Transmembrane signals and protooncogene induction evoked by carcinogenic metals and prevented by zinc.

Cd2+ provokes an immediate production of inositol trisphosphate and the release of Ca2+ from internal stores in human fibroblasts and some other mammalian cells. Ni2+, Co2+, Fe2+, and Mn2+ evoke the release of stored Ca2+, but are less potent than Cd2+ (apparent K0.5 = 40 nM). Zn2+ and Cu2+ competitively inhibit Ca2+ release evoked by Cd2+ without affecting Ca2+ release by hormones such as bradykinin. Zn2+ has the same apparent Ki value (80-90 nM) towards the five agonist metals, which suggests that the metals interact with the same site. Many other divalent cations neither released stored Ca2+ nor affected Cd(2+)-evoked Ca2+ release. The agonist metals appear to activate phospholipase C via a G protein rather than a tyrosine kinase. The production of reactive oxygen species is probably not involved in Ca2+ release by the metals. Cd2+ and other stimuli that raise cytosolic-free Ca2+ induce cyclic (AMP) production, apparently by activating a calmodulin-dependent adenylyl cyclase. We suggest that an orphan receptor mediates the hormonelike responses to Cd2+ and the other agonist metals. The receptor is referred to as an orphan because its physiological stimulus is unknown. Growth of the fibroblasts in high Zn2+ desensitizes them to the five agonist metals without affecting Ca2+ release by bradykinin or histamine. A several hour incubation in culture medium with normal Zn2+ fully restores responsiveness to the five active metals. Growth in high Zn2+ appears to repress the synthesis of the putative orphan receptor because inhibitors of RNA or protein synthesis, or asparagine-linked glycosylation, prevented the restoration of metal responsiveness.(ABSTRACT TRUNCATED AT 250 WORDS)


Introduction
Cadmium and nickel are modern environmental contaminants that are toxic and carcinogenic (1)(2)(3). Industrial exposure, food, and cigarette smoking are the major sources of body cadmium and nickel (2). Cd2+ in whole blood is 5 to 15 times higher in smokers than in nonsmokers in nonoccupationally exposed adults (4). Cadmium avidly binds to polythiol groups in proteins such as metallothionein as well as zinc sites in metalloenzymes and transcription factors (5)(6)(7)(8). Although the sub-stitution of Cd2+ for Zn2+ in metalloenzymes and DNA-binding proteins may produce a functional enzyme, Cd2+ has no known biological role and is regarded as a xenobiotic (1,5,6). The functions of nickel are largely confined to enzyme systems of primordial organisms and their close relatives (9). Ni + is a cofactor of three bacterial enzymes-hydrogenases, CO dehydrogenase, and methyl-CoM reductase-as well as bacterial and plant urease (9).

Carcinogenicity of Cadmium and Nickel
Cadmium and nickel are carcinogenic in laboratory animals (1)(2)(3). Occupational exposure to nickel predisposes workers to lung and nasal cancer (1). Exposure of rats to an aerosol containing 25 pg/m3 CdCl2 produced a 50% incidence of lung tumors (10). A single subcutaneous injection of 40 pmole/kg CdCl2 in rats produced a high incidence of Leydig cell adenomas in the testes, prostatic neoplasia, and sarcomas at site of injection (11). Oral administration of CdCl2 to rats also potently induced tumors in the prostate, testes, and the hematopoietic system (12). Cadmium and nickel compounds are inactive or weakly active in gene mutation assays (2,13,14). Therefore, epigenetic mechanisms probably play a significant role in the carcinogenicity of Cd2' and Ni2+, although the mechanisms are not well understood.
In vitro treatment of fibroblasts or prostatic epithelial cells with CdCl2 produced transformed cell lines that are tumorigenic (15,16). Cultured skin fibroblasts from Indian muntjac are highly sensitive to the toxic effects of Cd2+ (17). Long-term exposure to low levels of Cd2+ produced transformed muntjac cells with normal karyotypes that were 58-fold more resistant to Cd2+ than the parental cells (17). The development of resistance to Cd2+ apparently occurs concurrently with transformation.
