Info

Note: Range of average values is from ref. 1.

adaptability to copper intake is supported, at least indirectly by other research, not everyone will agree that the recommended intake might not have some negative consequences, long term. It is true that actual intakes of copper by Americans and Europeans are mostly within the recommended range and that overt copper deficiency is rare. Overt deficiency (with severe developmental consequences) occurs when there is a genetic defect in the functioning of a copper transporter (ATP7A), resulting in Menkes disease or the milder Occipital Horn Syndrome (see Sections 1.2. and 1.5.). Marginal deficiency is another question, and until we know how to define this better, some concerns about adequacy of intake will continue.

1.2. Copper Absorption

In humans and other mammals where it has been tested, copper is absorbed primarily or exclusively by the small intestine. (There may be some absorption in the stomach, but the available data are contradictory [1].) The percentage of copper absorbed is inversely related to dose (1,5), particularly in the extremes; at low doses, most of the copper in the diet is absorbed; at high doses the percentage is much lower (only 10% in the case of rats given 10 times the normal intake [1]). In humans with normal intakes, 55-75% is absorbed; in rats, the average percentage is somewhat less (30-50%). These values are based on the appearance of radioactive or heavy isotopes of copper in blood and tissues. It should, however, be pointed out that in addition to the approx 1 mg of copper received daily in the diet, a great deal more copper enters the digestive tract daily from endogenous organs and tissues, and most of that is reabsorbed as well (1,5). In humans, saliva (0.33-0.45 mg), gastric juices (1 mg), the bile (2.4 mg), as well as pancreatic

(0.4-1.3 mg) and duodenal fluids (0.4-2.2 mg) daily contribute an average of between 4 and 7.5 mg of additional copper to the contents of the digestive tract, and all but about 1 mg of that copper is reabsorbed. The 1 mg or so not reabsorbed represents most of copper excreted daily by human adults. Very little copper exits the body by other routes (only 30-60 |g/d in the urine of adult humans) (1,5). This means that copper is very actively recycled between the digestive tract and body fluids and tissues (particularly the liver) and that dietary copper is only a small proportion of the total absorbed daily. Copper secreted via the bile is the least readily reabsorbed, thus, the bile is the main route for copper excretion for the mammal (see Section 1.5.).

As concerns the details of intestinal copper absorption, the mechanisms by which it enters the cells of the intestinal mucosa and crosses into the interstitial fluid and blood are still not well understood. However, important recent progress has been made in identifying transporters and some other proteins that are likely to be involved. Earlier studies, mainly with rodents, indicated that the uptake of Cu(II) across the brush border involved both a nonenergy-dependent saturable carrier, active at lower copper concentrations, and diffusion—at higher concentrations (1). At the basolateral membrane of the cell, it was surmised that transfer into the blood and interstitial fluid was energy dependent and more rate limiting for overall absorption, there being conditions in which much copper enters the enterocyte but does not necessarily go further (and accumulates). However, at high concentrations, there is increased uptake as well as overall absorption. For this, additional carrier systems and/or diffusion may come into play, mediating additional transfer of copper across the basolateral/serosal membrane. Thus, even though a decreased percentage of copper is absorbed at high doses, actual absorption is greater at higher than at lower doses. It may be that some of the carriers (or transport mechanisms) recruited at higher copper doses are those that specialize in the transport of other transition metals, such as Zn(II) and Fe(II/III): High doses of the latter metal ions can inhibit copper absorption, and vice versa (1,20,21), particularly at the basolateral end of the enterocyte. Rates of basolateral copper transport also vary with physiological condition, increasing with pregnancy and in cancer (1,22), and decreasing (at least in female rats) upon repeated treatment with estrogen (22). This implies that expression and/or deployment of copper transporters/transport systems can vary and is regulated. In the conditions cited, retention of copper by the enterocyte varies inversely with overall transport, accumulating when the serosal transfer system is less active, and not accumulating when it is. Copper retained in the mucosal cells will be released back into the digestive tract when these cells migrate to the tips of the villi and are sloughed off, thus contributing further to the pool of copper available for reabsorption and excretion.

