- Email: [email protected]
A high-affinity zinc-binding plasma protein in channel catfish (Ictalurus punctatus)
0306~4492/91 $3.00+ 0.00 0 1991Pergamon Press plc
Camp. Eiockm. Physiof.Vol. IOOC,No. 3, pp. 491494, 1991
F’rintedin
Great Britain
A HIGH-AFFINITY ZINC-BINDING PLASMA PROTEIN CHANNEL CATFISH (ICTALURUS PUNCTATUS)
IN
P. J. BENTLEY Department of Anatomy, Physiological Sciences and Radiology, College of Veterinary Medicine, North Carolina State University, Raleigh, NC 27606, U.S.A. (Telephone: 919-829-4267) (Received 14 December 1990)
Abstract-l. Channel catfish (Ictalurus punctatus) have a remarkably high concentration of zinc (Zn) in their blood serum, about 20 pg/ml. However, compared to mammals, the concentrations of Zn in their tissues are not remarkable. The serum Zn is dialyzable against a solution containing 1 mM EDTA. 2. Following separation of serum proteins by gel-filtration most of the Zn was recovered in a fraction with the same peak volume of elution for the Zn and protein concentrations and having a molecular weight similar to bovine serum albumin. 3. Binding of Zn to such sites was not changed by Cu*+, Cd2+, Car+, or La3+. N-ethylemaleimide (NEM) appeared to decrease binding slightly. 4. Equilibrium dialysis with a Scatchard plot analysis of these data suggested that a single set of binding sites was present on the protein(s) with K,, of 2 x 10m5M. There were binding sites for 35 x 10e8 M Zn/mg protein. 5. Parallel equilibrium dialysis measurements using human, rabbit and chicken albumins indicated that the catfish serum protein(s) had a much higher affinity and binding capacity for Zn than the albumins in these species. 6. The catfish Zn serum-binding protein may be an albumin. The possible physiological significance of such a serum protein in freshwater fish is discussed.
associated with a protein fraction consistent in molecular weight with a catfish serum albumin. The Zn is bound with a much higher binding affinity, and has a greater number of binding sites than observed in mammalian and chicken albumins.
INTRODUCTION
Zinc (Zn) is an essential trace metal element (Underwood, 1977). Its deficiency results in disorders of growth, reproduction and the immune system. Such effects reflect its role in many enzymatic processes and the interactions between macromolecules (Cunnane, 1988). In mammals, normal serum concentrations of Zn are about 1 pug/ml. Most of this Zn is bound to serum proteins, especially albumins (Cunnane, 1988). Elevated concentrations of Zn occur in the serum when large dietary excesses of the mineral are consumed, and this increase is accompanied by rises in the Zn levels in various tissues, especially liver, kidneys and bone (Allen et al., 1983). Zinc has been shown to be an essential dietary constituent for normal growth in channel catfish (Gatlin and Wilson, 1983). These catfish (I. punctutus) are a freshwater species that are native to the southeastern United States, where they are extensively farmed as a source of food. Compared to mammals, they have a remarkably high concentration of Zn in their serum, over 20pg/ml (Gatlin and Wilson, 1983, and this paper). Other species of freshwater fish, including rainbow trout, lake trout, walleye and whitefish, have also been shown to have high plasma Zn concentrations (Bettger et al., 1987). Thus, high plasma or serum Zn concentrations may be a widespread phenomenon in freshwater fish. However, (this paper), tissue levels of Zn in catfish were not found to be remarkable and, indeed, they were similar to those in normal rabbits. The Zn in rainbow trout plasma is mostly bound to plasma proteins and is dialyzable (Bettger ef al., 1987). In catfish, most of this Zn (this paper) was found to be
MATERIALS AND
METHODS
Channel catfish, I. puncratus, raised for commercial purposes were obtained from several sources in North Carolina (see Table 1). Blood was collected from severed caudal vessels of fish that had been anesthetized with MS 222 (0.1%) (Sigma Chemical Co., St. Louis, MO) or had been electrically stunned. The serum was separated after clotting had occurred. Tissues for Zn analysis were collected from freshly purchased fish killed with the MS 222, using instruments that had been washed overnight in saturated EDTA and thoroughly rinsed in distilled water. They were weighed and stored at -20°C in polystyrene sample cups to await analysis. Analysis of zinc The Zn was analyzed by flame atomic absorption spectrophotometry (Perkin-Elmer, Model 2380), using ai electrodeless lamn for Zn absorbing at 213.9 nM and an air-acetylene flame. Standards were prepared from a certified Zn reference standard (Fisher Scientific, Fairlawn, NJ) and diluted with either distilled water, for serum analyses, or 0.5 N hydrochloric (HCl) acid for the tissue analyses. The HCl was “Baker-Instra-analyzed” for trace metal work (J. T. Baker Chemical Co., Phillipsburg, NJ). The tissues were ashed overnight in quartz crucibles at 520°C. The ash was dissolved in 1 N HCl and diluted to aive a final concentration of acid of 0.5 N. All glassware and the crucibles were cleaned with chromic-sulphuric acid (Fisher Scientific, Fairlawn, NJ) and thoroughly rinsed with distilled water before use. Zinc was
491
P. J. BENTLEY
492