Environmental Health Perspectives is coupled to an enzyme, phospholipase C (PLC), via an heterotrimeric GTP-binding protein composed of subunits a, f, and y; b) two second messengers, inositol trisphosphate (OP3( and diacylglycerol (DAG) are produced simultaneously by the hydrolysis of phosphatidylinositol bisphosphate (PIP2) when Cd2+ binds to a Zn2+ site in the external domain of the receptor; c) IP3 opens an intracellular Ca channel which releases Ca from the endoplasmic reticulum; and d) DAG activates protein kinase C (PKC) which phosphorylates an actin crosslinking protein called myristoylated alanine-rich C-kinase substrate (MARCKS) and translocates from the cytoplasm to the plasma membrane.
The diagram depicts the activation of Ca/calmodulin (CaM) and mitogen-activated (MAP) protein kinases, which occur in human fibroblasts stimulated with bradykinin or epidermal growth factor (99, 100). It is not yet known whether orphan receptor stimuli affect these two kinases. Cd2, induces "immediate/early" protooncogenes (egr-1 and c-myc) (33;Smith,unpublished). The diagram shows the activation of Ca/CaM-activated adenylyl cyclase (Type adenylyl cyclase), which apparently causes the cAMP increases produced by Cd2 , calcium ionophores, and bradykinin in human dermal fibroblasts (53).

Transmembrane Signaling and Cell Transformation
A variety of mitogenic stimuli (e.g., neuropeptides and peptide growth factors) trigger receptors that activate phospholipase C (18)(19)(20)(21)(22). Phospholipase C activation concomitantly produces inositol trisphosphate (1P3), which releases stored Ca2 , and diacylglycerol (DAG), which activates protein kinase C (PKC) as illustrated in Figure 1. Heterotrimeric G proteins belonging to the Gq class regulate the ,B isoform of phospholipase C (23). Malignant transformation by several different oncogenes causes alterations in the phosphoinositide pathway (24,25). One mechanism responsible for the transformed phenotype may be persistently elevated levels of diacylglycerol (25,26). Expression of a continuously activated mutant form of the a subunit of Gq transforms NIH3T3 cells (27).
We have proposed that Cd2+ may promote tumor development by fortuitously triggering an orphan receptor (28). Figure   1 summarizes the key features of the orphan receptor hypothesis. An orphan receptor is one for which the physiological stimulus is unknown. The putative orphan receptor was provisionally called a "Cd2+ receptor" because Cd2+ was the most potent stimulus known (28), although it was realized that Cd2+ was a xenobiotic and therefore not the physiological stimulus. Whether or not the site of action of the metals is a cell surface orphan receptor, the hormone-like responses to the metals appear to be unprecedented and remarkable with respect to metal potency and specificity. Here we review the evidence that Cd + fortuitously activates an orphan receptor which raises cytosolic free Ca2+ ([Ca +]j) and activates certain protein kinases, including PKC, the target of tumor promoting phorbol esters (22). Additionally, we discuss recent observations on the role of the orphan receptor in protooncogene induction by cadmium.  Ni2+, Fe2+, Mn2+) that mobilized stored Ca2+, which was determined by assaying 45Ca2+ efflux at 10-sec intervals (28)(29)(30). Ni2+, Co2+, Fe2+, and Mn2+ are 6, 7, 17, and 380 times less potent than Cd2+ (28,30;Smith,unpublished towards each of the five "agonist" metals, (Smith, unpublished data). Therefore, the metals appear to bind to the same site. Many other divalent metals including Ca2+, Mg2+ Ba2+, Sr2+, Be2+, and Pd2+ neither release stored Ca2+ nor inhibit Ca2+ release evoked by Cd2+ (28)(29)(30). Additionally several monovalent cations had no effect on Ca2+ release.
