- Email: [email protected]
Antiserum specific for the intact isoform-3 of metallothionein
J. Biochem. Biophys. Methods 63 (2005) 43 – 52 www.elsevier.com/locate/jbbm
Antiserum specific for the intact isoform-3 of metallothionein Abigail M. Tokheima, Ian M. Armitageb,T, Bruce L. Martina,b,T a
Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, MN 55455, United States b Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 6-155 Jackson Hall, 321 Church Street, SE, Minneapolis, MN 55455, United States Received 13 February 2004; accepted 30 January 2005
Abstract The recombinant form of isoform-3 of mouse brain metallothionein (MT3) was used as an antigen to immunize rabbits and raise MT3-selective antiserum. The antiserum was essentially specific for MT3 with 100-fold greater sensitivity for MT3 compared to MT1 or MT2. Immunonblot analysis of whole mouse brain homogenates showed that MT3 was present only in the fraction retained by a 30,000-Da cut-off filter. The antiserum was used to immunoprecipitate MT3 from mouse brain extracts of Swiss Webster mice and provided evidence that MT3 was a member of a macromolecular complex of greater than 30,000 Da mass in brain. An ELISA was developed using purified, recombinant mouse brain Cd7-MT3 as the antigen and used to quantify MT3 in mouse brain extracts. The concentration of MT3 was found to be 3.0 F 0.8 Ag/ml or approximately 3.5 Ag/g mouse brain (wet weight). D 2005 Elsevier B.V. All rights reserved. Keywords: Metallothionein-3, antiserum; Metallothionein-3, neuronal isoform
Abbreviations: MT1, isoform 1 of metallothionein; MT2, isoform 2 of metallothionein; MT3, isoform-3 of metallothionein; Cd7-MT3, MT3 containing 7 gram-equivalents of cadmium. T Corresponding authors. Bruce L. Martin is to be contacted at Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 6-155 Jackson Hall, 321 Church Street, SE, Minneapolis, MN 55455. Tel.: +1 612 625 5988. Ian M. Armitage is to be contacted at Tel.: +1 612 624 5977. E-mail addresses: [email protected] (I.M. Armitage)8 [email protected] (B.L. Martin). 0165-022X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jbbm.2005.01.003
44
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
1. Introduction Metallothioneins are a family of low-molecular weight, cysteine rich, metal ion binding proteins that serve as a thermodynamic sink for biological zinc along with other transition metal ions [1]. Structural and other biophysical studies have contributed to establishing the presence of distinct metal clusters in the protein [2]. Reports also have established the kinetic lability of the metal ions bound to MT1 including the specific transfer of these metal ions between different metal forms of MT and zinc depleted metallo-enzymes/proteins [1]. Of the four MT isoforms, isoform-3 is specific to neuronal tissue. The isoforms share high sequence identity and, for the three structurally characterized isoforms (MT1, MT2, and MT3), have conserved metal ion binding clusters in two distinct domains. The primary physiological roles of MT1 and MT2 remain elusive, but metallothioneins are reported to modulate three fundamental processes: (1) the modulation of oxygen-free radicals and nitric oxide; (2) apoptosis and; (3) the binding and exchange of heavy metals. All three physiological processes are believed to play a role in several diseases and specific associations between MT and several diseases have been made [1,3]. The situation with MT1 and 2 is in contrast to that of the tissue specific isoform, MT3. MT3 was discovered in the last decade in the central nervous system and has a specific activity, the inhibition of the growth and survival of neuronal cultures. MT3 was originally designated Growth Inhibitory Factor [4–6]. Although structurally conserved, the unique biological function of MT3 indicates the existence of different signaling pathways involving this isoform. Indeed, MT3 has been specifically implicated in neuronal processes. The isoforms, MT3 and MT1, respond differently to zinc deprivation of cells in culture with only MT1 being down-regulated in these cells [7]. Transcriptional regulation of MT3 levels is differentially regulated compared to MT1 and MT2 [8–11]. These data suggest that MT3 is not involved in metal ion homeostasis, as the MT1 and MT2 isoforms seem to be. A common ligand for all MTs is zinc, an important factor in multiple neuronal processes. Also, there are numerous reports of a link between zinc and Alzheimer’s Disease. Zinc can bind to both the h-amyloid peptide and the amyloid precursor protein, and the toxicity of h-amyloid is modulated by zinc [12,13]. Low levels of Zn2+ induce neurotrophic effects of h-amyloid and high levels of Zn2+ support the neurotoxic effects. These neurotrophic and neurotoxic effects were reported to be antagonized by the presence of MT3, but not MT1 or MT2 [14]. In neurons of transgenic mice overexpressing MT3, MT3 and Zn2+ were found co-localized indicating MT3 as a possible mediator of these events [15,16]. MT3 was found to be down regulated at the mRNA level in tissue from Alzheimer’s patients [17–20], although other studies have not confirmed this result [16,21]. These data indicate a possible connection of MT3 with neuronal function and Alzheimer’s Disease although no causative relationship can be inferred. We have started experiments to characterize the neuronal function of MT3. We report here the development of a selective tool for probing the levels and interactions of MT3 in brain tissue.
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
45
2. Experimental procedures 2.1. Materials Authentic MT1 and MT2 were purchased from Sigma-Aldrich as was the pNPP substrate (Sigma 104 substrate) for the alkaline phosphatase ELISA reactions. The Amplified Alkaline Phosphatase Immun-blot kit was purchased from Bio-Rad and contained goat anti-rabbit secondary antibody; biotin; streptavidin; biotinylated alkaline phosphatase; and the substrates for alkaline phosphatase (AP color development components). The Slot Blot apparatus was purchased from Bio-Rad. The wash buffer was also purchased from Bio-Rad as a 10 stock solution (10 TBS). ELISA compatible microtiter plates were from Costar and purchased through Fisher Scientific. 2.2. Preparation of anti-MT3 antiserum Antiserum to recombinant, mouse Cd7-MT3 has been prepared using the services of the Research Animal Resources group of the University of Minnesota. As outline in Fig. 1, two female New Zealand white rabbits were immunized using purified, recombinant mouse brain Cd7-MT3 (0.35 mg) emulsified in complete Freund’s adjuvant with a preimmune serum sample obtained before immunization. Three booster injections were given using the recombinant Cd7-MT3 (0.175 mg each boost) emulsified in incomplete Freund’s adjuvant. Rabbit serum was collected at each boost and, subsequently, at biweekly intervals (12–15 ml). Initial Western immunoblots using antiserum from the first
Fig. 1. Protocol for antibody generation. Details of the protocol for generation of the anti-metallothionein-3 antiserum are given in the text.