Whether and how nutritional copper status regulates copper absorption by the enterocyte has not been examined much at the cellular level. The existing data suggest the possibility of a biphasic response to copper dose. With the polarized Caco2 cell monolayer model, it was found that pretreat-ment with excess copper enhanced uptake and overall transport of 1 |M 64Cu (23), a response that is opposite to what would be expected for homeostasis. We have recently confirmed these results (Zerounian and Linder, unpublished) and examined, as well, the effects of copper deprivation. Cells depleted of copper by three different chelators responded by markedly increasing their uptake and overall transport of copper (24). Others have reported that copper deficiency or excess did not influence the expression of CTR1 (25), a transporter thought to be involved at least in some aspect of copper absorption (see later in this section).

Within the mucosal cell, most newly absorbed copper (about 80%) is retained in the cytosol, and most of that is bound to metallothioneins and/or proteins of similar size (1). Metallothioneins are small cysteine-rich proteins that store copper and other divalent metal ions (notably those of Zn, Mn, and Cd) (see refs. 1-3). As the affinity of these proteins for Cu(II) is higher than for most other abundant metal ions [notably Zn(II)], the incoming copper will displace these ions. If metallothionein concentrations in the mucosal cells are high (as when induced with high intakes of zinc), the binding of copper to metallothionein will interfere with its transfer across the serosal surface. Thus, high

Fig. 1. Possible mechanisms of copper uptake and transport in the intestinal cell (see text for details). Starting at the brush border (top of the cell model), copper may enter through CTR1 and DMT1. Copper is picked up on the inside of the brush-border membrane by several chaperone carriers that deliver the metal ion to specific sites: HAH1/ATOX1 to MNK in the trans-Golgi network (TGN); COX17 to the mitochondria; CCS to superoxide dismutase (SOD). Glutathione (GSH) and metallothionein (MT) may also pick up the copper directly, possibly to shuttle it to the basolateral end of the enterocyte. There, it might be released to albumin and transcuprein in the circulation (perhaps via CTR1). Alternatively, copper arriving at the ATP7A (MNK protein in the TGN) may be pumped into developing vesicles for exocytic release.

Fig. 1. Possible mechanisms of copper uptake and transport in the intestinal cell (see text for details). Starting at the brush border (top of the cell model), copper may enter through CTR1 and DMT1. Copper is picked up on the inside of the brush-border membrane by several chaperone carriers that deliver the metal ion to specific sites: HAH1/ATOX1 to MNK in the trans-Golgi network (TGN); COX17 to the mitochondria; CCS to superoxide dismutase (SOD). Glutathione (GSH) and metallothionein (MT) may also pick up the copper directly, possibly to shuttle it to the basolateral end of the enterocyte. There, it might be released to albumin and transcuprein in the circulation (perhaps via CTR1). Alternatively, copper arriving at the ATP7A (MNK protein in the TGN) may be pumped into developing vesicles for exocytic release.

doses of zinc are used to inhibit intestinal copper absorption in Wilson disease, where accumulation of excess copper is a problem (see Section 1.5.).

With regard to the potential transporters and carrier systems for intestinal absorption, several genes and gene products have emerged in the last several years that are likely to participate, although many functional details still need to be verified or established. Most of these mammalian copper transport genes and their products were identified by using sequence information from analogous genes in yeasts (8,12,26-28). CTR1, cloned in humans (29) and more recently in mice (25,30), is homologous to (and can substitute for) yeast ctr1 in complementation experiments (29). The corresponding ubiquitously expressed 190-amino-acid human transmembrane protein (about 22 kDa, with 3 transmembrane regions, and a hydrophilic, histidine and methionine-rich N-terminal domain external to the cell) is a candidate for the copper transporter in the brush border of intestinal cells. However, it seems likely that it is also in the basolateral membrane, because copper can and will enter the enterocyte from the blood. Transfection experiments have confirmed that hCTR1 promotes copper uptake into mammalian cells (30-32) [in addition to yeast (29)]. In yeast, it appears that uptake is coupled with efflux of two K+ ions (8,33) and that the reduced form [Cu(I)] is transported. CTR1 might thus provide for the facilitated diffusion of copper across the brush border, even at low copper concentrations. (It might even work in both directions, also facilitating release of excess copper into the gastrointestinal tract. We know that copper can flow through the intestinal wall in both directions, although, normally. the net direction is toward the blood [1].) There are two major mRNA transcripts (2 and 5.5 kb) expressed in human and mouse intestine (and most other tissues), which might target the transporter to different parts of the cell. In rats, however, one of the forms greatly predominates, the other predominating in the liver and hypothalamus (25). Expression of the mammalian homolog (hCTR2) of another (low-affinity) yeast transporter (ctr2) has also been detected at the mRNA level in most tissues, although its greatest expression was in placenta, and there was very little in liver and intestine (8,29). However, it was ineffective in complementing copper transport-deficient yeasts.