not detected in any of the vessels used in the analyses during the normal periods of their use.. Measurement of zinc binding to serum proteins Separation of serum proteins was performed on a gelfiltration column (C16/1&, Pharmacia-LKB Biotechnology, Piscatawav. NJ) nacked with Senhadex G-200. The elution fluid cot&red l?lOmM NaCl buffered with Tris-HCI to give a pH of 7.4, and 0.02% sodium azide as a preservative. The background Zn concentration in this solution was about 0.01 pg/ml. Fluid was passed in an ascending direction up the column with the aid of a peristaltic pump. The flow rate was about 8 ml/hr and fractions were collected at 30-min intervals with the aid of a fraction collector. The spectrophotometric absorption of the eluate was continuously monitored at 280nM to indicate the presence of proteins. The Zn concentration of the fractions was measured by flame atomic absorption spectrophotometry. The protein concentration of serum and the collected fractions for dialysis was determined by the method of Lowry et al. (1951). Equilibrium dialysis of the serum proteins was performed by placing the samples (1.5 ml) in dialysis bags (MWCO 3500, Spectropor 3, Fisher Scientific, Fairlawn, NJ) and incubating them (in polyethylene containers) in a large volume (200 ml) of the stirred elution fluid for 24 hr at 5°C. Equilibration between this solution and the solution in the dialysis bag was found to occur in this time. Binding of Zn was determined from the final Zn concentration in the dialysate and its protein concentration, and the final Zn concentration in the external dialysis solution. The rabbit tissues were obtained from male New Zealand white rabbits obtained from A. C. Daley’s Kabbitry, Farmville, NC. They were fed standard rabbit chow. They were killed in an atmosphere of 100% carbon dioxide, and the tissues were collected and analyzed in the same fashion as described for the catfish. Treatment of results The graphs, including the Scatchard plots (Scatchard, 1949) w>th-their linear regression lines were plotted with a Zenith (Model 48) computer using Sigma Plot CV 3.10 software (Graphic Software Systems Inc, Sausalito, CA). RESULTS
Serum zinc concentration in channel catfish Serum was collected from catfish from several locations in North Carolina and found to have mean concentrations of Zn ranging from about 14 to 21 pg/ml (Table 1). Nearly all of this Zn was found to be dialyzable. In samples of the serum collected from the Raleigh fish, only 3.3 f 0.4% (6) of the Zn remained associated with the serum sample after 24 hr incubation in the “elution” solution at pH 7.4 containing 1 mM EDTA. Catfish serum was subjected to gel-filtration on Sephadex G-200 (see Materials and Methods). A total of 74.3 f 7.1% (8) of the Zn in the serum samples applied to the column was recovered in a fraction with a volume of elution (I’,) of 122.6 f 1.5 ml. The Zn concentration “peak” concentration was almost identical with a major protein peak detected spectrophotometrically, V, 120.6 f 1.Oml. Bovine serum albumin that was used as a standard protein for comparison had a V, of 1255ml. Another small peak of Zn concentration with a V, of 82.7 k 0.37 (in the void volume of the column) contained 0.5 + 0.13% (8) of the total Zn in the serum. Detectable levels of Zn were not found in other fractions of the eluate.
Table 1. Serum concentrations of zinc in channel catfish (Ictalurus puncmrus) obtained from different locations in North Carolina Location Zn &g/ml) Gamer* 19.1 *0.64(10) Hillsborough? 14.2 f 0.63 (6) Raleigbt 21.1 f 0.43 (8) Kernersvill@ 16.4 k 0.56 (IO) 18.7 (36) Aydeo II *Aquaculture Advisory Service (fingerlings). tB. Grubb, farm pond (mature adults). SCollege of Veterinary Medicine longterm holding tanks, (mature adults). $Blue Ridge Hatchery (fingerlings). IlCarolina Catfish Classics (mature adults, pooled sample from 36 fish). Results are as means k SE for the number of fish given in parentheses.
Tissue concentrations
of zinc in catjish
Tissues from the group of catfish from “Gamer” (Table 1) were analyzed for their Zn content (Table 2). Included, for comparison, are a series of analyses of rabbit tissues also performed in this laboratory. Generally, it can be seen that the tissue Zn concentrations in the two species were similar. Notable differences were much higher concentrations of Zn in catfish skin but lower concentrations in their bone. Apart from their high serum concentrations of Zn, the catfish tissue Zn concentration did not display any indications that they were in a hyperzincemic condition compared to the rabbits. Binding of zinc to catjish serum proteins
The observed “hyperzincemia” in catfish, as compared to mammals, could be reflecting qualitative and quantitative differences in Zn binding to the serum proteins. The concentration of serum proteins in the catfish was found to vary somewhat. Thus, in the “Garner” fish, it was found to be 6.1 + 0.30 g/100 serum (6), while in the “Hillsborough” fish, from a farm pond, it was 3.6 + 0.25 g/lOOml (6). It was noteworthy that the serum Zn concentrations in the “Hillsborough” fish were also less than in the “Garner” fish but they were, according to mammalian It, thus, criteria, still markedly hyperzincemic. seemed possible that the hyperzincemia may be reflecting a higher binding affinity for Zn by the Table 2. Tissue concentrations of zinc in channel catfish and laboratory rabbits Zn (j~g/g wet wt) Catfish* Rabbit Serum 19.1 f 0.68 (10) 1.6 + 0.074 Blood 16.8 f 0.48 (10) Skeletal muscle 8.2 f 0.53 (10) 7.1 +0.4 Skin 74.4 f 4.00 (12) 5.7 + 0.66 Liver 24.7 f 0.59 (IO) 29.1 k 1.26 Brain 9.3 * 0.22 (IO) 10.2 * 0.44 Kidney 31.5 k 2.45 (12) 22.7 + 0.76 Gill arch 25.6 k 1.21 (7) Bone 58.94 1.17(9) 115 f 6.7 Total body 29.8 f 0.77 (7) 40t *Fish from Aquaculture Advisory Service, Garner, North Carolina. tFrom Spray and Widdowson (1951). Results are as means f SE for the numbers of fish given in parentheses. There were 9 rabbits in each group of analyses.
