Serum Cholesterol Levels in Rats Fed Thirteen Trace Elements1 HENRY A. SCHROEDER* Department of Physiology, Dartmouth Medical School, Hanover, New Hampshire, and Brattleboro Memorial Hospital, Brattleboro, Vermont ABSTRACT To ascertain differences in serum cholesterol levels as possibly affected by trace elements in the drinking water, rats were given soluble salts of zirconium, vanadium, niobium, chromium, nickel, cadmium, germanium, tin, lead, arsenic, and antimony at 5 ppm and selenium and tellurium at 2 ppm from the time of weaning until 11 to 30 months of age. The feeding of chromium at 1 ppm was associated with suppressed levels in males but not in females; 5 ppm appeared to be required for this effect in females. The lowest serum cholesterol levels were observed in groups given niobium, chromium and nickel, the highest in groups fed tellurium, with all animals receiving 1 ppm chromium. Significant differences in values of the 2 sexes appeared for 7 elements. It is possible that chromium, nickel and niobium exert anti-cholestero- Downloaded from jn.nutrition.org by guest on May 6, 2011 genic properties, whereas tellurium may be cholesterogenic. Mean circulating cholesterol levels in tunity to measure serum cholesterol and human beings tend to rise with age. This glucose levels. The present report con tendency is exaggerated in persons living in cerns 13 trace elements fed to rats from industrialized countries and is less evi the time of weaning. Differences in serum dent in more primitive societies. The pos cholesterol associated with element and sibility arose that some abnormal trace age are shown. element to which civilized man is exposed and which accumulates in his tissues with METHODS age might influence the synthesis of cho The environmental conditions in the lab lesterol or interfere with its catabolism. oratory, the precautions taken to avoid Conversely, some modern nutritional prac metallic contamination, the low metal diet tices might promote marginal deficiency of and the double-deionized drinking water an essential trace element on which ho- used have been reported in detail (6, 7). meostasis of cholesterol depends. The trace element content of the diet, Curran ( 1) showed that rat liver, in vi which was composed of seed rye, powdered tro and in vivo, synthesized cholesterol skim milk and corn oil, and the estimated less well in the presence of vanadyl ions total intakes, are shown in table 1. To the and better in the presence of trivalent chro basic drinking water, which contained zinc mium and manganous ions than liver un- (50 ppm), manganese (10 ppm), copper exposed to trace elements. Other trace (5 ppm), cobalt (1 ppm), molybdenum metals in the first transitional series were (1 ppm) and chromium (1 ppm), was relatively inert. Curran's rats were fed a added one of the following at 5 ppm ele commercial ration, which at that time was ment; zirconium sulfate, sodium niobate, marginally deficient in chromium (2). The antimony potassium tartrate, sodium ger- action of chromium feeding in stabilizing manate, stannous chloride, sodium arsen- serum cholesterol at a low level in male ite, vanadyl sulfate, nickelous acetate, rats has been demonstrated (3, 4); asso cadmium acetate, lead acetate. Sodium ciated with this effect was virtual abolition of spontaneous aortic plaques and accumu Received for publication November 1, 1967. 1Supported by Public Health Service Research Giant lation of aortic lipids (4). no. HE-05076 from the National Heart Institute; Con Studies on the Ufe-term effects of a num tract DA 2595 from the U.S. Army; CIBA Pharma ceutical Products, Inc.; and the Selenium-Tellurium ber of trace elements given in low doses to Development Association. 2Present address : 9 Beunont Avenue, Brattleboro, rats and mice (5) have provided an oppor Vermont 05301. 94: J. NOTBITION, '68. 