The potency order of the "agonist" and "antagonist" metals is similar to the Irving-Williams stability order (Cu2 > Cd2 > Zn2+> Ni2+> Co2+ > Fe2+> Mn2+) for the coordination of divalent metals by compounds containing both nitrogen and oxygen donors (9). Notably, the relative potencies of the metals span a 400-fold range from Cd2+ to Mn2+ as indicated above, which is similar to the range of the stability constants of the metals for model compounds containing both amino and carboxyl groups (9).
Cd2+ and the other four active metals also evoke Ca2+ release in human neuroblastoma cells and dog coronary endothelial cells (30,31). The potency order of the metals in the neuroblastoma and endothelial cells is the same as in human dermal fibroblasts (28)(29)(30)(31)  evokes the release of stored Ca2' in human lung fibroblasts and human aortic and intestinal smooth muscle cells (31). Cd2, fails to release stored Ca2, in rat aortic myocytes, rat skin fibroblasts, and human A431 cells (31 (28,30). A prior incubation with bradykinin, which depletes the IP3-sensitive Ca2+ store, abolished the effect of Cd2+ on [Ca2+]i (28). The initial spike produced by Cd2+ is followed by a sustained [Ca2+], increase, which is dependent on external Ca2+ and probably is caused by Ca2+ influx (28,30).
The addition of 5 or 10 IM Zn2+ just prior to 1 jiM Cd2+ prevented Cd2+ from increasing [Ca2+]i without affecting the [Ca2+]i response to hormones such as bradykinin (28,33). Rinsing the cells with a physiologic salt solution fully restored the [Ca2+]i response to a subsequent addition of Cd2+ (28,33). The rapid reversibility inhibition by Zn2+ is consistent with the competitive mechanism of Zn2+ inhibition discussed above. Cd2+ and the other active metals evoke net Ca2' efflux similarly to bradykinin or angiotensin (34,35). The net Ca2+ efflux is probably caused by the plasma membrane Ca2`ATPase of human fibroblasts (28,34). The endoplasmic reticulum probably does not rapidly reaccumulate the released Ca2`because of the prolonged active state of the IP3-gated Ca2+ channel. Verbost and coworkers have shown that Cd2+ inhibits the Ca2+ ATPase of inside out red cells by binding to a site in its cytoplasmic domain (36). The fact that Cd2`produces a net decrease in the total Ca2+ content of fibroblasts indicates that Cd2+ neither inhibits the plasma membrane Ca2+ ATPase nor markedly increases Ca2+ diffusion down its several thousand fold electrochemical gradient.
Initially we considered the production of reactive oxygen species to be an attractive mechanism of Ca2+ mobilization evoked by Cd2+ and the other active metals. The following observations, however, indicated that the production of reactive oxygen is not involved in the release of stored Ca2+ by the metals. First, production of reactive oxygen species by xanthine oxidase or addition of H202 (0.11 mM) failed to release stored Ca2+ in human fibroblasts (28). Second, agents that quench reactive oxygen (superoxide dismutase, mannitol) or antioxidants (butylated hydroxyanisole or butylated hydroxytoluene) had no effect on Ca2+ release evoked by Cd2+ or Fe2+ (28). Trump and coworkers (38) have reported that [Ca2+]i increases in renal epithelial cells play a role protooncogene induction by oxidative stress. The role of reactive oxygen species in protooncogene induction by Cd2+ has not yet been addressed in human fibroblasts, although, as indicated below, protooncogene induction correlates with the Ca2+-mobilizing response to the metals. The following observations suggest that the "agonist" and "antagonist" metals trigger the release of stored Ca2+ by binding to an external site on the cell surface. First, there was no detectable lag between the addition of 0.1 pM CdCl2 and the [Ca2+]i increase as might be expected for an external site of action (28,30,39). Second, loading the cells with a heavy metal chelator (N,N,N,N'tetrakis-(2-pyridylmethyl)-ethylenediamine) did not delay the onset or decrease the extent of Cd2+-evoked 45Ca2+ efflux (28,30). Third, no intracellular Cd2+ or Zn2+ was detected with fura-2 (28,30), whose fluorescence is exquisitely sensitive to these metals (28,32,40). Fourth, as described below, a cell surface sialoprotein appears to mediate metal responsiveness (40,41). It is unlikely that Cd2+ and the other metals that evoke the release of stored Ca2+ directly activate phospholipase C. Cd2+ potently inhibits one isoform of phospholipase C and has no effect on another isoform (42). Furthermore, phospholipase C is an intracellular enzyme which does not span the plasma membrane (43).

