Effects of cadmium on nuclear protein kinase C.

Cadmium is a carcinogen whose genotoxicity is only weak. Besides its tumor-initiating capacity, cadmium may be tumor-promoting, since it interferes with several steps of cellular signal transduction. We have investigated effects of cadmium(II) on protein kinase C (PKC), which is a key enzyme in the control of cellular growth and differentiation. Tumor-promoting phorbol esters cause an activation and translocation of PKC from the cytosol to the plasma membrane and to the nucleus of mammalian cells. In mouse 3T3/10 T 1/2 fibroblasts, cadmium(II) potentiated the effect of phorbol ester on nuclear binding and activation of PKC. Furthermore, in a reconstituted system consisting of rat liver nuclei and rat brain PKC, cadmium stimulated the binding of the enzyme to a 105-kDa protein. We propose a model in which cadmium(II) substitutes for zinc(II) in the regulatory domain of PKC, thus rendering the putative protein-protein binding site exposed. Further work is required to elucidate the potential role of the nuclear PKC binding protein(s) in the control of cell proliferation.


Introduction
Cadmium is carcinogenic in experimental animals but its genotoxicity is weak. Hence, epigenetic mechanisms have to be considered. In this article we focus on effects of cadmium(II) ions on signal transduction in mammalian cells, which may shift the balance between differentiation and proliferation to the latter state. Cadmium ions have been shown to affect the following levels in cellular signal transduction: -Signal triggering at the plasma membrane: CdCl2 at concentrations starting from as low as 0.1 pM total cadmium evokes the inositol trisphosphate signal (1). *Increase in free intracellular Ca2+: CdCl2 interferes with endoplasmatic reticulum Ca2+ ATPase and Ca2+ uptake into intracellular vesicles (2). *Activation and translocation of protein kinase C (PKC): the activating effect of tumor-promoting phorbol esters is enhanced by CdCl2 (3).
*Activation of oncogene expression: CdCI2 induces the transcription of the cellular oncogenes c-jun and c-myc in rat myeloblasts at a concentration of 5 pM total cadmium (4).
Since protein phosphorylation is involved on all levels of cellular regulation and protein kinase C is a key enzyme in the control of cellular proliferation, we have to consider metal interactions with this enzyme and especially with its nuclear translocation and activity.
Protein kinase C consists of at least nine isotypes, of which the three major species (a, l, and y) have closely related structures, each containing a regulatory and a catalytic domain ( Figure 1). The conserved region C1 of the regulatory domain is essential for the binding of the second messenger diacylglycerol or a tumor promoting phorbol ester, both of which induce translocation of the enzyme to cellular structures (5). Furthermore, the conserved region C1 has two Cys6His2 consensus sequences (5), which putatively constitute zinc fingers, since the enzyme contains four zinc atoms H2N1fcc in cysteine-and histine-liganded form (6). Additional zinc ions have been shown to enhance the activity and translocation of PKC to the plasma membrane (7,8) or to cooperate with phorbol ester in translocation of the enzyme to the cytoskeleton (9)(10)(11). With cadmium ions, activating or inhibiting effects on the activity of PKC have been reported (3). Since we have been interested in the nuclear role of the enzyme, we have studied the modulation of nuclear translocation and activity of PKC by cadmium. Translocation of this enzyme to the cell nucleus has been reported previously (12)(13)(14)(15), and endogeneous location of PKC in the nucleus has been documented (16,17).

Methods
Protein kinase C was purified from bovine brain as described previously (18 culture was done by standard procedures and cell nuclei were isolated according to Malviya et al. (19). Protein kinase C was assayed according to Newton and Koshland (20), phorbol ester binding was analyzed as described by Leach and Blumberg (21), and protein was determined with bicinchoninic acid after Hill and Straka (22). Free metal ion concentrations were adjusted by a metal buffer consisting of 50 mM Tris-HCI, pH 7.4, 0.5 mM EDTA, 0.5 mM EGTA, and 2 mM nitrilotriacetic acid, and calculated by a computer program of Fabiato (23) with stability constants from Smith and Martell (24). Metal analysis was performed by electrothermal atomic absorption spectroscopy as described previously (2). SDS polyacrylamide gel electrophoresis was done following Lammli (25), Western blotting was carried out according to Masmoudi et al. (17), and the overlay assay for protein binding of PKC was performed as described by Mochley-Rosen et al. (26). The detailed procedure for the binding assay has been published previously (27).  Table 2). The nuclear pellet alone did not bind the phorbol ester, hence we could take the amount of PDBu bound as the quantity of the enzyme bound to the nuclear protein fraction. The binding of PKC was enhanced by the addition of 0.1 nM free Cd2+ or 1.0 nM free Zn2+ ions. Figure 3 shows the dependence of this enhancing effect on the concentration of free Cd2+. At the lowest concentration of 10-11 M free Cd2+, 1 jiM free Ca2+ cooperated in the effect of Cd2+ on enzyme-protein binding; but over a wide range of Cd 2+ concentrations the effect was independent of the presence of a 2+ micromolar Ca concentration. A most pertinent issue is, of course, the nature of the nuclear proteins that bind PKC. In a first approach, an "overlay assay" was performed. The nuclear protein fraction was subjected to SDS-polyacrylamide gel electrophoresis and transferred onto nitrocellulose membrane strips. These were incubated with purified PKC (Table 3). In the absence of phosphatidylserine and phorbol ester TPA, no binding was detected. In the presence of the cofactors, PKC binding to a 50-kDa protein was detected, which was further intensified 2+ when 0.1 nM free Cd + was pre2sent. In the presence of cofactors and Cd , PKC was attached also to a 105-kDa protein.