475 476 HENRY A. SCHROEDER TABLE 1 Berkeley Medical Instrument spectropho- Trace element content of diet and drinking water tometer;5 a few were measured by the and estimated daily intakes 1J! method of AbeU et al. (11). Duplicate and replicate analyses agreed within 3%. Analyses of the diet for trace elements wt/day50.9654.4444.720.967.9635.9637.6435.7036.9236.6836.2037.7635.014.3014.96 body were made by a number of standard meth ZirconiumVanadiumNiobiumChromiumChromiumChromiumNickelCadmiumGermaniumTinLeadArsenicAntimonySeleniumTelluriumDietV-9/92.663.24 ods, colorimetrie for zirconium, vanadium, niobium, chromium, nickel, germanium, arsenic, tin, lead and selenium, and atomic absorption spectrophotometric for chro mium, cadmium, antimony and tellurium. RESULTS Serum cholesterol levels in rats given various amounts of chromium are shown in table 2. Sex differences appeared. Males fed 1.0 to 5.0 ppm in drinking water had 'Based on 6 g food and 7 g water Ingested/100 g lower mean values than those given none Downloaded from jn.nutrition.org by guest on May 6, 2011 body weight/day. in water. Females given 5.0 ppm in water 2These intakes are approximate but are comparable Gams in weight of young rats were similar, weights or about 2 ppm in food and water had of adult animals remained fairly constant, and varia lower values than those given 1.0 ppm or tions among groups were not observed. The intake of water was measured on control rats for one year- less. The difference in male and female marked variations in other groups were not evident 3Not detected. values at 1.0 ppm was significant (P < 0.001). According to these data, female selenate and sodium tellurite were added rats apparently require more chromium at 2 ppm element; sodium selenite at than males in order that cholesterol in 3 ppm.3 In addition, the basic water was serum be suppressed. Fourth generation given without chromium or with 5 ppm as chromium-deficient young rats of both the acetate. sexes had higher values than second gen Random-bred pregnant female rats were eration young rats which were assumed to purchased from a supplier4 and their off be less deficient. spring weaned at 21 to 23 days. Groups In table 3 are the mean serum choles of 52 or more of each sex, four to a cage, terol levels of rats fed 13 trace elements in were given one of the trace elements until drinking water, arranged according to the natural death occurred. Each group was values of males. Again sex differences ap placed in separate wooden racks, and cages peared in 7 cases. Male values were higher were not interchanged between groups. in the groups given vanadium, chromium, In addition, 4 generations of rats were selenium (P < 0.025), and in the controls bred without exposure to chromium other without chromium (P < 0.05). Female than the small amount in the diet, as re values were higher in those given cad ported (8). The second and fourth genera mium (P < 0.05), germanium and tin (P tions were used in these experiments. < 0.001). More or less consistently low Thirty brood females obtained from the values were noted in the groups of both supplier, which had been fed a diet con taining adequate chromium, were also ' In a previous publication on the effects of selenite studied. Their offspring were not believed and selenate on mice and rats (14), the concentration of selenium in the selenite given in water was errone to be deficient in chromium according to ously reported as 2 ppm. The supplier of sodium selenite (Nutritional Biochemicals Corporation, Cleve analyses of tissues of similarly bred ani land) stated in a letter that this product was mals (9). Na2SeO3'5H2O. Nearly 2 years later the compound was found by D. V. Frost, in our laboratory, to be Blood was obtained from the tail of the anhydrous, and was then so admitted to be by the warmed rat by cutting with a razor blade, supplier. As the hydrated form contains 45.7% selenium and the hydrated form 30%, the rats and and centrifuged. Serum cholesterol was mice given selenite actually received 3 ppm selenium. 4 Blue Spruce Farms, Altamont, New York. measured by the method of Huang et al. 5Berkeley Medical Instruments. New England X-Ray (10), using a premixed reagent and a and Electronic Equipment, Brookline, Massachusetts 02146. CHOLESTEROL IN RATS FED TRACE ELEMENTS 477 TABLE 2 Serum cholesterol levels in rats given chromium >Agedays204127761660689250Â«405709msr/100 ml78.5 Young *F2FÂ«MatureChromiumppm00012Â±55Males'No.11812231212Agedays204115761662405510mg/100ml86.0 Â±3.9Â»114.0 Â± 2.5109.6 Â±5.0Â«122.9 494.5 Â± 2.7 Â±8.277.5 Â±11.2116.0Â± Â± 2.1s86.2 6.0108.6 Â± 4.283.3 772.1 Â± 1.7 Â±3.2Â»77.0 Â± 5.3'63.0 Â±6.6sNo.12101219730128Females Â± 4.2Â« l Male and female values differ significantly: 2.0 ppm group, P < 0.025; 1.0 ppm group, P < 0.001; 5.0 ppm groups, P < 0.025. * Fa and Fi are rats bred from increasingly deficient mothers. 