Cadmium Activates Protein Kinase C
We have observed that a 2 min incubation of human dermal fibroblasts with 1 pM CdCI2 increased the incorporation of [32p] into myristoylated alanine-rich C-kinase 2 C°e grl -actin Figure 6. Induction of c-myc and egr-1 by Cd2+, platelet-derived growth factor (PDGF), fetal bovine serum (FBS), or phorbol myristate acetate (PMA). Human fibroblasts were incubated in a physiologic salt solution containing glucose (28) for 1 hr before adding 2 pM CdCI2, 10 ng/ml PDGF, 10% (v/v) FBS, or 0.1 pM PMA. Two hr later, total RNA was extracted and size fractionated on an agarose-formaldehyde gel. The RNA was transferred to a nylon membrane and hybridized to a c-myc cDNA probe that was [32p1 labeled by the Klenow large fragment of DNA polymerase primed with random hexamers. The membrane was stripped and reprobed for ,-actin as a control for RNA quantity and quality.  Figure 7. Key features of a putative orphan receptor that is activated by cadmium and inhibited by zinc.
substrate (MARCKS) as determined by immunoprecipitation and two-dimensional gel electrophoresis (Chen and Smith, unpublished data). The phosphorylation of MARCKS, an actin cross-linking protein (44), is a prominent and widespread response of mammalian cells to mitogenic stimuli (18,45). The increase in MARCKS phosphorylation evoked by Cd2+ was similar to that produced by phorbol myristate acetate (PMA) or bradykinin. In contrast to Cd2 , ZnCl2 did not affect MARCKS phosphorylation. These findings indicate that Cd2+ activates PKC in vivo because MARCKS is known to be a specific substrate of PKC in human fibroblasts (46,47). Presumably, PKC is activated by DAG produced by bradykinin or stimulation of the putative orphan receptor by Cd2 . PKC has a Zn2+ binding domain (48), and Zn2+ apparently modulates the interaction of the kinase with the plasma membrane (49,50). A 1-hr incubation of mouse fibroblasts with 50 pM Cd2+ had no effect on PKC activity, but it potentiated the association of PMA-activated PKC with the nucleus (51). Cd2+ does not evoke Ca2+ release in mouse fibroblasts (Swiss 3T3 cells) (Smith, unpublished data). The influence of Cd2+ on the association of PMA-activated PKC with the nucleus of mouse fibroblasts appears to be mechanistically unrelated to the activation of PKC by Cd in human fibroblasts.

Cadmium Increases Cyclic AMP Production
Bradykinin evokes cyclic AMP (cAMP) production in human fibroblasts (52). In these cells, CdCl2 (2 pM) increases cAMP production similarly to bradykinin (53). Ni2+ and Fe2+ also increased cAMP, whereas Zn2+ did not. Zn2+ blocked the effect of Cd2+, but not that of bradykinin, on cAMP production (53). Additionally, growth of the cells in high Zn2+ reversibly abolished cAMP production by Cd2+ without affecting the bradykinin response (53).