Another candidate transporter in the brush border that must be considered is DMT1/Nramp2, a divalent metal transporter that clearly carries Fe(II), Zn(II) and Mn(II) (34-36). Ubiquitously expressed in tissues and present in the intestinal brush border, its transport of metal ions is proton coupled and dependent on the membrane potential, as shown by the large inward current evoked in DMT1/DCT1-transfected oocytes upon their exposure to Fe(II), at pH 5.5, and loss of most of this current at pH 7.5 (34). Cu(II) and the ions of Zn, Mn, Cd, and Co were also effective in evoking current. Whether and/or when this large (561 amino acid) protein, with 12 transmembrane segments, is actually involved in normal copper transport has not yet been fully established. A role in iron absorption, as well as homeostatic regulation of its expression by iron within the intestine, has been well established (37,38); although the 3'UTR of two forms of the mRNA (39) contain an iron-responsive element (IRE), regulation by iron is mainly through changes in rates of transcription (39,40). In the Caco2 cell intestinal model system, we found no competition between Cu(II) and Fe(II) or Zn(II) for uptake of either 64Cu(II) or 59Fe(II) at equimolar concentrations (41). However, preliminary evidence indicates that iron deficiency enhances not just iron but also copper uptake (N.R. Zerounian and M.C. Linder, unpublished) and that inactivation of DMT1 mRNA by antisense oligomers reduces uptake of both metal ions in polarized Caco2 cells (42). The results suggest that DMT1 can mediate copper uptake when the metal ion is present in sufficient concentrations and/or when other metal ions are absent.

Several other genes and gene products involved in copper transport within mammalian cells and across their cell membranes have been identified, most of which are ubiquitously expressed, including by intestinal enterocytes. Assuming that these proteins have the same kinds of function in intestinal cells as they do in other cell types, one can imagine one or more pathways by which copper can travel from the brush border to the basolateral membrane and across into the blood, during intestinal absorption. This is summarized in Fig. 1. Along with the putative membrane transporters already described (CTR1, DMT1), one of the major discoveries initially made in yeast and then verified for mammalian cells is that there is a battery of what are now called "copper chaperone proteins" that carry copper to specific intracellular sites and enzymes. This fits with (and further substantiates) the concept that copper ions are prevented from being free in solution by being bound tightly to specific proteins and handed directly from one protein to another. (The same concept has emerged from studies of copper transport proteins in the blood plasma [see Section 1.3.].) Thus, copper entering cells will be carried by HAH1/ATOX1 (corresponding to yeast Atx1) to the trans-Golgi network (TGN) (or related vesicles), delivering copper to the P-type ATPases located there (8,43,44). In the case of the enterocyte, it would be ATP7A or MNK (the protein defective in Menkes disease); in the liver, it would be ATP7B or WND (the protein defective in Wilson disease). MNK and WND are thought to then transfer the copper into the TGN, for insertion into copper-dependent proteins that are secreted and for release of other forms of copper from specific cells. In the case of the enterocyte, we can imagine this exocytotic, energy-dependent system to be important for release of copper to the blood via exocytosis (Fig. 1). (In the case of the hepatocyte, WND is involved in the release of copper into bile canaliculi [see Sections 1.4. and 1.5.].)

Another chaperone, CCS (Fig. 1), delivers copper to the Cu/Zn superoxide dismutase (SOD) in the cytoplasm (8,45,46), which protects cells against superoxide radicals (see Section 2.1.2.). A third, COX17, takes copper to the mitochondria (8,47,48), where it is required for cytochrome-c oxidase, the terminal enzyme in respiration. Mediating insertion of copper into the latter enzyme may be the mammalian homologs of Sco1 and Sco2, which, in yeast, are necessary for assembly of this complex molecule (8). Some of the entering copper will also associate with cytosolic metallothionein(s) (MTs) (Fig. 1). As already related, radioisotope studies have shown that as it enters the enterocyte, most of the copper is in the cytoplasm. Most of that is with protein(s) of the size of MT (apparent molecular weight [Mr] about 10,000, actual molecular weight about 6000), and traces with something of the size of SOD (about 35 kDa) (1,49,50). [Little or none of the radioactive copper (or of copper measured directly) has been found to elute with components of the size of amino acids (1,49-52).] Because the chaperones Atx1 and Cox17 (at least in yeast) are almost the same size as MT (73 and 69 amino acids, respectively, vs 61 for MT), it seems likely that some of the radioactive copper that was thought to be with MT is actually attached to these chaperones. A portion of the newly entered radioactivity in the enterocytes is also associated with larger organelles and vesicles.