3 SE. Â«Differs from Fs group, P < 0.001. 5Differs from mature group, no chromium, P < 0.001. Downloaded from jn.nutrition.org by guest on May 6, 2011 *Breeders purchased while pregnant, food containing chromium. * Differs from mature groups, 1.0 ppm 660 and 689 days old, P < 0.001. 8 Differs from 2.0 ppm group, P < 0.01. D Â» iffers from 2.0 ppm group, P < 0.001. TABLE 3 Serum cholesterol levels in mature rats given various trace elements MalesGermaniumNickelNiobiumControls 1FemalesNo.mg/100 value 'mg/100 ml63.4 ml107.6Â±75.3 Â±75.3 0.001nsusâ€”ns< Â±75.7 Â±78.6 Â±77.5 Â±116.0Â±105.3 (Cr Â±82.5 1)TinChromiumLeadCadmiumZirconiumArsenicVanadiumAntimonySelenium Â±86.2 Â±72. Â±86.6 0.02ns< 1Â±103.7Â±113.0Â±100.7 Â±88.6 Â±89.7 0.05<0.01< Â±91.0Â±91.6Â±97.6 Â±109.5 0.05< Â±67.9 0.005< Â±97.0 Â±109.6Â±1 0.001< Â±90.4 (VI)Selenium 0.001< Â±90.2 (IV)TelluriumControls Â±110.4Â±94.5 Â±122.9 11.4 0.001< 0.00112121119121212121213101012121212Agedays57832984966058040576075 Â±3.2*220.127.116.11.18.104.22.168Â«5.610.15.14.96.13.08.17Â«< Â±22.214.171.124.03.55.38.5Â«9.0Â«9.07.49.25.65.03.59.911.1 (no Cr)No.121212231212141012121110121212Agedays557342849662520405695750893804761790445405761P 1Significance of differences from controls. Â«SE. *Not significant. Â«Methodof Abell et al. (11), previously reported (4),for comparison. sexes given niobium, chromium, and Female control levels were the highest nickel, and high values in those fed tel of any group. Other elevated values (> lurium. 100 mg/100 ml) occurred in the groups Compared with the controls given 1.0 given zirconium, cadmium, germanium, ppm chromium, values for males were es tin, lead, arsenic and tellurium. Low val pecially elevated in groups fed tellurium, ues (< 80 mg/100 ml) were found in the selenium, antimony and no chromium and groups given vanadium, niobium, chro slightly elevated in the groups fed zirco mium and nickel. nium and vanadium. Cholesterol levels in Because two of these elements, chro males given germanium were suppressed. mium and vanadium, have been shown to 478 HENRY A. SCHROEDER affect cholesterol metabolism in vitro and values were higher in rats given all of the in vivo ( 1, 4 ), comparisons were made as elements except niobium and nickel. to significant differences in circulating cho Comparisons of levels of male rats fed lesterol between rats fed these metals and vanadium with those given other elements other elements (table 4). In relation to the showed significantly elevated values only chromium-fed group, levels were higher in in rats fed selenium and tellurium. Signifi male rats given antimony, selenium and cantly depressed values were observed in tellurium and lower in those fed niobium, those given niobium, nickel and germa nickel and germanium. In females, mean nium. In females, all values, except those TABLE 4 Significance of differences in serum cholesterol levels of rats given various trace elements compared with those given chromium and vanadium Males Females P Penee P P z+ enee Â¡ value 2+ i value enee 1 *+ value enee ' 2+ value CrControls, Controls, no 0.001- < 0.001- < 0.05+ < 0.05+ < Downloaded from jn.nutrition.org by guest on May 6, 2011 1 ppm < 0.02nsÂ»ns- < 0.005ns- 0.001+< CrZirconiumVanadiumNiobiumNickelCadmiumGermaniumTinLeadArsenicAntimonySeleniumTelluriumChromiumDiffer- 0.001+< 0.01nsnsns+ < < 0.01nsns+ 0.01- < 0.005- < 0.01ns-< < 0.005ns- 0.001+ < 0.001+ < 0.001â€” < < 0.001nsnsnsns+ 0.001+< 0.001+< nsnsns+ 0.001+< 0.001+ < 0.001+ < 0.005+ < 0.001+< 0.001+< 0.02+ < 0.01+ < 0.02+ < 0.001+ < 0.001+ < 0.01+ < 0.02+ < < < 0.005VanadiumDiffer- 0.005ChromiumDiffer- < 0.001VanadiumDiffer- 0.005 < 1 Plus sign indicates that the value is larger; minus sign, smaller than that for chromium or vanadium. 2 Significance of difference in the value from that of chromium or vanadium (see table 2). 