BIOLOGICAL TRACE ELEMENT RESEARCH 3, 83-98 (1981)
Interactions Among Nickel, Copper, and Iron in Rats Growth, Blood Parameters, and Organ Wt/Body Wt Ratios FORREST H. NIELSEN* AND THOMAS J. ZIMMERMAN
United States Department of Agriculture, Science and Education Administration, Grand Forks Human Nutrition Research Center, and Department of Biochemistry, University of North Dakota, Grand Forks, North Dakota 58202 Received December 16, 1980; Accepted January 20, 1981
Abstract In two fully-crossed, three-way, two-by-three-by-three, factorially arranged experiments, female weanling rats were fed a basal diet supplemented with iron at 15 and 45 Ixg/g, nickel at 0, 5, and 50 ~g/g, and copper at either 0, 0.5, and 5 ix/g (Expt. 1) or 0, 0.25, and 12 txg/g (Expt. 2). A gram of basal diet contained in Expt. 1 approximately 16 ng of nickel, 2.3 Ixg of iron, and 0.47 I~g of copper; and in Expt. 2, 20 ng of nickel, 1.3 txg of iron, and 0.39 ~g of copper: Expt. 1 was terminated at 11 weeks, and Expt. 2 at 8 weeks because, at those times, some rats fed no supplemental copper and the i.'gh level of nickel began to lose weight, or die from heart rupture. The findings demonstrated that relationships are complex among nickel, copper, and iron. Nickel interacted with copper and this interaction was influenced by dietary iron. Signs of copper deficiency were more severe when nickel was supplemented to the diet provided that copper deprivation was neither very severe nor mild. Iron deprivation apparently enhanced the antagonism by exacerbating copper deficiency. Signs of copper deficiehcy that were made more severe by nickel supplementation were depressed weight gain (Expt. 2), hematocrit (Expt. 1), hemoglobin, and plasma alkaline phosphatase activity; and elevated ratios of heart wt/body wt, kidney wt/body wt, and liver wt/body wt. Because nickel and copper have similar physical and chemical properties, the interactions 9 1981 by The Humana Press Inc. All rights of any nature whatsoever reserved. 0163~.984/81/0600-0083503.20
83
84
NIELSEN AND ZIMMERMAN
between those two elements were probably the result of isomorphous replacement of copper by nickel at various functional sites that interfered with some biological processes. Index Entries: Nickel, interaction with copper and iron; copper, interaction with nickel and iron; iron, interaction with nickel and copper; nickel-copper-iron interactions; rats, Ni, Cu, and Fe interactions in; growth in rats, and Ni, Cu, and Fe interactions; blood parameters in rats, Fe, Ni, and Cu interactions and; organ weight in rats, and Ni, Cu, and Fe interactions; body weight in rats, and Ni, Cu, and Fe interactions in.
Introduction Hill and Matrone (1) used a coordination chemistry approach, drawing freely from valence bond, molecular orbital, and ligand field theories, to develop the hypothesis that ions with similar valence shell electronic structures would competitively interact in biological systems. According to that hypothesis, nickel would interact with either copper or iron, because nickel would have similar chemical properties to those elements, thus forming similar complexes. A number of reports show that nickel interacts with iron. Data suggest that nickel and iron interact both synergistically and antagonistically. In the synergistic relationship, nickel apparently enhanced or was required for iron absorption (2-4). The antagonistic, or competitive, relationship demonstrated that severe iron deficiency apparently was more detrimental to nickel-supplemented (20 Ixg/g of diet) than to nickel-deficient fats; growth was more severely depressed and perinatal mortality was higher in supplemented rats (5). A few reports indicated that nickel competitively interacts with copper. Schroeder et al. (6) found that dietary nickel supplementation decreased lung and spleen copper. Spears et al. (7) reported that supplementing nickel (20 Ixg/g of diet) to rats alleviated copper deficiency signs of growth retardation, and depressed hematocrit and hemoglobin values. They speculated that nickel substituted for copper at certain biological sites, thus preferentially sparing copper for some vital functions. This speculation was supported by the finding that nickel tended to decrease the copper content in some tissues of copper-deficient rats. Evidence for a noncompetitive interaction between nickel and copper includes findings that nickel deprivation depressed the copper level in liver, spleen, and kidney of rats (8,9) and in liver of sheep (10). Because copper and iron form different types of complexes, any interaction between them would most likely be noncompetitive, or occur where cationic properties, but not complex structure, are important. Most evidence to date supports that view and was reviewed recently by Chan and Rennert (11). The preceding suggests that there are complex relationships among nickel, copper, and iron in biological systems. Our primary objective in the following studies was to confirm an interaction between copper and nickel by using
NI, CU, AND FE INTERACTIONS
85
factorially arranged experiments. However, because iron nutriture apparently affects the response of rats to nickel deprivation (2-5), iron was included as an additional factor. As a result, our secondary objective became confirming the complex relationship among nickel, copper, and iron.
