Citation
Platinum(II) metallothioneins

Material Information

Title:
Platinum(II) metallothioneins
Creator:
Bongers, Jacob, 1959-
Copyright Date:
1989
Language:
English

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Jacob Bongers. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
21071746 ( OCLC )
AGZ1066 ( LTUF )
0030455991 ( ALEPH )

Downloads

This item has the following downloads:


Full Text
PLATINUM(II) METALLOTHIONEINS

By
JACOB BONGERS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA

1989




To my parents




ACKNOWLEDGMENTS

The Plasma II was purchased with a Biomedical Research Support grant funded by the National Institutes of Health and administered by the Division of Sponsored Research, University of Florida. Partial support of this project was provided by a Milheim Foundation grant and the Division of Sponsored Research, University of Florida.
I thank David Richardson for helping me in all
aspects of my work, and I appreciate his passionate interest in chemistry. John Bell deserve thanks for collaborating with us on several projects and undertaking the difficult job of isolating and purifying the protein used in this study. I also thank Joanne Lopez for her valuable technical assistance. I thank Frances Goodman for her kind assistance in operating and maintaining the instruments, and for many helpful discussions about both of our research projects. Dr. Roy King is acknowledged for his valuable assistance and I owe a special debt of gratitude to Cyndy Walton for her expert assistance in the NMR spectroscopy. My thanks go out to the faculty and students of the inorganic division for their fine spirit of cooperation.

iii




TABLE OF CONTENTS

page
ACKNOWLEDGEMENTS....................................... iii
LIST OF TABLES......................................... v
LIST OF FIGURES........................................ vi
ABSTRACT............................................... viii
CHAPTERS
1 INTRODUCTION....................................... 1
Metallothionein.................................. 1
Binding of Platinum(II) to Metallothioneins...... 8
2 PROTEIN DETERMINATION BY ICP-AES...................... 13
Accuracy and Applications........................ 16
Precision and Sensitivity........................ 23
Advantages of the Method......................... 24
Materials and Methods............................ 25
3 BINDING OF PLATINUM(II) TO APOMETALLOTHIONEIN...... 29
Basic Physical and Chemical Properties of
Platinum(II) Adducts of Apometallothionein..... 29
Optical Studies.................................. 37
Nuclear Magnetic Resonance Spectra.................. 56
Binding Kinetics................................. 68
Conclusions...................................... 72
Materials and Methods............................ 76
4 BINDING OF PLATINUM(II) TO NATIVE METALLOTHIONEIN.. 83
Reactions of Native and Cadmium
Metallothioneins with Tetrachloroplatinate..... 84
Reactions of Native Metallothioneins
with Cisplatin Derivatives..................... 89
Conclusions..................................... 104
Materials and Methods............................ 111
5 GENERAL CONCLUSIONS............................... 115
REFERENCES....... ................................ ...... 118
BIOGRAPHICAL SKETCH.................................... 130

iv




LIST OF TABLES

page
2-1. Comparison of Spectrophotometric Molar
Absorptivities of Proteins Based
on ICP-AES of Sulfur with Published
Values Based on Other Methods
of Protein Determination........................ 17
2-2. Metal Binding Stoichiometries and Metal
Contents of Metalloproteins Determined
by ICP-AES of Sulfur and Metals................ 21
4-1. ICP-AES Analysis of the Products From
Reactions of Cd7MT and Native Equine
MT with K2PtCl4................................85
4-2. ICP-AES of Products From the Reaction of Native
Rat Liver MT-2 with cis-Pt(NH3)2(NO3)2......... 94

v




LIST OF FIGURES

1-1. Amino acid sequences of equine kidney MT-1A, rat liver MT-1, and rat liver MT-2............ 2
1-2. Crystallographic structure of rat liver MT-2.... 4
1-3. Crystallographic structures of the metalthiolate clusters of rat liver MT-2........... 5
3-1. Elution profiles of Cd7MT and Pt7MT on a Sephadex G-75 size-exclusion column
and native MT on a G-50 column.................. 31
3-2. Stokes' radii of Pt7MT and Cd MT determined by using Sephadex G-75 and G-50
size-exclusion columns.......................... 32
3-3. Reactions of DTNB and iodoacetate with thiolates and kinetics of reactions
with Pt7MT....................................... 36
3-4. Electronic absorption spectra of apoMT, Zn7MT, Cd7MT, and Pt7MT......................... 38
3-5. Spectrophotometric titration of apoMT with K2PtCl4 at pH 7................................. 40
3-6. Titration of partially platinated apoMT adducts with DTNB................................ 44
3-7. Relative molecular orbital energies estimated for square-planar Pt(II) complexes............ 46
3-8. Titration of apoMT with K2PtCl4 as followed by CD......................................... 51
3-9. CD spectra of apoMT and Pt7MT in the far ultraviolet.......................... 54
3-10. CD spectra of native rat liver MT............... 55

vi




3-11. 1H NMR spectra of native rat liver MT-2
and rat liver apoMT-2.......................... 58
3-12. 1H spin-echo NMR spectra of native rat
liver MT-2 and rat liver apoMT-2.............. 61
3-13. 1H NMR spectrum and spin-echo NMR spectrum of
rat liver Pt7MT-2............................... 63
3-14. Kinetics of the K2PtCl4-apoMT reaction
at pH 7.5....................................... 70
3-15. Kinetics of the K2PtCl4-apoMT reaction
at pH 2......................................... 71
4-1. Reaction of rat liver MT-2 with excess
cis-Pt(NH3) (NO3) at the ten minute mark,
as seen in ?he up ield region of the H
NMR spectrum..................................... 90
4-2. Reaction of rat liver MT-2 with excess
cis-Pt(NH3)2(NO 3) at the twelve hour
mark, as seen in he 1H NMR spectrum.......... 91
4-3. Elution profile of the products from the
reaction of native rat liver MT-2 with
excess Pt(NH3)2(N03)2........................... 93
4-4. Cleavage of disulfide by sulfite followed by
reaction of the resulting thiolate with TNBS.. 96
4-5. EPR line shapes of cis-Pt(ATMPO)2C12 and the
ATMPO ligand.................................... 98
4-6. HTGF AAS of chromatographic fractions from the
reaction of native MT with one mol equiv of
cis-Pt(ATMPO)2C 2'.............................. 100
4-7. HTGF AAS of chromatographic fractions from the
reaction of native MT with four mol equiv of
cis-Pt(ATMPO)2C2............................... 102
4-8. Possible products from the reactions of
K2PtCl4 with Cd7MT and native MT.............. 105
4-9. Three-metal B cluster of native MT and a
hypothetical three-Pt B cluster............... 108

vii




Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
PLATINUM(II) METALLOTHIONEINS By
Jacob Bongers
May 1989
Chairman: David E. Richardson Major Department: Chemistry
Platinum(II) adducts of the metal ion chelating protein, metallothionein (MT), were prepared and characterized by spectroscopic and other methods. ApoMT reacts with excess K2PtCl4 at pH 7 to form a monomeric adduct containing 6.95 0.24 mol Pt/mol MT as determined by inductively coupled plasma atomic emission spectrometry (ICP-AES). Polymeric adducts containing 17 2 mol Pt/mol MT are formed at pH 2. The Stokes' radius of Pt7MT is consistent with a two-domain structure containing two independent metal-thiolate cluster regions similar to that of native MT.
Spectrophotometric and spectropolarimetric titrations of apoMT with K2PtCl4 at neutral pH suggest that the first three mol equiv of Pt2+ ions form three Pt(cys)4 2- units, and that the binding of the remaining four mol equiv of Pt2+

viii




ions involves formation of bridging thiolates and Pt(II)thiolate clusters. This binding sequence was confirmed by titration of the partially platinated apoMT adducts with the thiolate assay reagent, DTNB. The biphasic kinetics of Pt(II) binding to apoMT are also consistent with this interpretation.
Reactions of Cd7MT and native MT with K2PtCl4, in the
presence of air, result in the liberation of three mol equiv of the originally bound metal ions and the binding of one mol equiv of Pt 2+. When air is excluded from the reaction, three mol equiv of metal ions are lost and four mol equiv of Pt2+ are bound. Pt(II) preferentially displaces metal ions from the 3-metal B-cluster region in reactions of native MT with derivatives of the antitumor drug, cis-Pt(NH3)2C12, as shown by analytical and 1H NMR data. Binding to native MT of Pt(II) from cis-Pt(ATMPO)2C12, resulted in translabilization of the spin-labeled ATMPO amine ligand, and the analytical results suggest binding of Pt(II) to the Nterminal methionine.
Possible Pt(II)-thiolate cluster structures for
Pt7MT and mixed-metal Pt(II) metallothioneins are described. Implications for the metabolism of platinum anticancer drugs are discussed. A method for the determination of micromolar protein concentrations and the metal binding stoichiometries of metalloproteins by ICP-AES of sulfur is also described.

ix




CHAPTER 1
INTRODUCTION
Metallothionein
Metallothioneins are low-molecular-weight cytosolic
proteins with extremely high metal and sulfur contents [14]. These proteins occur in natural systems ranging from microorganisms to humans. Metallothionein was discovered in 1957 by Margoshes and Vallee in a search for a tissue constituent responsible for the natural accumulation of cadmium in animal tissues [5]. Although the biological role of metallothionein is not known for certain it is thought to be involved in the regulation of the essential trace elements Zn and Cu. Metallothionein also binds various heavy metals ions in vivo including Cd(II), Hg(II), and Pb(II), and biosynthesis of the protein is stimulated by the presence of certain of these metal ions. This interesting phenomenon suggests a role in the sequestering of toxic metal ions.
Mammalian forms of MT are single polypeptides of 61
amino acid residues with molecular weights ranging from 6500 to 7000 Daltons depending on the metallic composition. The primary structures of the three metallothioneins employed in this study are shown in Figure 1-1 [6, 7]. Most vertebrate

1




2

Equine Kidney Rat Liver Rat Liver

MT-1A MT-1 MT-2

20
-T C A G S C K C K E
-T C S S S C G C K N
-S C A G S C K C K Q
40
-G G C A R C A Q G C
-V G C S K C A Q G C
-V G C A K C S Q G C

1 10
Ac-M D P N C S C P T G G S C Ac-M D P N C S C S T G G S C Ac-M D P N C S C A T D G S C
30
C R C T S C K K S C C S C C PC K C T S C K K S C C S C C PC K C T S C K K S C C S C C P50 60 V C K G A S D K C S C C A-OH V C K G A S D K C T C C A-OH I C K E A S D K C S C C A-OH

Figure 1-1. Amino acid sequences of equine kidney MT-1A [6], rat liver MT-1 [7], and rat liver MT-2 [7]. The oneletter amino acid code is: A, alanine; C, cysteine; D, aspartic acid; E, glutamic acid; G, glycine; I, isoleucine; K, lysine; M, methionine; N, asparagine; P, proline; Q, glutamine; S, serine; T, threonine; V, valine.




3

tissues contain two major "isoforms" of metallothionein designated MT-1 and MT-2 according to the order of elution by ion exchange chromatography. Some samples of MT-1 may be further resolved into subforms MT-1A, MT-1B, and so on. All mammalian forms of MT contain 20 cysteine residues (CH2SH side chain) at identical positions in the sequence and an Nterminal acetylated methionine (CH2CH2SCH3 side chain) giving a total of 21 sulfur atoms. The complete lack of residues with aromatic side chains is another unusual feature of MT.
Native MT binds a total of 7 divalent metal ions,
primarily Cd2+ and Zn2+ and often contains minor amounts of Cu+, Fe2+, Hg2+, Pb2+, and other metal ions depending on the source and method of isolation. Although MT from adult tissue contains predominantly Cd2+ and Zn2+, MT from fetal tissue generally contains predominantly Cu+ and Zn2+
The crystallographic structure of rat liver MT-2 [8] is shown in Figure 1-2. This structure determination was made at a resolution of 230 pm and is currently being refined with a new data set at 190 pm [9]. The 7 metal ions are bound exclusively to the sulfur atoms of the 20 cysteine residues in two distinct metal-thiolate clusters. Residues 1-29 enfold the 3-metal 9-cysteine cluster, which is referred to as the B domain. Residues 30-61 comprise the a domain, which enfolds the 4-metal 11-cysteine cluster. The unusual structure of MT has been described as a monolayer of




4

010
6 G40
0
055
Figure 1-2. Crystallographic structure of rat liver MT-2. Shown are the a carbons of each residue, the B carbons and sulfur atoms of each cysteine residue, and the Cd and Zn atoms. Reproduced from Furey et al. Science 1986, 231, 707 [8].




5

241
21 1
SNZbo

b 44
,4 8q!:Z3 C= ==gq
'52 4 1
04F3

Figure 1-3. Expanded views of the structures of (a) the B cluster domain and (b) the a cluster domain of rat liver MT-2. Shown are the cysteine sulfur atoms and the Cd and Zn atoms. Reproduced from Furey et al. Science 1986, 231, 707 [8].




6

polypeptide wrapping around the metal-thiolate clusters which take the place of the hydrophobic core of residues in a larger protein [8].
A detailed view of the crystallographic structures of the metal-thiolate clusters is shown in Figure 1-3. The B cluster consists of 3 metal ions, 3 bridging thiolate ligands, and 6 terminal thiolate ligands. The metal ions and bridging thiolates form a 6-membered ring in a chair conformation. The adamantane-like structure of the a cluster is composed of 4 metal ions, 5 bridging thiolates, and 6 terminal thiolates.
The distribution of Cd2+ and Zn2+ ions among the 7
binding sites (as observed by 113Cd NMR) is not a uniform one as shown in Figure 1-3. The relative numbers of Cd2+ and Zn2+ ions in a particular MT molecule and the distribution of these metals ions among the 7 binding sites vary from one molecule to the next. Apparent dissociation constants for Cd2+ and Zn2+ in MT (K = [(cys)33-][M2+ [(cys)3M]) have been estimated to be on the order of 10-15 M and 10~11 M respectively [10], based on spectrophotometric titrations of Cd7MT and Zn7MT with protons. Cd7MT can be prepared via displacement of Zn2+ by adding excess Cd2+ salt to solutions of the native protein. Isolation of native MT in high yield is generally accomplished by administering a series of CdCl2 injections to the animal so as to induce biosynthesis of MT prior to isolation. MT obtained in this




7

way contains both Cd2+ and Zn2+. The Cd2+ is localized preferentially in the a domain and the more weakly bound Zn2+ in the B domain.
Synthetic metal MT adducts have been prepared by three methods. Adducts may be prepared in vivo by injecting the animal with solutions of the metal ions of interest prior to isolation of the protein. The in vitro displacement of metal ions from native MT by added metal ions is another means of obtaining synthetic metal MT adducts. The third method entails reconstituting the apo (metal-free) form of MT with the desired metal ions in vitro.
MT adducts with Cu(I) [11], Hg(II) [12-14], Pb(II) [15], Au(I) [16], Bi(III) [17], Pt(II) [18], and other metals have all been prepared by the in vivo route. This procedure usually results in the recovery of mixed-metal MT adducts.
Metal MT adducts prepared by in vitro displacement of bound Cd2+ and Zn2+ from native and Zn7MT include adducts with Cd(II) [19-23], Hg(II) [24, 25], Au(I) [26], and Pt(II) [18, 27]. Waalkes and coworkers have surveyed the ability of 16 different metals ions to displace Zn2+ from Zn7MT [28]. Both mixed metal and single metal adducts have been prepared by this method.
Synthetic metal MT adducts prepared by reconstituting apoMT with various metals have been studied to the greatest extent. ApoMT is obtained by acidifying solutions of




8

native MT. At low pH the cysteine thiolates are protonated, and the metal ions liberated are removed from the protein solution by chromatography or exhaustive dialysis. Protocols for obtaining metallated derivatives of apoMT generally involve adding an excess of metal salt and then raising the pH to above neutral. The product or products are then desalted on a size-exclusion column. As the pH approaches the pKa values of the cysteine sulfhydryls (ca. 8), the thiolates become unstable with respect to rapid air-oxidation to cystine disulfides. Reconstitutions are therefore performed under a nitrogen or inert gas atmosphere to prevent this undesirable denaturation. M7MT adducts prepared from apoMT include adducts with Co(II) [29, 30], Ni(II) [29], Fe(II) [31], Cu(I) [32-34], Zn(II) [33, 35], Cd(II) [21, 23, 33, 35-39], Pb(II) [40], Hg(II) [40], and Bi(III) [40]. Neilson and coworkers have surveyed the binding of 18 different metal ions to apoMT [41].
Binding of Platinum(II) to Metallothionein
The potent anticancer agent cis-dichlorodiammineplatinum(II) (cisplatin) [42, 43] has been the subject of intense study since its antitumor activity was discovered [44] by Rosenberg et al. Most of this research has been concerned with the binding of cisplatin and analogous compounds to polynucleotides [45] as the mode of action of




9

these platinum drugs probably involves binding to DNA. The binding of platinum drugs to polypeptides and proteins has been studied to a far lesser extent [46, 47]. This aspect of the physiological action of cisplatin is important in understanding the toxicological effects of the drug. Cisplatin is known to cause extensive kidney damage as its primary toxicity, and this nephrotoxicity is one of the major dose-limiting factors in its therapeutic use [48]. In addition, the sequestering of Pt(II) from cisplatin by MT may be responsible for some of the reduction in the anticancer effect of the drug in repeated dose protocols.
Metallothionein is present in relatively high
concentration in the cytosol of the kidney and is a probable binding site for platinum anticancer agents because of its abundance of cysteine residues and the N-terminal methionine. Platinum(II) is known to form very stable complexes with sulfur donor ligands [49]. The binding of Pt(II) to the sulfur bearing residues cysteine and methionine has been documented by protein crystallography wherein Pt(II) salts are used as heavy-atom labels [50].
The biodistribution of Pt in the tissues of
experimental animals administered cisplatin intravenously has been described [51-53]. Platinum is rapidly cleared from the blood following injection, with filterable Pt in the plasma of rats declining to near zero after several hours [51, 53]. Tissue levels of Pt decline slowly over




10

about one week post-injection to leave several percent of the initial dose retained by tissue [51]. Cisplatin nephrotoxicity takes several days to manifest maximum severity during which the low molecular weight reaction products formed in the plasma are excreted into the urine and reabsorbed into the kidneys. By far the highest tissue concentration of Pt during metabolism of cisplatin is found in the kidney [51-53].
Sharma and Edwards found that at the time of maximum Pt accumulation in the rat kidney (ca. 24 hours post-injection) about 60% of cellular Pt in the kidney is localized in the cytosol [54]. This cytosolic Pt was partitioned into three main fractions by size-exclusion chromatography. Low molecular weight species (MW < 1000) accounted for 52% of the total Pt, 23% was found in the high molecular weight fraction (MW > 20000), and the remaining 25% eluted in a "metallothionein-like" fraction (6000 < MW > 12000) [54]. These observations and similar results reported by other workers [55, 56] suggest a possible role for MT in the detoxification of cisplatin and its metabolites.
Zelazowski and coworkers demonstrated that the total in vivo bound Pt in the kidneys of cisplatin treated rats can be significantly increased via induction of MT biosynthesis by preinjection with CdCl2 [57]. MT-like sizeexclusion fractions are composed of MT-1 and MT-2 as confirmed by gel electrophoresis. The authors also found




11
that binding of Pt from cisplatin in vivo and in vitro was associated with a large decrease in the Zn content (ca. 50%) of the native protein but only trivial Cd2+ losses. Binding of Pt(II) to native MT thus involves displacement of the relatively weakly bound Zn2+ ions.
Pt(II) may induce the biosynthesis of MT as has been documented for certain other heavy metal ions. However, several groups have presented evidence that stimulation of MT biosynthesis by Pt(II) is very minimal or nonexistent [57, 58].
Bakka et al. found that cultured cells with naturally high MT contents were resistant to the toxic effects of cisplatin [59]. This observation extends the hypothesis of heavy metal detoxification by MT and raises the interesting question of whether elevated MT concentrations might be used to alleviate the toxic effects of cisplatin and increase its effectiveness as an anticancer drug. However, tumor cells with higher MT concentrations might also be more resistant to the anticancer effects of cisplatin thus reducing the effectiveness of the drug. In another recent series of experiments, mutant tumor cells with elevated MT contents were implanted in mice and tumor growth following i.v. administration of cisplatin compared to a control group of mice implanted with wild-type cells [60]. Tumor volume was reduced by 80% in tumors from the parent cells, whereas the tumors from MT-rich cells were almost completely resistant.




12

Human carcinoma cells with high levels of MT, induced by chronic exposure to heavy metals and various anticancer drugs, are also known to acquire significant resistance to cisplatin [61].
Naganuma et al. have reported an increase in survival rates of mice injected with MT inducing metals prior to injection of a lethal dose of ciplatin [62, 63]. The greatest protection was imparted by preinjection with Cu(II) and Bi(III). This protective effect is presumably due to the binding of Pt(II) to MT but no direct relationship with the increased MT levels has been unequivocally established. A reduction in the lethal and renal toxicity of cisplatin without impairment of its antitumor activity in mice by preinjection with the MT inducer Bi(N0393 has been described recently [64].
The following chapters describe in vitro Pt(II) MT adducts from the reaction of K2PtCl4 with apoMT and the reactions of K2PtCl4 and cisplatin derivatives with Cd7MT and native MT. This study is the first attempt to examine these adducts on the molecular level, and its primary aim is to provide synthetic models of the in vivo cisplatin MT metabolites reviewed above. The incorporation of squareplanar Pt2+ ions in sites normally occupied by tetrahedral Cd2+ and Zn2+ ions is also of interest concerning the degree to which the structures of metal MT adducts are determined by the metal versus the polypeptide component of the adduct.




CHAPTER 2
PROTEIN DETERMINATION BY ICP-AES
The analytical method used to obtain bound metal
stoichiometries and bound metal contents of metal MT adducts is described in some detail in this section as these are central results in the following study. The distinction between metal stoichiometries and metal contents is drawn here as some mixed metal MT adducts are heterogeneous with respect to the bound metal ions. In the native protein for example, Cd2+ ions preferentially occupy sites in the a domain while Zn2+ ions tend to reside mainly in the B domain, but these two metal ions, as well as the lesser amounts of Cu+ and other bound metal ions, are not uniformly distributed across the binding sites of the protein. True binding stoichiometries can be obtained solely from analytical data only for single metal adducts for which stoichiometric binding occurs. For mixed metal adducts, on the other hand, true stoichiometries cannot be determined without some further information beyond the analytical data.
The analysis of metal MT adducts has traditionally
involved the determination of the bound metal or metals by

13




14
one method and quantification of the polypeptide component by independent means. Metals can be conveniently assayed by either atomic absorption spectrometry (AAS) or inductively coupled plasma atomic emission spectrometry (ICP-AES). Determination of the polypeptide component, however, presents a more difficult analytical problem.
Protein concentrations are usually obtained on a
routine basis by measuring the absorbance of micromolar solutions at the absorption maximum occurring at ca. 280 nm. This band arises from absorption by the aromatic side chains of tyrosine and tryptophan residues. The molar absorptivity must, of course, be arrived at by the use a molecular weight and some primary method of determining total protein. An accurate molecular weight is preferably obtained from amino acid sequence data, but a crude value obtained by hydrodynamic methods, such as size-exclusion chromatography or sedimentation, may have to suffice in some cases. The numerous existing methods of total protein determination include Nessler and Kjeldahl total nitrogen determinations, various direct colorimetric methods such as the Biuret, and the most simple and commonly used gravimetric or "dry weight" methods.
Unfortunately the ultraviolet spectrum of apoMT is
devoid of a 280 nm peak due to its peculiar lack of aromatic residues. The spectrum contains an intense band at 190 nm arising from the strong v-v* and weak n-v* amide transitions




15

of the polypeptide backbone and the lowest thiol transition of the cysteine side chains near 195 nm [2]. The most commonly used method of estimating MT concentrations is to measure the absorbance of apoMT in 0.01 M HCl at 220 nm using a molar absorptivity of 48200 M~1 cm~1 [65]. This value was obtained by a wet chemical thiolate assay. The problems associated with the spectrophotometry of the sharply rising shoulder of this peak have been described thoroughly by Wetlaufer [66].
We have developed a method of obtaining micromolar
protein concentrations and bound metal stoichiometries by ICP-AES of sulfur [67] which overcomes the shortcomings of traditional methods. Temperatures in an ICP torch are sufficiently high to excite many nonmetallic elements and several analytically useful emission lines in the vacuum ultraviolet have been observed [68]. Of particular interest to this study is the prominent sulfur line at 180.73 nm, which has an estimated detection limit of 15 pig/L [69]. Proteins with known contents of the sulfur bearing residues cysteine and methionine can thus be quantified at the pg level by ICP-AES, provided the instrument is equipped with evacuated or otherwise oxygen-free optics.
ICP-AES determinations of total sulfur in tissue
samples [70, 71] and the use of an ICP-AES instrument as a sulfur detector in the HPLC of tissue extracts [72] have been reported. The purpose of the following experiments is




16

to demonstrate the viability of pumping micromolar purified protein solutions directly into the nebulizer of the ICP-AES instrument and determining protein concentrations based on the emission of sulfur at 180.73 nm.
Accuracy and Applications
Elemental analysis of proteins by ICP-AES has
traditionally been restricted to metals in acid digested samples of metalloproteins. We initially sought to demonstrate that accurate absolute concentrations of proteins, regardless of the presence or absence of bound metals, can be obtained based on sulfur content by pumping buffered micromolar solutions directly into the nebulizer without any prior sample preparation, and using aqueous solutions of Na2SO4 as sulfur standards. The protein solutions, although relatively dilute, do have slightly different physical properties than the Na2SO4 standards, such as viscosity and surface tension, which could lead to differences in aerosol formation in the nebulizer and systematic errors in the results. One test of accuracy is to compare spectrophotometric molar absorptivities determined by ICP-AES of sulfur with published absorptivities based on traditional methods of protein determination (Table 2-1).