46
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
collection showed positive reactivity with all three isoforms, but the antiserum became more selective for MT3 with the additional booster injections. Subsequently, screening of the antiserum was done using slot blots with authentic samples of MT1, MT2, and MT3 as controls. 2.3. Western immunoblots Briefly, protein samples were separated by SDS gel electrophoresis and proteins electro-transferred onto nitrocellulose (or similar) membrane in tris–glycine buffer. The membrane was blocked with 5% instant milk in tris buffered saline, pH 7.4 (TBS; 20 mM Tris, pH 7.4; and 150 mM NaCl), washed with TBS, and incubated with antiserum in TBS. After incubation with the primary antibody, the blot was washed with TBS and positive immunoreactivity detected using the Amplified Alkaline Phosphatase Immun-Blot assay kit from BioRad. This method exploits the high affinity of the interaction between streptavidin and biotin. Once washed, the blot was incubated for 1 h at room temperature with biotinylated goat anti-rabbit secondary antibody. This antibody recognizes and binds to rabbit proteins, such as the anti-MT3 antibody raised in rabbits. After incubation with secondary antibody, the blot was washed two times with TBS plus 0.05% Tween-20 (TTBS) for 15 min followed by one wash with TBS for 15 min. A solution containing a complex of streptavidin and biotinylated alkaline phosphatase was added to the blot and incubated for 1 h at room temperature. The blot was again rinsed with TBS plus Tween-20 and TBS as described. The presence of MT3 was detected by monitoring the color development after exposure to the substrate for the alkaline phosphatase-conjugated secondary antibody [22,23]. For alkaline phosphatase, the substrate used was comprised of a mix of bromochloroindolyl chloride phosphate (BCIP) and nitroblue tetrazolium (NBT). In the presence of alkaline phosphatase, BCIP was hydrolyzed with the resulting product forming an insoluble complex with NBT and stains the blot (blue to black) in the position of the alkaline phosphatase linked antibody, and therefore, the antigen. Only bands interacting with the primary antibody will show positive immunoreactivity. 2.4. Slot blot for the identification of MT3 The presence of MT3 in brain extracts was measured using Slot blots and an ELISA assay. For the Slot blots, 200-Al brain homogenates were applied to pre-cut nitrocellulose sheets (0.45 Am, Bio-Rad) using the Bio-Rad Slot-Blot apparatus. Homogenate protein was allowed to bind under gentle vacuum to the nitrocellulose supported by three sheets of Bio-Dot SF filter paper after a prewash of the nitrocellulose with 100 Al TBS. Each well was then washed with 200 Al of TBS. After washing, the membrane was removed and blocked using 5% milk in TBS for 1 h at room temperature. The milk was removed and the membrane washed 2 with TTBS for 5 min and 1 with TBS for 5 min. The MT3 primary antibody was applied and incubated overnight at 4 8C with rocking. After overnight incubation, the primary antibody was removed and the membrane washed as described. Detection was done using the reagents
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
47
of the Amplified Alkaline Phosphatase Immun-Blot Kit from Bio-Rad Laboratories as already described. 2.5. ELISA for the quantification of MT3 in brain extracts Samples of brain extract (40 Al of brain supernatants and 160 Al TBS) were added into wells of a Costar 96 well microtiter plate designed for protein binding. The plate was incubated overnight at 4 8C to maximize protein binding. A vacuum pipette was used to remove the supernatant and each well was washed three times with 0.2 ml of TBS with each wash incubated for 5 min. After the washes, 0.2 ml of the primary antibody (rabbit anti-MT3) was added to each well and incubated at room temperature for 1 h. The wash steps were repeated and 0.2 ml of biotinylated goat anti-rabbit secondary antibody was added to each well. This was incubated for 30 min at room temperature, followed by the wash steps. A solution (0.2 ml) of streptavidin with biotinylated alkaline phosphatase was added to each well and incubated for 30 min at room temperature. After the incubation and subsequent wash steps, the microtiter plate was inserted into a Molecular Devices Spectramax 340 microtiter plate reader and the plate incubated for 30 s at the reaction temperature (26 8C). A substrate solution (0.2 ml of 10 mM pNPP) was added and the reaction monitored for 30 min at 410 nm as the remaining alkaline phosphatase hydrolyzed the pNPP introduced. The amount of pNPP hydrolysis was proportional to the amount of alkaline phosphatase bound that reflects the amount of antigen present in the well. The ELISA assay of MT3 in mouse brains was accomplished in triplicate wells and compared to a standard curve established using authentic purified, recombinant mouse brain MT3 (0.30 mg/ml).
