Cu,Zn-superoxide dismutase is lower and copper chaperone CCS is higher in erythrocytes of copper-deficient rats and mice

Discovery of a sensitive blood biochemical marker of copper status would be valuable for assessing marginal copper intakes. Rodent models were used to investigate whether erythrocyte concentrations of copper,zinc-superoxide dismutase (SOD), and the copper metallochaperone for SOD (CCS) were sensitive to dietary copper changes. Several models of copper deficiency were studied in postweanling male Holtzman rats, male Swiss Webster mice offspring, and both rat and mouse dams. Treatment resulted in variable but significantly altered copper status as evaluated by the presence of anemia, and lower liver copper and higher liver iron concentrations in copper-deficient compared with copper-adequate animals. Associated with this copper deficiency were consistent reductions in immunoreactive SOD and robust enhancements in CCS. In most cases, the ratio of CCS:SOD was several-fold higher in red blood cell extracts from copper-deficient compared with copper-adequate rodents. Determination of red cell CCS:SOD may be useful for assessing copper status of humans.     

Development of copper deficiency in rats fed fructose or starch: weekly measurements of copper indices in blood

Copper deficiency was induced in weanling rats fed diets whose sole source of carbohydrates was starch or fructose for 7 weeks. Conventional parameters of copper status, plasma copper concentrations, ceruloplasmin activity, and erythrocyte superoxide dismutase (SOD) activity were longitudinally monitored weekly to follow the development of the deficiency and to correlate these indices with the degree of severity of the deficiency. Although 30% of the rats fed a copper-deficient fructose diet died and no deaths occurred in rats fed the copper-deficient starch diet, plasma copper, ceruloplasmin, and SOD activities were reduced to a similar extent in all rats fed copper-deficient diets regardless of the type of dietary carbohydrate. Thus, none of the indices used accurately reflected the greater degree of deficiency or mortality in rats fed the fructose diet deficient in copper. The results of the present study underscore the need for more sensitive tests or alternative parameters to assess copper status in living animals.     

Copper modulates the degradation of copper chaperone for Cu,Zn superoxide dismutase by the 26 S proteosome

Copper chaperones are copper-binding proteins that directly insert copper into specific targets, preventing the accumulation of free copper ions that can be toxic to the cell. Despite considerable advances in the understanding of copper transfer from copper chaperones to their target, to date, there is no information regarding how the activity of these proteins is regulated in higher eukaryotes. The insertion of copper into the antioxidant enzyme Cu,Zn superoxide dismutase (SOD1) depends on the copper chaperone for SOD1 (CCS). We have recently reported that CCS protein is increased in tissues of rats fed copper-deficient diets suggesting that copper may regulate CCS expression. Here we show that whereas copper deficiency increased CCS protein in rats, mRNA level was unaffected. Rodent and human cell lines cultured in the presence of the specific copper chelator 2,3,2-tetraamine displayed a dose-dependent increase in CCS protein that could be reversed with the addition of copper but not iron or zinc to the cells. Switching cells from copper-deficient to copper-rich medium promoted the rapid degradation of CCS, which could be blocked by the proteosome inhibitors MG132 and lactacystin but not a cysteine protease inhibitor or inhibitors of the lysosomal degradation pathway. In addition, CCS degradation was slower in copper-deficient cells than in cells cultured in copper-rich medium. Together, these data show that copper regulates CCS expression by modulating its degradation by the 26 S proteosome and suggest a novel role for CCS in prioritizing the utilization of copper when it is scarce.     

Leukotriene B4 synthesis and metabolism by neutrophils and granule-free cytoplasts.

Copper-deprived sycamore (Acer pseudoplatanus) cells do not excrete molecules of active laccase in their culture medium. In the range of 2-100 micrograms of copper initially present per litre of nutrient solution, the total laccase activity measured in the cell suspensions at the end of the exponential phase of growth was closely proportional to the amount of added copper. However, copper-deprived cells excreted the laccase apoprotein (laccase without copper) at the same rate as copper-supplied cells excreted the active, copper-containing, laccase. When the culture medium was initially supplied with limiting amounts of copper, the active laccase was excreted until all copper molecules were metabolized. Thereafter, the laccase apoprotein was excreted. Consequently, at the end of the exponential phase of growth, the cell supernatants contained a mixture of apoprotein and copper-containing laccase. After purification and concentration, this mixture of copper-containing laccase (blue) and laccase apoprotein (slightly yellow) showed a yellow-green colour. Under copper-limiting culture conditions an equivalent decrease of Type 1, Type 2 and Type 3 Cu2+ was observed. Addition of copper to copper-deficient enzyme solutions does not result in a recovery of the enzyme activity. However, when added to copper-deficient sycamore-cell suspensions, copper induced a recovery of the excretion of active enzyme, at a normal rate, within about 10 h. The first molecules of active laccase were excreted after 3-4 h.     