17

Table 2-1. Comparison of Spectrophotometric Molar Absorptivities of Proteins Based on ICP-AES of Sulfur with Published Values Based on Other Methods of Protein Determination.
Proteina E/nm E/104 M1 cm-1
ICPb Litc Ref
Cd7metallothionein equine kidney
MW 6.88 kD
21 mol S/mol 73

250

10.9 0.2

11.4 H

Ferricytochrome c equine kidney MW 12.4 kD 4 mol S/mol

410 280

11.8 0.3
2.62 0.06

11.20 B 10.95 B 10.60 B
2.40 D 2.32 B 2.12 D

Ferrimyoglobin equine MW 17.5 kD 2 mol S/mol

74

75

76 77
78,79
80 78,79 81

82

17.9 0.3

3.24 0.05

409

280

18.8 18.6 17.1 16.0

H
C
C
H

83
84 85
86,79

87 85

3.2 H 3.12 C




18

Table 2-1--continued.

Proteina x/nm E/104 M1 cm-1 ICPb Litc Ref

Ribonuclease A bovine pancreas MW 13.7 kD 12 mol S/mol

1.13 0.02

1.19 A
1.14 D,E
1.13 H 1.06 H 1.04 D 1.01 H 0.98 H

Trypsin bovine pancreas MW 24.0 kD 14 mol S/mol

3.72 0.04

278

89,79 90,79
91,79 92,93
94,93 95
96,79

280

4.4
4.14
4.00 3.98 3.86 3.86
3.79
3.74 3.74 3.72 3.72 3.69 3.69
3.67
3.46

97
98,99
100,79 101,79 102,79 103,93
104,93 105,93 81,93 106,79 107,79 108,93 109,93 110,93 111,93 112,93

H
D
F
D
A
H
H D,F
D
H
H
G
H A,D
H

88




19

Table 2-1--continued.

Proteina )A/nm E/104 M-1 cm1
ICPb Litc Ref
Chymotrypsinogen A bovine
MW 25.6 kD 12 mol S/mol 113
282 5.30 0.11 5.30 F 114
5.20 D 115,79 5.20 D 116,79
5.12 H 117 5.12 H 118 5.04 D 119 4.13 D 120
280 5.28 0.11 5.32 D 80 5.17 H 121 5.15 H 122
5.11 H 94,79
4.61 F 123

a Molecular weights and sulfur contents were
amino acid sequences.

taken from

b (mean of sample mean of blank) (standard deviation)
for 6 replicate measurements of both sample and blank.
c Most of the e values in this column yore obtained from
reported specific absoyptivities (Alcm) using the
relationship: e = MW Alcm/10. Those values reported in the original literature as molar absorptivities (E) are
underlined. Also shown is the analytical method upon
which the published value is based: amino acid analysis,
A; colorimetric Fe determination, B; colorimetric heme
determination, C; dry weight, D; Nessler N determination,
E; Kjeldahl N determination, F; refractometry, G; no
explicit reference to analytical method given, H.




20

While the list of published molar absorptivities in
Table 2-1 is not an exhaustive one, it is an attempt to list as many literature values as possible in a nonselective manner [73-123]. The spread in the distribution of literature values for a particular protein is to be expected as each value was determined by different workers using the various analytical methods listed.
In general, very close agreement is found between the
experimental molar absorptivities and the central tendencies of the published values. According to these comparisons the ICP-AES method of protein quantification via determination of sulfur is an accurate one, at least within the bounds of the concentration range and conditions described in the experimental section.
Probably the most useful application of the method is for the determination of metalloprotein binding stoichiometries. Bound metal stoichiometries are traditionally obtained by determination of the metallic component by atomic absorption or ICP-AES, and determination of the polypeptide component by independent means such as dry weight or total nitrogen determinations. These same stoichiometries, however, can be obtained by an ICP-AES determination of both components by using a single sample.
Shown in Table 2-2 are bound metal stoichiometries of three homogeneous proteins and metal contents of two heterogeneous metallothioneins as determined by ICP-AES of




21

Table 2-2. Metal Binding Stoichiometries and Metal Contents of Metalloproteins Determined by ICP-AES of Sulfur and Metals.
Protein Metal Mol Metal/Mola
Native Metallothionein Cd 3.02 0.15 rat liver Zn 3.18 0.10 Cu 0.26 0.03
total 6.46 0.18
Native Metallothionein Cd 5.04 0.36 equine kidney Zn 0.94 0.08 Cu 0.63 0.23
total 6.61 0.41
Cd7Metallothionein Cd 6.95 0.040 equine kidney
Ferricytochrome c Fe 0.955 0.028 equine heart
Ferrimyoglobin Fe 0.900 0.04b equine

a (mean of sample mean of blank)
for 6 replicate measurements of

(standard deviation) both sample and blank.

b Spectrophotometric measurements indicated that this
sample contained 90% holoprotein and 10% apoprotein.




22

sulfur and metals. Native metallothioneins have variable metal contents depending on the source and method of isolation. This variability is illustrated by the samples of MT from rat liver and equine kidney in Table 2-2. Concentrations of native MT cannot be satisfactorily determined by a metals assay because of this variability in metal contents. Concentrations can be accurately determined by a sulfur assay, however, as all mammalian metallothioneins have identical sulfur contents (21 mol S/mol MT). The sum of the Cd, Zn, and Cu contents for either native MT falls slightly below the maximum binding capacity of MT (seven divalent metal ions), but this probably results from the binding of minor amounts of Fe, Pb, Hg, and other metals not included in the assays.
The experimentally determined Fe content of
ferrimyoglobin in Table 2-2 falls short of the true stoichiometry by about 10%. Spectrophotometric absorbance measurements of this sample taken at the Soret band at 409 nm arising from absorption by the iron containing heme moiety, and the band at 280 nm arising from aromatic side chains in the polypeptide portion of the molecule, also indicated the presence of 10% apoprotein, which lacks heme. The ferricytochrome c sample, that was purified extensively by ion-exchange chromatography, yielded an experimental result that is within 5% of the ideal stoichiometry.




23

Precision and Sensitivity
The standard deviations of the sulfur analyses in Table 2-1 are within 2% of the mean for 6 replicate measurements. The metal binding stoichiometries in Table 2-2 fall within 4-5% of the mean due to the uncertainties associated with both sulfur and metal determinations. This level of precision is acceptable for most applications particularly when one considers the rather poor reproducibility of many other existing methods of protein determination. If one wishes to make very fine distinctions (such as whether a certain protein binds 10 or only 9 metal ions) this level of precision becomes intolerable.
Small gains in precision might be realized by using
optimized instrumental parameters for each element instead of compromise parameters, a pulseless syringe pump, improvement of power supply tolerance, and other refinements. The careful use of an internal standard [124] to mitigate precision-limiting flicker-noise in the nebulizer [125] may also improve precision.
The sensitivity of any destructive technique of protein determination is a general concern. We routinely obtain satisfactory results for samples with sulfur concentrations of 1 mg/L and higher. The volume of sample required is between 1 and 4 mL, which corresponds to the consumption of ten to several hundred micrograms of protein depending on




24
the molar sulfur content of the protein. While the sample volumes required for a pneumatic nebulizer are about two orders of magnitude greater than those required for a graphite furnace, they are not at all prohibitive in terms of the milligram scale on which protein research is often conducted.
Flow injection analysis (FIA) coupled with emerging total-consumption (direct injection) sample introduction systems could, in principle, reduce sample volumes from several milliliters to several microliters and greatly increase the applicability of ICP-AES to proteins [126]. FIA utilizes the transient signals generated by injecting small discrete sample "plugs" into a carrier stream and thus eliminates the wasteful dead-volume of the conventional continuous-flow mode of sample presentation. Experimental sample introduction systems with high transport efficiencies have been described and include a miniature concentric nebulizer installed in the injecter tube of a conventional ICP torch [127], and a thermospray vaporizer [128].
Advantages of the Method
The main advantages of protein determination by ICP-AES of sulfur, besides the high selectivity and multielement capability of atomic spectrometry in general, are the generality and rapidity of the method and the ease of sample




25

preparation. The method is applicable to any protein having a known content of sulfur. The only significant spectral interferences with the sulfur line at 180.73 nm are wing-overlaps of Ca and Mn lines which have intensities about two orders of magnitude smaller on a molar basis [129]. Perchloric acid digestion, dry ashing, and other time-consuming and error-prone preparatory steps are not necessary. A promising application of the method is the rapid and precise determination of metal binding stoichiometries in metalloproteins by determining both protein and metals for a single sample by a single method. In addition, ICP-AES is a useful method for determination of sulfur content of proteins following quantification of total protein by other methods.
Materials and Methods
Preparation of Samples and Standards
Horse kidney and rat liver metallothioneins were
isolated and purified by a literature method [130]. Cd7MT was prepared by adding excess CdCl2 to a solution of native equine MT. An excess of the reducing agent dithioerythritol was added to ensure complete reduction of cysteine side-chains. After 24 hours the reaction mixture was fractionated on a Sephadex G-50 column. Horse heart cytochrome c was obtained from Sigma Chemical Company and




26

purified by carboxymethyl cellulose ion-exchange chromatography. The other proteins used in this investigation were obtained as lyophilized powders from Sigma and used without further purification. The ribonuclease sample was dissolved in 10 mM Tris/HCl buffer at pH 7.8. All other protein samples were dissolved in 5 mM phosphate buffer at pH 7.2. All samples were concentrated by ultrafiltration (YM5 membrane, Amicon Corp.) and then passed through a 0.2 pm filter (Gelman Sciences Inc.). The ultrafiltrates were used as blanks for both ICP-AES and spectrophotometric measurements. The metallothionein samples used for ICP-AES measurements had concentrations between 0.75-1.6 mg/L in sulfur. All other protein samples used were about 5 mg/L in sulfur.
Sulfur standards were solutions made from anhydrous Na2SO4. Standard solutions of the metallic elements were purchased from Aldrich Chemical Company.
Instrumentation
Elemental analyses were made by using a Perkin-Elmer
Plasma II Emission Spectrometer equipped with a Perkin-Elmer Series 7000 computer. The one-meter Ebert monochromator (1800 1/mm holographically ruled grating) is sealed in a heavy-walled aluminum tube maintained at 5 mTorr. The path from the argon ICP source to the optical components was purged with a 5 L/min flow of nitrogen. The sample




27

introduction system is comprised of a peristaltic pump (12 rollers, silicone rubber pump tubing) and a cross-flow nebulizer [131]. Argon flow to the nebulizer is regulated by a mass flow controller. The following compromise operating parameters were employed:
RF Power 1200 W
Viewing Height 12 mm above coil
Sample Delivery Rate 1 mL/min
Argon Flow Rates
plasma 15 L/min auxiliary 1 L/min nebulizer 1 L/min
Absorbance measurements were recorded on an IBM 9430
UV-Visible spectrophotometer. Samples were diluted to give absorbance values between 0.5 and 1.0 in a 1 cm path-length quartz cell.
Measurements
Individual ICP-AES measurements were obtained by
scanning a 0.020 nm window centered on the emission line of interest and recording intensity counts at the wavelength of maximum emission. For multielement analyses the monochromator slews from one emission line to the next in a sequential fashion. The emission lines scanned were




28
S I 180.73, Fe II 259.94, Cd II 214.44, Zn I 213.86, and Cu I 324.75 nm. Six replicate measurements were recorded for each element of each standard, sample, and blank. Linear calibrations were obtained immediately prior to running each protein sample and corresponding ultrafiltrate blank.




CHAPTER 3
PLATINUM(II) BINDING TO APOMETALLOTHIONEIN
The first goal of the research was to investigate Pt(II) MT adducts prepared from mixtures of apoMT and K2PtCl4. This chapter describes the preparation and spectroscopic characterization of such adducts. The dynamics of the binding process are also examined in some detail. Spectra and other experimental results obtained for the synthetic Pt(II) adducts are compared, wherever practical, with corresponding data obtained for the widely studied native form of the protein.
Basic Physical and Chemical Properties of
Platinum(II) Adducts of Apometallothionein
A change from a colorless to a yellow solution occurs when micromolar solutions of apoMT at neutral pH are mixed with excess K2PtCl4. The resulting adduct contains 6.95 0.24 mol Pt/mol MT as determined by ICP-AES determinations of platinum and sulfur. The ICP-AES method was used, unless stated otherwise, for all the elemental analyses reported in this work, and is described in detail in Chapter 2. Dozens of reconstitutions have been performed and the

29




30

resulting products invariably contain 7 mol Pt/mol MT. Once formed, Pt7MT is air-stable, and solutions have been stored for up to 6 months at 60C without loss of bound Pt2+ or chromatographic evidence of aggregation.
Pt7MT elutes as a single symmetrical peak on a Sephadex G-75 size-exclusion column (Figure 3-la). Dimeric, trimeric, and higher oligomeric species are also isolated when the reactants are mixed at higher concentrations. These oligomers are also found to contain approximately 7 mol Pt/mol MT. A similar elution profile is often observed for solutions of native MT, and the oligomerization is probably due to a small number of intermolecular disulfide linkages formed by partial oxidation of the cysteines during isolation or storage [19]. Oxidation of the cysteines can be reversed by 2-mercaptoethanol and other reducing agents. Sephadex G-75 is the chromatographic gel of choice for MT, but even G-50 resin provides sufficient resolution to separate monomers from oligomeric species (Figure 3-1b).
The Stokes' radius of native MT, as determined by sizeexclusion chromatography, is equivalent to that of a much higher molecular weight globular protein. Stokes' radius is defined as the radius of an equivalent hydrodynamic sphere. The large Stokes' radius of MT is due to the nonglobular elongated shape of the molecule (Figure 1-2). Pt7MT has a slightly larger Stokes' radius (ca. 1.80 nm) than Cd7MT or native MT (both ca. 1.60 nm). These radii were determined




31

a
Pt7MT
Cd7MT

100 1
Elution

20

i
140

Volume

/

160
mL

1
180

b

monomer
dimer trimer, etc.

60 80
Elution Volume

/

100
mL

Figure 3-1. Elution profiles of (a) Cd7MT and Pt7MT on a Sephadex G-75 size-exclusion column and (b) native MT on a G-50 column. The eluant was 3.5 mM phosphate buffer, pH
7.2, and the column was monitored by the absorbance at 229 nm (A229).

a)

80

M

40




32

0 ovalbumin
0.8
chymotrypsinogen A o a
0.4 myoglobin Pt7MT ---
CdMT
0.0
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2
R. / nm
1.4
ovalbumin
CP
- 1.chymotrypeinogen A myoglobin Cd1mT0 ridonucease A
0.6
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 R. / nm
Figure 3-2. Stokes' radii of Pt MT and Cd7MT determined by using calibrated Sephadex (a) G-75 and (b) G-50 sizeexclusion columns.




33

were determined by using Sephadex G-50 and G-75 columns calibrated against globular proteins with known Stokes' radii (Figure 3-2) [132].
The Stokes' radius of Pt7MT is consistent with
retention of the two domain structure of the native protein and Cd7MT. This is necessary but not sufficient evidence that Pt7MT has the two domain structure, or any fixed conformation, as we have estimated the Stoke's radius of a hypothetical random coil polypeptide of 61 residues to be very close to radii obtained for Pt7MT and Cd7MT (27].
We are not certain if Pt7MT is homogenous with respect to charge. Ion-exchange chromatography on DEAE cellulose against a 0.15 M KCl gradient gave a single extremely broad elution band. Nondenaturing polyacrylamide gel electrophoresis gave a similar result. The adduct recovered from the ion-exchange procedure contained 5.67 0.34 mol Pt/mol MT. The tenacious binding of Pt7MT to the ionexchange and electrophoresis media, and the loss of roughly one mol equiv of Pt2+, might be the result of a slow binding of Pt2+ ions to the pendant amino groups on the media. It is possible that one of the bound Pt2+ ions in Pt7MT is relatively exposed to solvent, possibly a Pt2+ ion bound to the N-terminal methionine. Binding of Pt(II) to the exposed methionine residues of proteins is well known [47] and Pt(II) salts have been used extensively as heavy atom labels by protein crystallographers [50].




34

Further experimental evidence supporting a bound Pt2+ at the methionine (met-1) was obtained by examining the extent to which Cd2+ and Pt2+ bind to apoMT chemically modified at the methionine residue. The N-terminal methionine of apoMT was converted to methionine sulfoxide, a much poorer ligand for Pt(II), by an oxygen atom exchange reaction with excess dimethylsulfoxide [133] to yield the modified MT and dimethylsulfide as the other product. Reconstitution with CdCl2 and K2PtCl4 followed by ICP analysis of the adducts showed that while the modified apoMT still bound the full complement of 7 Cd2+ ions (7.10 0.29 mol Cd/mol MT), the binding of Pt2+ was indeed decreased by approximately one mol equiv (5.79 0.60 mol Pt/mol MT). This evidence alone is not sufficient to prove that Pt2+ binds to met-1, but it is highly unlikely that this would not occur based on the known chemistry of Pt(II) and this amino acid.
The bound metal ions in Pt7MT are not liberated from
the protein under acidic conditions (as occurs for Cd7MT and other adducts). Dialysis of Pt7MT against 1 M HCl did not result in the loss of bound Pt2+ and no further binding occurs at this pH following the addition of excess K2PtCl The average binding constant for Pt7MT must be many orders of magnitude higher than that of Cd7MT which remains fully metallated only at pH > 5 [134].




35
The great stability of the Pt2+-to-cysteine-thiolate
complexes is further evidenced by the observation that apoMT reacts with excess K2PtCl4 at pH 2 to form polymeric adducts containing 17 1 mol Pt/mol MT. The degree of binding indicates that an essentially one-to-one Pt-to-cysteine coordination is occurring. The binding of slightly less than 20 Pt2+ ions and the polymeric nature of the products can both be attributed to intermolecular S-Pt-S bridging. As the amino acid sequence of MT contains three pairs of adjacent cysteines (Figure 1-1) it is also possible that a small proportion of the Pt2+ ions are coordinated in a bidentate fashion to two adjacent cysteines. The apparently random mode of binding at low pH may be a result of the much slower rate of binding or the reported random coil structure of apoMT at low pH [135].
Pt7MT is also quite stable to displacement of Pt2+ ions by other electrophiles, such as the oxidizing agent 5,5'dithiobis(2-nitrobenzoic acid) (DTNB) and the alkylating agent iodoacetate (Figure 3-3). At neutral pH, Pt7MT reacts with the cysteine assay reagent DTNB at an initial rate ca.
8 times slower than found for the reaction of the apo protein under the same conditions [37]. lodoacetate reacts with Pt7MT ca. 15 times slower than reported for the reaction of iodoacetamide with apoMT under the same conditions. These rate reductions are nearly twice those found for the reactions of Cd7MT with DTNB and iodoacetamide




NO2
COO
S
O cO
OO 0 co

ICH C o

+ RS

DTNB 10
5
0
0 2 4 6 Time / min

-6 10~0
C) 5

0

0

200 400 Time / min

Figure 3-3. Shown are (a) reactions of DTNB and iodoacetate with thiolates and (b) time courses for the reactions of 1.0 mM DTNB with 1.7 pM Pt7MT and 1.0 mM iodoacetate with 9.0 pM Pt7MT at neutral pH.

36

+ RS

sO -,
10
S

4

+

s
S
11O 2

a

2
U

RSCH2COO + I
ICH2COO~

b

600

3




37
[37], which is consistent with the known decrease in reactivity of the cysteine thiolates with stronger metal-tosulfur bonds. Roughly half of the 20 cysteines in Pt7MT reacted with either reagent. ICP analysis of the products showed that no Pt2+ ions were displaced from the protein in either of the two reactions.
All seven Pt2+ ions are liberated from Pt7MT, however, by dialysis against excess cyanide to form highly stable Pt(CN)42. Complete removal of Pt2+ from 35 p.M Pt7MT by equilibrium dialysis in 0.1 M NaCN at pH 7.0 was confirmed by both spectrophotometric and ICP-AES measurements of the products. ApoMT was recovered from this dialysis in its completely reduced form, i.e. no characteristic cystine disulfide peak at 280 nm was observed in the ultraviolet spectrum. Air-oxidation of cysteine pairs to cystine disulfide groups is presumably followed by redox reaction with CN~ yielding cys~ and cys-SCN [136]. Dialysis of Pt7MT against CN~ or other ligands with high affinity for Pt(II) under acidic conditions could be used to recover intact apoMT from Pt7MT.
Optical Spectra
The electronic absorption spectrum of Pt7MT consists of a broad region of absorption similar to the spectra of other metal MT adducts (Figure 3-4). The spectrum is consistent




38

20-

'-4
S
C)
-4
.'ji
0 'I

w

15-

105-

04
20

apo Zn Cd Pt

0

250

300

i5
350

i
400

nm

Figure 3-4. Electronic absorption spectra of apoMT, Zn7MT, Cd7MT, and Pt7MT.




39

with the presence of Pt2+ to cysteine thiolate coordination as well as possible coordination of Pt2+ to the thioether sulfur of met-1. As illustrated in Figure 3-4, Pt7MT follows the general trend of the absorption spectra of MT adducts with heavier metals to extend to longer wavelengths [2]. The spectra of Cd7MT and Zn7MT are similar in shape and intensity to those of Cd(II) and Zn(II) complexes of 2-mercaptoethanol which are arise from thiolate-to-metal charge-transfer transitions [134]. The spectrum of Pt7MT also contains weak d-d ligand field bands in addition to intense thiolate-to-Pt charge-transfer components.
The electronic absorption spectrum of Pt7MT is similar to the spectra of other metal MT adducts in that it is complex and difficult to interpret. Some interesting information about Cd7MT, however, has been obtained from a spectrophotometric titration of apoMT with Cd2+ ions [38]. We have performed a similar experiment to study the binding of Pt2+ ions to apoMT.
Figure 3-5 shows the spectra of partially platinated Pt(II) apoMT adducts recorded for a series of apoMT solutions equilibrated with 1-8 mol equiv of K2PtCl4 at pH
7.5. Also shown are the incremental difference spectra obtained by subtracting from each spectrum the spectrum preceding it. The simple approach of adding increments of K2PtCl4 to a single solution of apoMT was not practical here as Pt(II) binding to apoMT at these low concentrations is




40

20
210 a
0
200 250 300 350 400
X nrn
2-1
3-2
- b
6-5
1 6
0
200 250 300 350 400 X / nm
Figure 3-5. Spectrophotometric titration of apoMT with K PtCl at pH 7. Shown are (a) the spectra resulting from the addition of 0 to 8 mol equiv of K PtCl to apoMT and (b) the incremental difference spectra obtaineA by the subtracting the preceding spectrum from each spectrum.