Materials and Methods Female weanling Sprague-Dawley rats (Gibco Animal Resources Laboratory, Madison, WI)* were weighed individually upon arrival and housed three per all-plastic cage measuring 50 x 24 x 16 cm and located inside a laminar flow rack (Lab Products, Carfield, NJ). The rats were assigned to groups of six in a fully-crossed, three-factor, two-by-three-by-three experiment in a completely randomized factorial arrangement of treatments. The levels of copper, iron, and nickel supplemented to the basal diet were the variables. In Expt. 1, the basal diet was supplemented with copper at 0, 0.5, and 5 ixg/g; iron, 15 (inadequate) and 45 p,g/g; and nickel, 0, 5, and 50 p,g/g. In Expt. 2, the basal diet was supplemented with copper at 0, 0.25, and 12 jxg/g; iron, 15 and 45 p,g/g; and nickel, 0, 5, and 50 p~g/g. Nickel was supplemented as NiC12-3H20 (Ultrapure grade, Alfa Inorganics, Beverly, MA). Copper w a s C u S 0 4 " 5 H 2 0 (Puratronic grade, Alfa Inorganics, Beverly, MA). Iron was a mixture of 60% Fe2(SO4)3 9 nil20 and 40% FeSO4-nH~O that was prepared from iron sponge (Puratronic grade, Alfa Inorganics, Beverly, MA) and sulfuric acid (Ultrex grade, J.T. Baker Chemical Co., Phillipsburg, N J). The rats had access to deionized water (Super Q System, Millipore Corp., Bedford, MA) in plastic cups. Fresh food in plastic cups was provided each day. Plastic equipment and cleaning procedures were described (5,12,13). Absorbent paper under the cages caught droppings and was changed every other day. Room temperature was maintained at 23~ Automatically controlled room lighting provided 12 h of light and 12 h of darkness. Animals were weighed weekly. The diets were mixed 3 days before the start of each experiment and about bi= weekly, thereafter. They were stored at - 16~ in tightly capped plastic containers. The air-dried basal diets with copper omitted (2) contained, per gram, about 16 ng of nickel, 2.3 p.g of iron, and 0.47 p~g of copper in Expt. 1; and 20 ng of nickel, 1.3 p~g of iron, and 0.39 ~g of copper in Expt. 2, as determined by atomic absorption spectrometric methods (2). The rats were fed their respective diets for 11 weeks in Expt. 1 and 8 weeks in Expt. 2. The experiments were terminated at those times because some rats fed no supplemental copper and 50 p,g of nickel/g of diet began to lose weight and *Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the US Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable.
86
NIELSEN AND ZIMMERMAN
die from heart rupture. The rats were weighed, then decapitaed subsequent to ether anesthesia and cardiac exsanguination with a heparin-coated needle and syringe. The liver, heart, one kidney, and the spleen (Expt. 2) were removed, blotted dry, and weighed immediately. Microhematocrits were determined on heart blood. Hemoglobin levels were determined by cyanmethemoglobin method (Total Hemoglobin Test Kit, Sigma Chemical Co., St. Louis, MO). Plasma was analyzed as follows: (a) alkaline phosphatase, with p-nitrophenyl phosphate as substrate (Phosphatase Test Kit, Sigma Chemical Co., St. Louis, MO), and (b) cholesterol, by enzymatic assay (Cholesterol Reagent Set, Boehringer Mannheim Biochemical, Indianapolis, IN). Data were treated by three-way analysis of variance and comparisons between individual treatment means and groups were evaluated for significance by the Scheff6 test with a per experiment rate of 0.05 (14). In the following, all significant differences described between treatment means or groups were found so by the Scheff6 test.