3. Results and discussion In the course of studies associated with evaluating the role of MT3 in Alzheimer’s Disease, standard Western blotting techniques were used to screen for the presence of MT3 in brains from a mouse model (Tg 2576 mouse strain) of AD. Initial attempts were made using an antibody raised in rabbits against recombinant human MT3. Human antipeptide antiserum to a peptide of a unique sequence of MT3 corresponding to residues 52–64 [20] were obtained from Dr. Paul Fraser (Center for Research in Neurodegenerative Diseases, University of Toronto) and used to probe for metallothionein-3. Unfortunately, no immunoreactivity was detected on the blot (not shown). The absence of immunoreactivity was attributed to the 3 amino acid differences between human and mouse in the peptide sequence used to prepare the antiserum (see Table 1). The
Table 1 Sequence for residues 52 to 64 of metallothionein-3 from mouse and human Mouse Human
KGEEGAKAEAEK KGGEAAEAEAEK
48
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
corresponding region in mouse is also designated. Thus, effort was made to develop antisera against recombinant mouse MT3. 3.1. Characterization of anti-metallothionein-3 antiserum Recombinant mouse MT3 was used to immunize White New Zealand rabbits. Antiserum was developed and collected as described. To test for the specificity of the antiserum, immunoblots were done to compare immunoreactivity with authentic MT1, MT2, and MT3. MT1 and MT2 were commercial preparations from rabbit liver. Authentic MT3 was recombinant, mouse Cd7-MT3 previously characterized [24]. Initial collections of the antiserum showed reactivity with all three isoforms, although greater with MT3, as determined by Western immunoblots (not shown). Following subsequent, booster injections of MT3 as the antigen, the antiserum showed considerable selectivity for MT3 compared to MT1 and MT2. A representative slot blot analysis is shown in Fig. 2. Little immunoreactivity was detected even with 10 Ag MT1 or MT2 but 0.1 Ag MT3 was detected, a 100-fold lower amount. The antiserum was used in subsequent experiments essentially as a MT3-specific antiserum. The available anti-MT3 antibody preparation was also found to immunoprecipitate MT3 whether as the purified protein (Fig. 3A) or from a cellular extract (Fig. 3B). Immunoprecipitation of pure, recombinant, mouse brain MT3 showed that as little as 0.30 Ag could be immunoprecipitated as detected by subsequent Western blotting (Fig. 3A). Samples of Swiss Webster mice brain extracts in the appropriate immunoprecipitation buffer were incubated overnight with antibody at 4 8C. Immobilized Protein Aagarose was added and incubated for 1 h at room temperature. After incubation, the sample was centrifuged with the Protein A-agarose and anything associated being pelleted. The pellet was washed with 10 mM MOPS, pH 7.3 and re-pelleted and washed two additional times. MT3 and associated proteins were eluted from the immobilized Protein A-agarose by boiling for 30 s in SDS gel sample buffer for analysis by SDS gel electrophoresis and Western immunoblotting for MT3. Fig. 3B shows the resulting Western blot with the immunoreactive band at the position consistent with authentic recombinant, mouse brain MT3. This showed that MT3 could be immunoprecipitated
Fig. 2. Slot blot analysis of antiserum specificity. The indicated quantities of authentic MT1 and MT2 from rabbit liver, and recombinant, mouse MT3 were applied to nitrocellulose using the slot blot procedure described. The nitrocellulose was processed, and the antigen detected as described in the text.
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
49
Fig. 3. Western immunoblot of MT3 immunoprecipitation. Panel A shows the results of immunoprecipiatation experiments using pure recombinant, mouse brain MT3. Different amounts of protein from 0.3 to 2.5 Ag were incubated with anti-MT3 antiserum and incubated overnight at 4 8C. The sample was processed as described and the immunoprecipitated pellet harvested for analysis by Western blotting. Shown are the bands detected by Western blotting. In Panel B, brain extract from Swiss Webster mice was prepared and used as a source for MT3. Antiserum was added and the sample incubated and processed as described in the text to yield protein associated with immobilized Protein A. Immunoprecipitated proteins were harvested by boiling in SDS–polyacrylamide gel electrophoresis sample buffer and the proteins separated by electrophoresis on as 12% minigel. After separation, the proteins were electrotransferred to nitrocellulose, blotted with anti-MT3 antiserum, and detected as described. MT3 was detected in immunoprecipitated samples as indicated by the arrow in the figure.
from mouse brains. Other proteins immunoprecipitated as part of a complex with MT3 are invisible by this method. 3.2. Identification of MT3 associated proteins Preliminary experiments have been done to provide evidence that MT3 was in a protein complex in mouse brain and to identify interacting proteins. Brain extract from a Swiss Webster mouse was fractionated according to molecular weight using the Centriplus-30 and Centriplus-10 centrifugal concentrators. Three fractions were developed: (1) proteins with MW greater than 30 kDa; (2) proteins with MW between 10 kDa and 30 kDa; and (3) proteins with MW less than 10 kDa. Only the fraction of MW greater than 30 kDa showed immunoreactivity with the anti-MT3 antiserum after SDS gel electrophoresis and Western blotting (not shown). Samples of these same brain fractions were also immunoprecipitated
50
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
and analyzed by mass spectrometry. Numerous peaks (not shown) were identified by MALDI-TOF mass spectrometry from the immunoprecipitated pellet from a sample retained by a membrane with a nominal 30,000 molecular weight limit cutoff. Identification of associated proteins has been accomplished using mass spectrometry methods after immunoaffinity isolation [25]. 3.3. Quantification of MT3 in mouse brain extract using an ELISA assay A MT3 specific ELISA assay was developed and used to measure the amount of MT3 in whole mouse brains. Fig. 4 shows the standard curve for the MT3 ELISA. The standard curve was sufficiently sensitive to measure 0.001 Ag of metallothionein-3. Using this ELISA assay, an average value of 0.003 F 0.0008 mg/ml was determined for MT3. For the same mice, total protein was found to be 10.0 F 2.3 mg/ml. The value for MT3 represents 0.03% of total protein in these samples. This value is equivalent to approximately 3.5 Ag MT3 per gram brain tissue (wet weight). Using this ratio and considering an average mass of a human brain of approximately 1300 g, a human brain would contain approximately 4.5 mg MT3. Using the incorporation of 119Cd2+, Erickson et al. [16] estimated that MT3 was present in human frontal cortex at a concentration of 20 Ag MT3 per gram tissue. Although not described, it was presumed this represents the wet weight of cortex tissue. Cerebral cortex is one region of the brain enriched in metallothionein-3 so it is not surprising that the relative abundance of MT3 found in cortex tissue is higher than our value from homogenized whole brain. 3.4. Summary Antiserum essentially specific for isoform-3 of mouse metallothionein has been raised and shown to be applicable to immunoblotting, immunoprecipitation, and ELISA
Fig. 4. Standard curve for MT3 ELISA assay. Shown is the ELISA standard curve generated using authentic, recombinant mouse brain MT3. Protein (from 0.001 to 0.10 Ag MT3) was added to each well and incubated overnight at 4 8C and detection was done as described. The rate of the alkaline phosphatase conjugated to the secondary antibody provides an indirect measure of the amount of primary antibody, and hence, the amount of MT3 present in the well.