41
very slow requiring about an hour to ensure equilibrium for each step. These long reaction times would be troublesome in the latter type of experiment as solutions of the apoMT at neutral pH are exceedingly sensitive to rapid airoxidation.
Binding of the first mol equiv of Pt2+ to apoMT results in an absorption peak centered at about 250 nm. A very intense shoulder of absorption appears at ca. 250 nm after binding of the second mol equiv of Pt2+. The incremental difference spectrum for the third mol equiv of Pt2+ contains two broad maxima at 250 and 300 nm. Mol equiv 4-7 result in absorption peaks of diminishing intensity at ca 280-300 nm. These data confirm the earlier finding that apoMT becomes saturated after binding 7 mol equiv of Pt2+
Wilner and coworkers observed a 6 nm red shift in the incremental difference spectra when apoMT was titrated with more than three mol equiv of Cd(II) [38]. A decrease in the intensity of the absorption in the later stages of the titration (>3 mol equiv Cd2+) was also observed. Good and Vasak also report a red shift in the electronic absorption spectra recorded during step-wise incorporation of Fe(II) into apoMT [31]. These red-shifts as the titration of apoMT exceeds three mol equiv of metal ions have been attributed to the onset of thiolato bridging.
The first three metal ions presumably bind to three separate sets of terminal thiolate ligands to form three




42
M(cys)42- units. Binding of the remaining four metal ions involves the formation of bridging thiolates from the remaining uncoordinated thiolates as well as transformation of some terminal thiolates to bridging thiolates [38]. Supporting evidence for this inferred mechanism was provided by spectra of Cd(II) complexes with the synthetic tetrathiol dodecapeptide, P-12 (Ac-P-C-Ornithine-C-P-Q-C-Q-C-R-R-V)
[38]. Titration of P-12 with Cd(II) yields sharp breaking points at 1:1 and 2:3 molar peptide-to-Cd ratios presumably due to the formation of a monomeric mononuclear complex and a dimeric trinuclear complex, respectively. The band maximum of the monomeric complex red-shifts ca. 6 nm as excess Cd2+ is added and bridging thiolates are created upon formation of the trimeric complex [38].
Further evidence for this mechanism of cluster
formation comes from an EPR study of Co(II) binding to apoMT done by Vasak and Kagi [137]. A linear increase in the room temperature EPR signal intensity occurs with titration of apoMT with 3-4 mol equiv of Co(II) ions followed by a linear decrease in intensity to zero with the addition of further ions to the saturation point of 7 mol equivalents. The EPR spectrum of Co7MT is characteristic of high spin tetrahedral Co(II) and is observed only at liquid helium temperatures. The loss of paramagnetism at higher temperatures is due to antiferromagnetic coupling between thiolato bridged Co(II) ions. The EPR titration data also




43
suggest that the first 3-4 Co(II) ions bind to separate sets of terminal thiolates and that bridging and cluster formation occur after this initial process.
We have obtained evidence that the above mechanism is operative in the binding of Pt(II) to apoMT by use of the colorimetric thiolate assay reagent, DTNB. As discussed in the previous section, DTNB reacts many times slower with thiolates coordinated to Pt2+ than with free thiolates. This difference in reactivity makes it possible to determine the concentrations of both coordinated and free thiolates at each step in the titration of apoMT with Pt2+. This experiment is not amenable to the study of Cd2+ and Zn2+ binding to apoMT as bound Cd2+ and Zn2+ do not sufficiently slow down the reaction of DTNB with thiolate.
Figure 3-6 shows the results obtained when each of the solutions used in the spectrophotometric titration were allowed to react for one minute with a 315:1 molar excess of DTNB:MT under anaerobic conditions. It is clear that the first 3 mol equiv of Pt2+ coordinate to 4 cysteine thiolates per Pt2+. Binding of the remaining 4 mol equiv of Pt2+ ions occurs by coordination to progressively fewer thiolate ligands per Pt2+ ion. Roughly two thiolates are found to react with DTNB during the time permitted after the apo protein is saturated with the full complement of seven Pt2+ ions. This residual reactivity is due either to two uncoordinated thiolates in Pt7MT or two relatively reactive




44
20
191817
516S15
-,14
1312
011
)10
9)
Q> 8
7
CO,
>5
C4
1
0
0 1 2 3 4 5 6 7 5
Mol Equivalents K2PtCl4
Figure 3-6. Titration of the partially platinated apoMT adducts from each step of the spectrophotometric titration with the thiolate assay reagent, DTNB.




45
coordinated thiolates being present. The reactivity of DTNB with Pt2+ bound thiolates will vary depending on many factors, primarily the bond strength of the individual group and the accessibility to the reagent solution. It is possible, for instance, that two bridging thiolates in Pt7MT are relatively exposed to the surrounding medium.
The changes in the maxima of the ultraviolet spectra that occur upon the addition of more than 3 mol equiv of Pt2+ to apoMT can be accounted for by the formation of bridging thiolates and the transformation of terminal thiolates to bridging thiolates. In attempting to explain the apparent red-shift of the ligand-to-metal chargetransfer (LMCT) absorption maximum that coincides with this change in the mode of Pt2+ binding, it is helpful to compare the electronic spectra of Pt(II) tetrahalides (PtX4 2-) with the spectra of their dihalo bridged dimers (Pt2X62-)'
Figure 3-7 is a molecular orbital diagram for a PtX42complex showing the three allowed LMCT transitions. The LMCT spectra of the square-planar tetrahalide complexes of Pt(II), Pd(II), and Au(III) consist of two intense bands separated by about 9000-12000 cm-1 [138]. The first, lowest energy, absorption band (E ; 1 X 104 M-1 cm-1) originates from the two closely spaced Ma* Lv transitions and the second, more intense, band (e z 4 X 104 M-1 cm-1) originates from the Mu* Lc transition. For the tetrachloro and tetrabromo complexes, the energy of the first allowed band




46

Pt

Pt L4

a (. 6p
b (o) d 2 y 2 6s
y d
eg (tr) dxz, dyz
, -. a Ig( a) d 2
I -2 5d
bS 2u ( iT" )
a2, TT
b
e g (TT b
b
b, 6
e~ y

bI
ag(b

-4, \I

L
S I'
10
SI I
It I
I,,
IS,
1.11
II It
I~
I I11t I SISIS S
II
~ It, 5'
.1515 II
II
Ii S
5,
I '~~Ii *SS~ I, IS
1 <4 iT III- 55
I ,li
- 11)1
SI I,
- It, t It
( 1~1 I/I
SI 15 I~,I ~ I'I
* III
*1
/ I
S ~tI
I II
'~' 0

Figure 3-7. Relative molecular orbital energies estimated for square-planar Pt(II) complexes.




47
increases in the order: AuBr4~ < PdBrg2- < AuCl4~ < PdCl42< PtBr42 < PtCl42. For PtX4 2- the Ma* La band usually occurs at wavelengths too low to be experimentally observed.
The dihalo bridged dimers of Pt(II) and Pd(II)
typically have intense absorption bands which lie 30-60 nm below the first LMCT bands of the corresponding monomers [139-142]. This behavior was observed for (Bu4N)2Pd2C16 and (Bu4N)2Pd2Br6 in glasses at 77 K in an early study by Mason and Gray [141] who speculated that these bands arose from the transfer of nonbonding r electrons from the bridging ligands to Pt.
However, more recent polarized crystal spectra of (Et4N)2Pt2Br6 [139] show that this first LMCT band must arise from transfer of r electrons from the terminal ligands. In order to understand this observation it is necessary to consider separately the two sets of molecular orbitals formed by the bridging and terminal ligands in the D2h dimer. Defining the x-axis as normal to the plane of the molecule, one finds that while the px orbitals of the bridging (Pxb) and terminal (pxt) bromines both form pure out-of-plane v MO's, the pyb and Pzb orbitals are involved in a bonding only, and cannot be combined into a and in-plane r MO's as obtained for pyt and pzt. Since the first intense band in the crystal spectrum of (Et4N)2Pt2Br6 (317 nm, 9000 M-1 cm-1) is the only strong band in z




48
polarization up to 270 nm, it must assigned to Mu* Lv transitions originating from the terminal bromines [139].
The occurrence of the first Ma* Lv LMCT band of (Et4N)2Pt2Br6 at a wavelength 49 nm longer than that of K2PtBr4 [143] can be explained by using a very simple argument [139]. The Pt atoms in the dimer should be more electronegative than the Pt atom in the monomer because there are fewer bromides per Pt in the dimer to donate electronic charge. It is not surprising, then, that the first LMCT transition in the dimer should occur at lower energy than in the monomer as it is well known, from the work of Jorgensen [144], that the energy of this band can be correlated with the difference in electronegativities of metal and ligand. Delocalization within the central Pt(p-X)2Pt region may also stabilize LMCT excited states of the dimers. The red-shifting and decrease in intensity of the band maxima of the incremental difference spectra in the later stages of the spectrophotometric titration of apoMT with K2PtCl4 (Figure 3-5a) are completely consistent with a decrease in the number of terminal thiolates versus bridging thiolates and a lowering in energy of the Ma* Lv LMCT transitions for these terminal thiolates due to presence of bridging thiolates.
It is also possible that some shifting of the LMCT bands to longer wavelengths results from a distortion of the ligand geometry at the Pt centers from D4h to D2d




49
symmetry. The lowest lying LMCT band of the tetrahedrally distorted Pt(II) complex with the sterically hindered PEt3 ligand (P-Pt-P angles are ca. 1500), to take an extreme example, is found to be red shifted about 30 nm from the corresponding band for the planar bis(1,2-diethylphosphino)ethane complex [145]. However, it is not necessary to invoke this effect, and large distortions from square-planar geometry in these Pt(II) MT adducts do not appear to be energetically feasible because of the very large ligand field stabilization energies that would have to be overcome. Distortion of PtBr4 2- (A, = 2.75 eV, A2 = 0.74 eV, A3 = 0.44 eV) to the tetrahedral limit (A ~ 1.22 eV = 4A,/9), for example, would theoretically require 2.71 eV [138]. By comparison, the Gibbs' energy of a typical intramolecular hydrogen bond in a protein is only about 0.20 eV in aqueous solution [146]. The strongest hydrogen bonding is found in the hydrophobic interiors of proteins [146], and the interior of MT is unusual in that it contains metal-thiolate clusters rather than a hydrophobic core of amino acid residues. It seems reasonable, then, to expect that the structure of a metal MT adduct is largely dictated by the energetics of metal-thiolate cluster formation. However, the net forces of intermolecular hydrogen bonding exerted by the surrounding polypeptide chain of metal MT adducts are likely to cause slight distortions, even in the case of Pt(II).




50
The circular dichroism spectra of apoMT containing 0 to 7 mol equiv of bound Pt2+ (Figure 3-8) are more difficult to interpret than the corresponding electronic absorption spectra. Ellipticity bands in the spectra of metal MT adducts contain contributions from both the conformationally dependent configurational chirality of the polypeptide and charge-transfer from the chiral cysteine thiolate ligands to the metal ions. Optical activity above 240 nm, however, contains only contributions from the asymmetric chargetransfer components [147].
The CD spectrum of apoMT at neutral pH is similar to its electronic absorption spectrum in that it is nearly devoid of spectral information except for a sharply decreasing negative ellipticity band beginning at about 240 nm and extending into the far ultraviolet. As shown in Figure 3-8, the addition of the first mol equiv of Pt2+ brings about a single positive ellipticity band centered at about 250 nm. Addition of mol equiv 2-3 result in a significant drop in the (+) intensity of the spectrum but very little change in its shape. The (+) intensity of the 250 nm maximum rises abruptly after the addition of 4 mol equiv of Pt2+ and remains fairly unchanged through 6 mol equiv, although there is some broadening of the spectrum in the region 250-300 nm. Binding of the seventh and final mol equiv of Pt2+ causes a large increase in the (+) intensity




51

4
-42
0
3
--1
C\2 Q) 2
C)
L j 0225 250 275 300 325 350 X / nm
Figure 3-8. Titration of apoMT with 0 to 7 mol equiv of K2PtCl4 as followed by circular dichroism.




52
of the spectrum giving a broad shoulder of ellipticity similar to that of the electronic absorption spectrum.
Abrupt changes in CD spectra upon addition of the
fourth mol equiv of metal ions have also been observed for partially metallated adducts of apoMT with Cd2+ and Zn2+ ions [21]. One interpretation of the spectral changes observed for Pt(II) binding is that major changes in the conformation of the protein occur during binding of the first 3 mol equiv of Pt2+ ions, which entails the coordination of 4 cysteine ligands for each Pt2+ ion as shown by the DTNB assays. The fact that the CD features undergo practically no change at all for adducts containing 4-6 mol equiv of Pt2+ is attributable to the inclusion of these additional bound ions into preformed metal-thiolate clusters via the formation of bridging thiolates. This interpretation relies heavily upon the common assumption that the CD spectra of metal MT adducts in general are ligand based, i.e. changes in rotary strength are linked to configurational changes in the cysteine ligands which occur with metal binding [2]. A similar lack of CD spectral changes following the addition of 4 mol equiv of metal ions has been found for the titration of apoMT with Cd2+ ions [21].
The large (+) increase in the CD intensity observed for the binding of the seventh and final mol equiv of Pt2+ is somewhat peculiar. We have suggested that one of the Pt2+




53
ions in Pt7MT is bound to the N-terminal methionine residue, and it is reasonable to expect that this neutral ligand is thermodynamically and kinetically less reactive than the cysteine thiolates [49]. It is expected, then, that binding to methionine will occur only after saturation of the cysteine thiolate ligands. The binding of the seventh Pt2+ ion to met-1 would produce a new chiral center at the thioether sulfur, and possibly a conformational alteration of the first several residues in this portion of the molecule, giving rise to a significant change in the CD spectrum.
The far-UV CD spectra of apoMT (Figure 3-9) and native MT (Figure 3-10) both contain a strong negative ellipticity band near 200 nm. This band is similar to that observed for random coil polypeptides and is a reflection of the lack of secondary structure in both the apo and metallated forms of MT as confirmed by the crystallographic structure (Figure 1-2). The native protein contains essentially no a-helix or B-sheet regions. The crystallographic structure does show a considerable number of B-turns which were also indicated by an earlier Raman and infrared absorption study which found MT to be a "predominantly turn-containing protein" [148]. The lower intensity of the 200 nm ellipticity band for native MT as compared with apoMT probably results from an increase in the number of B-turns in the folded metallated form of MT. A decrease in the




54

0
Pt7MT opoUT
0
0 -10
,8-2O
-3 0
40
r-I 30200 220 240
X / nm
Figure 3-9. Circular dichroism spectra of apoMT and Pt7MT in the far ultraviolet.




55

6
0
-66
0
U
Q)
10
-6
-12
190 220 250 280
X / nm
Figure 3-10. Circular dichroism spectra of native rat liver MT in 3 mM phosphate buffer pH 7.2.




56
intensity of this band for Pt7MT compared to apoMT probably also results from a similar structural organization during Pt2+ binding.
Nuclear Magnetic Resonance Spectra
NMR spectroscopy has become increasingly important in the study of the structure and dynamics of metallothionein and proteins in general. Otvos and Armitage correctly deduced the organization of the metal-thiolate clusters in MT by selective decoupling of the 113Cd NMR spectrum of 113Cd7MT [149]. Their structural model was reported seven years prior to the publication of the crystallographic structure, and was of central importance to MT research in the intervening years. Wuthrich and coworkers at the ETH in Zurich have employed recently developed two-dimensional 1H techniques to provide complete sequence-specific assignments for the 1H spectra of rabbit liver and rat liver Cd7metallothioneins [150, 151] and construct models for the solution structures of metallothioneins [152, 153].
MT is a relatively small biopolymer and thus amenable to study by 1H NMR. However, the signal dispersions in 1H spectra of MT are not as great as in most other proteins due to the lack of ring currents from aromatic residues. The principle difficulties in obtaining high quality 1H NMR spectra of proteins are the relatively low molar solubility




57
of proteins, the large number of resonance lines occurring within a narrow chemical shift range, and the broad line widths due to slow tumbling rates. These difficulties have been overcome to large degree by the higher fields and better probe designs of modern instruments.
The 300 MHz 1H spectra recorded for the apo and
native forms of rat liver MT-2 are shown in Figure 3-11. Because D20 solutions were employed, the solventexchangeable amide protons, which resonate in the range 7-10 ppm, were not observed. The spectra of apoMT and native MT are similar in their gross features. The resonances in the region 3.8-4.6 ppm are predominantly those of the a-protons. The intense peak at ca. 3.0 ppm contains the methylene resonances of the twenty cysteine residues. The sharp singlets at 2.09 and 2.03 ppm derive from the methyl protons of the methionine residue and its N-acetyl group, respectively. An envelope of lysine 6-CH2 resonances is centered at ca. 1.7 ppm. The methyl doublets of the alanine residues are grouped near 1.4 ppm, and those of the threonine residues near 1.2 ppm. The group of peaks near
1.0 ppm and below arise from methyl resonances of the aliphatic side chains of valine and isoleucine residues.
The spectrum of apoMT (Figure 3-lla) is practically identical to that of a mixture of the constituent amino acids [154] indicating that apoMT at low pH is essentially a random coil polypeptide with all amino acid residues sharing




58

a

*-~ 7 'A

9

<4~

) b

Figure 3-11. 300 MHz 1H NMR spectra of (a) native rat liver MT-2 in 20 mM phosphate buffer pH 7.2 and (b) rat liver apoMT-2 in dilute acetic acid pH 2 (the multiplet at ca. 2.05 ppm, between the two sharp singlets, is from HD CO2D). Resolution enhancement was employed (see Materials and Methods).

-v
1~

j

3 5 30

I15 1 PP




59

the same average environment. The spectrum of the Cd2+ and Zn2+ containing native MT (Figure 3-11b), on the other hand, shows noticeable perturbations within the resonance envelope for a particular amino acid, with the most striking effects revealed within the alanine and threonine methyl regions.
A disperse lysine 6-CH2 proton envelope (1.6-1.8 ppm)
appears in the spectrum of the folded metallated form of the protein. The lysines in apoMT, on the other hand, appear to share the same average environment and give rise to a single well defined quintet at 1.68 ppm. MT has a high proportion of lysine residues (CH2CH2CH2CH2NH2 side chain), located almost exclusively adjacent to cysteines in positions which are highly conserved among the amino acid sequences of all mammalian metallothioneins (cf. Figure 1-1). Kojima [73] suggested that these lysines, which are protonated at physiological pH, serve to neutralize the net negative charge borne by the metal-thiolate clusters thereby stabilizing the molecule. Vasak, Kagi, and coworkers [155-157] have provided experimental evidence which support this type of interaction involving the lysine residues.
In the spectrum of apoMT the cysteine B-CH2 resonances (40 protons) centered at about 3.0 ppm overlap with E-CH2 resonances of the 8 lysines (16 protons). In the spectrum of native MT the cysteine B-CH2 resonances are spread out over range of about 0.5 ppm under the base of this peak [151]. However, a comparison of the area under the 3.0 ppm




60
peak with that of the 0.9-1.1 ppm envelope (valine and isoleucine methyls, 12 protons) confirms that the majority of the cysteine B-CH2 resonances are still centered at 3.0 ppm in native MT.
The chemical shifts are sensitive to the metallic
elements present. Mitsumori and Tohyama [158] have shown that when the Zn2+ ions in native MT (67:37 Cd:Zn ratio) are displaced by Cd2+ ions, to form Cd7MT, the apparent triplets at 1.2 and 1.3 ppm coalesce into two doublets. The heterogenous metal composition of native MT results in two slightly different environments for threonines 9 and 27 causing a superimposition of a pair of doublets for each residue giving the two "pseudo-triplet" shapes. The native MT sample whose spectrum appears in Figure 3-11 contained
3.18 0.10 mol Zn/mol MT and the presence of bound Zn2+ is apparent from the pseudo-triplets of the two threonine residues.
The nature of the metal binding sites in rat liver MT-2 is revealed in a comparison of the 1H spin-echo spectra of apoMT and native MT (Figure 3-12). Spin-echo experiments generally employ the pulse sequence 900-r-1800-r-echo.
x y
The echo has the appearance of two free induction decays (FID's) placed back to back. The FID which follows the echo maximum may be recorded and Fourier transformed to obtain the conventional spectrum. Each peak in the resolved echo, however, is modulated at a frequency which is determined by




61

a

;~?; ,~ *,
* *~d'~~".'\' .2~ti.....< II

b

Figure 3-12. 300 MHz 1H spin-echo NMR spectra of (a) native rat liver MT-2 in 20 mM phosphate buffer pH 7.2 and
(b) rat liver apoMT-2 in dilute acetic acid pH 2. Resolution enhancement was employed (see Materials and Methods).




62
the J value of the given nucleus. The phase and amplitude of each peak in the spin-echo spectrum, then, is a function of this modulation due to spin-spin coupling and the length of T.
Note that the phase of the cysteine B-methylene
envelope (ca. 3.0 ppm) in the spin-echo spectrum of native MT (Figure 3-12a) is reversed from that of apoMT (Figure 312b). The phase of a spin-echo resonance line is sensitive to spin-spin coupling [159], so coordination of metal ions to the cysteine residues must alter the coupling of these protons to each other and the cysteine a protons. The only other apparent phase reversal is for the peak farthest upfield, which is that of the methyl protons of valine 39. This is peculiar as binding of metal ions to the isopropyl side chain of this residue does not occur, but an interaction of the carbonyl oxygen or amide nitrogen with a metal ion cannot be ruled out. This spin-echo experiment was adapted from a similar experiment reported by Galdes et al. for equine kidney MT (160].
The conventional 1H NMR spectrum and spin-echo spectrum of Pt7MT are shown in Figure 3-13. The familiar resonance envelopes for the different amino acid spin systems, that are seen for the apo and native protein, are also found in about the same chemical shift ranges in the 1H spectrum of Pt7MT. Unfortunately, even very concentrated samples of Pt7MT gave only a very weak signal. The spectrum is of




63

a

I
V
~

b

Figure 3-13. 300 MHz 1H (a) NMR spectrum and (b) spin-echo NMR spectrum of 0.6 mM rat liver Pt7MT-2 (6.96 0.41 mol Pt/mol MT) in 20 mM phosphate buffer pH 7.2. Resolution enhancement was employed (see Materials and Methods).

\.~ ~ I~~4'

~!

g




64
relatively poor quality due to the low signal to noise ratio. The low S/N may be the result of heterogeneity in the binding of Pt2+ to individual MT apoMT molecules, slower T relaxation rates due to the larger size of Pt7MT compared to apoMT and other metal MT adducts or more restricted intramolecular motion due to Pt2+ binding, or some other unknown reason. A slight improvement in signal strength was obtained by using longer delay times between pulses, which indicates that relaxation processes are at least part of the problem in this case.
Coordination of the cysteine residues to 195Pt (natural abundance: 32%) causes a 32% reduction the intensity of the B-CH2 resonances. The 195Pt satellite peaks are broadened beyond detection at this high applied field as a consequence of chemical shift anisotropy [161]. The 3.0 ppm region in the spectrum of Pt7MT (Figure 3-13a) shows very little intensity, even taking into account coupling of the cysteine CH2 protons to 195Pt. This is not necessarily due to heterogeneity in Pt7MT binding sites, however, as these resonances are spread over a large range (ca. 0.5 ppm) in the homogeneous adduct, Cd7MT [151].
Examination of the spectra for apoMT and native MT
(Figure 3-11) reveals that the cysteine B-CH2 peak in the spectrum of native MT is shifted about 0.1 ppm downfield from that found for apoMT. This same deshielding effect is seen also in Figure 3-13a for Pt7MT. What was observed as a




65
single broad 3.0 ppm resonance region in the spectrum of native MT occurs as three separate peaks in the spectrum of Pt7MT. Binding of the cysteines to Pt2+ may cause more pronounced deshielding of some the B-CH2 protons than occurs with Cd2+ and Zn2+ binding. This effect is consistent with the greater electron-withdrawing tendency of Pt(II) versus Cd(II) or Zn(II), as seen in the electronic absorption spectra.
The two sharp methyl singlets from the acetylated
methionine residue which appear in the other MT spectra are noticeably absent in the spectrum of Pt7MT. There is a relatively sharp peak occurring at 2.05 ppm which can be assigned to the acetyl group methyl but the methyl resonance of met-1 is either shifted to another region of the spectrum or broadened beyond detection.
Gummin and coworkers have recently investigated the NMR properties of the complex formed between K2PtCl4 and N-acetylmethionine [162]. The CH3S peak (1.96 ppm) shifts downfield by 0.31 ppm upon complexation of Pt2+ by the sulfur atom and splits into a "1:4:1" pattern due to 195Pt coupling. The two 195Pt satellites were broadened beyond detection in the 300 MHz spectrum. The methyl singlet of the acetyl group was unaffected by complexation. These results are in agreement with the earlier finding by Boswell, et. al, that upon complexation with K2PtCl4 the




66
CH3S resonance line of this compound is shifted 0.30 ppm downfield from 2.04 ppm for the free ligand [163].
From the above results it is expected that coordination of Pt2+ to met-1 of MT would cause the CH3S peak at 2.09 ppm to vanish while leaving unchanged the CH3C(O) peak at 2.03 ppm. This is exactly what is observed in the spectrum of Pt7MT, strongly suggesting that coordination of Pt2+ to the thioether sulfur of met-1 does occur in this adduct. However, predicting the new location of the CH3S peak in the spectrum of Pt7MT is problematic. The direction and extent of the shift is dependent on the particular protein molecule in which the methionine residue resides. In the complex formed by K2PtCl4 and met-65 of tuna ferrocytochrome c, for example, this peak is actually removed 0.15 ppm upfield from the position of this peak in the free protein [163].
Bell and coworkers have recently examined the 500 MHz 1H spin-echo NMR spectra of cell culture media before and after addition of K2PtCl4, cisplatin, and Pt(en)Cl2 [164]. They report the disappearance of the methionine CH3S peak at
2.15 ppm and the appearance of a new peak at ca. 2.71 ppm after addition of each of the above reagents. It is interesting, in light of this report, that a fairly sharp peak is observed at 2.75 ppm in the spectrum of Pt7MT. This peak may be that of the methionine CH3S coordinated to Pt2+ but this peak may lie elsewhere in the spectrum, perhaps obscured by other resonances.