Growth in high Zn2+ appears to selectively and reversibly repress the synthesis of the orphan receptor (see the next section). The metal specificity of cAMP production is the same as the metal specificity of Ca2+ mobilization. Calcium ionophores (A23187 or ionomycin) also markedly increase cAMP production. Raising [Ca2+]i may increase cAMP in human fibroblasts via the Ca2+-calmodulin activated (Type I) adenylyl cyclase as depicted in Figure 1 (54), although the expression of the Type I cyclase has not yet been directly demonstrated in these cells. By contrast to human fibroblasts, Cd2+ has no effect on cAMP in dog coronary endothelial cells, although Cd2+ raises [Ca2+]i in these cells (53).
Other stimuli that raise [Ca2+]i in the endothelial cells also had no affect on cAMP, suggesting that they lack the Type I adenylyl cyclase.

Selective Desensitization of Fibroblasts to Cd2+ and Other Stimuli of the Putative Orphan Receptor
Growth of human fibroblasts in culture medium containing 100 1iM Zn2+ selectively and reversibly desensitizes them to Cd2+ (55). Note that the desensitization produced by growth in high Zn2+ is mechanistically distinct from competitive inhibition by Zn2+. Removing the Zn2+ immediately reverses competitive inhibition; however, a 10-hr incubation in culture medium is required to restore Cd2+ responsiveness to cells that have been grown in high Zn2+ (55). Figure 4 depicts a plausible mechanism that may account for the desensitization-repression of the synthesis of the putative orphan receptor. Growth in high Zn2+ reversibly abolished the [Ca2+] response to Cd2+ without affecting the [Ca2+]i response to bradykinin ( Figure 4) (55 ). 45Ca2+ efflux and [3H]inositol phosphate determinations also showed that growth in high Zn2+ reversibly and selectively desensitized the cells to Cd2+ (55). Growth in high Zn2+ almost abolished Cd2+-evoked production of [3H]inositol mono-, bis-, and trisphosphate and had no effect on bradykininevoked [3H]inositol phosphate production ( Figure 3) (55). Growth in high Zn2+ nearly prevented the stimulation of 45Ca2+ efflux by Cd2+ and had no effect on the stimulation of efflux by bradykinin or histamine (55). The half-time for the disappearance of Cd2+ responsiveness after adding 100 liM Zn2+ was 17 hr (55).
Inhibition of RNA or protein synthesis with actinomycin D or cycloheximide, or asparagine-linked glycosylation with tunicamycin (56) prevented the restoration of Cd2' responsiveness. Notably, tunicamycin B1 blocked the restoration of Cd2' responsiveness at 0.1 pg/ml, which only slightly affected leucine incorporation into protein (Chen and Smith, unpublished data). Brefeldin A, which reversibly and selectively disrupts Golgi stacks and prevents post-translational processing of nascent peptides (57,58), blocked the restoration of Cd2' responsiveness (41). Half-maximal inhibition of the restoration of Cd2+ responsiveness occurred at -10 ng/ml brefeldin A. The subsequent removal of brefeldin A and incubation in culture medium for 8 hr or more fully restored Cd2' responsiveness. Adding Zn2+ back to the culture medium at the time of brefeldin A removal prevented the restoration of Cd2+ responsiveness (41). These findings suggest that asparagine-linked glycosylation is required for the restoration of Cd2' responsiveness to cells that have been grown in high Zn2+.
Zn2+ transiently induces metallo thionein and heat shock proteins in mammalian cells (59)(60)(61). It seems unlikely that these proteins are responsible for desensitizing the cells to Cd2+ and the other stimuli of the putative orphan receptor (55), although this possibility has not been excluded.