Increase of Cu,Zn-superoxide dismutase activity during differentiation of human K562 cells involves activation by copper of a constantly expressed copper-deficient protein.

Cu,Zn-superoxide dismutase activity, expressed on the basis of cell number, increased by 50% during sodium butyrate-induced differentiation of human K562 erythroleukemia cells. The increased enzyme activity was found to be concomitant with constant Cu,Zn-superoxide dismutase mRNA and immunoreactive protein levels and was accompanied by a rise in intracellular copper and glutathione. Incubation of K562 cell homogenates with copper caused an increase of Cu,Zn-superoxide dismutase activity which reached the levels observed after differentiation in the presence of sodium butyrate. The same treatment led to no significant activity increase in homogenates derived from differentiated cells. Externally added ceruloplasmin increased both intracellular copper levels and Cu,Zn-superoxide dismutase activity in undifferentiated cells to a level comparable with that observed after induction of differentiation. Both increments were abolished by depletion of cell glutathione. Cu,Zn-superoxide dismutase purified from control cells had both a lower kcat and a lower copper content than the enzyme purified from differentiated cells. From these data we conclude that: 1) Cu,Zn-superoxide dismutase is present in K562 cells also under the form of a less active copper-deficient enzyme, 2) the extent of enzyme activation is regulated post-translationally by differential delivery of copper as a function of differentiation stage, and 3) glutathione is likely to play a role in delivering copper to the copper-deficient protein in intact K562 cells.     

In vivo apoprotein catabolism of high density lipoproteins in copper-deficient, hypercholesterolemic rats

High density lipoprotein (HDL) apoprotein catabolism was examined in male Sprague-Dawley rats deficient in dietary copper. Twenty-four rats were randomly divided into two groups: copper-adequate (control, 5 mg of copper/kg diet) and copper-deficient (0.6 mg of copper/kg diet). After 5 weeks, animals were administered a tracer dose of iodinated HDL protein previously isolated from donor rats that were subjected to the same dietary treatments as the test animals. Copper-deficient rats exhibited a 54% increase in plasma volume and a 26% increase in HDL protein concentration above controls. Consequently, the intravascular pool of total HDL protein was increased 2-fold. The fractional catabolic rate of total HDL protein was similar between groups. However, because of the increased intravascular HDL pool in copper-deficient animals, the absolute catabolic rate was greater (640 +/- 49 micrograms/hr vs 316 +/- 12 micrograms/hr in controls). Tissue uptake of total HDL protein in copper-deficient rats tended to be greater in the kidneys, spleen, and testes compared with controls; the heart exhibited a significant 2.3-fold increase. In contrast, the catabolic rate of HDL protein in the liver and adrenal gland were not different between treatment groups. That an obligatory increase in HDL protein uptake was not observed in the liver and adrenal gland (organs which are sensitive to and can further metabolize cholesterol) suggests that these organs may be regulated, possibly contributing to the observed hypercholesterolemia in this model. These data imply that total HDL apoprotein catabolism is increased in response to the increased intravascular pool of HDL in copper-deficient rats.     

Interaction Between Dietary Carbohydrate and Copper Nutriture on Lipid Peroxidation in Rat Tissues

The effects of the interactions between dietary carbohydrates and copper deficiency on superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px) activities and their roles in peroxidative pathways were investigated. Weanling rats were fed diets deficient in copper and containing either 62% starch, fructose, or glucose. Decreased activity of SOD was noted in all rats fed the copper-deficient diets regardless of the nature of dietary carbohydrate. However, the decreased activity was more pronouced in rats fed fructose. Feeding the fructose diets decreased the activity of GSH-Px by 25 and 50% in the copper-supplemented and copper-deficient rats, respectively, compared to enzyme activities in rats fed similar diets containing either starch or glucose. The decreased SOD and GSH-Px activities in rats fed the fructose diet deficient in copper were associated with increased tissue per-oxidation and decreased hepatic adenosine triphosphate (ATP). When the fructose in the diet of copper-deficient rats was replaced with either starch or glucose, tissue SOD and GSH-Px activities were increased and these increases in enzyme activity were associated with a tendency toward reduced mitochondrial peroxidation when compared to the corre-sponding values for rats fed fructose throughout the experiment Dietary fructose aggrevated the symptoms associated with copper deficiency, but starch or glucose ameliorated them. The protective effects were more pronounced with starch than with glucose.     