67
The spin-echo spectrum of Pt7MT in Figure 3-13b is similar to that of the native protein (Figure 3-12b) in that the phases of the various resonance regions are the same. The only apparent difference is in the phase of the methyl doublet of the isopropyl side chain of valine 39 that appears farthest upfield in the two spectra. The phase of this peak is positive in the Pt7MT and apoMT spin-echo spectra but, but as noted earlier in this section, it is negative in the native MT spectrum. The large positive cysteine 3.0 ppm peak confirms that cysteines are involved in Pt2+ binding.
195Pt NMR appeared to be a very promising method of studying the structure of Pt7MT at the outset of this investigation. The high natural abundance, good receptivity, and extremely large known chemical shift range of this spin 1/2 nucleus make it possible, in principle, to determine the number of distinct Pt-thiolate clusters and how the Pt2+ ions in these clusters are arranged by the spin-coupling in the spectrum as was done for Cd7MT by using
-13Cd NMR [165, 166]. Unfortunately no 195Pt signal was observed from concentrated samples of Pt7MT using either the 200 or 300 MHz instruments. Attempts by Ismail and Sadler to observe 195Pt signals from Pt2+ bound to other macromolecules also ended in failure, and this was attributed to relaxation by chemical shift anisotropy [167].




68
A similar difficulty in obtaining 199Hg NMR spectra of Hg7MT has been encountered [25].
Binding Kinetics
Unlike the intensive research activity dedicated to
learning about the structures and equilibrium properties of MT adducts, the kinetics of metal binding to apoMT is an essentially unexplored area. The main reasons for this lack of interest probably include the difficulty in modeling a complex system containing multiple binding sites and the prohibitively rapid binding rates of the physiologically prevalent Cd2+ and Zn2+ ions. The extreme inertness of Pt(II) complexation can be used to advantage in this regard as it was in the previously described titration of partially platinated MT derivatives with DTNB.
Changes in the ultraviolet absorption spectrum provide a convenient albeit crude measure of the extent of reaction of K2PtCl4 with apoMT. The complex nature of the spectrum does not translate into a simple relationship between intensity at a given wavelength and concentration of reactants. It provides only a very rough measure of the concentration of bound Pt2+ ions.
Absorbance was monitored versus time at 280 nm for a micromolar solution of apoMT mixed with a 50-fold molar excess of K2PtCl4 at room temperature. This wavelength was




69
chosen as the absorption at 280 nm increases in a nearly linear fashion versus mol equiv of bound Pt(II) (Figure 35b). The data may be fitted as two pseudo-first-order processes to a logarithmic plot of (A, A), which is proportional to the concentration of apoMT, versus time (Figure 3-14a). This treatment is consistent with either consecutive or parallel reactions leading to the formation of Pt7MT. The plot of log(A. A), in Figure 3-14b was obtained by subtracting from the biphasic plot the contribution of the slow step II, i.e. log(A00 A),,, after extrapolating step II to zero time.
One interesting observation about this treatment of the data is that the ratio of A. for step I to that of step II is 0.32/0.75 = 0.43 = 3.0/7.0. The obvious interpretation is that the fast phase is the binding of the first 3 Pt2+ ions to the apoMT molecule, which occurs by the formation of three Pt(cys)4 2- moieties, and that incorporation of the remaining 4 Pt2+ ions, which must proceed via the formation of some bridging thiolates, is the slower step II. It is reasonable to suppose that binding of the first 3 Pt2+ ions would occur more rapidly due to the abundance of solventexposed thiolate ligands resulting from the greater freedom of motion of the loosely coiled polypeptide during the early stages of binding. The number of binding sites available to an incoming Pt2+ ion during the later stages of binding is




70

-0.2
-0.3
< -0.4
-0.5
0
-0.6
-0.7
0

-0.5 -

0~
0

-1.0
-1.5 -

-2.0
0

a
0
0
10 20 30 40 Time / min

b

0

5

10
Time / min

1
15

Figure 3-14. Shown are (a) the time course of the reaction of 10 p.M apoMT with a 50-fold molar excess of K PtCl in 15 mM phosphate buffer pH 7.5, and (b) the plot obiainea by subtracting the contribution to the absorbance of the slow component of the reaction, A11. At, is the absorbance at 280 nm estimated for infinite time.




71

-0.7 1

-0.8
-0.9
% -1.0
o -1.2
-1.3
-1.4
-1.5
0

0
0

20 40 60 80
Time / min

Figure 3-15. Time course of the reaction of 10 p.M apoMT with a 50-fold molar excess of K2PtCl4 in 0.01 M HCl. A0 is the absorbance at 280 nm estimated for infinite time.

I 1 100 120




72
more restricted, on a purely statistical basis, than during the initial formation of the Pt(cys)4 2- units.
This interpretation assumes that the binding of each
Pt2+ ion produces an equal increase in the absorbance at 280 nm. Of course, this assumption about the additiveness of bound Pt2+ absorptivities is only approximately correct as may be seen by the titration data in Figure 3-5.
The slower binding of excess Pt2+ to apoMT at low pH
does not exhibit biphasic kinetics. The data yield a single pseudo-first-order process (Figure 3-15). As was seen in the 1H NMR spectra, apoMT at low pH is essentially a random coil and this is reflected in the binding of roughly 17 Pt2+ ions per MT to form a polymeric mixture.
Conclusions
At the outset of this project it was anticipated that Pt(II) would probably form a quite stable adduct with apoMT at neutral pH. The retention of bound Pt2+ at very low pH was not anticipated. Nor was it expected that seven squareplanar Pt(II) ions would be chelated per MT molecule, which happens to be the same stoichiometry found for Cd2+, Zn2+, and other divalent metals with tetrahedral stereochemistry. The apparent similarities between Pt7MT and these other adducts, besides the stoichiometry, are somewhat surprising in light of the two very different coordination geometries.




73
The large Stokes' radius of Pt7MT suggests an
elongated two-domain structure similar to that of native MT. The electronic absorption and CD spectra obtained in the step-wise titration of apoMT with K2PtCl4 follow a pattern very similar to that reported for the titration of apoMT with CdCl2 [38]. However, the large change in the CD spectrum attending the binding of the seventh Pt2+ ion has no analogue in the case of Cd2+ binding [21], and this spectral change suggests Pt2+ binding at the N-terminal methionine which is not a binding site for Cd2+. The DTNB assays of the partially platinated apoMT samples indicate the initial formation of three Pt(cys)4 2- units followed by incorporation of the remaining Pt2+ ions in bridged metalthiolate clusters. A similar binding sequence has been proposed for the formation of Cd7MT [38].
That such cluster formation should occur in Pt7MT is
not surprising in view of the great propensity of classical thiol ligands to act as bridging ligands in their Pt(II) complexes. Simple thiols react with Pt(II) to form amorphous [Pt(p-RS)2]n polymers (R = CH3, CH3CH2' C6H5' etc.) [49]. The fundamental differences between Pt7MT and adducts of apoMT with the lighter metallic elements, in addition to the different stereochemistries, are the much greater strength of the Pt2+ to sulfur bonds, the much slower rate of Pt2+ binding, and the probable binding of one the seven Pt2+ ions at met-1.




74
The process of metal ion binding at neutral pH to form the fully metallated form of MT from the apo form takes a course intermediate between two extremes. The first extreme is that of a rigid apoMT with preformed binding sites into which the metal ions are held. The second extreme is that of a random coil polypeptide that wraps around the metal thiolate complexes. ApoMT is essentially a random coil at neutral pH as its 1H NMR spectrum is nearly identical to that of a mixture of the component amino acids [135]. The binding of Pt(II) to apoMT is probably a fairly random process under kinetic control. If the initial formation of Pt(cys)42- units is sufficiently faster than the subsequent phase of binding which entails thiolato bridging, as suggested by the kinetic data in Figure 3-14, it is not necessary that there be any thermodynamically favored binding sites in order for the type of spectral changes in Figure 3-5b to be observed.
It is possible, for instance, that the binding of 7 Cd(II) ions to apoMT at neutral pH is a random process producing a distribution of kinetic products followed by intramolecular reorganization of the Cd(II)-thiolate bonds to form a single homogeneous Cd7MT as the thermodynamic product. 113Cd NMR studies of Cd7MT [39] and metal ion exchange between Cd7MT and Zn7MT [20] and a recent variable temperature CD study of Cd(II) binding to Zn7MT by Stillman




75
and Zelazowski [22] all suggest that metal ions are mobile within a cluster.
Although the initial binding process and formation of metal-thiolate clusters to form Pt7MT appears quite similar to those observed for adducts with Cd(II) and other metal ions, it is possible that the arrangement of Pt(II) ions among the 20 cysteine residues in the equilibrium product at room temperature differs from one molecule to the next due to the extreme kinetic inertness of Pt(II) which could prohibit the type of intramolecular rearrangement described above. The 1H NMR spectrum of Pt7MT (Figure 3-13a) is similar to the spectrum of native MT in its gross features but has a relatively weak signal strength even for very concentrated samples, which suggests this type of inhomogeneity. The weak 1H NMR signal may also be a consequence of unusual relaxation properties of Pt7MT or other unknown factors however.
It was hoped that the application of 195Pt NMR would reveal the structural arrangement of these Pt(II)-thiolate clusters. This has not been realized due to unforeseen experimental difficulties. We must instead garner other small pieces of chemical and physical evidence to build a structural model. The discussion of a structural model for Pt7MT is postponed until the end of Chapter 4, wherein additional data is presented regarding the arrangement of




76
Pt2+ ions in Pt7MT and mixed metal Pt(II) adducts prepared via displacement of Cd2+ and Zn2+ from Cd7MT and native MT.
Materials and Methods
General Procedures
Equine kidney MT-1A and rat liver MT-1 and MT-2 were isolated and purified according to a literature procedure [130]. ApoMT was prepared by exhaustive dialysis of the native protein against 0.01 M HCl. A small amount of the metal-chelating agent diphenylthiocarbazone in CC14 was suspended in the 0.01 M HCl.
All chemicals employed were reagent grade or better. Deionized water and buffer solutions were passed through a column of Chelex chelating resin to remove trace metal contaminants. Dialysis tubing (Spectrapor MW cut-off: 3500) was boiled repeatedly in deionized water and EDTA solution prior to use.
Manipulations of oxygen-sensitive materials were done in a glove-box with nitrogen atmosphere and oxygen furnace. Solutions were thoroughly degassed on a vacuum line prior to transfer to the glove-box. Absorption and CD spectra were recorded on an IBM 9430 UV-visible spectrophotometer and a Jasco J-500 spectropolarimeter, respectively.
Sephadex G-50 and G-75 size-exclusion columns were calibrated for Stokes' radius against ovalbumin (R0 =




77
2.73 nm), chymotrypsinogen A (2.09), myoglobin (1.88), ribonuclease A (1.64), and blue dextran (co). R, was plotted against (-logKd), where Kd is the elution coefficient [132]. The eluant was 3.5 mM phosphate buffer, pH 7.25, and the elution was monitored continuously by absorbance at 229 nm. Flow rate was controlled by a calibrated peristaltic pump. Protein containing fractions were concentrated by using an Amicon YM5 or YM2 ultrafiltration membrane.
Analytical Method
Elemental Analyses were done by using a Perkin-Elmer
Plasma II Emission spectrometer equipped with a Perkin-Elmer Series 7000 computer. The emission lines scanned were S I 180.73 nm, Cd II 214.44, Zn I 213.86, 324.75 and Pt II 214.420 nm. All protein samples were run side-by-side with ultrafiltrate or dialysate blanks. The instrumentation and methodology are described in Chapter 2.
Preparation of Pt(II) ApoMT Adducts
Samples of Pt7MT were prepared by adjusting solutions of apoMT in 0.01 M HCl to pH 7 with K2HPO4 followed by the addition of a 10 to 100-fold molar excess of K2PtCl4. After standing for 12 to 24 hours, the reaction mixtures were fractionated on a Sephadex G-50 or G-75 size-exclusion column, and the MT containing fractions pooled and




78
concentrated by using an Amicon YM2 or YM5 ultrafiltration membrane.
To study the extent of Pt binding to apoMT at low pH, solutions of apoMT in 0.01 M HCl were allowed to react with excess K2PtCl4 over 24 hours. The extent of binding was determined by ICP-AES after stripping off unreacted Pt2+ by ultrafiltration and desalting the equilibrium mixture on a size-exclusion column at neutral pH.
Binding of Pt(II) to ApoMT Modified at Methionine
The N-terminal methionine of apoMT was converted to methionine sulfoxide by using the method described by Schechter [133]. ApoMT in 1 M HCl was allowed to react with 1 M dimethyl-sulfoxide for 30 minutes. The resulting dimethyldisulfide and unreacted DMSO were removed by repeated dialysis against 0.01 M HCl. Cd(II) and Pt(II) adducts of this modified apoMT were prepared in the glovebox at neutral pH as by using the procedure described in the preceding section. The Cd(II) and Pt(II) adducts of the modified apoMT were assayed for Cd, Pt, and S by ICP-AES.
Reactions of Pt7MT with DTNB and Iodoacetate
Buffered deaerated solutions of 3.8 pM Pt7MT (1 mL) and freshly prepared 2.0 mM 5,5'-dithiobis(2-nitrobenzoic acid) (1 mL) were mixed in a cuvet in the glove-box. The cuvet was sealed and transferred to the sample compartment of the




79
UV-vis spectrophotometer and the increase in absorbance at 412 nm versus time recorded. The absorbance of a blank containing all reagents except protein was subtracted from the absorbance of the sample. The final pH was 7.4.
The extent of reaction at equilibrium was determined by passing the equilibrium mixture through an ultrafiltration membrane and then measuring the absorbance of the ultrafiltrate at 412 nm versus that of a blank. The extent of reaction was calculated using the molar absorptivity of the resulting dianion of 2-nitro-5-thiobenzoic acid, 14150 M-1 cm-1 at 412 nm [168].
The rate and extent of the reaction of 9.0 p.M Pt7MT
with 1.0 mM iodoacetate were obtained at pH 7.2 by measuring the release of I- in a stirred cell with a calibrated Iselective electrode (Lazar Res. Lab. IS-146).
Reaction of Pt7MT with Cyanide
1.00 mL of 35 p.M Pt7MT was dialyzed against 4.00 ml of
0.1 M NaCN in 0.1 M phosphate buffer pH 7.0 for 12 hrs during which the yellow color inside the dialysis bag was gradually lost. The Pt was found to be evenly distributed between of the contents of the bag and the surrounding solution by ICP-AES analysis. Absorption bands characteristic of Pt(CN)4 2- were found in the ultraviolet spectrum of the solution.




80
Spectrophotometric Titration of Rat Liver ApoMT-1 with Pt(II) and DTNB
A series of 3.5 p.M apoMT solutions containing from
0 to 10 molar equivalents of K2PtCl4 were prepared in the glove-box. ApoMT solutions in 0.01 M HCl were adjusted to pH 7 with K2HPO4 and then diluted to 0.65 mL after addition of K2PtCl4 increments. The reaction mixtures were left standing for three hours to equilibrate. A solution of 12 mM DTNB buffered at pH 7.5 was also prepared in the glove-box.
The equilibrium mixtures were transferred to a 1 cm cuvet which was sealed with a greased septum. A 100 pL syringe charged with 60 p.L of the DTNB (5,5'-dithiobis(2nitrobenzoic acid)) solution was then mounted atop the cuvet, with the needle protruding through the septum, using a short length of tightly fitting rubber tubing. The DTNB reagent was held aloft over the protein solution by a small dead-space. This assembly was then quickly transferred to the sample compartment of the spectrophotometer.
After recording the spectrum of the equilibrium
mixture, the DTNB in the syringe was discharged into the cuvet. A spectrum was recorded one minute after mixing. The cuvet and syringe were then cleaned and returned to the glove-box and the above procedure repeated for each of the remaining samples. All spectra were recorded at slow scan speed, with air in the reference beam, and using the same




81
cuvet. Spectra of blanks containing all reagents except protein and K2PtCl4 were subtracted from sample-spectra. Free thiolate concentrations were estimated using the reported molar absorptivity of the dianion of TNB (2-nitro5-thiobenzoic acid), 14150 M~1 cm-1 at 412 nm [168].
Titration of ApoMT with Pt(II) Followed by CD
A series of 10 p.M apoMT solutions containing 0 to 10
mol equiv of K2PtCl4 were prepared in the glove-box. ApoMT solutions in 0.01 M HCl were adjusted to pH 7.5 with NaOH and K2HPO4 and then diluted to volume after addition of K2PtCl4 increments.
CD spectra were recorded for the samples using four
labeled 1 cm cuvets sealed with greased septa. CD spectra of blanks containing all reagents except protein and K2PtCl4 were recorded in the four cuvets and these blank-spectra were subtracted from the corresponding sample-spectra taken in the same cuvet. Five replicate scans were recorded for each sample and blank and signal-averaged.
NMR Spectra
Samples were prepared by lyophilizing solutions of
protein in dilute phosphate buffer and then dissolving the resulting solids in D20. This procedure was performed twice for each sample in order to remove residual HDO. The samples were then transferred to 1 mm sample tubes.




82
All H NMR spectra were run at 220C on a Varian VXR 300 spectrometer. Chemical shifts were referenced to external DSS (3-(trimethylsilyl)-l-propanesulfonate). The HDO signal was suppressed via Ti discrimination, i.e. by applying a 1800 pulse and 2 s delay prior to each 900 pulse. Resolution enhancement was obtained, for the spectra indicated, by weighting the FID by exp(t/K), where K 0.5, and a Gaussian apodization function.
Spin echo spectra were obtained by the Carr-Purcell
method [169]. The pulse sequence 900-r-1800-7-collect was used with T = 50 ms.
Binding Kinetics
A syringe containing 80 iL of 4.95 mM K2PtCl4 was mounted atop a cuvet containing 720 p L of 11.1 pM apoMT buffered at pH 7.5 (50:1 molar ratio Pt:apoMT) as described above for the spectrophotometric titrations. This assembly was transferred from the glove-box to the spectrophotometer. The contents of the syringe were discharged into the cuvet and the increase in absorbance versus time at 280 nm was recorded on a strip-chart recorder.
A similar experiment was conducted for a solution of apoMT in 0.01 M HCl. ApoMT is stable with respect to airoxidation at this pH so the manipulations were performed on the bench top.




CHAPTER 4
DISPLACEMENT OF CADMIUM AND ZINC FROM
METALLOTHIONEIN BY PLATINUM(II)
The preceding chapter described experiments with
simple systems consisting of solutions apoMT and K2PtCl4. In this chapter the more complex and biologically relevant problem of the formation of platinum(II) metallothioneins by displacement of Cd2+ and Zn2+ ions from native MT is considered. The first problem to be examined will be the reaction of K2PtCl4 with Cd7MT. This is followed by a discussion of the binding of Pt(II) from K2PtCl4 and the derivatives of the anticancer drug, cisplatin, to heterogenous native metallothioneins, which contain predominantly Cd2+ and Zn2+ as well as lesser amounts of Cu+. The results presented in Chapter 3 for the binding of Pt(II) to apoMT are used in an attempt to understand the adducts obtained by displacement of metal ions from the above adducts and a structural model for Pt(II)-thiolate clusters is described. Finally, implications for the metabolism of cisplatin in vivo are outlined.

83




84
Reactions of Native and Cadmium
Metallothioneins with Tetrachloroplatinate
An experiment involving reactions of K2PtCl4 with
Cd7MT gave somewhat unexpected results [27]. Our original intention was to attempt to produce Cd7MT with a single Pt2+ bound to the thioether sulfur of the N-terminal met-1 residue. We assumed that the Cd2+ ions in Cd7MT were bound strongly enough to prevent the Pt2+ ions from reacting with the cysteine residues. It was found, however, that the addition of one mol equiv of K2PtCl4 or a slight molar excess of K2PtCl4 resulted in the liberation of 3 mol equiv of Cd2+ ions from Cd7MT (Table 4-1). In the presence of air a single Pt2+ ion was bound whereas in the exclusion of air
4 Pt2+ ions were bound to produce a Pt4Cd4MT.
The retention of 4 mol equiv of Cd2+ ions in all the
above cases strongly suggests that Pt(II) is displacing Cd2+ from the 3-Cd B cluster and leaving the 4-Cd a cluster intact. This result is a reasonable one as Cd2+ shows a marked preference for the a cluster in native MT and tends to accumulate in this cluster as opposed to the more weakly bound Zn2+ which occupies the B cluster almost exclusively. Furthermore, it is known that the B cluster region in Cd7MT and native MT is susceptible to attack by the strong electrophiles DTNB and iodoacetamide while the a region are generally inaccessible to these reagents [37].




85

Table 4-1. ICP-AES of the Products From Reactions of Cd7MT and Native Equine MT with K2PtCl4.
Reactants Metal Contents of Products (conditions)a
Metal Mol/Mol MTb

5.5 p.M Cd7MT + 5.8 p.M K2PtCl4 (in air)

8.2 p.M Cd7MT + 85 (in air)
6.9 p.M Cd MT + 55 (under nitrogen)

p.M K2 PtCl4 p.M K2PtCl4

native MT

28 p.M native MT + 990 p.M K2PtCl4 (under nitrogen)

Cd
Pt
total

3.82 1.11
4.93

Cd 4.01 Pt 1.16 total 5.17

Cd
Pt
total
Cd
Zn Cu
total

4.28 3.71 7.99
5.04 0.94 0.63 6.61

+
+
+
+
+
+
i+ i+ t+ i+
+

0.87 0.25 0.91
0.12 0.04 0.13
0.25
0.22 0.33
0.36 0.08
0.23
0.43

monomeric products
Cd 3.33 0.49 Zn 0.25 0.10 Cu 0.94 0.21 Pt 3.62 0.42 total 8.14 0.69
oligomeric products Cd 5.04 0.55 Zn 0.18 0.11 Cu 0.10 0.14 Pt 5.27 0.53 total 10.59 i 0.78

a All reactions mixtures were at pH 7.2, room temperature. b (mean of sample mean of blank) (standard deviation)
for 6 replicate measurements of both sample and blank.