Results The first sign of nickel-copper interaction occurred during the last week of each experiment. In Expt. 1, three rats fed the diet without additional copper and with 50 ~g Ni/g began to lose weight--two in the group fed 15 p~g Fe/g, and one in the group fed 45 p~g Fe/g diet. In Expt. 2, six rats, three in each group described above, began to lose weight and two actually died (one from each group) from heart rupture before the experiment was terminated. In severely copperdeficient rats fed the dietary nickel supplements of 0 and 5 p~g/g, none died or lost weight. Of the severely copper-deficient rats, those fed 50 p~g of nickel/g of diet were growing just as well, or slightly better, than those fed 0 and 5 p~g/g up until the final week. Thus, there was no evidence that nickel interacted with copper to depress final body weight (Tables 1 and 2). Weight gain during the final weeks apparently was a good indicator of poor growth. With that parameter, the interaction between nickel and copper was significant in Expt. 2 but not in Expt. 1. As expected, both copper and iron deprivation depressed growth in Expts. 1 and 2. Tables 1 and 2 show that both copper and iron deprivation depressed, whereas nickel deprivation did not affect, the hemoglobin level. The low iron diet apparently depressed hemoglobin levels more markedly in Expt. 2 than in Expt. 1, especially when the diet was low in copper. Thus, some interactions differed slightly between experiments. In Expt. 1, nickel supplementation depressed hemoglobin in rats fed the diets supplemented with no copper and 15 p~g of iron/g. At the higher level of dietary iron, hemoglobin was apparently depressed in the severly copper-deficient rats only by the 50 p~g/g nickel supplement. In Expt. 2, nickel supplementation did not further depress the hemoglobin that was already extremely low in rats fed the low-iron diet with no copper supplement.
NI, CU, AND FE INTERACTIONS
87
Nonetheless, the interaction between nickel and copper was significant because the 50 I-zgNi/g diet significantly depressed the hemoglobin in rats fed diets supplemented with no copper and 45 i~g Fe/g. The fact that nickel affected the hemoglobin in severely copper-deficient rats at one level of iron supplementation, but not at the other, is reflected by the significant nickel x copper • iron interaction in Expt. 2. Because anemia in copper-deficient rats was more severe in rats at 15 ~g/g than at 45 I~g/g Fe, the interaction between copper and iron was significant. The interaction between nickel and iron did not agree between experiments because the response of the low iron rats fed no supplemental copper to nickel did not agree. In Expt. 2, the interaction between nickel and iron was significant because nickel supplementation elevated the hemoglobin in rats fed the low, but not the high, level of iron. In Expt. 1, the opposite apparently was true. When nickel was supplemented to the low-iron diets the hemoglobin was depressed in the severely copper-deficient rats, and this depression was not balanced by the slight elevation in hemoglobin in copper-supplemented rats. When the higher level of iron was fed, average hemoglobin was slightly higher for nickel-supplemented than for nickel-deprived rats. The hematocrit findings were similar to those for hemoglobin. In both experiments, copper deficiency depressed, and dietary iron did not affect, plasma alkaline phosphatase activity (Tables 1 and 2). The activity was also markedly depressed by 50 ~g/g of dietary nickel in rats fed no supplemental copper, regardless of dietary iron. In the copper-supplemented rats of both experiments, the depression of the activity was less marked, and in one group (low iron, high copper) was slightly elevated by nickel supplementation. Thus, the interaction between copper and nickel was significant. In Expt. 2, the different response to nickel supplementation by copper-sufficient (12 ~g Cu/g diet) rats fed either low or adequate iron probably was the major factor in the slight, but significant, interaction between nickel and iron. The depression of plasma alkaline phosphatase activity was slightly greater in the low- than in the high-iron, copper-deficient rats and probably caused the slight interaction between copper and iron found in Expt. 2. No three-way interaction between nickel, copper, and iron was found in either experiment. Tables 3 and 4 show that the heart wt/body wt ratio was higher in copperdeficient than in copper-supplemented rats. In Expt. 1, nickel supplementation of the severely copper-deficient rats accentuated this sign of deficiency, regardless of iron supplementation. Thus, the interaction between nickel and copper was significant. In Expt. 2, nickel supplementation elevated the heart wt/body wt ratio in copper-deficient rats fed the high level of dietary iron, and apparently elevated the ratio in rats fed 0.25 l~g Cu/g and the tow iron diet. The latter, however, was balanced by the finding that the ratio was higher in nickel-deprived than in nickel-supplemented rats fed no supplemental copper. Thus, on the average, dietary nickel affected the heart wt/body wt ratio when the high iron, but not when the low iron, diet was fed. Thus, the three-way interaction between nickel, copper, and iron was significant in Expt. 2. The heart wt/body wt ratio
88
NIELSEN
AND
ZIMMERMAN
%
a-,
. . . . . . . . . . . . . . .