Zinc-binding protein
493
Table 3. Zinc binding to catfish serum protein fraction* Equilibrium concentration Zinc bound of Zn in dialysis fluid &g/mg serum protein) Wmf) 0.13 2.175 0.078 0.32 4.24 f 0.148 0.70 7.77kO.153 1.8 12.62k 0.463 2.9 15.67kO.159 4.1 16.55k 0.258 8.5 18.32+ 0.548 24.0 23.56f 0.266 *For details see Materials and Methods. Results are as mean + SE for 6 measurements. Zinc Equilibrium Concentration wW
Fig. 1. Relationship of zinc bound to serum protein and the equilibrium concentration of the mineral in catfish (A), and (for serum albumins) rabbit (a), human (0) and chicken (A). Each point is the mean of 6 determinations. The standard errors were too small to display on a graph to this scale, but the values for the catfish are given in Table 3.
catfish serum proteins. The catfish (from “Ayden”) serum albumin-like fraction was, thus, collected from the gel-filtration column. It was pooled for studies of Zn binding by equilibrium dialysis (see Materials and Methods). A parallel comparison of Zn binding to commercial (Sigma Chemical Co., St. Louis, MO) preparations of rabbit, human and chicken serum albumins was also performed. The results are summarized in Fig. 1. It is readily apparent that the catfish serum protein preparation bound Zn far more readily than the mammalian or bird albumins. In addition, it had much higher binding capacity for Zn. A Scatchard plot analysis of the binding properties of the catfish serum protein to Zn was performed (Fig. 2, Table 3). It can be seen that there appeared to be a major single set of binding sites that, from the intercept with the x-axis, suggested binding sites for about 35 x lO_rM Zn/mg Protein (N) with an equilibrium dissociation constant (Kn) of 2 x 10msM. A similar plot (not shown) for the human serum albumin suggested the presence of two major sets of binding sites with N values of 3 x 10m8M Zn/mg protein and 11 x lo-*M Zn/mg protein and a K,, respectively, of 0.8 x 10m4M and 2.5 x 10m4M. 2r
Thus, as indicated in Fig. 1, the catfish serum protein has a much higher affinity for binding Zn, as well as a large binding capacity. Eflects of N-ethylemaleitnide (NEM) and various divalent ions on binding of zinc to catfish plasma proteins
In order to help describe the nature and specificity of the Zn-binding sites in the catfish serum, protein binding of Zn was measured in the presence of NEM, which irreversibly alkylates sulphydryl groups, and various other divalent ions and La3+ (Table 4). It can be seen that the presence of the various ions made little or no difference to the Zn binding. N-ethylemaleimide appeared to decrease binding slightly, by about 25%. DISCUSSION
Compared to mammals, channel catfish, and several other species of freshwater fish, exhibit a natural “hyperzincemia” as defined by relatively high levels of Zn in their serum as compared to mammals (Gatlin and Wilson, 1983; Bettger et al., 1987). This condition, however, appears to be a normal physiological one in such fish. Tissue concentrations of Zn, thus, do not indicate that channel catfish have excesses of this essential trace metal element. The high serum levels of Zn can be accounted for by the presence of a serum protein, or proteins, that, compared to mammals, have a remarkably high binding affinity for zinc. This binding protein(s) elutes as a single component by gel-filtration and appears to have a molecular weight slightly higher than bovine serum albumin (67 kDa), which was used as a standard protein in such separations. Table 4. Effects of N-ethylemaleimide (NEM), Cd2+, Cu’+, Ca2+ and La’+ on bindine of zinc to catfish serum orotein Conditions NEM, IO-‘M
0 0
0
10 r,
20 Zn Bound x IO’ Mlmg
0
30
40
Protein
Fig. 2. A Scatchard plot of binding of zinc to catfish serum protein(s) compared to the equilibrium zinc concentration. For details see Materials and Methods. Each point is the mean for 6 determinations using the data given in Table 3.
Cd’+, Cd*+, Cu2+, Cu’+, Ca*+, La)+.
lo-‘M lo-‘M lo-‘M W4 M lo-* M 102M
Equilibrium Zn cont. (rgiml) 0.39 3.0 0.35 0.39 0.36 0.39 0.38 0.40
Zn binding Control’ (pg Zn/mg protein) 3.31 + 0.106 11.49 f 0.200 4.55 + 0.217 4.16 k 0.125 5.07 + 0.183 3.81 Ifr0.039 4.85 + 0.150 4.44 f 0.078
4.49 15.74 4.51 4.89 4.61 4.89 4.60 4.98
*Equilibrium concentrations of Zn are not precisely reproducible in such dialysis experiments. Thus, the control binding values are predicted ones from bracheting between closest control values (see Table 3) assuming a linear relationship. Results are as mean + SE for 6 measurements.