86
Templeton and coworkers have found that 113Cd NMR lines attributed to Cd2+ ions in the B cluster are lost before those attributed to Cd2+ ions in the a cluster during the displacement of Cd2+ from Cd7MT by mercuribenzoate [170].
Rapid air oxidation of cysteine to cystine disulfide apparently occurs during the disruption of the B cluster by entering Pt2+ ions. The ultraviolet spectrum of the product has absorption in the region 250-320 nm that is much too intense to be accounted for by the metal chromophores alone. Some of this absorption must arise from cystine disulfide groups which have absorption maxima at 270-280 nm. This oxidation of the cysteine side chains would abolish metal binding in this cluster. The single mol equiv of Pt2+ which is found in the final product is probably bound to the thioether sulfur of met-1, which does not undergo airoxidation under these conditions. The reaction of Pt2+ ions with the cysteine sulfurs must be much faster than the reaction with methionine sulfur as the Cd2+ ions are lost and oxidation of the cysteines occurs even for the equimolar mixture of K2PtCl4 and Cd7MT.
The reaction of excess K2PtCl4 with Cd7MT under
nitrogen to produce Pt4Cd4MT suggests that the B cluster region can accommodate 3 Pt2+ ions, assuming that the remaining Pt2+ is bound to met-1. Extending this line of reasoning, the arrangement of Pt2+ ions in Pt7MT may be that




87

of two separate 3Pt-thiolate cluster regions with the seventh Pt2+ ion bound at the methionine.
The products from the anaerobic reaction of excess
K2PtCl4 with native MT were divided about equally between an oligomeric fraction and a monomeric fraction when chromatographed on a size exclusion column. The Stokes' radius of the monomeric products is 0.2 nm larger than that of original native MT, i.e. the same as that of Pt7MT (ca. 1.8 nm). The oligomeric fraction was composed of a large dimer peak and progressively smaller trimer, tetramer, and higher oligomer peaks, which were pooled for analysis by ICP-AES (Table 4-1). The monomeric products show a 75% deficiency in Zn content but only a 35% deficiency in Cd content as compared to the native MT used as starting material. This result is consistent with displacement of metal ions occurring predominantly in the B cluster as it is known that Zn2+ ions are localized in this region of the native protein. The metal contents of the monomeric fraction are similar to those obtained for the anaerobic reaction of Cd7MT with K2PtCl4. In both cases 4 mol equiv of Pt2+ ions are gained and 3 mol equiv of the initially bound metal ions are lost.
The metal contents of the oligomeric fraction are quite different than those found for the monomeric fraction. The total mol metal/mol MT is very high and although most of the Zn2+ has been displaced, none of the Cd2+ is lost. The high




88

content of total bound metals suggests that extensive disruption of the usual metal cluster structure and conformation of MT has occurred in this fraction. Such a catastrophic structural change could certainly make the affected MT molecules susceptible to oligomerization via intermolecular S-Pt-S cross-linking. The relatively high concentrations of the reactants and the large excess of K2PtCl4 employed may also promote the formation of oligomeric species.
Examination of the data in Table 4-1 also reveals
that while a net loss of Zn2+ and, to a lesser extent, Cd2+ occurs, essentially all of the Cu originally bound by native MT is retained in the products. It is also interesting that nearly all of this Cu ends up in the monomeric fraction of the products. It is known that Cu is more tightly bound by MT than Cd2+, presumably as Cu(I) [41]. Studies by Winge and coworkers indicate that Cu+ is preferentially bound by the B cluster region of MT [32] as originally suggested by Briggs and Armitage [171]. It is possible that those molecules of native MT in which the metal ions occupying the B cluster are all or nearly all Cu+ ions are much less reactive with Pt(II) due to the presence of very tightly bound Cu(I). The results in Table 4-1 are consistent with this proposition but the Cu concentrations in the native MT employed are rather small relative to the experimental errors in the ICP-AES data.




89
Reactions of Native Metallothioneins
with Cisplatin Derivatives
The analytical data obtained for the reactions of Cd7MT and native MT with K2PtCl4 strongly suggest that Pt(II) is preferentially displacing metal ions from the B clusters of these reactants. It is desirable, however, to obtain spectroscopic data from which precise structural information about the protein molecule can be derived. The use of 1H NMR spectroscopy and the complete set of assignments available for native rat liver MT-2 [151] provide this opportunity. Changes in the resonance lines of individual amino acid residues of MT in the upfield region of the 1H NMR spectrum of rat liver MT-2 ten minutes following the addition of excess cis-Pt(NH3)2(N03)2 are shown in Figure 4-1. This derivative was used instead of cisplatin because of the prohibitively low solubility of the latter.
Note that the resonance lines of residues 8, 9, and 27, all of which lie in the B domain of MT (residues 1-29), are significantly broadened during the early stages of the reaction while the lines of residues in the a (residues 30-61) domain are relatively unaffected. This confirms that Pt(II) is preferentially displacing metal ions from the B cluster as originally suggested by the analytical data.
The changes in the upfield region of the spectrum and a slight decrease in the intensity of the cysteine B-CH2 line




90

~tA~ <>Kp
IA
i/I

6 1. L2

-0pp. PP.M

Figure 4-1. 300 MHz 1H NMR spectra of (a) 0.33 mM native rat liver MT-2 in 20 mM phosphate buffer pH 7 and (b) the same sample 10 min after the addition of a 20-fold molar excess of cis-Pt(NH3)2(NO3 ). Selected resonance lines are labeled with the names of t9e amino acid residues from which they arise.

~\4
k

/4

K

I

a




91

a

K
'(till V
>w)

\tttt > ~j At

32 33 23 25 24 22 20 0

b
I '4
'2 *5' (~ ~t*I
p
F I v01.kirIv
I)

a 256 2 4 232 23C

S F3 34 2 2 33 ,

Figure 4-2. 300 MHz H NMR spectra of (a) 0.33 mM native rat liver MT-2 in 20 mM phosphate buffer pH 7 and (b) the same sample 12 hrs after the addition of a 20-fold molar excess of cis-Pt(NH3)2(NO3)2.

I>
~ '0

1 2

0 8 IPM




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ESXW0RGK5_8DFK2U INGEST_TIME 2017-07-11T22:16:15Z PACKAGE UF00101017_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

PLATINUM (II) METALLOTHIONEINS By JACOB BONGERS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1989

PAGE 2

To my parents

PAGE 3

ACKNOWLEDGMENTS The Plasma II was purchased with a Biomedical Research Support grant funded by the National Institutes of Health and administered by the Division of Sponsored Research, University of Florida. Partial support of this project was provided by a Milheim Foundation grant and the Division of Sponsored Research, University of Florida. I thank David Richardson for helping me in all aspects of my work, and I appreciate his passionate interest in chemistry. John Bell deserve thanks for collaborating with us on several projects and undertaking the difficult job of isolating and purifying the protein used in this study. I also thank Joanne Lopez for her valuable technical assistance. I thank Frances Goodman for her kind assistance in operating and maintaining the instruments, and for many helpful discussions about both of our research projects. Dr. Roy King is acknowledged for his valuable assistance and I owe a special debt of gratitude to Cyndy Walton for her expert assistance in the NMR spectroscopy. My thanks go out to the faculty and students of the inorganic division for their fine spirit of cooperation.

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGEMENTS iii LIST OF TABLES V LIST OF FIGURES vi ABSTRACT viii CHAPTERS 1 INTRODUCTION 1 Metallothionein 1 Binding of Platinum(II) to Metallothioneins 8 2 PROTEIN DETERMINATION BY ICP-AES 13 Accuracy and Applications 16 Precision and Sensitivity 23 Advantages of the Method 24 Materials and Methods 25 3 BINDING OF PLATINUM(II) TO APOMETALLOTHIONEIN 29 Basic Physical and Chemical Properties of Platinum(II) Adducts of Apometallothionein 29 Optical Studies 37 Nuclear Magnetic Resonance Spectra 56 Binding Kinetics 68 Conclusions 72 Materials and Methods 76 4 BINDING OF PLATINUM(II) TO NATIVE METALLOTHIONEIN.. 83 Reactions of Native and Cadmium Metallothioneins with Tetrachloroplatinate 84 Reactions of Native Metallothioneins with Cisplatin Derivatives 89 Conclusions 104 Materials and Methods Ill 5 GENERAL CONCLUSIONS 115 REFERENCES 118 BIOGRAPHICAL SKETCH 130 iv

PAGE 5

LIST OF TABLES page 2-1. Comparison of Spectrophotometry Molar Absorptivities of Proteins Based on ICP-AES of Sulfur with Published Values Based on Other Methods of Protein Determination 17 2-2. Metal Binding Stoichiometries and Metal Contents of Metalloproteins Determined by ICP-AES of Sulfur and Metals 21 4-1. ICP-AES Analysis of the Products From Reactions of CdyMT and Native Equine MT with K 2 PtCl 4 85 4-2. ICP-AES of Products From the Reaction of Native Rat Liver MT-2 with cis-Pt (NH 3 ) 2 (N0 3 ) 2 94

PAGE 6

LIST OF FIGURES page 1-1. Amino acid sequences of equine kidney MT-1A, rat liver MT-1 and rat liver MT-2 2 1-2. Crystal lographic structure of rat liver MT-2.... 4 1-3. Crystallographic structures of the metalthiolate clusters of rat liver MT-2 5 3-1. Elution profiles of Cd y MT and Pt y MT on a Sephadex G-75 size-exclusion column and native MT on a G-50 column 31 3-2. Stokes' radii of PtyMT and CdyMT determined by using Sephadex G-75 and G-50 size-exclusion columns 32 3-3. Reactions of DTNB and iodoacetate with thiolates and kinetics of reactions with Pt ? MT 36 3-4. Electronic absorption spectra of apoMT, Zn y MT, CdyMT, and Pt y MT 38 3-5. Spectrophotometric titration of apoMT with K 2 PtCl 4 at pH 7 40 3-6. Titration of partially platinated apoMT adducts with DTNB 44 3-7. Relative molecular orbital energies estimated for square-planar Pt(II) complexes 46 3-8. Titration of apoMT with K 2 PtCl 4 as followed by CD 51 3-9. CD spectra of apoMT and Pt ? MT in the far ultraviolet 54 3-10. CD spectra of native rat liver MT 55

PAGE 7

3-11. 1 H NMR spectra of native rat liver MT-2 and rat liver apoMT-2 58 3-12. 1 H spin-echo NMR spectra of native rat liver MT-2 and rat liver apoMT-2 61 3-13. 1 H NMR spectrum and spin-echo NMR spectrum of rat liver Pt y MT-2 6 3 3-14. Kinetics of the K 2 PtCl 4 -apoMT reaction at pH 7.5 70 3-15. Kinetics of the K 2 PtCl 4 -apoMT reaction at pH 2 71 4-1. Reaction of rat liver MT-2 with excess cis -Pt(NH 3 ) 2 ( N0 3 ) o at tlle ten m i nut e mark, as seen in the upfield region of the 1 H NMR spectrum 90 4-2. Reaction of rat liver MT-2 with excess 91 cis-Pt(NH 3 ) 2 (N0 3 ) 2 at the twelve hour mark, as seen in the 1 H NMR spectrum. 4-3. Elution profile of the products from the reaction of native rat liver MT-2 with excess Pt(NH 3 ) 2 (N0 3 ) 2 93 4-4. Cleavage of disulfide by sulfite followed by reaction of the resulting thiolate with TNBS.. 96 4-5. EPR line shapes of cis -Pt( ATMPOKCl-, and the ATMPO ligand 98 4-6. HTGF AAS of chromatographic fractions from the reaction of native MT with one mol equiv of cis-Pt(ATMPO) 2 Cl 2 100 4-7. HTGF AAS of chromatographic fractions from the reaction of native MT with four mol equiv of cis-Pt(ATMPO) 2 Cl 2 102 4-8. Possible products from the reactions of K 2 PtCl 4 with Cd ? MT and native MT 105 4-9. Three-metal & cluster of native MT and a hypothetical three-Pt 6 cluster 108 VII

PAGE 8

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PLATINUM(II) METALLOTHIONEINS By Jacob Bongers May 1989 Chairman: David E. Richardson Major Department: Chemistry Platinum(II) adducts of the metal ion chelating protein, metallothionein (MT) were prepared and characterized by spectroscopic and other methods. ApoMT reacts with excess K 2 PtCl 4 at pH 7 to form a monomeric adduct containing 6.95 0.24 mol Pt/mol MT as determined by inductively coupled plasma atomic emission spectrometry (ICP-AES). Polymeric adducts containing 17 2 mol Pt/mol MT are formed at pH 2. The Stokes' radius of PtyMT is consistent with a two-domain structure containing two independent metal-thiolate cluster regions similar to that of native MT. Spectrophotometric and spectropolarimetric titrations of apoMT with K 2 PtCl 4 at neutral pH suggest that the first three mol equiv of Pt 2+ ions form three Pt(cys) 4 2 ~ units, and that the binding of the remaining four mol equiv of Pt 2+

PAGE 9

ions involves formation of bridging thiolates and Pt(II)thiolate clusters. This binding seguence was confirmed by titration of the partially platinated apoMT adducts with the thiolate assay reagent, DTNB. The biphasic kinetics of Pt(II) binding to apoMT are also consistent with this interpretation Reactions of CdyMT and native MT with K 2 PtCl 4 in the presence of air, result in the liberation of three mol eguiv of the originally bound metal ions and the binding of one mol eguiv of Pt 2+ When air is excluded from the reaction, three mol eguiv of metal ions are lost and four mol eguiv of Pt 2+ are bound. Pt(II) preferentially displaces metal ions from the 3-metal 6-cluster region in reactions of native MT with derivatives of the antitumor drug, cis-Pt(NH 3 ) 2 Cl 2 as shown by analytical and 1 H NMR data. Binding to native MT of Pt(II) from cis-Pt(ATMPO) 2 Cl 2 resulted in translabilization of the spin-labeled ATMPO amine ligand, and the analytical results suggest binding of Pt(II) to the Nterminal methionine. Possible Pt( II) -thiolate cluster structures for PtyMT and mixed-metal Pt(II) metallothioneins are described. Implications for the metabolism of platinum anticancer drugs are discussed. A method for the determination of micromolar protein concentrations and the metal binding stoichiometries of metalloproteins by ICP-AES of sulfur is also described.

PAGE 10

CHAPTER 1 INTRODUCTION Metal lothionein Metallothioneins are low-molecular-weight cytosolic proteins with extremely high metal and sulfur contents [14 ] These proteins occur in natural systems ranging from microorganisms to humans. Metallothionein was discovered in 1957 by Margoshes and Vallee in a search for a tissue constituent responsible for the natural accumulation of cadmium in animal tissues [5]. Although the biological role of metallothionein is not known for certain it is thought to be involved in the regulation of the essential trace elements Zn and Cu. Metallothionein also binds various heavy metals ions in vivo including Cd(II), Hg(II), and Pb(II), and biosynthesis of the protein is stimulated by the presence of certain of these metal ions. This interesting phenomenon suggests a role in the seguestering of toxic metal ions. Mammalian forms of MT are single polypeptides of 61 amino acid residues with molecular weights ranging from 6500 to 7000 Daltons depending on the metallic composition. The primary structures of the three metallothioneins employed in this study are shown in Figure 1-1 [6, 7]. Most vertebrate

PAGE 11

1 10 Equine Kidney MT-1A Ac-M DPNCSCPTGGSC Rat Liver MT-1 Ac-M DPNCSCSTGGSC Rat Liver MT-2 Ac-M DPNCSCATDGSC 20 30 -TCAGSCKCKECRCTSCKKSCCSCCP -TCSSSCGCKNCKCTSCKKSCCSCCP -SCAGSCKCKQCKCTSCKKSCCSCCP 40 50 60 -GGCARCAQGCVCKGASDKCSCC A-OH -VGCSKCAQGCVCKGASDKCTCC A-OH -VGCAKCSQGCICKEASDKCSCC A-OH Figure 1-1. Amino acid sequences of equine kidney MT-1A [6], rat liver MT-1 [7], and rat liver MT-2 [7]. The oneletter amino acid code is: A, alanine; C, cysteine; D, aspartic acid; E, glutamic acid; G, glycine; I, isoleucine; K, lysine; M, methionine; N, asparagine; P, proline; Q, glutamine; S, serine; T, threonine; V, valine.

PAGE 12

3 tissues contain two major "isoforms" of metal lothionein designated MT-1 and MT-2 according to the order of elution by ion exchange chromatography. Some samples of MT-1 may be further resolved into subforms MT-lA, MT-lB, and so on. All mammalian forms of MT contain 20 cysteine residues (CH 2 SH side chain) at identical positions in the seguence and an Nterminal acetylated methionine (CH 2 CH 2 SCH 3 side chain) giving a total of 21 sulfur atoms. The complete lack of residues with aromatic side chains is another unusual feature of MT. Native MT binds a total of 7 divalent metal ions, primarily Cd 2+ and Zn 2+ and often contains minor amounts of Cu + Fe 2+ Hg 2+ Pb 2+ and other metal ions depending on the source and method of isolation. Although MT from adult tissue contains predominantly Cd 2+ and Zn 2+ MT from fetal tissue generally contains predominantly Cu and Zn" The crystallographic structure of rat liver MT-2 [8] is shown in Figure 1-2. This structure determination was made at a resolution of 2 30 pm and is currently being refined with a new data set at 190 pm [9]. The 7 metal ions are bound exclusively to the sulfur atoms of the 20 cysteine residues in two distinct metal-thiolate clusters. Residues 1-29 enfold the 3-metal 9-cysteine cluster, which is referred to as the 6 domain. Residues 30-61 comprise the a domain, which enfolds the 4-metal 11-cysteine cluster. The unusual structure of MT has been described as a monolayer of

PAGE 13

Figure 1-2. Crystal lographic structure of rat liver MT-2 Shown are the a carbons of each residue, the 6 carbons and sulfur atoms of each cysteine residue, and the Cd and Zn atoms. Reproduced from Furey et al. Science 1986, 231 707 [8].

PAGE 14

Figure 1-3. Expanded views of the structures of (a) the R cluster domain and (b) the a cluster domain of rat liver MT-2. Shown are the cysteine sulfur atoms and the Cd and Zn atoms. Reproduced from Furey et al. Science 1986, 231 707 [8].

PAGE 15

6 polypeptide wrapping around the metal-thiolate clusters which take the place of the hydrophobic core of residues in a larger protein [8]. A detailed view of the crystallographic structures of the metal-thiolate clusters is shown in Figure 1-3. The R cluster consists of 3 metal ions, 3 bridging thiolate ligands, and 6 terminal thiolate ligands. The metal ions and bridging thiolates form a 6-membered ring in a chair conformation. The adamantane-like structure of the a cluster is composed of 4 metal ions, 5 bridging thiolates, and 6 terminal thiolates. The distribution of Cd 2+ and Zn 2+ ions among the 7 binding sites (as observed by 113 Cd NMR) is not a uniform one as shown in Figure 1-3. The relative numbers of Cd 2+ and Zn 2+ ions in a particular MT molecule and the distribution of these metals ions among the 7 binding sites vary from one molecule to the next. Apparent dissociation constants for Cd 2+ and Zn 2+ in MT (K [ (cys) 3 3 ~] [M 2+ ]/ [(cys) 3 M ]) have been estimated to be on the order of 10~ 15 M and 10 -11 M respectively [10], based on spectrophotometric titrations of CdyMT and Zn y MT with protons. CdyMT can be prepared via displacement of Zn 2+ by adding excess Cd 2+ salt to solutions of the native protein. Isolation of native MT in high yield is generally accomplished by administering a series of CdCl 2 injections to the animal so as to induce biosynthesis of MT prior to isolation. MT obtained in this

PAGE 16

7 way contains both Cd 2+ and Zn 2+ The Cd 2+ is localized preferentially in the a domain and the more weakly bound Zn 2+ in the R domain. Synthetic metal MT adducts have been prepared by three methods. Adducts may be prepared in vivo by injecting the animal with solutions of the metal ions of interest prior to isolation of the protein. The in vitro displacement of metal ions from native MT by added metal ions is another means of obtaining synthetic metal MT adducts. The third method entails reconstituting the apo (metal-free) form of MT with the desired metal ions in vitro MT adducts with Cu(I) [11], Hg(II) [12-14], Pb(II) [15], Au(I) [16], Bi(III) [17], Pt(II) [18], and other metals have all been prepared by the in vivo route. This procedure usually results in the recovery of mixed-metal MT adducts Metal MT adducts prepared by in vitro displacement of bound Cd 2+ and Zn 2+ from native and Zn y MT include adducts with Cd(II) [19-23], Hg(II) [24, 25], Au(I) [26], and Pt(II) [18, 27]. Waalkes and coworkers have surveyed the ability of 16 different metals ions to displace Zn 2+ from Zn y MT [28]. Both mixed metal and single metal adducts have been prepared by this method. Synthetic metal MT adducts prepared by reconstituting apoMT with various metals have been studied to the greatest extent. ApoMT is obtained by acidifying solutions of

PAGE 17

8 native MT. At low pH the cysteine thiolates are protonated, and the metal ions liberated are removed from the protein solution by chromatography or exhaustive dialysis. Protocols for obtaining metallated derivatives of apoMT generally involve adding an excess of metal salt and then raising the pH to above neutral. The product or products are then desalted on a size-exclusion column. As the pH approaches the pK a values of the cysteine sulfhydryls (ca. 8), the thiolates become unstable with respect to rapid air-oxidation to cystine disulfides. Reconstitutions are therefore performed under a nitrogen or inert gas atmosphere to prevent this undesirable denaturation. M 7 MT adducts prepared from apoMT include adducts with Co (II) [29, 30], Ni(II) [29], Fe(II) [31], Cu(I) [32-34], Zn(II) [33, 35], Cd(II) [21, 23, 33, 35-39], Pb(II) [40], Hg(II) [40], and Bi(III) [40]. Neilson and coworkers have surveyed the binding of 18 different metal ions to apoMT [41]. Binding of Platinumdll to Metallothionein The potent anticancer agent cis-dichlorodiammineplatinum(II) (cisplatin) [42, 43] has been the subject of intense study since its antitumor activity was discovered [44] by Rosenberg et al Most of this research has been concerned with the binding of cisplatin and analogous compounds to polynucleotides [45] as the mode of action of

PAGE 18

9 these platinum drugs probably involves binding to DNA. The binding of platinum drugs to polypeptides and proteins has been studied to a far lesser extent [46, 47]. This aspect of the physiological action of cisplatin is important in understanding the toxicological effects of the drug. Cisplatin is known to cause extensive kidney damage as its primary toxicity, and this nephrotoxicity is one of the major dose-limiting factors in its therapeutic use [48], In addition, the seguestering of Pt(II) from cisplatin by MT may be responsible for some of the reduction in the anticancer effect of the drug in repeated dose protocols. Metallothionein is present in relatively high concentration in the cytosol of the kidney and is a probable binding site for platinum anticancer agents because of its abundance of cysteine residues and the N-terminal methionine. Platinum(II) is known to form very stable complexes with sulfur donor ligands [49]. The binding of Pt(II) to the sulfur bearing residues cysteine and methionine has been documented by protein crystallography wherein Pt(II) salts are used as heavy-atom labels [50]. The biodistribution of Pt in the tissues of experimental animals administered cisplatin intravenously has been described [51-53]. Platinum is rapidly cleared from the blood following injection, with filterable Pt in the plasma of rats declining to near zero after several hours [51, 53]. Tissue levels of Pt decline slowly over

PAGE 19

10 about one week post-injection to leave several percent of the initial dose retained by tissue [51]. Cisplatin nephrotoxicity takes several days to manifest maximum severity during which the low molecular weight reaction products formed in the plasma are excreted into the urine and reabsorbed into the kidneys. By far the highest tissue concentration of Pt during metabolism of cisplatin is found in the kidney [51-53]. Sharma and Edwards found that at the time of maximum Pt accumulation in the rat kidney (ca. 24 hours post-injection) about 60% of cellular Pt in the kidney is localized in the cytosol [54]. This cytosolic Pt was partitioned into three main fractions by size-exclusion chromatography. Low molecular weight species (MW < 1000) accounted for 52% of the total Pt, 2 3% was found in the high molecular weight fraction (MW > 20000), and the remaining 25% eluted in a "metallothionein-like" fraction (6000 < MW > 12000) [54]. These observations and similar results reported by other workers [55, 56] suggest a possible role for MT in the detoxification of cisplatin and its metabolites. Zelazowski and coworkers demonstrated that the total in vivo bound Pt in the kidneys of cisplatin treated rats can be significantly increased via induction of MT biosynthesis by preinjection with CdCl 2 [57]. MT-like sizeexclusion fractions are composed of MT-1 and MT-2 as confirmed by gel electrophoresis. The authors also found

PAGE 20

11 that binding of Pt from cisplatin in vivo and in vitro was associated with a large decrease in the Zn content (ca. 50%) of the native protein but only trivial Cd 2+ losses. Binding of Pt(II) to native MT thus involves displacement of the relatively weakly bound Zn 2+ ions. Pt(II) may induce the biosynthesis of MT as has been documented for certain other heavy metal ions. However, several groups have presented evidence that stimulation of MT biosynthesis by Pt(II) is very minimal or nonexistent [57, 58]. Bakka et al. found that cultured cells with naturally high MT contents were resistant to the toxic effects of cisplatin [59]. This observation extends the hypothesis of heavy metal detoxification by MT and raises the interesting guestion of whether elevated MT concentrations might be used to alleviate the toxic effects of cisplatin and increase its effectiveness as an anticancer drug. However, tumor cells with higher MT concentrations might also be more resistant to the anticancer effects of cisplatin thus reducing the effectiveness of the drug. In another recent series of experiments, mutant tumor cells with elevated MT contents were implanted in mice and tumor growth following i.v. administration of cisplatin compared to a control group of mice implanted with wild-type cells [60]. Tumor volume was reduced by 80% in tumors from the parent cells, whereas the tumors from MT-rich cells were almost completely resistant.