2=
~
o o
E
u~ ,,_t
9
O
ar
E
. ~. . . .
~
r-.i e,i c~ c~ o-: ,---.i .~ ~
e,i ,--,i ~
O
O
O
o
-1 E ~
._~
z.2 E
9
~2
r~
r,.p
O
~
O
O
~
O
O
~
O
~
O
O
~
O
O
~
NI,
CU,
AND
FE
89
INTERACTIONS
&
9
E r Cq
gg~ggg~'~ oo
o o o
r
d
~
(",1
~
"V---.
r.-,
0O
rd
d
c~
E un
o 0
0
dd
d d d
x
0d
~
o
~~
',D
~I
',,o o-1
~
rq
0
E
S
6o
zz~,
i~
i.r 1
o-,
-d~.v
0
.-- ~
',d
s eB
o")
P ~'~,
"E " - ' , ~ x
s ~~ x~ •. ~ - - 8
e-,
~
~
~ ~.~
U & Z Z Z U Z m
~ ~4
Interactions Among Nickel, Copper, and Iron in Rats Growth, Blood Parameters, and Organ Wt/Body Wt Ratios FORREST H. NIELSEN* AND THOMAS J. ZIMMERMAN
United States Department of Agriculture, Science and Education Administration, Grand Forks Human Nutrition Research Center, and Department of Biochemistry, University of North Dakota, Grand Forks, North Dakota 58202 Received December 16, 1980; Accepted January 20, 1981
Abstract In two fully-crossed, three-way, two-by-three-by-three, factorially arranged experiments, female weanling rats were fed a basal diet supplemented with iron at 15 and 45 Ixg/g, nickel at 0, 5, and 50 ~g/g, and copper at either 0, 0.5, and 5 ix/g (Expt. 1) or 0, 0.25, and 12 txg/g (Expt. 2). A gram of basal diet contained in Expt. 1 approximately 16 ng of nickel, 2.3 Ixg of iron, and 0.47 I~g of copper; and in Expt. 2, 20 ng of nickel, 1.3 txg of iron, and 0.39 ~g of copper: Expt. 1 was terminated at 11 weeks, and Expt. 2 at 8 weeks because, at those times, some rats fed no supplemental copper and the i.'gh level of nickel began to lose weight, or die from heart rupture. The findings demonstrated that relationships are complex among nickel, copper, and iron. Nickel interacted with copper and this interaction was influenced by dietary iron. Signs of copper deficiency were more severe when nickel was supplemented to the diet provided that copper deprivation was neither very severe nor mild. Iron deprivation apparently enhanced the antagonism by exacerbating copper deficiency. Signs of copper deficiehcy that were made more severe by nickel supplementation were depressed weight gain (Expt. 2), hematocrit (Expt. 1), hemoglobin, and plasma alkaline phosphatase activity; and elevated ratios of heart wt/body wt, kidney wt/body wt, and liver wt/body wt. Because nickel and copper have similar physical and chemical properties, the interactions 9 1981 by The Humana Press Inc. All rights of any nature whatsoever reserved. 0163~.984/81/0600-0083503.20
83
84
NIELSEN AND ZIMMERMAN
between those two elements were probably the result of isomorphous replacement of copper by nickel at various functional sites that interfered with some biological processes. Index Entries: Nickel, interaction with copper and iron; copper, interaction with nickel and iron; iron, interaction with nickel and copper; nickel-copper-iron interactions; rats, Ni, Cu, and Fe interactions in; growth in rats, and Ni, Cu, and Fe interactions; blood parameters in rats, Fe, Ni, and Cu interactions and; organ weight in rats, and Ni, Cu, and Fe interactions; body weight in rats, and Ni, Cu, and Fe interactions in.