494
P. J. BENTLEY
The present observations only allow speculation as to the precise nature of the catfish serum Zn-binding protein. The peaks of protein concentration activity, determined spectrophotometrically, and the Zn concentration corresponded, but it remains possible that other proteins of similar molecular weight were present in the collected gel-filtration fractions. Such proteins could be contributing to the observed metal binding. One possibility is transferrin (molecular weight 78 kDa) though human transferrin has been found to have an even lower binding affinity for Zn than human serum albumin (Charlwood, 1979). A Scatchard plot analysis (Scatchard, 1949) of the binding information suggested that there was one major type of binding site on the catfish serum protein, and this had a KD for Zn binding of 2 x lo-‘M and a number of binding sites (N) of 35 x lo-‘M/mg protein. For comparison, values were determined similarly for human serum albumin, which is the principal Zn-binding protein in human plasma. Two sets of sites were apparent, one with a KD of 0.8 x 10m4M and N of 3 x 10-8M/mg protein and the other, which had a lower affinity, a KD of 2.5 x 10e4 M and N of 11 x lOwEM/mg protein. The catfish serum protein, thus, had a substantially higher binding affinity and capacity for Zn than the human serum albumin. The suggested presence of a single type of binding site on the catfish serum protein favors the contribution of a single protein and, as the predominant one expected in this fraction is an albumin, the binding protein is most likely a catfish albumin. Zinc is known to have a special propensity to bind to imidazole groups but can also interact with sulphydryl groups and electrostatic anionic sites (Friedberg, 1974). Masking the sulphydryl groups with NEM had a small effect in reducing binding of the Zn to the serum protein. However, this effect could be an indirect one due to interference with access of Zn to other sites. Various ions, Cd’+, CU*+, Ca2+ and La3+, had no detectable effects on the binding of Zn to the serum proteins. The Cd*+ and Cu2+ would be expected to bind to sulphydryl groups while all the ions may interact with anionic sites. It would, thus, appear that the Zn binding sites are quite specific for this metal, possibly reflecting the importance of steric factors influencing access to such sites, or the involvement of nonsulphydryl, nonelectrostatic sites such as imidazole groups. There are some interesting possible physiological consequences of the presence of such a Zn-binding protein in catfish plasma. Such binding would appear to maintain the physiologically active “free” concentrations of Zn at low levels despite the high total plasma Zn concentration. Thus, any possible toxic effects of such high concentrations, of Zn in the plasma would be muted by its binding. Such binding may provide a “sink” that could protect the tissues from the effects of potentially toxic excesses of accumulated Zn. Whether such accumulations occur commonly in such fish is unknown, and is indeed doubtful, though it could occur in contaminated
aqueous environments. Such binding of Zn to plasma proteins could also provide storage sites for Zn that is accumulated from the diet and by uptake from the water in which they live. The latter has been shown to occur in rainbow trout (Spry and Wood, 1989), and also in these channel catfish (P. J. Bentley, unpublished observations). Binding to plasma proteins could, thus, provide a trap and a store for accumulated nutritional Zn prior to its orderly distribution to essential tissue sites. Retention of Zn that is accumulated by the fish may, thus, be enhanced and reflect an adaptation to life in a dilute aqueous environment. The binding of Zn to the catfish serum protein(s) has an interesting parallel with a human familial hyperzincemia identified by Smith et al. (1976). Serum zinc concentrations up to about 5 pg/ml (not as great as in catfish but up to 5 times normal human values) were observed in members of several generations of an American family. The plasma albumin of these individuals was found to exhibit an enhanced binding of Zn (Failla et al., 1982). Vertebrate albumins may, thus, have a propensity to change their Zn binding properties genetically, a facility that has been exploited in some freshwater fish. REFERENCES Allen J. G., Masters H. B., Peet R. L., Mullins K. R., Lewis R. D., Skirrow S. Z. and Fry J. (1983) Zinc toxicity in ruminants. J. camp. Path. 93; 3631377.. Bettaer W. J.. S~rv D. J.. Cockell K. A.. Cho C. Y. and H&on J. W. (1987) The distribution of zinc and copper in plasma and erythrocyte membranes if rainbow trout (Salmo gairdner). Comp. Biochem. Physiol. 87C, 445-451.
Charlwood P. A. (1979) The relative affinity of transferrin and albumin for zinc. Biochim. biophys. Acta Ssl, 260-265. Cunnane S. C. (1988) In Zinc: Clinical and Biochemical Significance, pp. 79-81; pp. 11-12. CRC Press, Boca Raton. FL. Failla M. L., van de Veerdonk M., Morgan W. T. and Smith J. C. (1982) Characterization of zinc-binding proteins of plasma in familial hyperzincemia. J. Lab. clin. Med. 100, 943-952. Friedberg F. (1974) Effects of metal binding on protein structure. Q. Rev. Biophys. 7, l-33. Gatlin D. M. and Wilson R. P. (1983) Dietary zinc requirement of fingerling channel catfish. J. Nutr. 113,630-635. Lowry 0. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-275. Scatchard G. (1949) The attractions of proteins for small molecules and ions. Ann. N. Y. Acad. Sci. 51, 66@672. Smith J. C., Zeller J. A., Brown E. D. and Ong S. C. (1976) Elevated plasma zinc. Science 193, 496-498. Spray C. M. and Widdowson E. M. (1951) The effect of growth and development on the composition of mammals. Br. J. Nutr. 4, 332-353. Spry D. J. and Wood C. M. (1989) A kinetic method for measurement of zinc influx in uiuo in the rainbow trout, and the effects of water-borne calcium on flux rates. J. exp. Biol. 142, 425446.