3 Not significant. TABLE 5 Changes with age in serum cholesterol in rats fed zirconium, niobium, antimony, lead and cadmium ElementZirconiumNiobiumAntimonyLeadCadmiumArsenic '< value value 1< Â± 3.8 Â»89.7Â± Â±1.8100.7Â±9.081.1Â±3.778.6 5.969.3 0.005ns'ns< 0.01nsnsnsnsns Â± 3.470.3 Â± 2.8975.7 Â± 2.776.7 Â±4.886.3 Â± 2.497.6 Â±4.492.6 4.972.3 Â± 0.001nsnsns< Â±5.6103.4 Â± 4.7479.3 Â± 6.4886.6 Â±7.12103.7 Â± 6.774.0 Â±10.0376.1 Â±8.5397.0 Â± 4.7485.6Â± 2.367.6 Â±8.986.5 0.05nsnsnsFemalesNo.1212121112101212101212Agedays2708932288491617905107504 Â± 2.8188.6 Â±9.80113.0 Â± 8.7091.6Â±10.591.0Â±10.1P Â±8.98P 4MalesNo.1212121012121011121412101012121212Agedays270893166655849125790300510695750300485510750770804mg/100ml69.0 i Significance of difference from first value in each group. * SE. N Â» ot significant. * For comparison of repetitive analyses on different rats of same group. CHOLESTEROL IN RATS FED TRACE ELEMENTS 479 of animals given niobium and nickel, were those on the right of the Table were more significantly elevated. likely to be associated with high serum Changes in circulating cholesterol levels cholesterol levels than were the transi with age are shown in table 5. No increase tional metals on the left. There was no was found in animals fed niobium, lead or significant correlation with atomic num cadmium, nor in females given antimony. ber. In the experience of this laboratory, Increases occurred in rats of both sexes which will be reported, the elements on fed zirconium and in males fed antimony. the right of the Table were more likely Data on chromium (table 1) show some to exert innate or overt toxicity in mice decrease with age, especially in females. and rats than were the transitional metals. The data on changes in serum choles DISCUSSION terol with age are incomplete. They sug The rat is a poor animal for the study gest, however, that age-linked increases of experimental hypercholesteremia, re can occur in marginal chromium defi quiring special diets and the feeding of ciency, in rats of both sexes fed zirconium cholesterol and saturated fats to develop and in males fed antimony. They suggest high serum levels. The present study con that such increases may not occur in rats cerns adequate diets and unsaturated fats given niobium, lead and cadmium. (corn oil), and high serum levels were When the groups fed tellurium were Downloaded from jn.nutrition.org by guest on May 6, 2011 only noted occasionally. Nevertheless, compared with the others, as in table 3, trends and tendencies appeared which significant differences appeared largely in might turn out to be significant in more males. Probabilities (P) of the differences susceptible mammals. being due to chance of the order of <0.005 On this basis, these data indicate that occurred in the zirconium, niobium, chro trivalent chromium given to rats in drink mium, nickel, germanium, tin, lead and ing water was associated with lowered control groups, of < 0.01 in the cadmium serum cholesterol levels, as shown previ group, of < 0.025 in the antimony group ously (4), that chromium deficiency in and < 0.05 in the arsenic group, all values young rats was associated with elevated being smaller. Only the selenium group levels, and that female rats appeared to did not so differ. In females, the vana require larger doses for this effect to ap dium, niobium, chromium, and nickel pear than males. The data show that, in groups differed (P < 0.005), as possibly general, higher levels occurred in male did the selenium group (P < 0.05), values rats fed the metalloid or nonmetallic ele being smaller than those of the tellurium ments in the "A" groups of the Periodic group. Table â€” arsenic, antimony, selenium and Because it was possible that differences tellurium â€” than in those fed metals in in cholesterol levels might be related to the transitional or "B" groups â€” vana factors other than the trace elements fed dium, niobium, nickel and cadmium â€”or to the animals, attempts were made to heavy metals in the "A" groups â€”tin and avoid unknown influences as far as practi lead. Female rats, however, also had cable. Males and females were strictly higher levels when they were given those comparable. Several groups of animals elements situated in the right of the Peri were bred and observed simultaneously as odic Table, beginning with cadmium, and follows: Controls (no chromium), chro lower levels when fed elements situated in mium (5 ppm), cadmium and lead; con- the left or "B" groups. trals (1 ppm chromium), arsenic, germa To test this impression, the elements nium and tin; lead (males), zirconium, were arranged in order of position in the niobium and antimony; chromium (5 Periodic Table from left to right and as ppm), vanadium, nickel, selenium and signed numbers from 2 (zirconium) to 27 tellurium. Each experiment required 4 (tellurium). Paired rank correlations with years for completion (5). As