Introduction Hill and Matrone (1) used a coordination chemistry approach, drawing freely from valence bond, molecular orbital, and ligand field theories, to develop the hypothesis that ions with similar valence shell electronic structures would competitively interact in biological systems. According to that hypothesis, nickel would interact with either copper or iron, because nickel would have similar chemical properties to those elements, thus forming similar complexes. A number of reports show that nickel interacts with iron. Data suggest that nickel and iron interact both synergistically and antagonistically. In the synergistic relationship, nickel apparently enhanced or was required for iron absorption (2-4). The antagonistic, or competitive, relationship demonstrated that severe iron deficiency apparently was more detrimental to nickel-supplemented (20 Ixg/g of diet) than to nickel-deficient fats; growth was more severely depressed and perinatal mortality was higher in supplemented rats (5). A few reports indicated that nickel competitively interacts with copper. Schroeder et al. (6) found that dietary nickel supplementation decreased lung and spleen copper. Spears et al. (7) reported that supplementing nickel (20 Ixg/g of diet) to rats alleviated copper deficiency signs of growth retardation, and depressed hematocrit and hemoglobin values. They speculated that nickel substituted for copper at certain biological sites, thus preferentially sparing copper for some vital functions. This speculation was supported by the finding that nickel tended to decrease the copper content in some tissues of copper-deficient rats. Evidence for a noncompetitive interaction between nickel and copper includes findings that nickel deprivation depressed the copper level in liver, spleen, and kidney of rats (8,9) and in liver of sheep (10). Because copper and iron form different types of complexes, any interaction between them would most likely be noncompetitive, or occur where cationic properties, but not complex structure, are important. Most evidence to date supports that view and was reviewed recently by Chan and Rennert (11). The preceding suggests that there are complex relationships among nickel, copper, and iron in biological systems. Our primary objective in the following studies was to confirm an interaction between copper and nickel by using
NI, CU, AND FE INTERACTIONS
85
factorially arranged experiments. However, because iron nutriture apparently affects the response of rats to nickel deprivation (2-5), iron was included as an additional factor. As a result, our secondary objective became confirming the complex relationship among nickel, copper, and iron.
Materials and Methods Female weanling Sprague-Dawley rats (Gibco Animal Resources Laboratory, Madison, WI)* were weighed individually upon arrival and housed three per all-plastic cage measuring 50 x 24 x 16 cm and located inside a laminar flow rack (Lab Products, Carfield, NJ). The rats were assigned to groups of six in a fully-crossed, three-factor, two-by-three-by-three experiment in a completely randomized factorial arrangement of treatments. The levels of copper, iron, and nickel supplemented to the basal diet were the variables. In Expt. 1, the basal diet was supplemented with copper at 0, 0.5, and 5 ixg/g; iron, 15 (inadequate) and 45 p,g/g; and nickel, 0, 5, and 50 p,g/g. In Expt. 2, the basal diet was supplemented with copper at 0, 0.25, and 12 jxg/g; iron, 15 and 45 p,g/g; and nickel, 0, 5, and 50 p~g/g. Nickel was supplemented as NiC12-3H20 (Ultrapure grade, Alfa Inorganics, Beverly, MA). Copper w a s C u S 0 4 " 5 H 2 0 (Puratronic grade, Alfa Inorganics, Beverly, MA). Iron was a mixture of 60% Fe2(SO4)3 9 nil20 and 40% FeSO4-nH~O that was prepared from iron sponge (Puratronic grade, Alfa Inorganics, Beverly, MA) and sulfuric acid (Ultrex grade, J.T. Baker Chemical Co., Phillipsburg, N J). The rats had access to deionized water (Super Q System, Millipore Corp., Bedford, MA) in plastic cups. Fresh food in plastic cups was provided each day. Plastic