Effect of propionyl-L-carnitine treatment on membrane phospholipid fatty acid turnover in diabetic rat erythrocytes

In this work we have examined the effect of the oral administration of propionyl-L-carnitine (PLC) on the membrane phospholipid fatty acid turnover of erythrocytes from streptozotocin-induced diabetic rats. A statistically significant reduction in radioactive palmitate, oleate, and linoleate, but not arachidonate, incorporation into membrane phosphatidylcholine (PC) of diabetic rat erythrocytes with respect to control animals was found. Changes in radioactive fatty acid incorporation were also found in diabetic red cell phosphatidylethanolamine (PE), though they were not statistically significant. Oral propionyl-L-carnitine (PLC) treatment of diabetic rats partially restored the ability of intact red cells to reacylate membrane PC with palmitate and oleate, and reacylation with linoleate was fully restored. The analysis of the membrane phospholipid fatty acid composition revealed a consistent increase of linoleate levels in diabetic rat red cells, and a modest decrease of palmitate, oleate and arachidonate. The phospholipid fatty acid composition of diabetic red blood cells was not affected by the PLC treatment. Lysophosphatidylcholine acyl-CoA transferase (LAT) specific activity measured with either palmitoyl-CoA or oleyl-CoA was significantly reduced in diabetic erythrocyte membranes in comparison to controls. In addition LAT kinetic parameters of diabetic erythrocytes were altered. The reduced LAT activity could be partially corrected by PLC treatment of diabetic rats. Our data suggest that the impaired erythrocyte membrane physiological expression induced by the diabetic disease may be attenuated by the beneficial activity of PLC on the red cell membrane phospholipid fatty acid turnover.Abbreviations LAT lysophosphatidylcholine acyl-CoA transferase - PC phosphatidylcholine - PE phosphatidylethanolamine - PLC propionyl-L-carnitine - STZ streptozotocin     

Role of band 3 tyrosine phosphorylation in the regulation of erythrocyte glycolysis.

Previous studies demonstrated that the in vitro tyrosine phosphorylation of the human erythrocyte anion transporter, band 3, prevented the binding of various glycolytic enzymes to the N terminus of the cytoplasmic tail. Since these enzymes are inhibited in their bound state, the functional consequences of band 3 tyrosine phosphorylation in the red cell should be to activate the enzymes and elevate glycolysis. We searched for various enhancers of band 3 tyrosine phosphorylation using a novel assay designed to measure the phosphotyrosine levels at the band 3 tyrosine phosphorylation/glycolytic enzyme-binding site. This assay measures the extent of phosphorylation of a synthetic band 3 peptide entrapped within resealed red cells. Using this assay, three distinct compounds, all mild oxidants, were found to stimulate the tyrosine phosphorylation of band 3. All three compounds were also found to elevate glycolytic rates in intact erythrocytes. Moreover, the antitumor drug adriamycin was found to coordinately prevent these agents from stimulating both band 3 tyrosine phosphorylation and erythrocyte glycolysis. These results suggest a possible function for a protein tyrosine kinase in human erythrocytes, to regulate glycolysis through the tyrosine phosphorylation of band 3.     

Different susceptibility of red cell membrane proteins to calpain degradation.

The presence of low levels of calpastatin activity in erythrocytes of hypertensive rats affects regulation of calpain activity so it is highly susceptible to activation within physiological fluctuations in [Ca2+]. Under identical conditions, in red cells of normotensive rats, calpain activation is efficiently controlled by the high levels of calpastatin activity, and a progressive increase in proteinase activity can only be observed in parallel with a decrease in the level of calpastatin. In intact erythrocytes from hypertensive rats exposed to small variations in [Ca2+], degradation of anion transport protein (band 3) and Ca(2+)-ATPase appears as a primary event indicating that these two transmembrane proteins are probably early recognized as targets of intracellular calpain activity. Furthermore, band 3 protein seems to be structurally modified in erythrocytes from hypertensive rats, as indicated by its increased susceptibility to degradation in the presence of 10-50 microM Ca2+. In addition, when exposed to progressive and limited increases in [Ca2+], erythrocytes from hypertensive rats, but not those from normotensive rats, show a high degree of fragility that can be restored to normal values by inhibition of calpain. These results indicate that, within fluctuations in [Ca2+] close to physiological values, regulation of calpain activity is efficiently accomplished in normal erythrocytes but is completely lost in cells from hypertensive animals. Regulation is of critical importance in maintaining normal structural and functional properties of selective red cell membrane and cytoskeletal proteins, among which band 3 and Ca(2+)-ATPase appear to be the substrates with highest susceptibility to digestion by calpain.     