Based on this system of copper transporters/chaperones, one might postulate that (Fig. 1), after crossing the brush border of the enterocyte, most of copper is normally shuttled to the TGN and into its channels through delivery of the copper by ATOX1/HAH1 to the MNK protein (ATP7A). Energy is required for pumping the copper into the TGN channels, from where it may be packaged into vesicles and migrate to the basolateral membrane, for exocytosis. (The form of copper that would be delivered is unknown.) This could be the major pathway for overall intestinal absorption in the normal state. However, one can also speculate that alternative pathways exist involving cytosolic chap-erones (perhaps still to be identified) that might carry copper to CTR1 in the basolateral membrane, for transfer to the blood along a chemical gradient. In addition, the delivery of copper to specific enzymes and organelles is not a "one-way street." Thus, either the same or as yet unknown chaper-ones must carry intracellular copper back to plasma membrane transporters, for exit from both ends of the enterocyte. (Turnover of intracellular, copper-dependent proteins is also occurring and will free copper that must be either reincorporated or released from the cells.)

Still another possibility to be considered (Fig. 1) is that glutathione (GSH) plays the role of a general chaperone for copper ions (7) and delivers it to CTR1 in the plasma membranes, but this has been difficult to document and remains to be further explored. Clearly, in vitro, GSH will reduce and bind Cu(I) as well as deliver it to MT and some copper-dependent apoenzymes, like SOD and hemocyanin (7). Cellular GSH levels are high (in the millimolar range) and inversely related to cellular copper concentrations. In a particular line of hepatoma cells (HAC) (53,54), fractionation (by size-exclusion chromatography on Superose 12B) of cytosol after 67Cu treatment showed that, initially, radioactivity associated with a Mr 4200 peak that eluted in the same position as a mixture of Cu(II) and GSH. Moreover, labeling of this component declined fairly rapidly and coincided with the increased labeling of a component of the Mr of metallothionein (54). (SOD labeling occurred later.) The inverse relationship between Cu and GSH levels in cells could reflect a propensity for Cu(II) to oxidize GSH, and indeed, excess copper promotes an increase in Se-dependent GSH peroxidase. However, as already described, most previous fractionations of intestinal cytoplasm did not find newly absorbed radiocopper with components in the size range of GSH-Cu (a tripeptide complex) or with a component of Mr of 4200 (using Sephadex G75). Also, Superose 12B is not a very good medium for separating components unless they are very different in shape and size. The column itself was equilibrated with buffer purged of air, but no evidence was provided that anaerobic fractionation made a difference, and the cell extract was certainly not anaerobic prior to its application. The radiocopper in the proposed GSH complex was most highly labeled 1 h after 67Cu administration and remained at that level for several hours thereafter, suggesting considerable stability. Clearly, this needs to be explored further. Most recently, observations that copper binds to chaperones after it has entered by a variety of means (that might involve different transporters) points to the possibility that recognition (and thus direct binding) of the chaperone to the membrane transporter might not be required (55). If so, perhaps GSH mediation is involved in that step; if not, one would expect that the copper transporters would have sites for chaperone binding (perhaps different ones for different chaperones); this remains to be examined. In any event, because there are specific chaperones, GSH may not normally be needed directly for copper distribution within the cell. However, it might be needed to restore the abilities of copper binding proteins (with thiol groups) to bind their copper; if the chaperones are not available for some reason, it might play a backup role. It might also provide electrons for reduction of copper during transport; in addition, it might even play the role of a generalized copper chaperone, for proteins that have no specific chaperone to provide them with the element.