equipment and cleaning procedures were described (5,12,13). Absorbent paper under the cages caught droppings and was changed every other day. Room temperature was maintained at 23~ Automatically controlled room lighting provided 12 h of light and 12 h of darkness. Animals were weighed weekly. The diets were mixed 3 days before the start of each experiment and about bi= weekly, thereafter. They were stored at - 16~ in tightly capped plastic containers. The air-dried basal diets with copper omitted (2) contained, per gram, about 16 ng of nickel, 2.3 p.g of iron, and 0.47 p~g of copper in Expt. 1; and 20 ng of nickel, 1.3 p~g of iron, and 0.39 ~g of copper in Expt. 2, as determined by atomic absorption spectrometric methods (2). The rats were fed their respective diets for 11 weeks in Expt. 1 and 8 weeks in Expt. 2. The experiments were terminated at those times because some rats fed no supplemental copper and 50 p,g of nickel/g of diet began to lose weight and *Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the US Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable.
86
NIELSEN AND ZIMMERMAN
die from heart rupture. The rats were weighed, then decapitaed subsequent to ether anesthesia and cardiac exsanguination with a heparin-coated needle and syringe. The liver, heart, one kidney, and the spleen (Expt. 2) were removed, blotted dry, and weighed immediately. Microhematocrits were determined on heart blood. Hemoglobin levels were determined by cyanmethemoglobin method (Total Hemoglobin Test Kit, Sigma Chemical Co., St. Louis, MO). Plasma was analyzed as follows: (a) alkaline phosphatase, with p-nitrophenyl phosphate as substrate (Phosphatase Test Kit, Sigma Chemical Co., St. Louis, MO), and (b) cholesterol, by enzymatic assay (Cholesterol Reagent Set, Boehringer Mannheim Biochemical, Indianapolis, IN). Data were treated by three-way analysis of variance and comparisons between individual treatment means and groups were evaluated for significance by the Scheff6 test with a per experiment rate of 0.05 (14). In the following, all significant differences described between treatment means or groups were found so by the Scheff6 test.
Results The first sign of nickel-copper interaction occurred during the last week of each experiment. In Expt. 1, three rats fed the diet without additional copper and with 50 ~g Ni/g began to lose weight--two in the group fed 15 p~g Fe/g, and one in the group fed 45 p~g Fe/g diet. In Expt. 2, six rats, three in each group described above, began to lose weight and two actually died (one from each group) from heart rupture before the experiment was terminated. In severely copperdeficient rats fed the dietary nickel supplements of 0 and 5 p~g/g, none died or lost weight. Of the severely copper-deficient rats, those fed 50 p~g of nickel/g of diet were growing just as well, or slightly better, than those fed 0 and 5 p~g/g up until the final week. Thus, there was no evidence that nickel interacted with copper to depress final body weight (Tables 1 and 2). Weight gain during the final weeks apparently was a good indicator of poor growth. With that parameter, the interaction between nickel and copper was significant in Expt. 2 but not in Expt. 1. As expected, both copper and iron deprivation depressed growth in Expts. 1 and 2. Tables 1 and 2 show that both copper and iron deprivation depressed, whereas nickel deprivation did not affect, the hemoglobin level. The low iron diet apparently depressed hemoglobin levels more markedly in Expt. 2 than in Expt. 1, especially when the diet was low in copper. Thus, some interactions differed slightly between experiments. In Expt. 1, nickel supplementation depressed hemoglobin in rats fed the diets supplemented with no copper and 15 p~g of iron/g. At the higher level of dietary iron, hemoglobin was apparently depressed in the severly copper-deficient rats only by the 50 p~g/g nickel supplement. In Expt. 2, nickel supplementation did not further depress the hemoglobin that was already extremely low in rats fed the low-iron diet with no copper supplement.