Underwood E. J. (1977) In Trace Elements in Animal and Human Nutrition. 4th edn, pp. 196-242. Academic Press, London.
Camp. Eiockm. Physiof.Vol. IOOC,No. 3, pp. 491494, 1991
F’rintedin
Great Britain
A HIGH-AFFINITY ZINC-BINDING PLASMA PROTEIN CHANNEL CATFISH (ICTALURUS PUNCTATUS)
IN
P. J. BENTLEY Department of Anatomy, Physiological Sciences and Radiology, College of Veterinary Medicine, North Carolina State University, Raleigh, NC 27606, U.S.A. (Telephone: 919-829-4267) (Received 14 December 1990)
Abstract-l. Channel catfish (Ictalurus punctatus) have a remarkably high concentration of zinc (Zn) in their blood serum, about 20 pg/ml. However, compared to mammals, the concentrations of Zn in their tissues are not remarkable. The serum Zn is dialyzable against a solution containing 1 mM EDTA. 2. Following separation of serum proteins by gel-filtration most of the Zn was recovered in a fraction with the same peak volume of elution for the Zn and protein concentrations and having a molecular weight similar to bovine serum albumin. 3. Binding of Zn to such sites was not changed by Cu*+, Cd2+, Car+, or La3+. N-ethylemaleimide (NEM) appeared to decrease binding slightly. 4. Equilibrium dialysis with a Scatchard plot analysis of these data suggested that a single set of binding sites was present on the protein(s) with K,, of 2 x 10m5M. There were binding sites for 35 x 10e8 M Zn/mg protein. 5. Parallel equilibrium dialysis measurements using human, rabbit and chicken albumins indicated that the catfish serum protein(s) had a much higher affinity and binding capacity for Zn than the albumins in these species. 6. The catfish Zn serum-binding protein may be an albumin. The possible physiological significance of such a serum protein in freshwater fish is discussed.
associated with a protein fraction consistent in molecular weight with a catfish serum albumin. The Zn is bound with a much higher binding affinity, and has a greater number of binding sites than observed in mammalian and chicken albumins.
INTRODUCTION
Zinc (Zn) is an essential trace metal element (Underwood, 1977). Its deficiency results in disorders of growth, reproduction and the immune system. Such effects reflect its role in many enzymatic processes and the interactions between macromolecules (Cunnane, 1988). In mammals, normal serum concentrations of Zn are about 1 pug/ml. Most of this Zn is bound to serum proteins, especially albumins (Cunnane, 1988). Elevated concentrations of Zn occur in the serum when large dietary excesses of the mineral are consumed, and this increase is accompanied by rises in the Zn levels in various tissues, especially liver, kidneys and bone (Allen et al., 1983). Zinc has been shown to be an essential dietary constituent for normal growth in channel catfish (Gatlin and Wilson, 1983). These catfish (I. punctutus) are a freshwater species that are native to the southeastern United States, where they are extensively farmed as a source of food. Compared to mammals, they have a remarkably high concentration of Zn in their serum, over 20pg/ml (Gatlin and Wilson, 1983, and this paper). Other species of freshwater fish, including rainbow trout, lake trout, walleye and whitefish, have also been shown to have high plasma Zn concentrations (Bettger et al., 1987). Thus, high plasma or serum Zn concentrations may be a widespread phenomenon in freshwater fish. However, (this paper), tissue levels of Zn in catfish were not found to be remarkable and, indeed, they were similar to those in normal rabbits. The Zn in rainbow trout plasma is mostly bound to plasma proteins and is dialyzable (Bettger ef al., 1987). In catfish, most of this Zn (this paper) was found to be
MATERIALS AND
METHODS
Channel catfish, I. puncratus, raised for commercial purposes were obtained from several sources in North Carolina (see Table 1). Blood was collected from severed caudal vessels of fish that had been anesthetized with MS 222 (0.1%) (Sigma Chemical Co., St. Louis, MO) or had been electrically stunned. The serum was separated after clotting had occurred. Tissues for Zn analysis were collected from freshly purchased fish killed with the MS 222, using instruments that had been washed overnight in saturated EDTA and thoroughly rinsed in distilled water. They were weighed and stored at -20°C in polystyrene sample cups to await analysis. Analysis of zinc The Zn was analyzed by flame atomic absorption spectrophotometry (Perkin-Elmer, Model 2380), using ai electrodeless lamn for Zn absorbing at 213.9 nM and an air-acetylene flame. Standards were prepared from a certified Zn reference standard (Fisher Scientific, Fairlawn, NJ) and diluted with either distilled water, for serum analyses, or 0.5 N hydrochloric (HCl) acid for the tissue analyses. The HCl was “Baker-Instra-analyzed” for trace metal work (J. T. Baker Chemical Co., Phillipsburg, NJ). The tissues were ashed overnight in quartz crucibles at 520°C. The ash was dissolved in 1 N HCl and diluted to aive a final concentration of acid of 0.5 N. All glassware and the crucibles were cleaned with chromic-sulphuric acid (Fisher Scientific, Fairlawn, NJ) and thoroughly rinsed with distilled water before use. Zinc was
491
P. J. BENTLEY
492