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
51
protocols. Development of an ELISA assay provided a means to quantify MT3 in mouse brain extracts and yielded a value of 0.003 F 0.0008 mg/ml or 3.5 Ag MT3 per gram brain tissue (wet weight). This value was not inconsistent with reported values of MT3 in human brain tissue [16].
Acknowledgments Marc A. Denn is thanked for technical assistance with the preparation of recombinant Cd7-MT3. This research was supported by funds from the Minnesota Medical Foundation (B.L.M.) and from the Minnesota Medical Foundation and University of Minnesota Graduate School (I.M.A.).
References [1] Klaassen CD. In: Klaassen CD, editor. Metallothionein IV. Basel7 Birkhauser Verlaag; 1999. [2] Zangger K, Armitage IM. Dynamics of interdomain and intermolecular interactions in mammalian metallothioneins. J Inorg Biochem 2002;88:135 – 43. [3] Simpkins CO. Metallothionein in human disease. Cell Mol Biol 2000;46:465 – 88. [4] Uchida Y, Ihara Y, Tomonaga M. Alzheimer’s disease brain extract stimulates the survival of cerebral cortical neurons from neonatal rats. Biochem Biophys Res Commun 1988;150:1263 – 7. [5] Uchida Y, Takio K, Titani K, Ihara Y, Tomonaga M. The growth inhibitory factor that is deficient in the Alzheimer’s disease brain is a 68 amino acid metallothionein-like protein. Neuron 1991;7:337 – 47. [6] Uchida Y, Tomonaga M. Neurotrophic action of Alzheimer’s disease brain extract is due to the loss of inhibitory factors for survival and neurite formation of cerebral cortical neurons. Brain Res 1989;481: 190 – 3. [7] Palmiter RD. Constitutive expression of metallothionein-III (MT-III), but not MT-I, inhibits growth when cells become zinc deficient. Toxicol Appl Pharmacol 1995;135:139 – 46. [8] Kobayashi H, Uchida Y, Ihara Y, Nakajima K, Kohsaka K, Miyatake T, Tsuji S. Molecular cloning of rat growth inhibitory factor cDNA and the expression in the central nervous system. Brain Res Mol Brain Res 1993;19:188 – 94. [9] Belloso E, Henandez J, Giralt M, Kille CD, Hidalgo J. Effect of stress on mouse and rat brain metallothionein I and II mRNA levels. Neuroendocrinology 1996;64:430 – 9. [10] Hidalgo J, Belloso E, Henandez J, Gasull T, Molinero A. Role of glucocorticoids on rat brain metallothionein-I and -III response to stress. Stress 1997;1:231 – 40. [11] Carrasco J, Hernandez J, Bluethmann H, Hidalgo J. Interleukin-6 and tumor necrosis factor-alpha type I receptor deficient mice reveal a role of IL-6 and TNF-alpha on brain metallothionein-I and -III regulation. Brain Res Mol Brain Res 1998;57:221 – 34. [12] Lovell MA, Xie C, Markesberry WR. Protection against amyloid beta peptide toxicity by zinc. Brain Res 1999;823:88 – 95. [13] Moreira P, Pereira C, Santos MS, Oliveira C. Effect of zinc ions on the cytotoxicity induced by the amyloid beta-peptide. Antioxid Redox Signal 2000;2:317 – 25. [14] Irie Y, Keung WM. Metallothionein-III antagonizes the neurotoxic and neurotrophic effects of amyloid beta peptides. Biochem Biophys Res Commun 2001;282:416 – 20. [15] Masters BA, Quaife CJ, Erickson JC, Kelly EJ, Froelick GJ, Zambrowicz BP, et al. Metallothionein III is expressed in neurons that sequester zinc in synaptic vesicles. J Neurosci 1994;14:5844 – 57. [16] Erickson JC, Sewell AK, Jensen LT, Winge DR, Palmiter RD. Enhanced neurotrophic activity in Alzheimer’s disease cortex is not associated with down-regulation of metallothionein-III (GIF). Brain Res 1994;649:297 – 304.
52
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
[17] Tsuji S, Kobayashi H, Uchida Y, Ihara Y, Miyatake T. Molecular cloning of human growth inhibitory factor cDNA and its down-regulation in Alzheimer’s disease. EMBO J 1992;11:4843 – 50. [18] Naruse T, Igarashi S, Furuya T, Kobayashi H, Miyatake T, Tsuji S. Structures of the human and mouse growth inhibitory factor-encoding genes. Gene 1994;144:283 – 7. [19] Carrasco J, Girault M, Molinero A, Penkowa M, Moos T, Hidalgo J. Metallothionein (MT)-III: generation of polyclonal antibodies, comparison with MT-I+II in the freeze lesioned rat brain and in a bioassay with astrocytes, and analysis of Alzheimer’s disease brains. J Neurotrauma 1999;16:1115 – 29. [20] Yu WH, Lukiw WJ, Bergeron C, Niznik HB, Fraser PE. Metallothionein III is reduced in Alzheimer’s disease. Brain Res 2001;894:37 – 45. [21] Amoureux MC, Van Gool D, Herrero MT, Dom R, Colpaert FC, Pauwels PJ. Regulation of metallothioneinIII (GIF) mRNA in the brain of patients with Alzheimer disease is not impaired. Mol Chem Neuropathol 1997;32:101 – 21. [22] Turner BM. The use of alkaline-phosphatase-conjugated second antibody for the visualization of electrophoretically separated proteins recognized by monoclonal antibodies. J Immunol Methods 1983; 63:1 – 6. [23] Derango R, Page J. The quantitation of coupled bead antibody by enzyme-linked immunosorbent assay. J Immunoass 1996;17:145 – 53. [24] Oz G, Zangger K, Armitage IM. Three-dimensional structure and dynamics of a brain specific growth inhibitory factor: metallothionein-3. Biochemistry 2001;40:11433 – 41. [25] Lahti DW, Hoekman JD, Tokheim AM, Martin BL, Armitage IM. Identification of mouse brain proteins associated with isoform 3 of metallothionein. Protein Science; in press.