Resolution of the paradox of red cell shape changes in low and high pH

The molecular basis of cell shape regulation in acidic pH was investigated in human erythrocytes. Intact erythrocytes maintain normal shape in the cell pH range 6.3-7.9, but invaginate at lower pH values. However, consistent with predicted pH-dependent changes in the erythrocyte membrane skeleton, isolated erythrocyte membranes evaginate in acidic pH. Moreover, intact cells evaginate at pH greater than 7.9, but isolated membranes invaginate in this condition. Labeling with the hydrophobic, photoactivatable probe 5-[125I]iodonaphthyl-1-azide demonstrated pH-dependent hydrophobic insertion of an amphitropic protein into membranes of intact cells but not into isolated membranes. Based on molecular weight and on reconstitution experiments using stripped inside-out vesicles, the most likely candidate for the variably labeled protein is glyceraldehyde-3-phosphate dehydrogenase. Resealing of isolated membranes reconstituted both the shape changes and the hydrophobic labeling profile seen in intact cells. This observation appears to resolve the paradox of the contradictory pH dependence of shape changes of intact cells and isolated membranes. In intact erythrocytes, the demonstrated protein-membrane interaction would oppose pH-dependent shape effects of the spectrin membrane skeleton, stabilizing cell shape in moderately abnormal pH. Stabilization of erythrocyte shape in moderately acidic pH may prevent inappropriate red cell destruction in the spleen.     

Interaction of diesel exhaust particles with human, rat and mouse erythrocytes in vitro

Inhaled ultrafine (nano) particles can translocate into the bloodstream and interact with circulatory cells causing systemic and cardiovascular events. To gain more insight into this potential mechanism, we studied the interaction of diesel exhaust particles (DEP) with human, rat and mouse erythrocytes in vitro. Incubation of erythrocytes with DEP (1, 10 or 100 μg/ml) for 30 min caused the highest hemolytic effect (up to 38%) in rats, compared to small but significant hemolysis in mice (up to 2.5%) and humans (up to 0.7%). Transmission electron microscopy of erythrocytes revealed the presence of variable degrees of ultrafine (nano)-sized aggregates of DEP either internalized and/or adsorbed onto the erythrocytes in the three species. A significant amount of DEP was found in rat and mouse (but not human) erythrocytes. Lipid erythrocyte susceptibility to in vitro peroxidation measured by malondialdehyde showed a significant and dose-dependent increase in erythrocytes of rats, but not humans or mice. Unlike in human erythrocytes, total antioxidant status (TAS) and superoxide dismutase (SOD) activity in rats were significantly and dose- dependently decreased. In mouse erythrocytes, DEP caused a decreased in SOD (at 10 μg/ml) and TAS (at 100 μg/ml) activities. In conclusion, DEP caused species-dependent erythrocyte hemolysis and oxidative stress, and were either taken up and/or adsorbed onto the red blood cells. Rat (and to a lesser degree mouse) erythrocytes were susceptible to DEP. Human erythrocytes showed the highest resistance to the observed effects. These species difference should be noted when using rats and mice blood as models for humans.     