not detected in any of the vessels used in the analyses during the normal periods of their use.. Measurement of zinc binding to serum proteins Separation of serum proteins was performed on a gelfiltration column (C16/1&, Pharmacia-LKB Biotechnology, Piscatawav. NJ) nacked with Senhadex G-200. The elution fluid cot&red l?lOmM NaCl buffered with Tris-HCI to give a pH of 7.4, and 0.02% sodium azide as a preservative. The background Zn concentration in this solution was about 0.01 pg/ml. Fluid was passed in an ascending direction up the column with the aid of a peristaltic pump. The flow rate was about 8 ml/hr and fractions were collected at 30-min intervals with the aid of a fraction collector. The spectrophotometric absorption of the eluate was continuously monitored at 280nM to indicate the presence of proteins. The Zn concentration of the fractions was measured by flame atomic absorption spectrophotometry. The protein concentration of serum and the collected fractions for dialysis was determined by the method of Lowry et al. (1951). Equilibrium dialysis of the serum proteins was performed by placing the samples (1.5 ml) in dialysis bags (MWCO 3500, Spectropor 3, Fisher Scientific, Fairlawn, NJ) and incubating them (in polyethylene containers) in a large volume (200 ml) of the stirred elution fluid for 24 hr at 5°C. Equilibration between this solution and the solution in the dialysis bag was found to occur in this time. Binding of Zn was determined from the final Zn concentration in the dialysate and its protein concentration, and the final Zn concentration in the external dialysis solution. The rabbit tissues were obtained from male New Zealand white rabbits obtained from A. C. Daley’s Kabbitry, Farmville, NC. They were fed standard rabbit chow. They were killed in an atmosphere of 100% carbon dioxide, and the tissues were collected and analyzed in the same fashion as described for the catfish. Treatment of results The graphs, including the Scatchard plots (Scatchard, 1949) w>th-their linear regression lines were plotted with a Zenith (Model 48) computer using Sigma Plot CV 3.10 software (Graphic Software Systems Inc, Sausalito, CA). RESULTS
Serum zinc concentration in channel catfish Serum was collected from catfish from several locations in North Carolina and found to have mean concentrations of Zn ranging from about 14 to 21 pg/ml (Table 1). Nearly all of this Zn was found to be dialyzable. In samples of the serum collected from the Raleigh fish, only 3.3 f 0.4% (6) of the Zn remained associated with the serum sample after 24 hr incubation in the “elution” solution at pH 7.4 containing 1 mM EDTA. Catfish serum was subjected to gel-filtration on Sephadex G-200 (see Materials and Methods). A total of 74.3 f 7.1% (8) of the Zn in the serum samples applied to the column was recovered in a fraction with a volume of elution (I’,) of 122.6 f 1.5 ml. The Zn concentration “peak” concentration was almost identical with a major protein peak detected spectrophotometrically, V, 120.6 f 1.Oml. Bovine serum albumin that was used as a standard protein for comparison had a V, of 1255ml. Another small peak of Zn concentration with a V, of 82.7 k 0.37 (in the void volume of the column) contained 0.5 + 0.13% (8) of the total Zn in the serum. Detectable levels of Zn were not found in other fractions of the eluate.
Table 1. Serum concentrations of zinc in channel catfish (Ictalurus puncmrus) obtained from different locations in North Carolina Location Zn &g/ml) Gamer* 19.1 *0.64(10) Hillsborough? 14.2 f 0.63 (6) Raleigbt 21.1 f 0.43 (8) Kernersvill@ 16.4 k 0.56 (IO) 18.7 (36) Aydeo II *Aquaculture Advisory Service (fingerlings). tB. Grubb, farm pond (mature adults). SCollege of Veterinary Medicine longterm holding tanks, (mature adults). $Blue Ridge Hatchery (fingerlings). IlCarolina Catfish Classics (mature adults, pooled sample from 36 fish). Results are as means k SE for the number of fish given in parentheses.
Tissue concentrations
of zinc in catjish
Tissues from the group of catfish from “Gamer” (Table 1) were analyzed for their Zn content (Table 2). Included, for comparison, are a series of analyses of rabbit tissues also performed in this laboratory. Generally, it can be seen that the tissue Zn concentrations in the two species were similar. Notable differences were much higher concentrations of Zn in catfish skin but lower concentrations in their bone. Apart from their high serum concentrations of Zn, the catfish tissue Zn concentration did not display any indications that they were in a hyperzincemic condition compared to the rabbits. Binding of zinc to catjish serum proteins
The observed “hyperzincemia” in catfish, as compared to mammals, could be reflecting qualitative and quantitative differences in Zn binding to the serum proteins. The concentration of serum proteins in the catfish was found to vary somewhat. Thus, in the “Garner” fish, it was found to be 6.1 + 0.30 g/100 serum (6), while in the “Hillsborough” fish, from a farm pond, it was 3.6 + 0.25 g/lOOml (6). It was noteworthy that the serum Zn concentrations in the “Hillsborough” fish were also less than in the “Garner” fish but they were, according to mammalian It, thus, criteria, still markedly hyperzincemic. seemed possible that the hyperzincemia may be reflecting a higher binding affinity for Zn by the Table 2. Tissue concentrations of zinc in channel catfish and laboratory rabbits Zn (j~g/g wet wt) Catfish* Rabbit Serum 19.1 f 0.68 (10) 1.6 + 0.074 Blood 16.8 f 0.48 (10) Skeletal muscle 8.2 f 0.53 (10) 7.1 +0.4 Skin 74.4 f 4.00 (12) 5.7 + 0.66 Liver 24.7 f 0.59 (IO) 29.1 k 1.26 Brain 9.3 * 0.22 (IO) 10.2 * 0.44 Kidney 31.5 k 2.45 (12) 22.7 + 0.76 Gill arch 25.6 k 1.21 (7) Bone 58.94 1.17(9) 115 f 6.7 Total body 29.8 f 0.77 (7) 40t *Fish from Aquaculture Advisory Service, Garner, North Carolina. tFrom Spray and Widdowson (1951). Results are as means f SE for the numbers of fish given in parentheses. There were 9 rabbits in each group of analyses.