Either way, another of the emerging themes in copper physiology (and also other areas of metabolism) is redundancy of function. Thus, the absence of a particular (and seemingly critical) transport protein or transporter is usually not lethal, and the phenotype can even be mild. For example, the knockout of ATOX1 (56) resulted in growth retardation and failure to thrive after birth, but development continued to some degree (intestinal absorption was probably not completely absent). (ATOX1 was particularly important to placental transfer [56].) Similarly, although intestinal copper absorption is greatly diminished in Menkes disease, it is not zero, and the difficulty in postnatal survival with Menkes is, in part, the result of gestational/developmental abnormalities arising from defective copper transport in other tissues (57).

1.3. Transport of Copper to Tissues and Its Uptake

After copper emerges on the "blood" side of the enterocyte, it immediately binds to proteins of the interstitial fluid and portal blood plasma. Two proteins with specific, high-affinity copper binding sites have been identified. One of these is albumin, which binds copper with the help of the the three amino acids at its N-terminus (including a histidine in the third position). High-affinity copper binding has been demonstrated for human (58,59) and bovine (58,60) albumin. The histidine near the N-terminus is missing in the albumins of some other vertebrates, including the dog and pig (1). This does not eliminate binding but lowers copper affinity about 10-fold (60). Albumin is, by far, the most abundant plasma protein and it can be calculated that there is sufficient albumin in 1 mL of plasma to bind 30 |g of Cu! In actuality, only about 100-150 ng/mL (or 10-12% of the total plasma copper) are bound to this protein at any time (10,61,62). At least in rats and humans, most of the rest is associated with two other proteins, ceruloplasmin, and a macroglobulin, transcuprein, first identified in rats (10,63,64). Together, transcuprein and albumin represent the bulk of the exchangeable copper pool in blood plasma (and presumably interstitial) fluid (1,5,9,10). Ceruloplasmin is not part of this pool, as its copper is not exchangeable or dialyzable. It also seems unlikely that histidine (or other amino acids) participate as Cu-amino acid complexes. There is no firm evidence that copper is bound to histidine or other amino acids in vivo, although it may exist as a trinuclear complex with albumin-copper (59). Histidine (100 |M) does accelarate the rate at which copper comes off albumin and transcuprein (1,9), but release is still very slow. Minutes after ingestion or injection of radioactive copper, virtually all of it is associated with macroglobulin and albumin in the plasma (Fig. 2A), and none is detected in the low-molecular-weight fractions where amino acids and small peptides elute in size-exclusion chromatography. (The same is true when plasma is labeled by direct addition of radioactive copper ions [63].) So soon after administration, none of the radioactive copper is with ceruloplasmin (Fig. 2A). This contrasts with the distribution of nonradioactive copper, the bulk of which, at any time, is with ceruloplasmin (about 70%) (61,62). Thus, about 12% each of the total copper is with albumin and transcuprein, and traces are with a variety of other enzymes, clotting factors, and some low-molecular-weight peptides (Table 2). As already indicated, off-rates for copper from transcuprein

Fig. 2. Distribution of radioactive copper and total copper among plasma membrane components fractionated by size-exclusion chromatography. (A) Copper-64 radioactivity (closed circles), actual copper (ppb, divided by 20; determined by furnace atomic absorption) (x), and proteins (absorbance at 280 nm) (solid line) in fractions eluted from Sephadex G150 after application of plasma from a rat 15 min after injection of 64Cu(II)-nitrilotriacetate (about 10 ng Cu). (B) Copper concentrations (determined by furnace atomic absorption) (closed circles) and proteins (absorbance at 280 nm) (solid line) in fractions eluted from Sephadex G150 after application of serum from a patient with hemochromatosis. (In such patients, there is somewhat less ceruloplasmin.)

Fig. 2. Distribution of radioactive copper and total copper among plasma membrane components fractionated by size-exclusion chromatography. (A) Copper-64 radioactivity (closed circles), actual copper (ppb, divided by 20; determined by furnace atomic absorption) (x), and proteins (absorbance at 280 nm) (solid line) in fractions eluted from Sephadex G150 after application of plasma from a rat 15 min after injection of 64Cu(II)-nitrilotriacetate (about 10 ng Cu). (B) Copper concentrations (determined by furnace atomic absorption) (closed circles) and proteins (absorbance at 280 nm) (solid line) in fractions eluted from Sephadex G150 after application of serum from a patient with hemochromatosis. (In such patients, there is somewhat less ceruloplasmin.)

or albumin are very slow, but the two proteins rapidly exchange copper with each other (1,10,64). This mimics the pattern of events in the cell, where intracellular copper chaperones directly exchange copper with their specific protein targets (44,66,67). The ability of copper to bind to transcuprein in the face of abundant albumin (with its high-affinity copper sites) speaks to its own high affinity for copper.