PAGE 21

12 Human carcinoma cells with high levels of MT, induced by chronic exposure to heavy metals and various anticancer drugs, are also known to acguire significant resistance to cisplatin [61] Naganuma et al. have reported an increase in survival rates of mice injected with MT inducing metals prior to injection of a lethal dose of ciplatin [62, 63]. The greatest protection was imparted by preinjection with Cu(II) and Bi(III). This protective effect is presumably due to the binding of Pt(II) to MT but no direct relationship with the increased MT levels has been uneguivocally established. A reduction in the lethal and renal toxicity of cisplatin without impairment of its antitumor activity in mice by preinjection with the MT inducer Bi(N0 3 ) 3 has been described recently [64 ] The following chapters describe in vitro Pt(II) MT adducts from the reaction of K 2 PtCl 4 with apoMT and the reactions of K 2 PtCl 4 and cisplatin derivatives with CdyMT and native MT. This study is the first attempt to examine these adducts on the molecular level and its primary aim is to provide synthetic models of the in vivo cisplatin MT metabolites reviewed above. The incorporation of sguareplanar Pt 2+ ions in sites normally occupied by tetrahedral Cd 2+ and Zn 2+ ions is also of interest concerning the degree to which the structures of metal MT adducts are determined by the metal versus the polypeptide component of the adduct.

PAGE 22

CHAPTER 2 PROTEIN DETERMINATION BY ICP-AES The analytical method used to obtain bound metal stoichiometries and bound metal contents of metal MT adducts is described in some detail in this section as these are central results in the following study. The distinction between metal stoichiometries and metal contents is drawn here as some mixed metal MT adducts are heterogeneous with respect to the bound metal ions. In the native protein for example, Cd 2+ ions preferentially occupy sites in the a domain while Zn 2+ ions tend to reside mainly in the & domain, but these two metal ions, as well as the lesser amounts of Cu + and other bound metal ions, are not uniformly distributed across the binding sites of the protein. True binding stoichiometries can be obtained solely from analytical data only for single metal adducts for which stoichiometric binding occurs. For mixed metal adducts, on the other hand, true stoichiometries cannot be determined without some further information beyond the analytical data. The analysis of metal MT adducts has traditionally involved the determination of the bound metal or metals by 13

PAGE 23

14 one method and quantification of the polypeptide component by independent means. Metals can be conveniently assayed by either atomic absorption spectrometry (AAS) or inductively coupled plasma atomic emission spectrometry (ICP-AES). Determination of the polypeptide component, however, presents a more difficult analytical problem. Protein concentrations are usually obtained on a routine basis by measuring the absorbance of micromolar solutions at the absorption maximum occurring at ca. 280 nm. This band arises from absorption by the aromatic side chains of tyrosine and tryptophan residues. The molar absorptivity must, of course, be arrived at by the use a molecular weight and some primary method of determining total protein. An accurate molecular weight is preferably obtained from amino acid sequence data, but a crude value obtained by hydrodynamic methods, such as size-exclusion chromatography or sedimentation, may have to suffice in some cases. The numerous existing methods of total protein determination include Nessler and Kjeldahl total nitrogen determinations, various direct colorimetric methods such as the Biuret, and the most simple and commonly used gravimetric or "dry weight" methods. Unfortunately the ultraviolet spectrum of apoMT is devoid of a 280 nm peak due to its peculiar lack of aromatic residues. The spectrum contains an intense band at 190 nm arising from the strong n-n and weak r\-n amide transitions

PAGE 24

15 of the polypeptide backbone and the lowest thiol transition of the cysteine side chains near 195 nm [2]. The most commonly used method of estimating MT concentrations is to measure the absorbance of apoMT in 0.01 M HC1 at 220 nm using a molar absorptivity of 48200 M _1 cm -1 [65]. This value was obtained by a wet chemical thiolate assay. The problems associated with the spectrophotometry of the sharply rising shoulder of this peak have been described thoroughly by Wetlaufer [66]. We have developed a method of obtaining micromolar protein concentrations and bound metal stoichiometries by ICP-AES of sulfur [67] which overcomes the shortcomings of traditional methods. Temperatures in an ICP torch are sufficiently high to excite many nonmetallic elements and several analytically useful emission lines in the vacuum ultraviolet have been observed [68]. Of particular interest to this study is the prominent sulfur line at 180.73 nm, which has an estimated detection limit of 15 p.g/L [69]. Proteins with known contents of the sulfur bearing residues cysteine and methionine can thus be quantified at the ng level by ICP-AES, provided the instrument is equipped with evacuated or otherwise oxygen-free optics. ICP-AES determinations of total sulfur in tissue samples [70, 71] and the use of an ICP-AES instrument as a sulfur detector in the HPLC of tissue extracts [72] have been reported. The purpose of the following experiments is

PAGE 25

16 to demonstrate the viability of pumping micromolar purified protein solutions directly into the nebulizer of the ICP-AES instrument and determining protein concentrations based on the emission of sulfur at 180.73 nm. Accuracy and Applications Elemental analysis of proteins by ICP-AES has traditionally been restricted to metals in acid digested samples of metalloproteins. We initially sought to demonstrate that accurate absolute concentrations of proteins, regardless of the presence or absence of bound metals, can be obtained based on sulfur content by pumping buffered micromolar solutions directly into the nebulizer without any prior sample preparation, and using agueous solutions of Na 2 S0 4 as sulfur standards. The protein solutions, although relatively dilute, do have slightly different physical properties than the Na 2 S0 4 standards, such as viscosity and surface tension, which could lead to differences in aerosol formation in the nebulizer and systematic errors in the results. One test of accuracy is to compare spectrophotometric molar absorptivities determined by ICP-AES of sulfur with published absorptivities based on traditional methods of protein determination (Table 2-1).

PAGE 26

17 Table 2-1 Comparison of Spectrophotometry Molar Absorptivities of Proteins Based on ICP-AES of Sulfur with Published Values Based on Other Methods of Protein Determination Protein 3 ~K /nm e/10 4 M 1 cm" 1 ICP D Lit c Ref Cd 7 metallothionein equine kidney MW 6.88 kD 21 mol S/mol 73 250 10.9 0.2 11.4 H 74 Ferricytochrome c equine kidney MW 12.4 kD 4 mol S/mol 75 410 11.8 0.3 11.20 B 76 10.95 B 77 10.60 B 78,79 280 2.62 0.06 2.40 D 80 2.32 B 78,79 2.12 D 81 Ferrimyoqlobin equine MW 17.5 kD 2 mol S/mol 82 409 17.9 0.3 280 3.24 0.05 18.8

PAGE 27

18 Table 2-1 — continued. Protein 3 *x /nm e/10 4 M 1 cm" 1 ICP D Lit c Ref Ribonuclease A bovine pancreas MW 13.7 kD 12 mol S/mol 88 278 1.13 0.02 1

PAGE 28

19 Table 2-1 — continued. Protein 3 ~A/nm e/10 4 M 1 cm" 1 ICP b Lit c Ref Chymotrypsinogen A bovine MW 25.6 kD 12 mol S/mol 113 282 280 30 0.11

PAGE 29

20 While the list of published molar absorptivities in Table 2-1 is not an exhaustive one, it is an attempt to list as many literature values as possible in a nonselective manner [73-123]. The spread in the distribution of literature values for a particular protein is to be expected as each value was determined by different workers using the various analytical methods listed. In general very close agreement is found between the experimental molar absorptivities and the central tendencies of the published values. According to these comparisons the ICP-AES method of protein guantif ication via determination of sulfur is an accurate one, at least within the bounds of the concentration range and conditions described in the experimental section. Probably the most useful application of the method is for the determination of metalloprotein binding stoichiometries. Bound metal stoichiometries are traditionally obtained by determination of the metallic component by atomic absorption or ICP-AES, and determination of the polypeptide component by independent means such as dry weight or total nitrogen determinations. These same stoichiometries, however, can be obtained by an ICP-AES determination of both components by using a single sample. Shown in Table 2-2 are bound metal stoichiometries of three homogeneous proteins and metal contents of two heterogeneous metallothioneins as determined by ICP-AES of

PAGE 30

21 Table 2-2. Metal Binding Stoichiometries and Metal Contents of Metalloproteins Determined by ICP-AES of Sulfur and Metals. Protein

PAGE 31

22 sulfur and metals. Native metallothioneins have variable metal contents depending on the source and method of isolation. This variability is illustrated by the samples of MT from rat liver and equine kidney in Table 2-2. Concentrations of native MT cannot be satisfactorily determined by a metals assay because of this variability in metal contents. Concentrations can be accurately determined by a sulfur assay, however, as all mammalian metallothioneins have identical sulfur contents (21 mol S/mol MT) The sum of the Cd, Zn, and Cu contents for either native MT falls slightly below the maximum binding capacity of MT (seven divalent metal ions), but this probably results from the binding of minor amounts of Fe, Pb, Hg, and other metals not included in the assays. The experimentally determined Fe content of ferrimyoglobin in Table 2-2 falls short of the true stoichiometry by about 10%. Spectrophotometric absorbance measurements of this sample taken at the Soret band at 409 nm arising from absorption by the iron containing heme moiety, and the band at 280 nm arising from aromatic side chains in the polypeptide portion of the molecule, also indicated the presence of 10% apoprotein, which lacks heme. The ferricytochrome c sample, that was purified extensively by ion-exchange chromatography, yielded an experimental result that is within 5% of the ideal stoichiometry.

PAGE 32

23 Precision and Sensitivity The standard deviations of the sulfur analyses in Table 2-1 are within 2% of the mean for 6 replicate measurements. The metal binding stoichiometries in Table 2-2 fall within 4-5% of the mean due to the uncertainties associated with both sulfur and metal determinations. This level of precision is acceptable for most applications particularly when one considers the rather poor reproducibility of many other existing methods of protein determination. If one wishes to make very fine distinctions (such as whether a certain protein binds 10 or only 9 metal ions) this level of precision becomes intolerable. Small gains in precision might be realized by using optimized instrumental parameters for each element instead of compromise parameters, a pulseless syringe pump, improvement of power supply tolerance, and other refinements. The careful use of an internal standard [124] to mitigate precision-limiting flicker-noise in the nebulizer [125] may also improve precision. The sensitivity of any destructive technique of protein determination is a general concern. We routinely obtain satisfactory results for samples with sulfur concentrations of 1 mg/L and higher. The volume of sample required is between 1 and 4 mL, which corresponds to the consumption of ten to several hundred micrograms of protein depending on

PAGE 33

24 the molar sulfur content of the protein. While the sample volumes required for a pneumatic nebulizer are about two orders of magnitude greater than those required for a graphite furnace, they are not at all prohibitive in terms of the milligram scale on which protein research is often conducted. Flow injection analysis (FIA) coupled with emerging total-consumption (direct injection) sample introduction systems could, in principle, reduce sample volumes from several milliliters to several microliters and greatly increase the applicability of ICP-AES to proteins [126]. FIA utilizes the transient signals generated by injecting small discrete sample "plugs" into a carrier stream and thus eliminates the wasteful dead-volume of the conventional continuous-flow mode of sample presentation. Experimental sample introduction systems with high transport efficiencies have been described and include a miniature concentric nebulizer installed in the injecter tube of a conventional ICP torch [127], and a thermospray vaporizer [128]. Advantages of the Method The main advantages of protein determination by ICP-AES of sulfur, besides the high selectivity and multielement capability of atomic spectrometry in general, are the generality and rapidity of the method and the ease of sample

PAGE 34

25 preparation. The method is applicable to any protein having a known content of sulfur. The only significant spectral interferences with the sulfur line at 180.73 nm are wing-overlaps of Ca and Mn lines which have intensities about two orders of magnitude smaller on a molar basis [129]. Perchloric acid digestion, dry ashing, and other time-consuming and error-prone preparatory steps are not necessary. A promising application of the method is the rapid and precise determination of metal binding stoichiometries in metal loproteins by determining both protein and metals for a single sample by a single method. In addition, ICP-AES is a useful method for determination of sulfur content of proteins following quantification of total protein by other methods. Materials and Methods Preparation of Samples and Standards Horse kidney and rat liver metallothioneins were isolated and purified by a literature method [130]. Cd 7 MT was prepared by adding excess CdCl 2 to a solution of native equine MT. An excess of the reducing agent dithioerythritol was added to ensure complete reduction of cysteine side-chains. After 24 hours the reaction mixture was fractionated on a Sephadex G-50 column. Horse heart cytochrome c was obtained from Sigma Chemical Company and

PAGE 35

26 purified by carboxymethyl cellulose ion-exchange chromatography. The other proteins used in this investigation were obtained as lyophilized powders from Sigma and used without further purification. The ribonuclease sample was dissolved in 10 mM Tris/HCl buffer at pH 7.8. All other protein samples were dissolved in 5 mM phosphate buffer at pH 7.2. All samples were concentrated by ultrafiltration (YM5 membrane, Amicon Corp.) and then passed through a 0.2 |xm filter (Gelman Sciences Inc.). The ultraf iltrates were used as blanks for both ICP-AES and spectrophotometric measurements. The metallothionein samples used for ICP-AES measurements had concentrations between 0.75-1.6 mg/L in sulfur. All other protein samples used were about 5 mg/L in sulfur. Sulfur standards were solutions made from anhydrous Na 2 S0 4 Standard solutions of the metallic elements were purchased from Aldrich Chemical Company. Instrumentation Elemental analyses were made by using a Perkin-Elmer Plasma II Emission Spectrometer equipped with a Perkin-Elmer Series 7000 computer. The one-meter Ebert monochromator (1800 1/mm holographically ruled grating) is sealed in a heavy-walled aluminum tube maintained at 5 mTorr. The path from the argon ICP source to the optical components was purged with a 5 L/min flow of nitrogen. The sample

PAGE 36

27 introduction system is comprised of a peristaltic pump (12 rollers, silicone rubber pump tubing) and a cross-flow nebulizer [131]. Argon flow to the nebulizer is regulated by a mass flow controller. The following compromise operating parameters were employed: RF Power 1200 W Viewing Height 12 mm above coil Sample Delivery Rate 1 mL/min Argon Flow Rates plasma 15 L/min auxiliary 1 L/min nebulizer 1 L/min Absorbance measurements were recorded on an IBM 9430 UV-Visible spectrophotometer. Samples were diluted to give absorbance values between 5 and 1 in a 1 cm path-length guartz cell. Measurements Individual ICP-AES measurements were obtained by scanning a 0.020 nm window centered on the emission line of interest and recording intensity counts at the wavelength of maximum emission. For multielement analyses the monochromator slews from one emission line to the next in a seguential fashion. The emission lines scanned were

PAGE 37

28 S I 180.73, Fe II 259.94, Cd II 214.44, Zn I 213.86, and Cu I 324.75 nm. Six replicate measurements were recorded for each element of each standard, sample, and blank. Linear calibrations were obtained immediately prior to running each protein sample and corresponding ultraf iltrate blank.

PAGE 38

CHAPTER 3 PLATINUM(II) BINDING TO APOMETALLOTHIONEIN The first goal of the research was to investigate Pt(II) MT adducts prepared from mixtures of apoMT and K 2 PtCl 4 This chapter describes the preparation and spectroscopic characterization of such adducts. The dynamics of the binding process are also examined in some detail. Spectra and other experimental results obtained for the synthetic Pt(II) adducts are compared, wherever practical, with corresponding data obtained for the widely studied native form of the protein. Basic Physical and Chemical Properties of Platinum ( lit Adducts of Apometallothionein A change from a colorless to a yellow solution occurs when micromolar solutions of apoMT at neutral pH are mixed with excess K 2 PtCl 4 The resulting adduct contains 6.95 0.24 mol Pt/mol MT as determined by ICP-AES determinations of platinum and sulfur. The ICP-AES method was used, unless stated otherwise, for all the elemental analyses reported in this work, and is described in detail in Chapter 2. Dozens of reconstitutions have been performed and the 29

PAGE 39

30 resulting products invariably contain 7 mol Pt/mol MT. Once formed, PtyMT is air-stable, and solutions have been stored for up to 6 months at 6C without loss of bound Pt 2+ or chromatographic evidence of aggregation. Pt 7 MT elutes as a single symmetrical peak on a Sephadex G-75 size-exclusion column (Figure 3-la). Dimeric, trimeric, and higher oligomeric species are also isolated when the reactants are mixed at higher concentrations. These oligomers are also found to contain approximately 7 mol Pt/mol MT. A similar elution profile is often observed for solutions of native MT, and the oligomerization is probably due to a small number of intermolecular disulfide linkages formed by partial oxidation of the cysteines during isolation or storage [19]. Oxidation of the cysteines can be reversed by 2-mercaptoethanol and other reducing agents. Sephadex G-75 is the chromatographic gel of choice for MT, but even G-50 resin provides sufficient resolution to separate monomers from oligomeric species (Figure 3-lb) The Stokes' radius of native MT, as determined by sizeexclusion chromatography, is eguivalent to that of a much higher molecular weight globular protein. Stokes' radius is defined as the radius of an eguivalent hydrodynamic sphere. The large Stokes' radius of MT is due to the nonglobular elongated shape of the molecule (Figure 1-2). PtyMT has a slightly larger Stokes' radius (ca. 1.80 nm) than CdyMT or native MT (both ca. 1.60 nm) These radii were determined

PAGE 40

31 en CN CM < Pt 7 MT Cd 7 MT 80 100 120 140 160 Elution Volume / ml_ 180 CM CM < monomer dimer trimer, etc 40 60 Elution 80 Volume / 100 ml_ Figure 3-1. Elution profiles of (a) Cd ? MT and Pt y MT on a Sephadex G-75 size-exclusion column and (b) native MT on a G-50 column. The eluant was 3 5 mM phosphate buffer, pH 7.2, and the column was monitored by the absorbance at 229 nm (A 22g )

PAGE 41

32 q ovalbumin a *i — i — i — i — i — i — i — i — i 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 R / nm 1.4 -i ovalbumin 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 R / nm Figure 3-2. Stokes' radii of Pt^MT and Cd y MT determined by using calibrated Sephadex (a) G-75 and (b) G-50 sizeexclusion columns.

PAGE 42

33 were determined by using Sephadex G-50 and G-75 columns calibrated against globular proteins with known Stokes' radii (Figure 3-2) [132]. The Stokes' radius of PtyMT is consistent with retention of the two domain structure of the native protein and CdyMT. This is necessary but not sufficient evidence that PtyMT has the two domain structure, or any fixed conformation, as we have estimated the Stoke 's radius of a hypothetical random coil polypeptide of 61 residues to be very close to radii obtained for PtyMT and CdyMT [27]. We are not certain if PtyMT is homogenous with respect to charge. Ion-exchange chromatography on DEAE cellulose against a 0.15 M KC1 gradient gave a single extremely broad elution band. Nondenaturing polyacrylamide gel electrophoresis gave a similar result. The adduct recovered from the ion-exchange procedure contained 5.67 + 0.34 mol Pt/mol MT. The tenacious binding of PtyMT to the ionexchange and electrophoresis media, and the loss of roughly one mol equiv of Pt 2+ might be the result of a slow binding of Pt 2+ ions to the pendant amino groups on the media. It is possible that one of the bound Pt 2+ ions in PtyMT is relatively exposed to solvent, possibly a Pt 2+ ion bound to the N-terminal methionine. Binding of Pt(II) to the exposed methionine residues of proteins is well known [47] and Pt(II) salts have been used extensively as heavy atom labels by protein crystallographers [50].

PAGE 43

34 Further experimental evidence supporting a bound Pt 2+ at the methionine (met-1) was obtained by examining the extent to which Cd 2+ and Pt 2+ bind to apoMT chemically modified at the methionine residue. The N-terminal methionine of apoMT was converted to methionine sulfoxide, a much poorer ligand for Pt(Il), by an oxygen atom exchange reaction with excess dimethylsulf oxide [133] to yield the modified MT and dimethylsulf ide as the other product. Reconstitution with CdCl 2 and K 2 PtCl 4 followed by ICP analysis of the adducts showed that while the modified apoMT still bound the full complement of 7 Cd 2+ ions (7.10 0.29 mol Cd/mol MT) the binding of Pt 2+ was indeed decreased by approximately one mol equiv (5.79 0.60 mol Pt/mol MT) This evidence alone is not sufficient to prove that Pt 2+ binds to met-1, but it is highly unlikely that this would not occur based on the known chemistry of Pt(II) and this amino acid. The bound metal ions in PtyMT are not liberated from the protein under acidic conditions (as occurs for CdyMT and other adducts) Dialysis of PtyMT against 1 M HC1 did not result in the loss of bound Pt 2+ and no further binding occurs at this pH following the addition of excess K 2 PtCl 4 The average binding constant for PtyMT must be many orders of magnitude higher than that of CdyMT which remains fully metallated only at pH > 5 [134].

PAGE 44

35 The great stability of the Pt 2+ -to-cysteine-thiolate complexes is further evidenced by the observation that apoMT reacts with excess K 2 PtCl 4 at pH 2 to form polymeric adducts containing 17 1 mol Pt/mol MT. The degree of binding indicates that an essentially one-to-one Pt-to-cysteine coordination is occurring. The binding of slightly less than 20 Pt ions and the polymeric nature of the products can both be attributed to intermolecular S-Pt-S bridging. As the amino acid seguence of MT contains three pairs of adjacent cysteines (Figure 1-1) it is also possible that a small proportion of the Pt 2+ ions are coordinated in a bidentate fashion to two adjacent cysteines. The apparently random mode of binding at low pH may be a result of the much slower rate of binding or the reported random coil structure of apoMT at low pH [135]. PtyMT is also guite stable to displacement of Pt 2+ ions by other electrophiles, such as the oxidizing agent 5,5'dithiobis(2-nitrobenzoic acid) (DTNB) and the alkylating agent iodoacetate (Figure 3-3). At neutral pH, PtyMT reacts with the cysteine assay reagent DTNB at an initial rate ca. 8 times slower than found for the reaction of the apo protein under the same conditions [37]. Iodoacetate reacts with PtyMT ca. 15 times slower than reported for the reaction of iodoacetamide with apoMT under the same conditions. These rate reductions are nearly twice those found for the reactions of CdyMT with DTNB and iodoacetamide

PAGE 45

36 COO coo + RS COO COO a ICH COO RS RS CH 2 COO 2

PAGE 46

37 [37], which is consistent with the known decrease in reactivity of the cysteine thiolates with stronger metal-tosulfur bonds. Roughly half of the 20 cysteines in PtyMT reacted with either reagent. ICP analysis of the products showed that no Pt 2+ ions were displaced from the protein in either of the two reactions. All seven Pt 2+ ions are liberated from PtyMT, however, by dialysis against excess cyanide to form highly stable Pt(CN) 4 2. Complete removal of Pt 2+ from 35 [iM PtyMT by eguilibrium dialysis in 0.1 M NaCN at pH 7.0 was confirmed by both spectrophotometric and ICP-AES measurements of the products. ApoMT was recovered from this dialysis in its completely reduced form, i.e. no characteristic cystine disulfide peak at 280 nm was observed in the ultraviolet spectrum. Air-oxidation of cysteine pairs to cystine disulfide groups is presumably followed by redox reaction with CN~ yielding cys~ and cys-SCN [136]. Dialysis of PtyMT against CN~ or other ligands with high affinity for Pt(II) under acidic conditions could be used to recover intact apoMT from PtyMT. Optical Spectra The electronic absorption spectrum of PtyMT consists of a broad region of absorption similar to the spectra of other metal MT adducts (Figure 3-4). The spectrum is consistent

PAGE 47

38 20 i 200 250 300 350 400 X / nm Figure 3-4. Electronic absorption spectra of apoMT, Zn y MT, Cd 7 MT, and Pt y MT.

PAGE 48

39 with the presence of Pt 2+ to cysteine thiolate coordination as well as possible coordination of Pt 2+ to the thioether sulfur of met-1. As illustrated in Figure 3-4, Pt y MT follows the general trend of the absorption spectra of MT adducts with heavier metals to extend to longer wavelengths [2]. The spectra of Cd 7 MT and Zn 7 MT are similar in shape and intensity to those of Cd(II) and Zn(II) complexes of 2-mercaptoethanol which are arise from thiolate-to-metal charge-transfer transitions [134]. The spectrum of Pt 7 MT also contains weak d-d ligand field bands in addition to intense thiolate-to-Pt charge-transfer components. The electronic absorption spectrum of PtyMT is similar to the spectra of other metal MT adducts in that it is complex and difficult to interpret. Some interesting information about CdyMT, however, has been obtained from a spectrophotometric titration of apoMT with Cd 2+ ions [38]. We have performed a similar experiment to study the binding of Pt 2+ ions to apoMT. Figure 3-5 shows the spectra of partially platinated Pt(II) apoMT adducts recorded for a series of apoMT solutions eguilibrated with 1-8 mol equiv of K 2 PtCl 4 at pH 7.5. Also shown are the incremental difference spectra obtained by subtracting from each spectrum the spectrum preceding it. The simple approach of adding increments of K 2 PtCl 4 to a single solution of apoMT was not practical here as Pt(II) binding to apoMT at these low concentrations is

PAGE 49

40 T 1 1 1 200 250 300 350 400 A / nm Figure 3-5. Spectrophotometry titration of apoMT with K 2 PtCl 4 at pH 7. Shown are (a) the spectra resulting from the addition of to 8 mol equiv of K^PtCl^ to apoMT and (b) the incremental difference spectra obtained by the subtracting the preceding spectrum from each spectrum.