Antiserum specific for the intact isoform-3 of metallothionein Abigail M. Tokheima, Ian M. Armitageb,T, Bruce L. Martina,b,T a
Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, MN 55455, United States b Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 6-155 Jackson Hall, 321 Church Street, SE, Minneapolis, MN 55455, United States Received 13 February 2004; accepted 30 January 2005
Abstract The recombinant form of isoform-3 of mouse brain metallothionein (MT3) was used as an antigen to immunize rabbits and raise MT3-selective antiserum. The antiserum was essentially specific for MT3 with 100-fold greater sensitivity for MT3 compared to MT1 or MT2. Immunonblot analysis of whole mouse brain homogenates showed that MT3 was present only in the fraction retained by a 30,000-Da cut-off filter. The antiserum was used to immunoprecipitate MT3 from mouse brain extracts of Swiss Webster mice and provided evidence that MT3 was a member of a macromolecular complex of greater than 30,000 Da mass in brain. An ELISA was developed using purified, recombinant mouse brain Cd7-MT3 as the antigen and used to quantify MT3 in mouse brain extracts. The concentration of MT3 was found to be 3.0 F 0.8 Ag/ml or approximately 3.5 Ag/g mouse brain (wet weight). D 2005 Elsevier B.V. All rights reserved. Keywords: Metallothionein-3, antiserum; Metallothionein-3, neuronal isoform
Abbreviations: MT1, isoform 1 of metallothionein; MT2, isoform 2 of metallothionein; MT3, isoform-3 of metallothionein; Cd7-MT3, MT3 containing 7 gram-equivalents of cadmium. T Corresponding authors. Bruce L. Martin is to be contacted at Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 6-155 Jackson Hall, 321 Church Street, SE, Minneapolis, MN 55455. Tel.: +1 612 625 5988. Ian M. Armitage is to be contacted at Tel.: +1 612 624 5977. E-mail addresses: [email protected] (I.M. Armitage)8 [email protected] (B.L. Martin). 0165-022X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jbbm.2005.01.003
44
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
1. Introduction Metallothioneins are a family of low-molecular weight, cysteine rich, metal ion binding proteins that serve as a thermodynamic sink for biological zinc along with other transition metal ions [1]. Structural and other biophysical studies have contributed to establishing the presence of distinct metal clusters in the protein [2]. Reports also have established the kinetic lability of the metal ions bound to MT1 including the specific transfer of these metal ions between different metal forms of MT and zinc depleted metallo-enzymes/proteins [1]. Of the four MT isoforms, isoform-3 is specific to neuronal tissue. The isoforms share high sequence identity and, for the three structurally characterized isoforms (MT1, MT2, and MT3), have conserved metal ion binding clusters in two distinct domains. The primary physiological roles of MT1 and MT2 remain elusive, but metallothioneins are reported to modulate three fundamental processes: (1) the modulation of oxygen-free radicals and nitric oxide; (2) apoptosis and; (3) the binding and exchange of heavy metals. All three physiological processes are believed to play a role in several diseases and specific associations between MT and several diseases have been made [1,3]. The situation with MT1 and 2 is in contrast to that of the tissue specific isoform, MT3. MT3 was discovered in the last decade in the central nervous system and has a specific activity, the inhibition of the growth and survival of neuronal cultures. MT3 was originally designated Growth Inhibitory Factor [4–6]. Although structurally conserved, the unique biological function of MT3 indicates the existence of different signaling pathways involving this isoform. Indeed, MT3 has been specifically implicated in neuronal processes. The isoforms, MT3 and MT1, respond differently to zinc deprivation of cells in culture with only MT1 being down-regulated in these cells [7]. Transcriptional regulation of MT3 levels is differentially regulated compared to MT1 and MT2 [8–11]. These data suggest that MT3 is not involved in metal ion homeostasis, as the MT1 and MT2 isoforms seem to be. A common ligand for all MTs is zinc, an important factor in multiple neuronal processes. Also, there are numerous reports of a link between zinc and Alzheimer’s Disease. Zinc can bind to both the h-amyloid peptide and the amyloid precursor protein, and the toxicity of h-amyloid is modulated by zinc [12,13]. Low levels of Zn2+ induce neurotrophic effects of h-amyloid and high levels of Zn2+ support the neurotoxic effects. These neurotrophic and neurotoxic effects were reported to be antagonized by the presence of MT3, but not MT1 or MT2 [14]. In neurons of transgenic mice overexpressing MT3, MT3 and Zn2+ were found co-localized indicating MT3 as a possible mediator of these events [15,16]. MT3 was found to be down regulated at the mRNA level in tissue from Alzheimer’s patients [17–20], although other studies have not confirmed this result [16,21]. These data indicate a possible connection of MT3 with neuronal function and Alzheimer’s Disease although no causative relationship can be inferred. We have started experiments to characterize the neuronal function of MT3. We report here the development of a selective tool for probing the levels and interactions of MT3 in brain tissue.
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
45
2. Experimental procedures 2.1. Materials Authentic MT1 and MT2 were purchased from Sigma-Aldrich as was the pNPP substrate (Sigma 104 substrate) for the alkaline phosphatase ELISA reactions. The Amplified Alkaline Phosphatase Immun-blot kit was purchased from Bio-Rad and contained goat anti-rabbit secondary antibody; biotin; streptavidin; biotinylated alkaline phosphatase; and the substrates for alkaline phosphatase (AP color development components). The Slot Blot apparatus was purchased from Bio-Rad. The wash buffer was also purchased from Bio-Rad as a 10 stock solution (10 TBS). ELISA compatible microtiter plates were from Costar and purchased through Fisher Scientific. 2.2. Preparation of anti-MT3 antiserum Antiserum to recombinant, mouse Cd7-MT3 has been prepared using the services of the Research Animal Resources group of the University of Minnesota. As outline in Fig. 1, two female New Zealand white rabbits were immunized using purified, recombinant mouse brain Cd7-MT3 (0.35 mg) emulsified in complete Freund’s adjuvant with a preimmune serum sample obtained before immunization. Three booster injections were given using the recombinant Cd7-MT3 (0.175 mg each boost) emulsified in incomplete Freund’s adjuvant. Rabbit serum was collected at each boost and, subsequently, at biweekly intervals (12–15 ml). Initial Western immunoblots using antiserum from the first
Fig. 1. Protocol for antibody generation. Details of the protocol for generation of the anti-metallothionein-3 antiserum are given in the text.