Zinc-binding protein
493
Table 3. Zinc binding to catfish serum protein fraction* Equilibrium concentration Zinc bound of Zn in dialysis fluid &g/mg serum protein) Wmf) 0.13 2.175 0.078 0.32 4.24 f 0.148 0.70 7.77kO.153 1.8 12.62k 0.463 2.9 15.67kO.159 4.1 16.55k 0.258 8.5 18.32+ 0.548 24.0 23.56f 0.266 *For details see Materials and Methods. Results are as mean + SE for 6 measurements. Zinc Equilibrium Concentration wW
Fig. 1. Relationship of zinc bound to serum protein and the equilibrium concentration of the mineral in catfish (A), and (for serum albumins) rabbit (a), human (0) and chicken (A). Each point is the mean of 6 determinations. The standard errors were too small to display on a graph to this scale, but the values for the catfish are given in Table 3.
catfish serum proteins. The catfish (from “Ayden”) serum albumin-like fraction was, thus, collected from the gel-filtration column. It was pooled for studies of Zn binding by equilibrium dialysis (see Materials and Methods). A parallel comparison of Zn binding to commercial (Sigma Chemical Co., St. Louis, MO) preparations of rabbit, human and chicken serum albumins was also performed. The results are summarized in Fig. 1. It is readily apparent that the catfish serum protein preparation bound Zn far more readily than the mammalian or bird albumins. In addition, it had much higher binding capacity for Zn. A Scatchard plot analysis of the binding properties of the catfish serum protein to Zn was performed (Fig. 2, Table 3). It can be seen that there appeared to be a major single set of binding sites that, from the intercept with the x-axis, suggested binding sites for about 35 x lO_rM Zn/mg Protein (N) with an equilibrium dissociation constant (Kn) of 2 x 10msM. A similar plot (not shown) for the human serum albumin suggested the presence of two major sets of binding sites with N values of 3 x 10m8M Zn/mg protein and 11 x lo-*M Zn/mg protein and a K,, respectively, of 0.8 x 10m4M and 2.5 x 10m4M. 2r
Thus, as indicated in Fig. 1, the catfish serum protein has a much higher affinity for binding Zn, as well as a large binding capacity. Eflects of N-ethylemaleitnide (NEM) and various divalent ions on binding of zinc to catfish plasma proteins
In order to help describe the nature and specificity of the Zn-binding sites in the catfish serum, protein binding of Zn was measured in the presence of NEM, which irreversibly alkylates sulphydryl groups, and various other divalent ions and La3+ (Table 4). It can be seen that the presence of the various ions made little or no difference to the Zn binding. N-ethylemaleimide appeared to decrease binding slightly, by about 25%. DISCUSSION
Compared to mammals, channel catfish, and several other species of freshwater fish, exhibit a natural “hyperzincemia” as defined by relatively high levels of Zn in their serum as compared to mammals (Gatlin and Wilson, 1983; Bettger et al., 1987). This condition, however, appears to be a normal physiological one in such fish. Tissue concentrations of Zn, thus, do not indicate that channel catfish have excesses of this essential trace metal element. The high serum levels of Zn can be accounted for by the presence of a serum protein, or proteins, that, compared to mammals, have a remarkably high binding affinity for zinc. This binding protein(s) elutes as a single component by gel-filtration and appears to have a molecular weight slightly higher than bovine serum albumin (67 kDa), which was used as a standard protein in such separations. Table 4. Effects of N-ethylemaleimide (NEM), Cd2+, Cu’+, Ca2+ and La’+ on bindine of zinc to catfish serum orotein Conditions NEM, IO-‘M
0 0
0
10 r,
20 Zn Bound x IO’ Mlmg
0
30
40
Protein
Fig. 2. A Scatchard plot of binding of zinc to catfish serum protein(s) compared to the equilibrium zinc concentration. For details see Materials and Methods. Each point is the mean for 6 determinations using the data given in Table 3.
Cd’+, Cd*+, Cu2+, Cu’+, Ca*+, La)+.
lo-‘M lo-‘M lo-‘M W4 M lo-* M 102M
Equilibrium Zn cont. (rgiml) 0.39 3.0 0.35 0.39 0.36 0.39 0.38 0.40
Zn binding Control’ (pg Zn/mg protein) 3.31 + 0.106 11.49 f 0.200 4.55 + 0.217 4.16 k 0.125 5.07 + 0.183 3.81 Ifr0.039 4.85 + 0.150 4.44 f 0.078
4.49 15.74 4.51 4.89 4.61 4.89 4.60 4.98
*Equilibrium concentrations of Zn are not precisely reproducible in such dialysis experiments. Thus, the control binding values are predicted ones from bracheting between closest control values (see Table 3) assuming a linear relationship. Results are as mean + SE for 6 measurements.