facilities were serum cholesterol levels were calculated. limited to 1,000 rats, additional groups For males, r = +0.486 (P ~ 0.05); for were bred and started, usually in the fall females, r = +0.622 (P < 0.01). Although or winter, when previous groups were the list of elements studied is incomplete, partly depleted through death. Analyses 480 HENRY A. SCHKOEDER of sera were made during the summers on noted in samples from the adult population several groups at weekly intervals. The of the United States (9) and the excess of diet was as uniform as feasible, although tellurium (12, 13) deserve study. annual variations in the trace element con tent of seed rye obtained from the same LITERATURE CITED area each year were possible, owing to 1. Curran, G. L. 1954 Effect of certain transi annual variations in rainfall. No consist tion group elements on hepatic synthesis of cholesterol in the rat. J. Biol. Chem., 230: 765. ent changes were found among the groups 2. Mertz, W. 1967 Biological role of chro analyzed at each interval which could be mium. Federation Proc., 26: 186. attributed to unknown factors. 3. Schroeder, H. A., W. H. Vinton, Jr. and J. J. The values obtained by the method of Balassa 1962 Effect of chromium, cad mium and lead on serum cholesterol of rats. Abell et al. (Il) in the first series (4) Proc. Soc. Exper. Biol. Med., 109: 859. were compared with those obtained by that 4. Schroeder, H. A., and J. J. Balassa 1965 of Huang et al. (10) which was an essen Influence of chromium, cadmium and lead tially similar but more simple method. on rat aortic lipids and circulating choles terol. Amer. J. Physiol., 209: 433. Mature males fed lead in the first series 5. Schroeder, H. A., J. J. Balassa and W. H. (4) had 79.3 Â±6.5 mg cholesterol/100 Vinton, Jr. 1965 Chromium, cadmium and ml; those in the second had 86.6 Â± mg6.5 lead in rats : Effects on life span, tumors and /100 ml. Mature males fed chromium had tissue levels. J. Nutr., 86: 51. Downloaded from jn.nutrition.org by guest on May 6, 2011 6. Schroeder, H. A., W. H. Vinton, Jr. and J. J. 77.0 Â± 6.6 mg/100 ml in the first series Balassa 1963 Effects of chromium, cad and 86.2 Â± 3.2 mg/100 ml in the second. mium and lead on the growth and survival Apparently the 2 methods gave values in of rats. J. Nutr., 80: 819. the same ranges for 2 sets of animals fed 7. Schroeder, H. A., W. H. Vinton, Jr. and J. J. Balassa 1963 Effect of chromium, cad the same metals. mium and other trace metals on the growth Because all these animals, except those and survival of mice. J. Nutr., 80: 39. in one control group, were given chro 8. Schroeder, H. A. 1966 Chromium deficiency in rats: A syndrome simulating diabetes mel- mium, it is evident that if an element in litus with retarded growth. J. Nutr., 88: 439. fluenced serum cholesterol levels, the ef 9. Schroeder, H. A., J. J. Balassa and I. H. fect was either additive to or antagonistic Tipton 1962 Abnormal trace metals in man. to that of chromium. Additive effects in Chromium. J. Chron. Dis., 15: 941. one sex might be postulated for niobium 10. Huang, R. C., C. P. Chen, V. Wefler and A. Raftery 1961 A stable reagent for the and nickel; antagonistic effects for tellu Liebermann-Burchard reaction. Clin. Chem., rium, selenium, antimony, vanadium and 7: 542. zirconium. 11. Abell, L. L., B. B. Levy, B. B. Brodie and Therefore, it is possible that a common F. E. Kendall 1952 A simplified method for the estimation of total cholesterol in property of nickel, niobium and chromium serum and demonstration of its specificity. may be to influence serum cholesterol of J. Biol. Chem., 195: 357. rats at low levels, and that tellurium in 12. Schroeder, H. A., J. Buckman and J. J. some way may raise it. The lowest stan Balassa 1967 Abnormal trace metals in man: tellurium. J. Chron. Dis., 20: 147. dard deviations, and therefore individual 13. International Commission on Radiological variations, were found among analyses of Protection 1960 Report of Committee II sera from rats fed these metals. Tissue on Permissible Dose for Internal Radiation levels of the various elements will be or (1959). Pergamon Press, Oxford, England. 14. Schroeder, H. A. 1967 Effects of selenate, have been reported (5). Application of selenite and tellurite on the growth and early these data to the chromium deficiency survival of mice and rats. J. Nutr., 92: 334.