Rodents express a different spectrum of macroglobulins in their blood plasma than do humans and most other mammals. Rat transcuprein appears to be arinhibitor3, a monomeric macroglobulin with a total molecular weight of about 200,000 (64). Macroglobulins are known to bind other proteins (including, but not confined to, proteases). As the apparent molecular weight of the 67Cu-labeled rat transcuprein is 270,000, and as another polypeptide of Mr 70,000 is found with radiocopper-labeled transcuprein, the copper carrier may be in the form of a complex. The identity of this 70,000 subunit and the question of whether it is required for copper transport are still being explored. arInhibitor3 and the main human macroglobulin (a2-macroglobulin) both have a highly homologous, histidine-rich region that seems like a good prospect for copper binding. Because human plasma, like rat plasma, shows binding of radioactive copper to a macroglobulin (in this case, a2-macroglobulin, as determined by immunoprecipitation; Goforth and Linder, unpublished), and analysis of actual copper in human serum confirms that 10-15% is bound to a large protein (61,62) (Fig. 2B), it seems likely that a2-macroglobulin plays the same role in nonrodent mammals as arinhibitor3 does in rodents (i.e., that the "transcuprein" of nonrodents is a2-macroglobulin). Copper appears to bind to a2-mac-roglobulin (1) and may do so with equal affinity to rat transcuprein (68). a2-Macroglobulin has also been identified as the main plasma carrier of Zn(II), although the two metal ions do not compete in their binding (63).

Upon binding to these two proteins in the portal blood, most of the incoming copper goes directly into the cells of the liver and some also to the kidney. (Very little of it initially goes to other tissues and organs.) The liver is the most important organ for copper homeostasis. Not only is it the main initial repository of incoming copper, but it also releases a good portion of that copper back into the plasma, bound to newly synthesized ceruloplasmin. In addition, it is the source of copper excreted in the bile. Once back in the blood plasma, the copper on ceruloplasmin is available for uptake by tissues throughout the rest of the organism (9,63). Just as most of the copper on albumin and transcuprein (and thus in the exchangeable blood plasma pool) appears to be targeted to the liver and kidney, so ceruloplasmin is probably the main source of copper for other tissues, under normal circumstances. This conclusion derives from studies in which (1) tissue uptake of copper given intravenously as ceruloplasmin, or ionic copper was compared (9,69,70); (2) uptake of ionic 64Cu/67Cu was followed over time and it was found that uptake by nonhepatic tissues only occurred after the appearance of radioactive copper in plasma ceruloplasmin (63); and (3) uptake of 67Cu from ceruloplasmin and turnover of ceruloplasmin-copper were followed and found to be enhanced by copper deficiency (10). Data from heavy isotope studies in humans are consistent with the concept that plasma ceruloplasmin distributes copper to tissues (16). Two phases in the distribution of newly absorbed copper to tissues can thus be distinguished—the first being distribution to the liver (and kidney) in the period when transcuprein and albumin are carrying the radioisotope, and the second being distribution to other tissues after radiocopper is released from the liver on plasma ceruloplasmin and the radioisotope is no longer bound to transcuprein and albumin.

This scenario does not prove that, in the absence of albumin or transcuprein, for example, copper will not still first enter the liver or that, in the absence of ceruloplasmin, tissues will be generally deprived of copper. From "natural" genetic knockouts of albumin (Nagase analbuminemic rats) and ceruloplasmin (human aceruloplasminemia), it is clear that albumin is not required for copper uptake by the liver and kidney (71) and that ceruloplasmin copper is not the only form of copper available to most nonhepatic tissues (72-74). Using various cell lines, we and others have shown that copper can be taken up from ceruloplasmin as well as from albumin, transcuprein, or Cu-dihistidine (72,75-77).

Table 2

Copper-Binding Components in Human Blood Plasma

Estimated contribution to total copper content

Table 2

Copper-Binding Components in Human Blood Plasma

Estimated contribution to total copper content

Ceruloplasmin

0 0

Post a comment