PAGE 50

41 very slow requiring about an hour to ensure equilibrium for each step. These long reaction times would be troublesome in the latter type of experiment as solutions of the apoMT at neutral pH are exceedingly sensitive to rapid airoxidation. Binding of the first mol equiv of Pt 2+ to apoMT results in an absorption peak centered at about 250 nm. A very intense shoulder of absorption appears at ca. 250 nm after binding of the second mol equiv of Pt 2+ The incremental difference spectrum for the third mol equiv of Pt 2+ contains two broad maxima at 250 and 300 nm. Mol equiv 4-7 result in absorption peaks of diminishing intensity at ca 280-300 nm. These data confirm the earlier finding that apoMT becomes saturated after binding 7 mol equiv of Pt 2+ Wilner and coworkers observed a 6 nm red shift in the incremental difference spectra when apoMT was titrated with more than three mol equiv of Cd(II) [38]. A decrease in the intensity of the absorption in the later stages of the titration (>3 mol equiv Cd 2+ ) was also observed. Good and Vasak also report a red shift in the electronic absorption spectra recorded during step-wise incorporation of Fe(II) into apoMT [31]. These red-shifts as the titration of apoMT exceeds three mol equiv of metal ions have been attributed to the onset of thiolato bridging. The first three metal ions presumably bind to three separate sets of terminal thiolate ligands to form three

PAGE 51

42 M(cys) 4 2 ~ units. Binding of the remaining four metal ions involves the formation of bridging thiolates from the remaining uncoordinated thiolates as well as transformation of some terminal thiolates to bridging thiolates [38]. Supporting evidence for this inferred mechanism was provided by spectra of Cd(II) complexes with the synthetic tetrathiol dodecapeptide, P-12 (Ac-P-C-Ornithine-C-P-Q-C-Q-C-R-R-V) [38]. Titration of P-12 with Cd(II) yields sharp breaking points at 1:1 and 2:3 molar peptide-to-Cd ratios presumably due to the formation of a monomeric mononuclear complex and a dimeric trinuclear complex, respectively. The band maximum of the monomeric complex red-shifts ca. 6 nm as excess Cd* is added and bridging thiolates are created upon formation of the trimeric complex [38]. Further evidence for this mechanism of cluster formation comes from an EPR study of Co (II) binding to apoMT done by Vasak and Kagi [137]. A linear increase in the room temperature EPR signal intensity occurs with titration of apoMT with 3-4 mol equiv of Co (II) ions followed by a linear decrease in intensity to zero with the addition of further ions to the saturation point of 7 mol equivalents. The EPR spectrum of Co 7 MT is characteristic of high spin tetrahedral Co(II) and is observed only at liquid helium temperatures. The loss of paramagnetism at higher temperatures is due to antiferromagnetic coupling between thiolato bridged Co (II) ions. The EPR titration data also

PAGE 52

43 suggest that the first 3-4 Co (II) ions bind to separate sets of terminal thiolates and that bridging and cluster formation occur after this initial process. We have obtained evidence that the above mechanism is operative in the binding of Pt(II) to apoMT by use of the colorimetric thiolate assay reagent, DTNB. As discussed in the previous section, DTNB reacts many times slower with thiolates coordinated to Pt 2+ than with free thiolates. This difference in reactivity makes it possible to determine the concentrations of both coordinated and free thiolates at each step in the titration of apoMT with Pt 2+ This experiment is not amenable to the study of Cd 2+ and Zn 2+ binding to apoMT as bound Cd 2+ and Zn 2+ do not sufficiently slow down the reaction of DTNB with thiolate. Figure 3-6 shows the results obtained when each of the solutions used in the spectrophotometric titration were allowed to react for one minute with a 315:1 molar excess of DTNBrMT under anaerobic conditions. It is clear that the first 3 mol equiv of Pt 2+ coordinate to 4 cysteine thiolates per Pt 2+ Binding of the remaining 4 mol equiv of Pt 2+ ions occurs by coordination to progressively fewer thiolate ligands per Pt 2+ ion. Roughly two thiolates are found to react with DTNB during the time permitted after the apo protein is saturated with the full complement of seven Pt 2+ ions. This residual reactivity is due either to two uncoordinated thiolates in PtyMT or two relatively reactive

PAGE 53

44 i 1 r 1 2 3 4 5 6 Mol Equivalents K 2 PtCI 4 Figure 3-6. Titration of the partially platinated apoMT adducts from each step of the spectrophotometry titration with the thiolate assay reagent, DTNB.

PAGE 54

45 coordinated thiolates being present. The reactivity of DTNB with Pt^ bound thiolates will vary depending on many factors, primarily the bond strength of the individual group and the accessibility to the reagent solution. It is possible, for instance, that two bridging thiolates in PtyMT are relatively exposed to the surrounding medium. The changes in the maxima of the ultraviolet spectra that occur upon the addition of more than 3 mol eguiv of Pt 2+ to apoMT can be accounted for by the formation of bridging thiolates and the transformation of terminal thiolates to bridging thiolates. In attempting to explain the apparent red-shift of the ligand-to-metal chargetransfer (LMCT) absorption maximum that coincides with this change in the mode of Pt 2+ binding, it is helpful to compare the electronic spectra of Pt(II) tetrahalides (PtX 4 2 ~) with the spectra of their dihalo bridged dimers (Pt 2 Xg ). Figure 3-7 is a molecular orbital diagram for a PtX 4 2complex showing the three allowed LMCT transitions. The LMCT spectra of the square-planar tetrahalide complexes of Pt(II), Pd(II), and Au(III) consist of two intense bands separated by about 9000-12000 cm -1 [138]. The first, lowest energy, absorption band (e 1 X 10 4 M_1 cm -1 ) originates from the two closely spaced Ma* hw transitions and the second, more intense, band (e 4 X 10 4 M _1 cm -1 ) originates from the Ma*
PAGE 55

46 Pt Pt L4 6p a ig( a ) /' • u ( a n • ) 2u ( n ") b lg ( o *) d„2. y 1 1 ,uv / ', ,;T\/' e < n •) d XI d, ,,' ,, U i xg ( o ") d,2 a 2g ( ) b 2u ( n") V, *<\ '' \ u ( n b ) m' > >'\ \ a 2u ( n b ) \\\ *\ e g ( it b ) b„ ( n) u ( b ) big ( b ) ., ( b > AAA > 1 '.' ',, x V .1 1 I '1 > 1 ;*> Figure 3-7. Relative molecular orbital energies estimated for sguare-planar Pt(II) complexes.

PAGE 56

47 increases in the order: AuBr 4 ~ < PdBr 4 2< AuCl 4 ~ < PdCl 4 2 ~ < PtBr 4 2 < PtCl 4 2 '. For PtX 4 2 ~, the Ma* La band usually occurs at wavelengths too low to be experimentally observed. The dihalo bridged dimers of Pt(II) and Pd(II) typically have intense absorption bands which lie 30-60 nm below the first LMCT bands of the corresponding monomers [139-142]. This behavior was observed for (Bu 4 N) 2 Pd 2 Cl 6 and (Bu 4 N) 2 Pd 2 Br 6 in glasses at 77 K in an early study by Mason and Gray [141] who speculated that these bands arose from the transfer of nonbonding n electrons from the bridging ligands to Pt. However, more recent polarized crystal spectra of (Et 4 N) 2 Pt 2 Br 6 [139] show that this first LMCT band must arise from transfer of n electrons from the terminal ligands. In order to understand this observation it is necessary to consider separately the two sets of molecular orbitals formed by the bridging and terminal ligands in the D 2h dimer. Defining the x-axis as normal to the plane of the molecule, one finds that while the p x orbitals of the bridging (p^) and terminal (p xt ) bromines both form pure out-of -plane n MO's, the p b and p zb orbitals are involved in a bonding only, and cannot be combined into a and in-plane ir MO's as obtained for p t and P z -j-Since the first intense band in the crystal spectrum of (Et 4 N) 2 Pt 2 Br 6 (317 nm, 9000 M -1 cm -1 ) is the only strong band in z

PAGE 57

48 polarization up to 270 nm, it must assigned to Ma* Lit transitions originating from the terminal bromines [139]. The occurrence of the first Ma* hn LMCT band of (Et 4 N) 2 Pt 2 Br 6 at a wavelength 49 nm longer than that of K 2 PtBr 4 [14 3] can be explained by using a very simple argument [139]. The Pt atoms in the dimer should be more electronegative than the Pt atom in the monomer because there are fewer bromides per Pt in the dimer to donate electronic charge. It is not surprising, then, that the first LMCT transition in the dimer should occur at lower energy than in the monomer as it is well known, from the work of Jorgensen [144], that the energy of this band can be correlated with the difference in electronegativities of metal and ligand. Delocalization within the central Pt(MX) 2 Pt region may also stabilize LMCT excited states of the dimers. The red-shifting and decrease in intensity of the band maxima of the incremental difference spectra in the later stages of the spectrophotometric titration of apoMT with K 2 PtCl 4 (Figure 3-5a) are completely consistent with a decrease in the number of terminal thiolates versus bridging k thiolates and a lowering in energy of the Ma Lit LMCT transitions for these terminal thiolates due to presence of bridging thiolates. It is also possible that some shifting of the LMCT bands to longer wavelengths results from a distortion of the ligand geometry at the Pt centers from D 4h to D 2d

PAGE 58

49 symmetry. The lowest lying LMCT band of the tetrahedrally distorted Pt(II) complex with the sterically hindered PEt 3 ligand (P-Pt-P angles are ca. 150), to take an extreme example, is found to be red shifted about 30 nm from the corresponding band for the planar bis(l 2-diethylphosphino) ethane complex [145]. However, it is not necessary to invoke this effect, and large distortions from square-planar geometry in these Pt(II) MT adducts do not appear to be energetically feasible because of the very large ligand field stabilization energies that would have to be overcome. Distortion of PtBr 4 2(^ =2.75 eV, & 2 = 74 eV / A3 = 44 eV ) to tne tetrahedral limit (a 1-22 eV = 4^/9), for example, would theoretically require 2.71 eV [138]. By comparison, the Gibbs' energy of a typical intramolecular hydrogen bond in a protein is only about 0.20 eV in aqueous solution [146]. The strongest hydrogen bonding is found in the hydrophobic interiors of proteins [146], and the interior of MT is unusual in that it contains metal-thiolate clusters rather than a hydrophobic core of amino acid residues. It seems reasonable, then, to expect that the structure of a metal MT adduct is largely dictated by the energetics of metal-thiolate cluster formation. However, the net forces of intermolecular hydrogen bonding exerted by the surrounding polypeptide chain of metal MT adducts are likely to cause slight distortions, even in the case of Pt(II)

PAGE 59

50 The circular dichroism spectra of apoMT containing to 7 mol equiv of bound Pt 2+ (Figure 3-8) are more difficult to interpret than the corresponding electronic absorption spectra. Ellipticity bands in the spectra of metal MT adducts contain contributions from both the conformationally dependent conf igurational chirality of the polypeptide and charge-transfer from the chiral cysteine thiolate ligands to the metal ions. Optical activity above 240 nm, however, contains only contributions from the asymmetric chargetransfer components [147]. The CD spectrum of apoMT at neutral pH is similar to its electronic absorption spectrum in that it is nearly devoid of spectral information except for a sharply decreasing negative ellipticity band beginning at about 240 nm and extending into the far ultraviolet. As shown in Figure 3-8, the addition of the first mol equiv of Pt 2+ brings about a single positive ellipticity band centered at about 250 nm. Addition of mol equiv 2-3 result in a significant drop in the (+) intensity of the spectrum but very little change in its shape. The ( + ) intensity of the 250 nm maximum rises abruptly after the addition of 4 mol equiv of Pt 2+ and remains fairly unchanged through 6 mol equiv, although there is some broadening of the spectrum in the region 250-300 nm. Binding of the seventh and final mol equiv of Pt 2+ causes a large increase in the (+) intensity

PAGE 60

51 O B T3 o CD o Cb 350 Figure 3-8. Titration of apoMT with to 7 mol equiv of K 2 PtCl 4 as followed by circular dichroism.

PAGE 61

52 of the spectrum giving a broad shoulder of ellipticity similar to that of the electronic absorption spectrum. Abrupt changes in CD spectra upon addition of the fourth mol equiv of metal ions have also been observed for partially metallated adducts of apoMT with Cd 2+ and Zn 2+ ions [21]. One interpretation of the spectral changes observed for Pt(II) binding is that major changes in the conformation of the protein occur during binding of the first 3 mol equiv of Pt 2+ ions, which entails the coordination of 4 cysteine ligands for each Pt 2+ ion as shown by the DTNB assays The fact that the CD features undergo practically no change at all for adducts containing 4-6 mol equiv of Pt 2+ is attributable to the inclusion of these additional bound ions into preformed metal-thiolate clusters via the formation of bridging thiolates. This interpretation relies heavily upon the common assumption that the CD spectra of metal MT adducts in general are ligand based, i.e. changes in rotary strength are linked to conf igurational changes in the cysteine ligands which occur with metal binding [2]. A similar lack of CD spectral changes following the addition of 4 mol equiv of metal ions has been found for the titration of apoMT with Cd 2+ ions [21]. The large ( + ) increase in the CD intensity observed for the binding of the seventh and final mol equiv of Pt 2+ is somewhat peculiar. We have suggested that one of the Pt^

PAGE 62

53 ions in PtyMT is bound to the N-terminal methionine residue, and it is reasonable to expect that this neutral ligand is thermodynamically and kinetically less reactive than the cysteine thiolates [49]. It is expected, then, that binding to methionine will occur only after saturation of the cysteine thiolate ligands. The binding of the seventh Pt 2+ ion to met-1 would produce a new chiral center at the thioether sulfur, and possibly a conformational alteration of the first several residues in this portion of the molecule, giving rise to a significant change in the CD spectrum. The far-UV CD spectra of apoMT (Figure 3-9) and native MT (Figure 3-10) both contain a strong negative ellipticity band near 200 nm. This band is similar to that observed for random coil polypeptides and is a reflection of the lack of secondary structure in both the apo and metallated forms of MT as confirmed by the crystallographic structure (Figure 1-2). The native protein contains essentially no a-helix or B-sheet regions. The crystallographic structure does show a considerable number of 6-turns which were also indicated by an earlier Raman and infrared absorption study which found MT to be a "predominantly turn-containing protein" [148]. The lower intensity of the 200 nm ellipticity band for native MT as compared with apoMT probably results from an increase in the number of 6-turns in the folded metallated form of MT. A decrease in the

PAGE 63

54 Figure 3-9. Circular dichroism spectra of apoMT and Pt y MT in the far ultraviolet.

PAGE 64

55 6 &0 0) -a
PAGE 65

56 intensity of this band for PtyMT compared to apoMT probably also results from a similar structural organization during Pt 2+ binding. Nuclear Magnetic Resonance Spectra NMR spectroscopy has become increasingly important in the study of the structure and dynamics of metallothionein and proteins in general. Otvos and Armitage correctly deduced the organization of the metal-thiolate clusters in MT by selective decoupling of the 113 Cd NMR spectrum of 113 CdyMT [149]. Their structural model was reported seven years prior to the publication of the crystallographic structure, and was of central importance to MT research in the intervening years. Wuthrich and coworkers at the ETH in Zurich have employed recently developed two-dimensional -"-H techniques to provide complete sequence-specific assignments for the 1 H spectra of rabbit liver and rat liver Cd 7 metallothioneins [150, 151] and construct models for the solution structures of metallothioneins [152, 153]. MT is a relatively small biopolymer and thus amenable to study by 1 H NMR. However, the signal dispersions in 1 H spectra of MT are not as great as in most other proteins due to the lack of ring currents from aromatic residues. The principle difficulties in obtaining high quality 1 H NMR spectra of proteins are the relatively low molar solubility

PAGE 66

57 of proteins, the large number of resonance lines occurring within a narrow chemical shift range, and the broad line widths due to slow tumbling rates. These difficulties have been overcome to large degree by the higher fields and better probe designs of modern instruments. The 300 MHz 1 H spectra recorded for the apo and native forms of rat liver MT-2 are shown in Figure 3-11. Because D 2 solutions were employed, the solventexchangeable amide protons, which resonate in the range 7-10 ppm, were not observed. The spectra of apoMT and native MT are similar in their gross features. The resonances in the region 3.8-4.6 ppm are predominantly those of the a-protons. The intense peak at ca. 3.0 ppm contains the methylene resonances of the twenty cysteine residues. The sharp singlets at 2.09 and 2.03 ppm derive from the methyl protons of the methionine residue and its N-acetyl group, respectively. An envelope of lysine 6-CH 2 resonances is centered at ca. 1.7 ppm. The methyl doublets of the alanine residues are grouped near 1.4 ppm, and those of the threonine residues near 1.2 ppm. The group of peaks near 1.0 ppm and below arise from methyl resonances of the aliphatic side chains of valine and isoleucine residues. The spectrum of apoMT (Figure 3-lla) is practically identical to that of a mixture of the constituent amino acids [154] indicating that apoMT at low pH is essentially a random coil polypeptide with all amino acid residues sharing

PAGE 67

58 ^V / 'I ,1 I : '! /X//A. 1/ vwv Figure 3-11. 300 MHz 1 H NMR spectra of (a) native rat liver MT-2 in 20 mM phosphate buffer pH 7.2 and (b) rat liver apoMT-2 in dilute acetic acid pH 2 (the multiplet at ca. 2.05 ppm, between the two sharp singlets, is from HD 2 C0 2 D). Resolution enhancement was employed (see Materials and Methods).

PAGE 68

59 the same average environment. The spectrum of the Cd 2+ and Zn 2+ containing native MT (Figure 3-llb) on the other hand, shows noticeable perturbations within the resonance envelope for a particular amino acid, with the most striking effects revealed within the alanine and threonine methyl regions. A disperse lysine 6-CH 2 proton envelope (1.6-1.8 ppm) appears in the spectrum of the folded metal lated form of the protein. The lysines in apoMT, on the other hand, appear to share the same average environment and give rise to a single well defined guintet at 1.68 ppm. MT has a high proportion of lysine residues (CH2CH2CH2CH2NH2 side chain), located almost exclusively adjacent to cysteines in positions which are highly conserved among the amino acid sequences of all mammalian metallothioneins (cf. Figure 1-1). Kojima [73] suggested that these lysines, which are protonated at physiological pH, serve to neutralize the net negative charge borne by the metal-thiolate clusters thereby stabilizing the molecule. Vasak, Kagi and coworkers [155-157] have provided experimental evidence which support this type of interaction involving the lysine residues. In the spectrum of apoMT the cysteine fc-CH 2 resonances (40 protons) centered at about 3.0 ppm overlap with e-CH 2 resonances of the 8 lysines (16 protons). In the spectrum of native MT the cysteine B-CH 2 resonances are spread out over range of about 0.5 ppm under the base of this peak [151]. However, a comparison of the area under the 3.0 ppm

PAGE 69

60 peak with that of the 0.9-1.1 ppm envelope (valine and isoleucine methyls, 12 protons) confirms that the majority of the cysteine B-CH 2 resonances are still centered at 3.0 ppm in native MT. The chemical shifts are sensitive to the metallic elements present. Mitsumori and Tohyama [158] have shown that when the Zn 2+ ions in native MT (67:37 Cd:Zn ratio) are displaced by Cd 2+ ions, to form CdyMT, the apparent triplets at 1.2 and 1.3 ppm coalesce into two doublets. The heterogenous metal composition of native MT results in two slightly different environments for threonines 9 and 27 causing a superimposition of a pair of doublets for each residue giving the two "pseudo-triplet" shapes. The native MT sample whose spectrum appears in Figure 3-11 contained 3.18 0.10 mol Zn/mol MT and the presence of bound Zn 2+ is apparent from the pseudo-triplets of the two threonine residues. The nature of the metal binding sites in rat liver MT-2 is revealed in a comparison of the 1 H spin-echo spectra of apoMT and native MT (Figure 3-12). Spin-echo experiments generally employ the pulse sequence 90-T-180-T-echo. x y The echo has the appearance of two free induction decays (FID's) placed back to back. The FID which follows the echo maximum may be recorded and Fourier transformed to obtain the conventional spectrum. Each peak in the resolved echo, however, is modulated at a frequency which is determined by

PAGE 70

61 1 u IaL /X^l i ',! I 1 W^/'liV ^*v 1 ^V*W^V^.V^ v^V>VV'Figure 3-12. 300 MHz 1 H spin-echo NMR spectra of (a) native rat liver MT-2 in 20 mM phosphate buffer pH 7.2 and (b) rat liver apoMT-2 in dilute acetic acid pH 2. Resolution enhancement was employed (see Materials and Methods )

PAGE 71

62 the J value of the given nucleus. The phase and amplitude of each peak in the spin-echo spectrum, then, is a function of this modulation due to spin-spin coupling and the length of 7" Note that the phase of the cysteine B-methylene envelope (ca. 3.0 ppm) in the spin-echo spectrum of native MT (Figure 3-1 2a) is reversed from that of apoMT (Figure 312b) The phase of a spin-echo resonance line is sensitive to spin-spin coupling [159], so coordination of metal ions to the cysteine residues must alter the coupling of these protons to each other and the cysteine a protons. The only other apparent phase reversal is for the peak farthest upfield, which is that of the methyl protons of valine 39. This is peculiar as binding of metal ions to the isopropyl side chain of this residue does not occur, but an interaction of the carbonyl oxygen or amide nitrogen with a metal ion cannot be ruled out. This spin-echo experiment was adapted from a similar experiment reported by Galdes et al. for equine kidney MT [160]. The conventional 1 H NMR spectrum and spin-echo spectrum of Pt 7 MT are shown in Figure 3-13. The familiar resonance envelopes for the different amino acid spin systems, that are seen for the apo and native protein, are also found in about the same chemical shift ranges in the 1 H spectrum of PtyMT. Unfortunately, even very concentrated samples of PtyMT gave only a very weak signal. The spectrum is of

PAGE 72

63 r^,V AsvAa Figure 3-13. 300 MHz 1 H (a) NMR spectrum and (b) spin-echo NMR spectrum of 0.6 mM rat liver Pt ? MT-2 (6.96 0.41 mol Pt/mol MT) in 20 mM phosphate buffer pH 7.2. Resolution enhancement was employed (see Materials and Methods).