Haemolysis and Perturbations in the Systemic Iron Metabolism of Suckling,Copper-Deficient Mosaic Mutant Mice – An Animal Model of Menkes Disease

The biological interaction between copper and iron is best exemplified by the decreased activity of multicopper ferroxidases under conditions of copper deficiency that limits the availability of iron for erythropoiesis. However, little is known about how copper deficiency affects iron homeostasis through alteration of the activity of other copper-containing proteins, not directly connected with iron metabolism, such as superoxide dismutase 1 (SOD1). This antioxidant enzyme scavenges the superoxide anion, a reactive oxygen species contributing to the toxicity of iron via the Fenton reaction. Here, we analyzed changes in the systemic iron metabolism using an animal model of Menkes disease: copper-deficient mosaic mutant mice with dysfunction of the ATP7A copper transporter. We found that the erythrocytes of these mutants are copper-deficient, display decreased SOD1 activity/expression and have cell membrane abnormalities. In consequence, the mosaic mice show evidence of haemolysis accompanied by haptoglobin-dependent elimination of haemoglobin (Hb) from the circulation, as well as the induction of haem oxygenase 1 (HO1) in the liver and kidney. Moreover, the hepcidin-ferroportin regulatory axis is strongly affected in mosaic mice. These findings indicate that haemolysis is an additional pathogenic factor in a mouse model of Menkes diseases and provides evidence of a new indirect connection between copper deficiency and iron metabolism.     

Immunochemical evidence for the transmembrane orientation of glycophorin A. Localization of ferritin-antibody conjugates in intact cells.

Antibodies were raised in rabbits to a 51-amino acid cyanogen bromide-derived peptide of human erythrocyte glycophorin A which has been shown to represent the C-terminal end of the 131-residue polypeptide chain. Antibodies prepared by immunoadsorption were found to be directed against a chymotryptic-derived peptide (residues 102 to 118) of glycophorin A but were unreactive with either intact or proteolytically modified red blood cells. No cross-reactivity was observed with glycophorin B of human or sialoglycoproteins prepared from red blood cells of other mammalian species. Ferritin-antibody conjugates of such sera were applied to thin sections of intact red blood cells (frozen or protein embedded) and were found to localize exclusively to sites distributed uniformly along the inner surfaces of the membrane. No staining was seen on sections prepared from red blood cells from other species nor on sections of human red cells pretreated with unconjugated antisera. These results provide additional evidence in intact, fixed human erythrocytes that glycophorin A has a transmembrane orientation.     

Low-pH association of proteins with the membranes of intact red blood cells. I. Exogenous glycophorin and the CD4 molecule

Glycophorin and CD4 proteins are tightly associated with intact human erythrocyte membranes after a short-time incubation at low pH (1-2 min, pH lower than 5, 37 degrees C). Flow cytometry and epifluorescence microscope observations showed that after incubation of red cells with fluorescein isothiocyanate (FITC) labeled glycophorin at pH values lower than 5, the erythrocyte membrane and subsequently formed ghost membranes were fluorescent. Unlabeled glycophorin was reacted with mouse erythrocytes using the same low-pH conditions. Flow cytometry and fluorescence microscopy showed that anti-glycophorin monoclonal antibodies were able to recognize the epitopes of glycophorin associated with the mouse erythrocytes. Kinetic experiments showed that the interaction of FITC-glycophorin with red cell membranes can be monitored by a decrease in the fluorescence intensity. Erythrocyte associated glycophorin was not removed from the membranes after 24 h incubation in human plasma (in vitro, 39 degrees C). A glycoprotein extract containing CD4 was isolated from a T4-lymphoma cell line (CEM). This protein extract was incubated with erythrocytes using the same low-pH conditions. Fluorescently labeled monoclonal antibodies against CD4 stained the red cells after association of CD4 with the membranes. Electron microscopy showed 10 nm immunoglobulin G-coated gold beads associated with CD4-bearing erythrocyte membranes after incubation with anti-CD4 antibodies and then with the gold beads. The potential use of the CD4-erythrocyte complex as a therapeutical agent against acquired immune deficiency syndrome (AIDS) is suggested.     

Effects of halothane, enflurane, and isoflurane on plasma and erythrocyte antioxidant enzymes and trace elements

Alterations of the normal redox balance in mammals might be attributed to increases of plasma free-radical concentrations and/or a disruption of the protective mechanisms. These conditions lead to damage to cellular structure by the mechanism of lipoperoxidation, particularly in the liver, kidney, and central nervous system. In this study, the effect of general anesthesia on the oxidative metabolism of human plasma and erythrocytes was investigated. Forty-five patients undergoing anesthesia by using halothane, enflurane, or isoflurane were included in this study. Blood samples were taken preoperatively, the first hour, the first day, and the third day after the operation. Superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px) enzyme activities and trace elements such as cofactor copper (Cu), zinc (Zn) and selenium (Se) levels were measured in plasma and red blood cells. Our results showed that halothane and enflurane administration increased the plasma GSH-Px activity and reduced zinc levels. In addition, they lowered SOD and GSH-Px activities and trace element levels on erythrocytes. Isoflurane had no effect on plasma antioxidant enzymes, but, similar to the other, isoflurane decreased the plasma zinc levels, erythrocyte SOD and GSH-Px activities and trace element levels. Gülhane Military Medical Academy     