46
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
collection showed positive reactivity with all three isoforms, but the antiserum became more selective for MT3 with the additional booster injections. Subsequently, screening of the antiserum was done using slot blots with authentic samples of MT1, MT2, and MT3 as controls. 2.3. Western immunoblots Briefly, protein samples were separated by SDS gel electrophoresis and proteins electro-transferred onto nitrocellulose (or similar) membrane in tris–glycine buffer. The membrane was blocked with 5% instant milk in tris buffered saline, pH 7.4 (TBS; 20 mM Tris, pH 7.4; and 150 mM NaCl), washed with TBS, and incubated with antiserum in TBS. After incubation with the primary antibody, the blot was washed with TBS and positive immunoreactivity detected using the Amplified Alkaline Phosphatase Immun-Blot assay kit from BioRad. This method exploits the high affinity of the interaction between streptavidin and biotin. Once washed, the blot was incubated for 1 h at room temperature with biotinylated goat anti-rabbit secondary antibody. This antibody recognizes and binds to rabbit proteins, such as the anti-MT3 antibody raised in rabbits. After incubation with secondary antibody, the blot was washed two times with TBS plus 0.05% Tween-20 (TTBS) for 15 min followed by one wash with TBS for 15 min. A solution containing a complex of streptavidin and biotinylated alkaline phosphatase was added to the blot and incubated for 1 h at room temperature. The blot was again rinsed with TBS plus Tween-20 and TBS as described. The presence of MT3 was detected by monitoring the color development after exposure to the substrate for the alkaline phosphatase-conjugated secondary antibody [22,23]. For alkaline phosphatase, the substrate used was comprised of a mix of bromochloroindolyl chloride phosphate (BCIP) and nitroblue tetrazolium (NBT). In the presence of alkaline phosphatase, BCIP was hydrolyzed with the resulting product forming an insoluble complex with NBT and stains the blot (blue to black) in the position of the alkaline phosphatase linked antibody, and therefore, the antigen. Only bands interacting with the primary antibody will show positive immunoreactivity. 2.4. Slot blot for the identification of MT3 The presence of MT3 in brain extracts was measured using Slot blots and an ELISA assay. For the Slot blots, 200-Al brain homogenates were applied to pre-cut nitrocellulose sheets (0.45 Am, Bio-Rad) using the Bio-Rad Slot-Blot apparatus. Homogenate protein was allowed to bind under gentle vacuum to the nitrocellulose supported by three sheets of Bio-Dot SF filter paper after a prewash of the nitrocellulose with 100 Al TBS. Each well was then washed with 200 Al of TBS. After washing, the membrane was removed and blocked using 5% milk in TBS for 1 h at room temperature. The milk was removed and the membrane washed 2 with TTBS for 5 min and 1 with TBS for 5 min. The MT3 primary antibody was applied and incubated overnight at 4 8C with rocking. After overnight incubation, the primary antibody was removed and the membrane washed as described. Detection was done using the reagents
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
47
of the Amplified Alkaline Phosphatase Immun-Blot Kit from Bio-Rad Laboratories as already described. 2.5. ELISA for the quantification of MT3 in brain extracts Samples of brain extract (40 Al of brain supernatants and 160 Al TBS) were added into wells of a Costar 96 well microtiter plate designed for protein binding. The plate was incubated overnight at 4 8C to maximize protein binding. A vacuum pipette was used to remove the supernatant and each well was washed three times with 0.2 ml of TBS with each wash incubated for 5 min. After the washes, 0.2 ml of the primary antibody (rabbit anti-MT3) was added to each well and incubated at room temperature for 1 h. The wash steps were repeated and 0.2 ml of biotinylated goat anti-rabbit secondary antibody was added to each well. This was incubated for 30 min at room temperature, followed by the wash steps. A solution (0.2 ml) of streptavidin with biotinylated alkaline phosphatase was added to each well and incubated for 30 min at room temperature. After the incubation and subsequent wash steps, the microtiter plate was inserted into a Molecular Devices Spectramax 340 microtiter plate reader and the plate incubated for 30 s at the reaction temperature (26 8C). A substrate solution (0.2 ml of 10 mM pNPP) was added and the reaction monitored for 30 min at 410 nm as the remaining alkaline phosphatase hydrolyzed the pNPP introduced. The amount of pNPP hydrolysis was proportional to the amount of alkaline phosphatase bound that reflects the amount of antigen present in the well. The ELISA assay of MT3 in mouse brains was accomplished in triplicate wells and compared to a standard curve established using authentic purified, recombinant mouse brain MT3 (0.30 mg/ml).