A Zn2+ Site May Mediate the Hormonelike Responses
We hypothesize that Cd2+ activates a putative orphan receptor by binding to a site that is normally occupied by Zn2+. Total Zn2+ in plasma ranges from 10 to 20 pM Zn2+ in adults (62). Most of the Zn2+ is loosely bound to plasma proteins, therefore, free Zn2+ may be 0.2 to 1 pM. Because the apparent affinity of the metal site for Zn2+ is -0.1 pM, based on its Ki for metalevoked Ca2+ release, the site would be occupied by Zn2+ at the levels present in plasma. Recall that Zn2+ does not elicit hormonelike responses, but rather competitively inhibits those evoked by Cd2+ and the other metals. Therefore, we speculate that Zn2+ plays a role in the binding of the physiologic (unknown) stimulus or in receptor internalization or cycling.
There does not appear to be any precedent for the occurrence of a Zn2+ site in the external domain of a cell surface receptor, although some cytoplasmic receptors (e.g., estrogen and glucocorticoid receptors) have Zn2+ finger motifs (63). Human growth hormone contains three ligands that coordinate Zn2+, which forms a dimer Volume 102, Supplement 3, September 1994 that is stabilized by the metal (64). Additionally, Zn2+ in the 10 to 50 1iM range stabilizes the binding of human growth hormone to the human prolactin receptor (65), which contributes one of the four ligands that coordinate the metal. Although the physiological significance of the interaction between growth hormone and the prolactin receptor is unclear, the hormone-receptor "zinc sandwich" is a model system in which Zn2+ modulates the binding of a polypeptide hormone to a nonphysiologic receptor. The receptors for prolactin and growth hormone do not have a high affinity site for Zn2+ or other metals (64,65). Therefore, these receptors do not mediate the hormonelike responses to Cd2+.
Two lines of evidence suggest that the metals interact with histidyl residues. First, decreasing extracellular, not intracellular, pH induces [3H]IP3 production and Ca2+ release in the same cell types that respond to Cd2+ (31). Moreover, Zn2+ desensitizes fibroblasts and endothelial cells to a decrease in external pH as well as to Cd2+ without affecting responsiveness to Ca2+mobilizing hormones (30,55). The imidazole group of histidine (pKa 6-7) is the principle functional group with a pKa near the external pH (6.4) which half-maximally induced Ca2+ release from internal stores (30,31). Histidine is the most common amino acid in Zn2+ sites (5). Second, dye-sensitized production of singlet oxygen almost abolished Ca2+ mobilization evoked by a decrease in external pH as well as Cd2+ and the other agonist metals without affecting Ca2' release evoked by bradykinin or thrombin (66). Histidine is the most sensitive amino acid to photooxidation.
Experiments with neuraminidase have provided further evidence that the Cd2+ receptor is a cell-surface sialoprotein. Incubating the cells with neuraminidase (0.075 /ml) decreased the binding of fluorescein-WGA to the cells by -60% (Chen and Smith, unpublished data). Notably the treatment had no effect on the stimulation of 45Ca2+ efflux by 0.2 PM Cd2+, but it markedly decreased the inhibition of Cd2+-stimulated efflux by WGA (41).

Genistein Blocks the [Ca2+]i
Response to Plateletderived Growth Factor (PDGF) but Not to Cd2+ or Bradykinin The isoflavone genistein selectively inhibits tyrosine kinases, such as those of the receptors for epidermal growth factor and PDGF (74,75). The PDGF receptor kinase activates phospholipase C-yl by phosphorylating certain tyrosine residues (76,77). Hill (79,80). The relative insensitivity of the Cd2+ and bradykinin responses to genistein suggests that a G protein coupled receptor, rather than one belonging to the tyrosine kinase family, mediates Ca2+ release by Cd2+. Recent studies with herbimycin A, a tyrosine kinase inhibitor (81), and staurosporine, which potently inhibits various classes of protein kinases (82), indicate that Ca2+ release by Cd2+ is not dependent on protein kinase activity.
c-Fos and c-jun are components of the AP-1 transcription factor, which mediates nuclear events elicited by extracellular stimuli (91). Phorbol esters, growth factors, and cytokines activate PKC and induce AP-1 responsive gene expression (91). c-fos and c-jun form a stable heterodimer via a coiled-coil interaction known as a leucine zipper. Phosphorylation regulates c-jun both positively and negatively (92,93). Thus, Cd2+ may induce protooncogenes by activating certain protein kinases.

