zn dispo sition shows that when the UFkt* attains sufficient concentration (> 20 jim) it increases the concentration of UF&,. During renal processing of the UF fraction a large portion of the Zn, but little, if any, of the NTA is resorbed. The increase in Zn resorption is a necessary condition for the initiation of PCT toxicity, and since the initial PCT toxicity is clearly a necessary but not sufficient condition for eventual renal tumorigenesis (Alden et al. 1980), doses of NTA that do not initiate the response will not increase the renal tumour inci dence relative to that of controls. The demonstration of a distinct threshold for an NTA effect on PCT toxicity certainly negates the validity of any math ematical model for extrapolation of possible toxicity including tumorigenicity at a dose below this threshold from results noted at d<Bes that clearly exceed this threshold. The results noted in chronic ingestion studies are in agreement with this model in that only doses of NTA that increase urinary Zn have resulted in any renal tubular ceil lesions and only extreme doses have been associated with renal tubular cell neoplasia (National Cancer Institute. 1977; Nixon et ai. 1972). In contrast, chronic exposure to NTA doses that do not alter dietary Zn disposition (z. 003% NajNTA.HiO) do not induce any renal toxicity even after 24 months of ingestion (National Cancer Institute 1977; Nixon 11 at. 1972k Finally, a calculation of the average exposure of man to NTA from drinking-water, based on measured values in Canada, where NTA has been used in deter gents since 1970 (Maiaiyandt. Williams & O'Grady. 1979k shows an average exposure of Ot jig'kg/day (35ml HjO/kg/day * 2-82pg NTA/10Jml H,0) If all of this daily ingested dose of NTA were in the plasma UF pool at one time it would amount to only 036 nequiv of NTA/kg or 0009 \da (assuming 40 ml UF/kgk more than three orders of magnitude less than the plasma UFvr* threshold of >20>im which is k7 SAlOWi t\0\|e 0H71 650 R. L Anderson necessary to increase the UFt,, level and initiate PCT toxicity. In reality, man shows low absorption of ingested NTA |c. 127;) and rapid urinary clearance (Budny A Arnold. 1973) so that the ingestion of 01 kg, day would result in UFNTA several orders of magnitude less than the UFNTA threshold established in rats. REFERENCES Alden. C L. <t Kanerva. R. L. (1980). Pathogenesis of renal corneal tumors in rau fed 27; trisodium mirilotnacetale monohydrate (NajNTA.H]Ol Presented at the Thirtyfirst Annual Meeting of the American College of Veterin ary Pathologists. New Orleans. LA. 2 December 1980. Aldeq. C. L,, Kanerva. R. L_ Anderson. R. L. A Adkins. A. G. (I98U Short-term effects of dietary nitrilotriacetic acid in the male rat kidney. Vet. Pathol. In press. Anderson. R. L. & Kanerva. R. L. (1978a). Effect of nilrilotriacetate (NTAl on cation balance in the rat. FJ Cosmet. Toxicol. 16. 563. Anderson. R. L. <1 Kanerva. R. L. (1978b). Hypcrcaldnuria and crystalluria during ingestion of dietary nilnlotnacetate. FJ Cosmet. Toxicol. 16, 569. Anderson. R. L. A Kanerva. R. L. (1979). Comparisons of response of Fischer-344 and Charles River rats to 15*. nitrilotriacetic acid and 27; trisodium nitrilotriacetate. monohydrate. FJ Cosmet. Toxicol. 17, 137. Budncy. J. A. A Arnold. J. D. (1973k Nitrilotriacetate (NTA): human metabolism and its importance in the total safety evaluation program. Toxic, appl. Pkarmac. 25. 48. Foley. P. D. Becking G- Muller. J_ Coyer. R. A_ Falk. H. L A Chernoff. N. (1977). Report to the Great Lakes Research Advisory Board of the International Joint Commission on the Health Implications of NTA. Hammond. P. B. A Belitcs. R. P. (1980k Metals. In Casarett and DoulFs Toxicalogr. The Basic Science of Poisons. 2nd Ed. Edited by J. Doull. C. D Klasscn A M. O. Amdur. p. 422. MacMillan Publishing Co.. Inc.. New York. NY. Mahaffey. K. R. A Goyer, R. A. (1972k Trisodium nitnlotriacetate in drinking water. Metabolic and renal effects in rau. Archs enctr. Hlth. 25. 271. Malaiyandi. M,, Williams. D. T. and O'Grady. R. (1979). A national survey of nitrilotriacetic add in Canadian drinking water. Enrir. Sci. Teekhol. 13. 59. Mcrski. J. A. (1980k Acute structural changes in renal tubular epithelium following administration of nitrilo triacetate. FJ Cosmet. Toxicol. 19. 463. Michael W. R. A Wakim. J. M. (1971 k Metabolism of nitrilotriacetic add (NTAk Toxic, appl. Pharmac. 18.407. MichaeL W. R. A Wakim. J. M. (1973k Effect of trisodium nitrilotriacetate (NajNTA) on the metabolism of selected metal ions. Toxic, appl. Pharmac. 24. 519. National Academy of Soences/Nauonal Research Coundl (1972k Nutrient Requirementsfor Laboratory Animals. 111. 2nd Ed. NAS/NRC. Washington. DC. National Cancer Institute (1977k Bioassays of Nitrilotriacetic add (NTA) and Nitrilotriacetic add. Trisodium Salt. Monohydrate (NajNTA.HjO) for Possible Carci- nogenidty (NCT-CG-TR-6k DHEW Publication No. (NIH) 77-806. Nixon. G. A, Buchler. E. V. A Niewenhuis. R. J. (1972k Two-year rat feeding study with trisodium nitriloiriacetate and iu calcium chelate. Toxic, appl. Phamuc. 21. 244. Shapiro. J- Chapman. P. J. Dick. R. L, Dillon. P. J. O'Mdia. C. R. Spade. A. A Leduc. G. (1978k Ecological Effccu of Non-phosphate Builders: Final Report on NTA. Report to the Great Lakes Research Advisory Board of the International Joint Commission. Thayer. P. S. A Kensler. C. J. (1973k Current status of the environmental and human safety aspecu of nitrilotriacctic add (NTAk CRC Crit. Ret. enrir. Control 3. 375. Upton. R. A. (1975k Simple and reliable method for serial sampling of blood from rats. J. pharm. Sci. 64. 112. Yunice. A. A_ King R. W. Jr. Kratkitpamtch. S_ Haygood. C. C. A Lindetnan. R. D. (197S) Urinary zinc excretion following infusions of zinc sulfate, cysteine, histidine, or glycine. Am. J. Physiol. 235, F40. Zbinden. G. (1971k Experimental Renal Toxidty. In The Kidney Morphology. Biochemistry, Physiology. Vol. 11. Edited by C. Rouiller A A. F. Muller. Academic Press. New York. NY. ^2U)Ul tfOVKVefcv 0V.`