3. Results and discussion In the course of studies associated with evaluating the role of MT3 in Alzheimer’s Disease, standard Western blotting techniques were used to screen for the presence of MT3 in brains from a mouse model (Tg 2576 mouse strain) of AD. Initial attempts were made using an antibody raised in rabbits against recombinant human MT3. Human antipeptide antiserum to a peptide of a unique sequence of MT3 corresponding to residues 52–64 [20] were obtained from Dr. Paul Fraser (Center for Research in Neurodegenerative Diseases, University of Toronto) and used to probe for metallothionein-3. Unfortunately, no immunoreactivity was detected on the blot (not shown). The absence of immunoreactivity was attributed to the 3 amino acid differences between human and mouse in the peptide sequence used to prepare the antiserum (see Table 1). The
Table 1 Sequence for residues 52 to 64 of metallothionein-3 from mouse and human Mouse Human
KGEEGAKAEAEK KGGEAAEAEAEK
48
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
corresponding region in mouse is also designated. Thus, effort was made to develop antisera against recombinant mouse MT3. 3.1. Characterization of anti-metallothionein-3 antiserum Recombinant mouse MT3 was used to immunize White New Zealand rabbits. Antiserum was developed and collected as described. To test for the specificity of the antiserum, immunoblots were done to compare immunoreactivity with authentic MT1, MT2, and MT3. MT1 and MT2 were commercial preparations from rabbit liver. Authentic MT3 was recombinant, mouse Cd7-MT3 previously characterized [24]. Initial collections of the antiserum showed reactivity with all three isoforms, although greater with MT3, as determined by Western immunoblots (not shown). Following subsequent, booster injections of MT3 as the antigen, the antiserum showed considerable selectivity for MT3 compared to MT1 and MT2. A representative slot blot analysis is shown in Fig. 2. Little immunoreactivity was detected even with 10 Ag MT1 or MT2 but 0.1 Ag MT3 was detected, a 100-fold lower amount. The antiserum was used in subsequent experiments essentially as a MT3-specific antiserum. The available anti-MT3 antibody preparation was also found to immunoprecipitate MT3 whether as the purified protein (Fig. 3A) or from a cellular extract (Fig. 3B). Immunoprecipitation of pure, recombinant, mouse brain MT3 showed that as little as 0.30 Ag could be immunoprecipitated as detected by subsequent Western blotting (Fig. 3A). Samples of Swiss Webster mice brain extracts in the appropriate immunoprecipitation buffer were incubated overnight with antibody at 4 8C. Immobilized Protein Aagarose was added and incubated for 1 h at room temperature. After incubation, the sample was centrifuged with the Protein A-agarose and anything associated being pelleted. The pellet was washed with 10 mM MOPS, pH 7.3 and re-pelleted and washed two additional times. MT3 and associated proteins were eluted from the immobilized Protein A-agarose by boiling for 30 s in SDS gel sample buffer for analysis by SDS gel electrophoresis and Western immunoblotting for MT3. Fig. 3B shows the resulting Western blot with the immunoreactive band at the position consistent with authentic recombinant, mouse brain MT3. This showed that MT3 could be immunoprecipitated
Fig. 2. Slot blot analysis of antiserum specificity. The indicated quantities of authentic MT1 and MT2 from rabbit liver, and recombinant, mouse MT3 were applied to nitrocellulose using the slot blot procedure described. The nitrocellulose was processed, and the antigen detected as described in the text.
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
49
Fig. 3. Western immunoblot of MT3 immunoprecipitation. Panel A shows the results of immunoprecipiatation experiments using pure recombinant, mouse brain MT3. Different amounts of protein from 0.3 to 2.5 Ag were incubated with anti-MT3 antiserum and incubated overnight at 4 8C. The sample was processed as described and the immunoprecipitated pellet harvested for analysis by Western blotting. Shown are the bands detected by Western blotting. In Panel B, brain extract from Swiss Webster mice was prepared and used as a source for MT3. Antiserum was added and the sample incubated and processed as described in the text to yield protein associated with immobilized Protein A. Immunoprecipitated proteins were harvested by boiling in SDS–polyacrylamide gel electrophoresis sample buffer and the proteins separated by electrophoresis on as 12% minigel. After separation, the proteins were electrotransferred to nitrocellulose, blotted with anti-MT3 antiserum, and detected as described. MT3 was detected in immunoprecipitated samples as indicated by the arrow in the figure.
from mouse brains. Other proteins immunoprecipitated as part of a complex with MT3 are invisible by this method. 3.2. Identification of MT3 associated proteins Preliminary experiments have been done to provide evidence that MT3 was in a protein complex in mouse brain and to identify interacting proteins. Brain extract from a Swiss Webster mouse was fractionated according to molecular weight using the Centriplus-30 and Centriplus-10 centrifugal concentrators. Three fractions were developed: (1) proteins with MW greater than 30 kDa; (2) proteins with MW between 10 kDa and 30 kDa; and (3) proteins with MW less than 10 kDa. Only the fraction of MW greater than 30 kDa showed immunoreactivity with the anti-MT3 antiserum after SDS gel electrophoresis and Western blotting (not shown). Samples of these same brain fractions were also immunoprecipitated
50
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
and analyzed by mass spectrometry. Numerous peaks (not shown) were identified by MALDI-TOF mass spectrometry from the immunoprecipitated pellet from a sample retained by a membrane with a nominal 30,000 molecular weight limit cutoff. Identification of associated proteins has been accomplished using mass spectrometry methods after immunoaffinity isolation [25]. 3.3. Quantification of MT3 in mouse brain extract using an ELISA assay A MT3 specific ELISA assay was developed and used to measure the amount of MT3 in whole mouse brains. Fig. 4 shows the standard curve for the MT3 ELISA. The standard curve was sufficiently sensitive to measure 0.001 Ag of metallothionein-3. Using this ELISA assay, an average value of 0.003 F 0.0008 mg/ml was determined for MT3. For the same mice, total protein was found to be 10.0 F 2.3 mg/ml. The value for MT3 represents 0.03% of total protein in these samples. This value is equivalent to approximately 3.5 Ag MT3 per gram brain tissue (wet weight). Using this ratio and considering an average mass of a human brain of approximately 1300 g, a human brain would contain approximately 4.5 mg MT3. Using the incorporation of 119Cd2+, Erickson et al. [16] estimated that MT3 was present in human frontal cortex at a concentration of 20 Ag MT3 per gram tissue. Although not described, it was presumed this represents the wet weight of cortex tissue. Cerebral cortex is one region of the brain enriched in metallothionein-3 so it is not surprising that the relative abundance of MT3 found in cortex tissue is higher than our value from homogenized whole brain. 3.4. Summary Antiserum essentially specific for isoform-3 of mouse metallothionein has been raised and shown to be applicable to immunoblotting, immunoprecipitation, and ELISA
Fig. 4. Standard curve for MT3 ELISA assay. Shown is the ELISA standard curve generated using authentic, recombinant mouse brain MT3. Protein (from 0.001 to 0.10 Ag MT3) was added to each well and incubated overnight at 4 8C and detection was done as described. The rate of the alkaline phosphatase conjugated to the secondary antibody provides an indirect measure of the amount of primary antibody, and hence, the amount of MT3 present in the well.