Role of the Orphan Receptor in Protooncogene Induction by Cadmium
Two paradigms have been used to evaluate whether orphan receptor stimulation contributes to protooncogene induction by cadmium. First, the pharmacologic specificity of agonist and antagonist metals was used to determine whether receptor activation correlated with protooncogene induction. Second, human fibroblasts were grown in high Zn2+ to selectively and reversibly desensitize them to orphan receptor stimuli. Both approaches have produced correlative data that support the view that Cd2+ induces protooncogene expression, at least in part, by activating the calcium-mobilizing orphan receptor.
Briefly, the following observations implicate the orphan receptor in protooncogene induction by Cd2+. Incubation of human fibroblasts with 0.2 to 2 pM CdCl2 markedly and transiently increased c-myc and egr-1 expression as determined by northern analysis (Figure 6) (33). Cd2+ -evoked c-myc expression was maximal at 2 hr and then gradually decreased to the level of control cells (33). egr-1 expression evoked by Cd2+ was also transient, but preceded the increase in c-myc by about 30 min. Other metals that stimulate the Environmental Health Perspectives putative orphan receptor, such as Co2+) Ni2+, and Fe2+, also increased c-myc and egr-1 transcripts. Zn2+ by itself had no effect on c-myc expression, but prevented c-myc induction by Cd2+ or Ni2+. Zn2+ had no effect on c-myc expression evoked by fetal bovine serum (Pijuan and Smith, unpublished data). It is noteworthy that ferrous iron stimulates the growth of human fibroblasts (Smith, unpublished data) as well as HeLa and mouse melanoma cells by a transferrin receptorindependent mechanism (94,95). Further work is needed to determine whether or not the hormonelike responses evoked by Fe2+ contribute to the mitogenic response.
Growth of the cells in high Zn2+ almost abolished the effect of Cd2 + (0.2 to 2 pM) on egr-1 and c-myc expression (Pijuan and Smith, unpublished data). Incubating the cells for 24 hr in the usual culture medium (Dulbecco's modified Eagle's medium containing 1% fetal bovine serum) completely restored the induction of egr-1 and c-myc by Cd2+. Growth of the cells in high Zn2+ had little or no effect on the induction of c-myc by platelet-derived growth factor, forskolin, or PMA. Thus, protooncogene induction correlates with the metal specificity of the orphan receptor as well as reversible manipulation of orphan receptor responsiveness by varying the Zn2+ level of the culture medium. Notably Zn2+ markedly decreases tumor induction by Cd2+ (96)(97)(98). Further work is needed to evaluate the roles of protein phosphorylation and the putative orphan receptor in protooncogene induction and mitogenic stimulation by divalent metals. Figure 7 summarizes the principal features of Ca2+ mobilization evoked by Cd2+ and the other active metals. The following are the criteria on which we base the hypothesis that Cd2+ triggers Ca2+ mobilization via an orphan recetor: a) the target which mediates Ca + release exhibits remarkable affinity and specificity for divalent metals; b) the active metals evoke an immediate and marked production of IP3 and other second messengers similarly to Ca2+ mobilizing hormones; c) the second messenger responses to the metals are cell-type specific; and d) the metals appear to act at an external site via a sialoprotein (28,30,31,39,41,55). Conclusive validation of the orphan receptor hypothesis awaits the cloning and expression of the putative receptor. The findings reviewed here indicate that two carcinogenic metals, Cd2+ and Ni2+, evoke hormonelike responses in certain mammalian cells, apparently by binding to a site on the cell surface which exhibits extraordinary metal affinity and specificity.