494
P. J. BENTLEY
The present observations only allow speculation as to the precise nature of the catfish serum Zn-binding protein. The peaks of protein concentration activity, determined spectrophotometrically, and the Zn concentration corresponded, but it remains possible that other proteins of similar molecular weight were present in the collected gel-filtration fractions. Such proteins could be contributing to the observed metal binding. One possibility is transferrin (molecular weight 78 kDa) though human transferrin has been found to have an even lower binding affinity for Zn than human serum albumin (Charlwood, 1979). A Scatchard plot analysis (Scatchard, 1949) of the binding information suggested that there was one major type of binding site on the catfish serum protein, and this had a KD for Zn binding of 2 x lo-‘M and a number of binding sites (N) of 35 x lo-‘M/mg protein. For comparison, values were determined similarly for human serum albumin, which is the principal Zn-binding protein in human plasma. Two sets of sites were apparent, one with a KD of 0.8 x 10m4M and N of 3 x 10-8M/mg protein and the other, which had a lower affinity, a KD of 2.5 x 10e4 M and N of 11 x lOwEM/mg protein. The catfish serum protein, thus, had a substantially higher binding affinity and capacity for Zn than the human serum albumin. The suggested presence of a single type of binding site on the catfish serum protein favors the contribution of a single protein and, as the predominant one expected in this fraction is an albumin, the binding protein is most likely a catfish albumin. Zinc is known to have a special propensity to bind to imidazole groups but can also interact with sulphydryl groups and electrostatic anionic sites (Friedberg, 1974). Masking the sulphydryl groups with NEM had a small effect in reducing binding of the Zn to the serum protein. However, this effect could be an indirect one due to interference with access of Zn to other sites. Various ions, Cd’+, CU*+, Ca2+ and La3+, had no detectable effects on the binding of Zn to the serum proteins. The Cd*+ and Cu2+ would be expected to bind to sulphydryl groups while all the ions may interact with anionic sites. It would, thus, appear that the Zn binding sites are quite specific for this metal, possibly reflecting the importance of steric factors influencing access to such sites, or the involvement of nonsulphydryl, nonelectrostatic sites such as imidazole groups. There are some interesting possible physiological consequences of the presence of such a Zn-binding protein in catfish plasma. Such binding would appear to maintain the physiologically active “free” concentrations of Zn at low levels despite the high total plasma Zn concentration. Thus, any possible toxic effects of such high concentrations, of Zn in the plasma would be muted by its binding. Such binding may provide a “sink” that could protect the tissues from the effects of potentially toxic excesses of accumulated Zn. Whether such accumulations occur commonly in such fish is unknown, and is indeed doubtful, though it could occur in contaminated
aqueous environments. Such binding of Zn to plasma proteins could also provide storage sites for Zn that is accumulated from the diet and by uptake from the water in which they live. The latter has been shown to occur in rainbow trout (Spry and Wood, 1989), and also in these channel catfish (P. J. Bentley, unpublished observations). Binding to plasma proteins could, thus, provide a trap and a store for accumulated nutritional Zn prior to its orderly distribution to essential tissue sites. Retention of Zn that is accumulated by the fish may, thus, be enhanced and reflect an adaptation to life in a dilute aqueous environment. The binding of Zn to the catfish serum protein(s) has an interesting parallel with a human familial hyperzincemia identified by Smith et al. (1976). Serum zinc concentrations up to about 5 pg/ml (not as great as in catfish but up to 5 times normal human values) were observed in members of several generations of an American family. The plasma albumin of these individuals was found to exhibit an enhanced binding of Zn (Failla et al., 1982). Vertebrate albumins may, thus, have a propensity to change their Zn binding properties genetically, a facility that has been exploited in some freshwater fish. REFERENCES Allen J. G., Masters H. B., Peet R. L., Mullins K. R., Lewis R. D., Skirrow S. Z. and Fry J. (1983) Zinc toxicity in ruminants. J. camp. Path. 93; 3631377.. Bettaer W. J.. S~rv D. J.. Cockell K. A.. Cho C. Y. and H&on J. W. (1987) The distribution of zinc and copper in plasma and erythrocyte membranes if rainbow trout (Salmo gairdner). Comp. Biochem. Physiol. 87C, 445-451.
Charlwood P. A. (1979) The relative affinity of transferrin and albumin for zinc. Biochim. biophys. Acta Ssl, 260-265. Cunnane S. C. (1988) In Zinc: Clinical and Biochemical Significance, pp. 79-81; pp. 11-12. CRC Press, Boca Raton. FL. Failla M. L., van de Veerdonk M., Morgan W. T. and Smith J. C. (1982) Characterization of zinc-binding proteins of plasma in familial hyperzincemia. J. Lab. clin. Med. 100, 943-952. Friedberg F. (1974) Effects of metal binding on protein structure. Q. Rev. Biophys. 7, l-33. Gatlin D. M. and Wilson R. P. (1983) Dietary zinc requirement of fingerling channel catfish. J. Nutr. 113,630-635. Lowry 0. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-275. Scatchard G. (1949) The attractions of proteins for small molecules and ions. Ann. N. Y. Acad. Sci. 51, 66@672. Smith J. C., Zeller J. A., Brown E. D. and Ong S. C. (1976) Elevated plasma zinc. Science 193, 496-498. Spray C. M. and Widdowson E. M. (1951) The effect of growth and development on the composition of mammals. Br. J. Nutr. 4, 332-353. Spry D. J. and Wood C. M. (1989) A kinetic method for measurement of zinc influx in uiuo in the rainbow trout, and the effects of water-borne calcium on flux rates. J. exp. Biol. 142, 425446.
Underwood E. J. (1977) In Trace Elements in Animal and Human Nutrition. 4th edn, pp. 196-242. Academic Press, London.