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
51
protocols. Development of an ELISA assay provided a means to quantify MT3 in mouse brain extracts and yielded a value of 0.003 F 0.0008 mg/ml or 3.5 Ag MT3 per gram brain tissue (wet weight). This value was not inconsistent with reported values of MT3 in human brain tissue [16].
Acknowledgments Marc A. Denn is thanked for technical assistance with the preparation of recombinant Cd7-MT3. This research was supported by funds from the Minnesota Medical Foundation (B.L.M.) and from the Minnesota Medical Foundation and University of Minnesota Graduate School (I.M.A.).
References [1] Klaassen CD. In: Klaassen CD, editor. Metallothionein IV. Basel7 Birkhauser Verlaag; 1999. [2] Zangger K, Armitage IM. Dynamics of interdomain and intermolecular interactions in mammalian metallothioneins. J Inorg Biochem 2002;88:135 – 43. [3] Simpkins CO. Metallothionein in human disease. Cell Mol Biol 2000;46:465 – 88. [4] Uchida Y, Ihara Y, Tomonaga M. Alzheimer’s disease brain extract stimulates the survival of cerebral cortical neurons from neonatal rats. Biochem Biophys Res Commun 1988;150:1263 – 7. [5] Uchida Y, Takio K, Titani K, Ihara Y, Tomonaga M. The growth inhibitory factor that is deficient in the Alzheimer’s disease brain is a 68 amino acid metallothionein-like protein. Neuron 1991;7:337 – 47. [6] Uchida Y, Tomonaga M. Neurotrophic action of Alzheimer’s disease brain extract is due to the loss of inhibitory factors for survival and neurite formation of cerebral cortical neurons. Brain Res 1989;481: 190 – 3. [7] Palmiter RD. Constitutive expression of metallothionein-III (MT-III), but not MT-I, inhibits growth when cells become zinc deficient. Toxicol Appl Pharmacol 1995;135:139 – 46. [8] Kobayashi H, Uchida Y, Ihara Y, Nakajima K, Kohsaka K, Miyatake T, Tsuji S. Molecular cloning of rat growth inhibitory factor cDNA and the expression in the central nervous system. Brain Res Mol Brain Res 1993;19:188 – 94. [9] Belloso E, Henandez J, Giralt M, Kille CD, Hidalgo J. Effect of stress on mouse and rat brain metallothionein I and II mRNA levels. Neuroendocrinology 1996;64:430 – 9. [10] Hidalgo J, Belloso E, Henandez J, Gasull T, Molinero A. Role of glucocorticoids on rat brain metallothionein-I and -III response to stress. Stress 1997;1:231 – 40. [11] Carrasco J, Hernandez J, Bluethmann H, Hidalgo J. Interleukin-6 and tumor necrosis factor-alpha type I receptor deficient mice reveal a role of IL-6 and TNF-alpha on brain metallothionein-I and -III regulation. Brain Res Mol Brain Res 1998;57:221 – 34. [12] Lovell MA, Xie C, Markesberry WR. Protection against amyloid beta peptide toxicity by zinc. Brain Res 1999;823:88 – 95. [13] Moreira P, Pereira C, Santos MS, Oliveira C. Effect of zinc ions on the cytotoxicity induced by the amyloid beta-peptide. Antioxid Redox Signal 2000;2:317 – 25. [14] Irie Y, Keung WM. Metallothionein-III antagonizes the neurotoxic and neurotrophic effects of amyloid beta peptides. Biochem Biophys Res Commun 2001;282:416 – 20. [15] Masters BA, Quaife CJ, Erickson JC, Kelly EJ, Froelick GJ, Zambrowicz BP, et al. Metallothionein III is expressed in neurons that sequester zinc in synaptic vesicles. J Neurosci 1994;14:5844 – 57. [16] Erickson JC, Sewell AK, Jensen LT, Winge DR, Palmiter RD. Enhanced neurotrophic activity in Alzheimer’s disease cortex is not associated with down-regulation of metallothionein-III (GIF). Brain Res 1994;649:297 – 304.
52
A.M. Tokheim et al. / J. Biochem. Biophys. Methods 63 (2005) 43–52
[17] Tsuji S, Kobayashi H, Uchida Y, Ihara Y, Miyatake T. Molecular cloning of human growth inhibitory factor cDNA and its down-regulation in Alzheimer’s disease. EMBO J 1992;11:4843 – 50. [18] Naruse T, Igarashi S, Furuya T, Kobayashi H, Miyatake T, Tsuji S. Structures of the human and mouse growth inhibitory factor-encoding genes. Gene 1994;144:283 – 7. [19] Carrasco J, Girault M, Molinero A, Penkowa M, Moos T, Hidalgo J. Metallothionein (MT)-III: generation of polyclonal antibodies, comparison with MT-I+II in the freeze lesioned rat brain and in a bioassay with astrocytes, and analysis of Alzheimer’s disease brains. J Neurotrauma 1999;16:1115 – 29. [20] Yu WH, Lukiw WJ, Bergeron C, Niznik HB, Fraser PE. Metallothionein III is reduced in Alzheimer’s disease. Brain Res 2001;894:37 – 45. [21] Amoureux MC, Van Gool D, Herrero MT, Dom R, Colpaert FC, Pauwels PJ. Regulation of metallothioneinIII (GIF) mRNA in the brain of patients with Alzheimer disease is not impaired. Mol Chem Neuropathol 1997;32:101 – 21. [22] Turner BM. The use of alkaline-phosphatase-conjugated second antibody for the visualization of electrophoretically separated proteins recognized by monoclonal antibodies. J Immunol Methods 1983; 63:1 – 6. [23] Derango R, Page J. The quantitation of coupled bead antibody by enzyme-linked immunosorbent assay. J Immunoass 1996;17:145 – 53. [24] Oz G, Zangger K, Armitage IM. Three-dimensional structure and dynamics of a brain specific growth inhibitory factor: metallothionein-3. Biochemistry 2001;40:11433 – 41. [25] Lahti DW, Hoekman JD, Tokheim AM, Martin BL, Armitage IM. Identification of mouse brain proteins associated with isoform 3 of metallothionein. Protein Science; in press.