NI, CU, AND FE INTERACTIONS
87
Nonetheless, the interaction between nickel and copper was significant because the 50 I-zgNi/g diet significantly depressed the hemoglobin in rats fed diets supplemented with no copper and 45 i~g Fe/g. The fact that nickel affected the hemoglobin in severely copper-deficient rats at one level of iron supplementation, but not at the other, is reflected by the significant nickel x copper • iron interaction in Expt. 2. Because anemia in copper-deficient rats was more severe in rats at 15 ~g/g than at 45 I~g/g Fe, the interaction between copper and iron was significant. The interaction between nickel and iron did not agree between experiments because the response of the low iron rats fed no supplemental copper to nickel did not agree. In Expt. 2, the interaction between nickel and iron was significant because nickel supplementation elevated the hemoglobin in rats fed the low, but not the high, level of iron. In Expt. 1, the opposite apparently was true. When nickel was supplemented to the low-iron diets the hemoglobin was depressed in the severely copper-deficient rats, and this depression was not balanced by the slight elevation in hemoglobin in copper-supplemented rats. When the higher level of iron was fed, average hemoglobin was slightly higher for nickel-supplemented than for nickel-deprived rats. The hematocrit findings were similar to those for hemoglobin. In both experiments, copper deficiency depressed, and dietary iron did not affect, plasma alkaline phosphatase activity (Tables 1 and 2). The activity was also markedly depressed by 50 ~g/g of dietary nickel in rats fed no supplemental copper, regardless of dietary iron. In the copper-supplemented rats of both experiments, the depression of the activity was less marked, and in one group (low iron, high copper) was slightly elevated by nickel supplementation. Thus, the interaction between copper and nickel was significant. In Expt. 2, the different response to nickel supplementation by copper-sufficient (12 ~g Cu/g diet) rats fed either low or adequate iron probably was the major factor in the slight, but significant, interaction between nickel and iron. The depression of plasma alkaline phosphatase activity was slightly greater in the low- than in the high-iron, copper-deficient rats and probably caused the slight interaction between copper and iron found in Expt. 2. No three-way interaction between nickel, copper, and iron was found in either experiment. Tables 3 and 4 show that the heart wt/body wt ratio was higher in copperdeficient than in copper-supplemented rats. In Expt. 1, nickel supplementation of the severely copper-deficient rats accentuated this sign of deficiency, regardless of iron supplementation. Thus, the interaction between nickel and copper was significant. In Expt. 2, nickel supplementation elevated the heart wt/body wt ratio in copper-deficient rats fed the high level of dietary iron, and apparently elevated the ratio in rats fed 0.25 l~g Cu/g and the tow iron diet. The latter, however, was balanced by the finding that the ratio was higher in nickel-deprived than in nickel-supplemented rats fed no supplemental copper. Thus, on the average, dietary nickel affected the heart wt/body wt ratio when the high iron, but not when the low iron, diet was fed. Thus, the three-way interaction between nickel, copper, and iron was significant in Expt. 2. The heart wt/body wt ratio
88
NIELSEN
AND
ZIMMERMAN
%
a-,
. . . . . . . . . . . . . . .
2=
~
o o
E
u~ ,,_t
9
O
ar
E
. ~. . . .
~
r-.i e,i c~ c~ o-: ,---.i .~ ~
e,i ,--,i ~
O
O
O
o
-1 E ~
._~
z.2 E
9
~2
r~
r,.p
O
~
O
O
~
O
O
~
O
~
O
O
~
O
O
~
NI,
CU,
AND
FE
89
INTERACTIONS
&
9
E r Cq
gg~ggg~'~ oo
o o o
r
d
~
(",1
~
"V---.
r.-,
0O
rd
d
c~
E un
o 0
0
dd
d d d
x
0d
~
o
~~
',D
~I
',,o o-1
~
rq
0
E
S
6o
zz~,
i~
i.r 1
o-,
-d~.v
0
.-- ~
',d
s eB
o")
P ~'~,
"E " - ' , ~ x
s ~~ x~ •. ~ - - 8
e-,
~
~
~ ~.~
U & Z Z Z U Z m
~ ~4