Effect of Camium and
This binding was specific to cadmium, since no binding to this band was detected in the absence of the metal (27). PKC binding to nuclear protein was assayed by measuring the association with 45 nM (3H)phorbol dibutyrate. Two units/ml of partially purified PKC from bovine brain were added to the Triton-insoluble fraction of nuclear material, and centrifuged as described elsewhere (27). The nuclear pellet as such had no detectable phorbol ester binding.

2+ 2+
The concentrations of free Zn and Cd , which were adjusted by a metal buffer, are given in the 'nme of Cadmium treatment (min) Figure 2 Cadmium potentiation of phorbol-ester induced increase in fixed nuclear PKC activity. Cells were incubated for the times given with 50 pM cadmium in a serum-free medium. Phorbol ester TPA was added to a final concentration of 100 nM 20 min prior to termination of the incubation. Nuclei were isolated and PKC was assayed as given in Methods. PKC activity extracted by EDTA is termed labile; the activity subsequently extracted by 1% Triton X-100 is termed fixed. With TPA alone in the absence of cadmium the specific activity of labile PKC was 92, that of fixed PKC 1774 pmole 32P/min/mg. olated PKC Protein kinase C purified from bovine brain (after the final elution by an imidazole gradient up to 60 mM) exhibited a ratio of 6.6-g atom zinc per mole of enzyme. Dialysis against 1 mM EGTA reduced the ratio to 4.1 zinc per PKC, as also reported by others (6,28). The zinc content was further reduced to 1.8 zinc per PKC by derivatization of histidines in the enzyme with diethylpyrocarbonate, and below 0.3 by subsequent reaction of cysteines with methylmethanethiosulphate. In conclusion, two of the four chelator-stable metal ions seem to be bound to histidines and cysteines and two are bound to cysteines only. Furthermore, even in the presence of imidazole, additional zinc ions are bound by PKC, which can be removed by the chelator EGTA.

Discussion
What mechanism of interaction of PKC with zinc and cadmium ions enhances nudear targeting? In agreement with other researchers, we assume that the two zinc ions in zinc finger motifs have high affinities to a combination of four liganding amino acid side chains (cysteines and histidines). Makowski and Sunderman have determined the dissociation constants of the zinc-finger protein transcription factor IIIA as 1.0 x 10-8 M for the zinc-protein and 2.8 x 106 M for the cadmium-protein complex (29). Hence, zinc finger sites of the regulatory domain of PKC probably will be occupied with Zn + permanently in the presence of physiologic zinc concentrations; and we propose a model for metal binding to PKC, in which the two types of zinc finger structures are formed by a Cys4 Bound PKC was detected with antibody against a partial sequence of PKC (v3-fragment). and a Cys2His2 coordination, respectively ( Figure 1). This assignment is in accordance with our analysis of ligands for chelator-stable zinc atoms in PKC. Since in addition to the two Cys6His2 sequence motifs there are four more histidines in the same regulatory domain of PKC, two of these histidines are proposed to bind the additional modulatory zinc or cadmium ions. The latter ions could enhance the stimulation by phorbol esters of the binding of PKC to the nuclear target proteins. This proposal is further supported by the presence of basic amino acid clusters in the same sequence of PKC which are related to the basic clusters found in other "nuclear targeting" sequences (30,31).

Conclusion
In conclusion, our study suggests that, in addition to genotoxic action, epigenetic mechanisms for the carcinogenicity of cadmium should be taken into account. Cadmium stimulates the nuclear targeting of PKC at the very low concentration of the free ion of 0.1 nM. Possibly, this could activate transcription factors by phosphorylation, which in turn could stimulate the expression of proliferation genes.