Enzymatic changes in rat erythrocytes with increasing cell and donor age: Loss of superoxide dismutase activity associated with increases in catalytically defective forms

Erythrocytes from young (6–8 month old) and old (28–31 month old) WF rats were separated into four age fractions by centrifugation on a discontinuous, isotonic, arabinogalactan density gradient. Specific activities of super-oxide dismutase (SOD) with respect to activity per unit hemoglobin (Hb) and activity per cell were determined for each cell age group. SOD activity was found to diminish with respect to erythrocyte age. More significantly, it was found that young erythrocytes of old animals already contain considerably reduced SOD activity as compared to cells of young animals. The level of SOD catalytic activity per unit enzyme antigen was also found to decrease with both increasing cell and animal age. Young cells of old animals contain significant amounts of catalytically altered molecules.     

Effects of bacterial endotoxin on protecting copper-deficient rats from hyperoxia

The administration of very low doses of bacterial endotoxin protects rats during exposure to hyperoxia and is associated with the induction of lung antioxidant enzyme activities. Copper-deficient rats have increased susceptibility to O2 toxicity, which may be related to their decreased lung superoxide dismutase activity (SOD) or decreased plasma ceruloplasmin concentrations. To determine whether endotoxin can protect against hyperoxia in this susceptible model, we exposed copper-deficient and control rats to a fractional inspiratory concentration of O2 greater than 0.95 for 96 h after pretreatment with 500 micrograms/kg of bacterial endotoxin or phosphate-buffered saline (PBS). Mortality in the copper-deficient and control rats given PBS and exposed to O2 for 96 h was 100%. Copper-deficient rats died significantly earlier during the exposure than controls. No mortality occurred in either group treated with endotoxin and hyperoxia despite the decreased activity of copper-dependent enzymes in the copper-deficient rats. Copper-deficient rats treated with endotoxin and exposed to hyperoxia did increase lung Cu-Zn-SOD activity, but activity remained below levels found in air-exposed controls. Mn-SOD activity was found to be induced above air-exposed controls in the copper-deficient rats treated with endotoxin and exposed to hyperoxia. Hyperoxic exposure resulted in a marked increase in plasma ceruloplasmin concentrations in the control rats, but no increases in ceruloplasmin occurred in the copper-deficient animals. Endotoxin protects copper-deficient rats from hyperoxia despite their decreased lung Cu-Zn-SOD activity, and decreased plasma ceruloplasmin.     

Copper regulation of ceruloplasmin in copper-deficient rats.

In copper-deficient rats, oral intubation of copper increases the rate of ceruloplasmin synthesis without affecting general synthesis of plasma or liver proteins. It also restores the enzyme from half to full activity. Copper given by injection at doses commonly employed has additional nonspecific effects on protein synthesis and in some strains of rats produces severe hemolysis. In contrast to deficient rats, in normal rats copper does not elevate plasma ceruloplasmin unless hemolysis also occurs. Thus, at least in deficiency, copper availability controls the rate of synthesis, acitvation, and plasma concentration of ceruloplasmin.     

Haemolytic activity in the serum of brown trout, Salmo trutta

Brown trout serum contains a natural, spontaneous, antibody-independent lytic activity and a haemolysin antibody complement-mediated lytic activity against unsensitized and trout antibody-sensitized sheep erythrocytes, respectively. The use of various activators and inactivators of the mammalian complement system demonstrated that trout serum possesses complement or complement-like components similar in activity to those present in the classical and alternative pathways found in mammals. A single injection of trout with sheep erythrocytes stimulated the production of antibody-secreting cells in lymphoid organs and increased the levels of natural haemolysins. A second injection of sheep erythrocytes further raised the haemolysin values and antibody-secreting cell counts. Serum complement from homologous or closely related fish species was more effective for use in the haemolysin and antibody-secreting cell assays than that from heterologous sources, except guinea pig. Based on physico-chemical properties, gel filtration and immunoelectrophoretic studies, natural and induced anti-sheep erythrocyte haemolysins were found to be similar molecules and are possibly high molecular weight IgM antibodies.