PAGE 73

64 relatively poor quality due to the low signal to noise ratio. The low S/N may be the result of heterogeneity in the binding of Pt 2+ to individual MT apoMT molecules, slower T^ relaxation rates due to the larger size of PtyMT compared to apoMT and other metal MT adducts or more restricted intramolecular motion due to Pt 2+ binding, or some other unknown reason. A slight improvement in signal strength was obtained by using longer delay times between pulses, which indicates that relaxation processes are at least part of the problem in this case. Coordination of the cysteine residues to 195 Pt (natural abundance: 32%) causes a 32% reduction the intensity of the B-CH 2 resonances. The 195 Pt satellite peaks are broadened beyond detection at this high applied field as a consequence of chemical shift anisotropy [161]. The 3.0 ppm region in the spectrum of Pt 7 MT (Figure 3-1 3a) shows very little intensity, even taking into account coupling of the cysteine CH 2 protons to 195 Pt. This is not necessarily due to heterogeneity in PtyMT binding sites, however, as these resonances are spread over a large range (ca. 0.5 ppm) in the homogeneous adduct, CdyMT [151]. Examination of the spectra for apoMT and native MT (Figure 3-11) reveals that the cysteine B-CH 2 peak in the spectrum of native MT is shifted about 0.1 ppm downfield from that found for apoMT. This same deshielding effect is seen also in Figure 3-13a for PtyMT. What was observed as a

PAGE 74

65 single broad 3.0 ppm resonance region in the spectrum of native MT occurs as three separate peaks in the spectrum of Pt 7 MT. Binding of the cysteines to Pt 2+ may cause more pronounced deshielding of some the B-CH 2 protons than occurs with Cd 2+ and Zn 2+ binding. This effect is consistent with the greater electron-withdrawing tendency of Pt(II) versus Cd(II) or Zn(II), as seen in the electronic absorption spectra. The two sharp methyl singlets from the acetylated methionine residue which appear in the other MT spectra are noticeably absent in the spectrum of PtyMT. There is a relatively sharp peak occurring at 2.05 ppm which can be assigned to the acetyl group methyl but the methyl resonance of met-1 is either shifted to another region of the spectrum or broadened beyond detection. Gummin and coworkers have recently investigated the NMR properties of the complex formed between K 2 PtCl 4 and N-acetylmethionine [162]. The CH3S peak (1.96 ppm) shifts downfield by 0.31 ppm upon complexation of Pt 2+ by the sulfur atom and splits into a "1:4:1" pattern due to 195 Pt coupling. The two 195 Pt satellites were broadened beyond detection in the 300 MHz spectrum. The methyl singlet of the acetyl group was unaffected by complexation. These results are in agreement with the earlier finding by Boswell, et. al that upon complexation with K 2 PtCl 4 the

PAGE 75

66 CH 3 S resonance line of this compound is shifted 0.30 ppm downfield from 2.04 ppm for the free ligand [163]. From the above results it is expected that coordination of Pt 2+ to met-1 of MT would cause the CH 3 S peak at 2.09 ppm to vanish while leaving unchanged the CH 3 C(0) peak at 2.03 ppm. This is exactly what is observed in the spectrum of PtyMT, strongly suggesting that coordination of Pt 2+ to the thioether sulfur of met-1 does occur in this adduct. However, predicting the new location of the CH 3 S peak in the spectrum of PtyMT is problematic. The direction and extent of the shift is dependent on the particular protein molecule in which the methionine residue resides. In the complex formed by K^PtCl, and met-65 of tuna ferrocytochrome c, for example, this peak is actually removed 0.15 ppm upfield from the position of this peak in the free protein [163]. Bell and coworkers have recently examined the 500 MHz H spin-echo NMR spectra of cell culture media before and after addition of K 2 PtCl 4 cisplatin, and Pt(en)C12 [164]. They report the disappearance of the methionine CH 3 S peak at 2.15 ppm and the appearance of a new peak at ca. 2.71 ppm after addition of each of the above reagents. It is interesting, in light of this report, that a fairly sharp peak is observed at 2.75 ppm in the spectrum of PtyMT. This peak may be that of the methionine CH 3 S coordinated to Pt^ but this peak may lie elsewhere in the spectrum, perhaps obscured by other resonances.

PAGE 76

67 The spin-echo spectrum of PtyMT in Figure 3-13b is similar to that of the native protein (Figure 3-12b) in that the phases of the various resonance regions are the same. The only apparent difference is in the phase of the methyl doublet of the isopropyl side chain of valine 39 that appears farthest upfield in the two spectra. The phase of this peak is positive in the PtyMT and apoMT spin-echo spectra but, but as noted earlier in this section, it is negative in the native MT spectrum. The large positive cysteine 3.0 ppm peak confirms that cysteines are involved in Pt 2+ binding. 195 Pt NMR appeared to be a very promising method of studying the structure of PtyMT at the outset of this investigation. The high natural abundance, good receptivity, and extremely large known chemical shift range of this spin 1/2 nucleus make it possible, in principle, to determine the number of distinct Pt-thiolate clusters and how the Pt 2+ ions in these clusters are arranged by the spin-coupling in the spectrum as was done for CdyMT by using 113 Cd NMR [165, 166]. Unfortunately no 195 Pt signal was observed from concentrated samples of PtyMT using either the 200 or 300 MHz instruments. Attempts by Ismail and Sadler to observe 195 Pt signals from Pt 2+ bound to other macromolecules also ended in failure, and this was attributed to relaxation by chemical shift anisotropy [167].

PAGE 77

68 A similar difficulty in obtaining 199 Hg NMR spectra of Hg 7 MT has been encountered [25]. Binding Kinetics Unlike the intensive research activity dedicated to learning about the structures and equilibrium properties of MT adducts, the kinetics of metal binding to apoMT is an essentially unexplored area. The main reasons for this lack of interest probably include the difficulty in modeling a complex system containing multiple binding sites and the prohibitively rapid binding rates of the physiologically prevalent Cd 2+ and Zn 2+ ions. The extreme inertness of Pt(II) complexation can be used to advantage in this regard as it was in the previously described titration of partially platinated MT derivatives with DTNB. Changes in the ultraviolet absorption spectrum provide a convenient albeit crude measure of the extent of reaction of K 2 PtCl 4 with apoMT. The complex nature of the spectrum does not translate into a simple relationship between intensity at a given wavelength and concentration of reactants It provides only a very rough measure of the concentration of bound Pt 2+ ions. Absorbance was monitored versus time at 280 nm for a micromolar solution of apoMT mixed with a 50-fold molar excess of K 2 PtCl 4 at room temperature. This wavelength was

PAGE 78

69 chosen as the absorption at 280 nm increases in a nearly linear fashion versus mol equiv of bound Pt(II) (Figure 35b) The data may be fitted as two pseudo-first-order processes to a logarithmic plot of (A^ A) which is proportional to the concentration of apoMT, versus time (Figure 3-14a) This treatment is consistent with either consecutive or parallel reactions leading to the formation of Pt 7 MT. The plot of log(A 00 A)j in Figure 3-14b was obtained by subtracting from the biphasic plot the contribution of the slow step II, i.e. log(A 00 A)jj, after extrapolating step II to zero time. One interesting observation about this treatment of the data is that the ratio of A^ for step I to that of step II is 0.32/0.75 = 0.43 = 3.0/7.0. The obvious interpretation is that the fast phase is the binding of the first 3 Pt 2+ ions to the apoMT molecule, which occurs by the formation of three Pt(cys) 4 2 ~ moieties, and that incorporation of the remaining 4 Pt 2+ ions, which must proceed via the formation of some bridging thiolates, is the slower step II. It is reasonable to suppose that binding of the first 3 Pt 2+ ions would occur more rapidly due to the abundance of solventexposed thiolate ligands resulting from the greater freedom of motion of the loosely coiled polypeptide during the early stages of binding. The number of binding sites available to an incoming Pt 2+ ion during the later stages of binding is

PAGE 79

70 -0.2 -i 10 20 30 Time / min -0.5 -, -1.0 < I en o -1.5 -2.0 5 10 Time / min — i 15 Figure 3-14. Shown are (a) the time course of the reaction of 10 nM apoMT with a 50-fold molar excess of K^PtCl^ in 15 mM phosphate buffer pH 7.5, and (b) the plot obtained by subtracting the contribution to the absorbance of the slow component of the reaction, Ajj. A^ is the absorbance at 280 nm estimated for infinite time.

PAGE 80

71 20 40 60 80 100 120 Time / min Figure 3-15. Time course of the reaction of 10 p.M apoMT with a 50-fold molar excess of K 2 PtCl 4 in 0.01 M HC1 A„, is the absorbance at 280 nm estimated for infinite time.

PAGE 81

72 more restricted, on a purely statistical basis, than during the initial formation of the Pt(cys) 4 2units. This interpretation assumes that the binding of each Pt* ion produces an equal increase in the absorbance at 280 nm. Of course, this assumption about the additiveness of bound Pt 2+ absorptivities is only approximately correct as may be seen by the titration data in Figure 3-5. The slower binding of excess Pt 2+ to apoMT at low pH does not exhibit biphasic kinetics. The data yield a single pseudo-first-order process (Figure 3-15). As was seen in the 1 H NMR spectra, apoMT at low pH is essentially a random coil and this is reflected in the binding of roughly 17 Pt 2+ ions per MT to form a polymeric mixture. Conclusions At the outset of this project it was anticipated that Pt(II) would probably form a quite stable adduct with apoMT at neutral pH. The retention of bound Pt 2+ at very low pH was not anticipated. Nor was it expected that seven squareplanar Pt(II) ions would be chelated per MT molecule, which happens to be the same stoichiometry found for Cd 2+ Zn 2+ and other divalent metals with tetrahedral stereochemistry. The apparent similarities between PtyMT and these other adducts, besides the stoichiometry, are somewhat surprising in light of the two very different coordination geometries.

PAGE 82

73 The large Stokes' radius of PtyMT suggests an elongated two-domain structure similar to that of native MT. The electronic absorption and CD spectra obtained in the step-wise titration of apoMT with K 2 PtCl 4 follow a pattern very similar to that reported for the titration of apoMT with CdCl 2 [38]. However, the large change in the CD spectrum attending the binding of the seventh Pt 2+ ion has no analogue in the case of Cd 2+ binding [21], and this spectral change suggests Pt 2+ binding at the N-terminal methionine which is not a binding site for Cd 2+ The DTNB assays of the partially platinated apoMT samples indicate the initial formation of three Pt(cys) 4 units followed by incorporation of the remaining Pt* ions in bridged metalthiolate clusters. A similar binding seguence has been proposed for the formation of CdyMT [38]. That such cluster formation should occur in PtyMT is not surprising in view of the great propensity of classical thiol ligands to act as bridging ligands in their Pt(II) complexes. Simple thiols react with Pt(II) to form amorphous [Pt(p,-RS) 2 ] polymers (R = CH 3 CH 3 CH 2 c e H 5/ etc.) [49]. The fundamental differences between PtyMT and adducts of apoMT with the lighter metallic elements, in addition to the different stereochemistries, are the much greater strength of the Pt 2+ to sulfur bonds, the much slower rate of Pt 2+ binding, and the probable binding of one the seven Pt 2+ ions at met-1.

PAGE 83

74 The process of metal ion binding at neutral pH to form the fully metal lated form of MT from the apo form takes a course intermediate between two extremes. The first extreme is that of a rigid apoMT with preformed binding sites into which the metal ions are held. The second extreme is that of a random coil polypeptide that wraps around the metal thiolate complexes. ApoMT is essentially a random coil at neutral pH as its 1 H NMR spectrum is nearly identical to that of a mixture of the component amino acids [135]. The binding of Pt(II) to apoMT is probably a fairly random process under kinetic control. If the initial formation of Pt(cys) 4 2 ~ units is sufficiently faster than the subsequent phase of binding which entails thiolato bridging, as suggested by the kinetic data in Figure 3-14, it is not necessary that there be any thermodynamically favored binding sites in order for the type of spectral changes in Figure 3-5b to be observed. It is possible, for instance, that the binding of 7 Cd(II) ions to apoMT at neutral pH is a random process producing a distribution of kinetic products followed by intramolecular reorganization of the Cd( II) -thiolate bonds to form a single homogeneous CdyMT as the thermodynamic product. 113 Cd NMR studies of Cd ? MT [39] and metal ion exchange between Cd 7 MT and Zn 7 MT [20] and a recent variable temperature CD study of Cd(II) binding to Zn ? MT by Stillman

PAGE 84

75 and Zelazowski [22] all suggest that metal ions are mobile within a cluster. Although the initial binding process and formation of metal-thiolate clusters to form Pt ? MT appears guite similar to those observed for adducts with Cd(II) and other metal ions, it is possible that the arrangement of Pt(II) ions among the 20 cysteine residues in the eguilibrium product at room temperature differs from one molecule to the next due to the extreme kinetic inertness of Pt(II) which could prohibit the type of intramolecular rearrangement described above. The 1 H NMR spectrum of Pt y MT (Figure 3-1 3a) is similar to the spectrum of native MT in its gross features but has a relatively weak signal strength even for very concentrated samples, which suggests this type of inhomogeneity. The weak 1 H NMR signal may also be a conseguence of unusual relaxation properties of Pt 7 MT or other unknown factors however. It was hoped that the application of 195 Pt NMR would reveal the structural arrangement of these Pt(II)-thiolate clusters. This has not been realized due to unforeseen experimental difficulties. We must instead garner other small pieces of chemical and physical evidence to build a structural model. The discussion of a structural model for Pt ? MT is postponed until the end of Chapter 4, wherein additional data is presented regarding the arrangement of

PAGE 85

76 Pt 2+ ions in PtyMT and mixed metal Pt(II) adducts prepared via displacement of Cd 2+ and Zn 2+ from Cd y MT and native MT, Materials and Methods General Procedures Equine kidney MT-1A and rat liver MT-1 and MT-2 were isolated and purified according to a literature procedure [130]. ApoMT was prepared by exhaustive dialysis of the native protein against 0.01 M HC1 A small amount of the metal-chelating agent diphenylthiocarbazone in CC1 4 was suspended in the 0.01 M HC1. All chemicals employed were reagent grade or better. Deionized water and buffer solutions were passed through a column of Chelex chelating resin to remove trace metal contaminants. Dialysis tubing (Spectrapor MW cut-off: 3 500) was boiled repeatedly in deionized water and EDTA solution prior to use. Manipulations of oxygen-sensitive materials were done in a glove-box with nitrogen atmosphere and oxygen furnace. Solutions were thoroughly degassed on a vacuum line prior to transfer to the glove-box. Absorption and CD spectra were recorded on an IBM 9430 UV-visible spectrophotometer and a Jasco J-500 spectropolarimeter respectively. Sephadex G-50 and G-75 size-exclusion columns were calibrated for Stokes' radius against ovalbumin (R Q =

PAGE 86

77 2.73 nm) chymotrypsinogen A (2.09), myoglobin (1.88), ribonuclease A (1.64), and blue dextran (o) R Q was plotted against (-logK d ), where K d is the elution coefficient [132]. The eluant was 3 5 niM phosphate buffer, pH 7.25, and the elution was monitored continuously by absorbance at 229 nm. Flow rate was controlled by a calibrated peristaltic pump. Protein containing fractions were concentrated by using an Amicon YM5 or YM2 ultrafiltration membrane. Analytical Method Elemental Analyses were done by using a Perkin-Elmer Plasma II Emission spectrometer equipped with a Perkin-Elmer Series 7000 computer. The emission lines scanned were S I 180.73 nm, Cd II 214.44, Zn I 213.86, 324.75 and Pt II 214.420 nm. All protein samples were run side-by-side with ultraf iltrate or dialysate blanks. The instrumentation and methodology are described in Chapter 2 Preparation of Pt(II) ApoMT Adducts Samples of PtyMT were prepared by adjusting solutions of apoMT in 0.01 M HC1 to pH 7 with K 2 HP0 4 followed by the addition of a 10 to 100-fold molar excess of K 2 PtCl 4 After standing for 12 to 24 hours, the reaction mixtures were fractionated on a Sephadex G-50 or G-75 size-exclusion column, and the MT containing fractions pooled and

PAGE 87

78 concentrated by using an Amicon YM2 or YM5 ultrafiltration membrane To study the extent of Pt binding to apoMT at low pH, solutions of apoMT in 0.01 M HC1 were allowed to react with excess K 2 PtCl 4 over 24 hours. The extent of binding was determined by ICP-AES after stripping off unreacted Pt 2+ by ultrafiltration and desalting the equilibrium mixture on a size-exclusion column at neutral pH. Binding of Pt(II1 to ApoMT Modified at Methionine The N-terminal methionine of apoMT was converted to methionine sulfoxide by using the method described by Schechter [133]. ApoMT in 1 M HC1 was allowed to react with 1 M dimethyl-sulf oxide for 30 minutes. The resulting dimethyldisulf ide and unreacted DMSO were removed by repeated dialysis against 0.01 M HC1. Cd(II) and Pt(II) adducts of this modified apoMT were prepared in the glovebox at neutral pH as by using the procedure described in the preceding section. The Cd(II) and Pt(II) adducts of the modified apoMT were assayed for Cd, Pt, and S by ICP-AES. Reactions of Pt y MT with DTNB and Iodoacetate Buffered deaerated solutions of 3.8 M.M PtyMT (1 mL) and freshly prepared 2.0 mM 5, 5'-dithiobis( 2-nitrobenzoic acid) (1 mL) were mixed in a cuvet in the glove-box. The cuvet was sealed and transferred to the sample compartment of the

PAGE 88

79 UV-vis spectrophotometer and the increase in absorbance at 412 nm versus time recorded. The absorbance of a blank containing all reagents except protein was subtracted from the absorbance of the sample. The final pH was 7.4. The extent of reaction at equilibrium was determined by passing the equilibrium mixture through an ultrafiltration membrane and then measuring the absorbance of the ultraf iltrate at 412 nm versus that of a blank. The extent of reaction was calculated using the molar absorptivity of the resulting dianion of 2-nitro-5-thiobenzoic acid, 14150 M -1 cm -1 at 412 nm [168]. The rate and extent of the reaction of 9.0 |j.M PtyMT with 1 mM iodoacetate were obtained at pH 7.2 by measuring the release of I~ in a stirred cell with a calibrated I" selective electrode (Lazar Res. Lab. IS-146). Reaction of Pt y MT with Cyanide 1.00 mL of 35 (j-M PtyMT was dialyzed against 4.00 ml of 0.1 M NaCN in 0.1 M phosphate buffer pH 7.0 for 12 hrs during which the yellow color inside the dialysis bag was gradually lost. The Pt was found to be evenly distributed between of the contents of the bag and the surrounding solution by ICP-AES analysis. Absorption bands characteristic of Pt(CN) 4 2 ~ were found in the ultraviolet spectrum of the solution.

PAGE 89

80 Spectrophotometric Titration of Rat Liver ApoMT-1 with Ptflll and DTNB A series of 3.5 \iM apoMT solutions containing from to 10 molar equivalents of K 2 PtC14 were prepared in the glove-box. ApoMT solutions in 0.01 M HC1 were adjusted to pH 7 with K 2 HP0 4 and then diluted to 0.65 mL after addition of K 2 PtCl 4 increments. The reaction mixtures were left standing for three hours to equilibrate. A solution of 12 mM DTNB buffered at pH 7.5 was also prepared in the glove-box. The equilibrium mixtures were transferred to a 1 cm cuvet which was sealed with a greased septum. A 100 ^L syringe charged with 60 |aL of the DTNB (5, 5'-dithiobis(2nitrobenzoic acid)) solution was then mounted atop the cuvet, with the needle protruding through the septum, using a short length of tightly fitting rubber tubing. The DTNB reagent was held aloft over the protein solution by a small dead-space. This assembly was then quickly transferred to the sample compartment of the spectrophotometer. After recording the spectrum of the equilibrium mixture, the DTNB in the syringe was discharged into the cuvet. A spectrum was recorded one minute after mixing. The cuvet and syringe were then cleaned and returned to the glove-box and the above procedure repeated for each of the remaining samples. All spectra were recorded at slow scan speed, with air in the reference beam, and using the same

PAGE 90

81 cuvet. Spectra of blanks containing all reagents except protein and K 2 PtCl 4 were subtracted from sample-spectra. Free thiolate concentrations were estimated using the reported molar absorptivity of the dianion of TNB (2-nitro5-thiobenzoic acid), 14150 M -1 cm -1 at 412 nm [168]. Titration of ApoMT with Ptdll Followed by CD A series of 10 p,M apoMT solutions containing to 10 mol equiv of K 2 PtCl 4 were prepared in the glove-box. ApoMT solutions in 0.01 M HC1 were adjusted to pH 7.5 with NaOH and K 2 HP0 4 and then diluted to volume after addition of K 2 PtCl 4 increments. CD spectra were recorded for the samples using four labeled 1 cm cuvets sealed with greased septa. CD spectra of blanks containing all reagents except protein and K 2 PtCl 4 were recorded in the four cuvets and these blank-spectra were subtracted from the corresponding sample-spectra taken in the same cuvet. Five replicate scans were recorded for each sample and blank and signal-averaged. NMR Spectra Samples were prepared by lyophilizing solutions of protein in dilute phosphate buffer and then dissolving the resulting solids in D 2 0. This procedure was performed twice for each sample in order to remove residual HDO. The samples were then transferred to 1 mm sample tubes.

PAGE 91

82 All -'-H NMR spectra were run at 22C on a Varian VXR 300 spectrometer. Chemical shifts were referenced to external DSS (3-(trimethylsilyl)-l-propanesulfonate) The HDO signal was suppressed via T 1 discrimination, i.e. by applying a 180 pulse and 2 s delay prior to each 90 pulse. Resolution enhancement was obtained, for the spectra indicated, by weighting the FID by exp(t/K), where K 0.5, and a Gaussian apodization function. Spin echo spectra were obtained by the Carr-Purcell method [169]. The pulse seguence 90-7-180-T-collect was used with r = 50 ms. Binding Kinetics A syringe containing 80 m-L of 4.95 mM K 2 PtCl 4 was mounted atop a cuvet containing 720 nL of 11.1 p,M apoMT buffered at pH 7.5 (50:1 molar ratio Pt:apoMT) as described above for the spectrophotometric titrations. This assembly was transferred from the glove-box to the spectrophotometer. The contents of the syringe were discharged into the cuvet and the increase in absorbance versus time at 280 nm was recorded on a strip-chart recorder. A similar experiment was conducted for a solution of apoMT in 0.01 M HC1 ApoMT is stable with respect to airoxidation at this pH so the manipulations were performed on the bench top.

PAGE 92

CHAPTER 4 DISPLACEMENT OF CADMIUM AND ZINC FROM METALLOTHIONEIN BY PLATINUM(II) The preceding chapter described experiments with simple systems consisting of solutions apoMT and K 2 PtCl 4 In this chapter the more complex and biologically relevant problem of the formation of platinum(II) metallothioneins by displacement of Cd 2+ and Zn 2+ ions from native MT is considered. The first problem to be examined will be the reaction of K 2 PtCl 4 with CdyMT. This is followed by a discussion of the binding of Pt(II) from K 2 PtCl 4 and the derivatives of the anticancer drug, cisplatin, to heterogenous native metallothioneins, which contain predominantly Cd 2+ and Zn 2+ as well as lesser amounts of Cu + The results presented in Chapter 3 for the binding of Pt(II) to apoMT are used in an attempt to understand the adducts obtained by displacement of metal ions from the above adducts and a structural model for Pt(II)-thiolate clusters is described. Finally, implications for the metabolism of cisplatin in vivo are outlined. 83

PAGE 93

84 Reactions of Native and Cadmium Metallothioneins with Tetrachloroplatinate An experiment involving reactions of K 2 PtCl 4 with CdyMT gave somewhat unexpected results [27]. Our original intention was to attempt to produce CdyMT with a single Pt 2+ bound to the thioether sulfur of the N-terminal met-1 residue. We assumed that the Cd 2+ ions in CdyMT were bound strongly enough to prevent the Pt 2+ ions from reacting with the cysteine residues. It was found, however, that the addition of one mol eguiv of K 2 PtCl. or a slight molar excess of K 2 PtCl 4 resulted in the liberation of 3 mol equiv of Cd 2+ ions from CdyMT (Table 4-1). In the presence of air a single Pt 2+ ion was bound whereas in the exclusion of air 4 Pt^ ions were bound to produce a Pt 4 Cd.MT. The retention of 4 mol equiv of Cd 2+ ions in all the above cases strongly suggests that Pt(II) is displacing Cd 2+ from the 3-Cd R cluster and leaving the 4-Cd a cluster intact. This result is a reasonable one as Cd 2+ shows a marked preference for the a cluster in native MT and tends to accumulate in this cluster as opposed to the more weakly bound Zn 2+ which occupies the & cluster almost exclusively. Furthermore, it is known that the 6 cluster region in CdyMT and native MT is susceptible to attack by the strong electrophiles DTNB and iodoacetamide while the a region are generally inaccessible to these reagents [37],

PAGE 94

85 Table 4-1. ICP-AES of the Products From Reactions of Cd y MT and Native Equine MT with K 2 PtCl 4 Reactants (conditions) 3 Metal Contents of Products Metal Mol/Mol MT 1 5.5 \M Cd y MT + 5.8 \M K 2 PtCl4 (in air) Cd Pt total 3.82 0.87 1.11 0.25 4.93 0.91 8.2 M-M Cd ? MT + 85 \M K 2 PtCl 4 (in air) Cd 4.01 0.12 Pt 1.16 0.04 total 5.17 0.13 6.9 |XM Cd ? MT + 55 |iM K 2 PtCl 4 (under nitrogen) Cd 4.28 0.25 Pt 3.71 0.22 total 7.99 0.33 native MT Cd 5.04 0.36 Zn 0.94 0.08 Cu 0.63 0.23 total 6.61 0.43 28 \M native MT + 990 |iM K 2 PtCl 4 (under nitrogen) monomeric products Cd

PAGE 95

86 Templeton and coworkers have found that 113 Cd NMR lines attributed to Cd 2+ ions in the B cluster are lost before those attributed to Cd 2+ ions in the a cluster during the displacement of Cd 2+ from CdyMT by mercuribenzoate [170]. Rapid air oxidation of cysteine to cystine disulfide apparently occurs during the disruption of the 6 cluster by entering Pt 2+ ions. The ultraviolet spectrum of the product has absorption in the region 250-320 nm that is much too intense to be accounted for by the metal chromophores alone. Some of this absorption must arise from cystine disulfide groups which have absorption maxima at 270-280 nm. This oxidation of the cysteine side chains would abolish metal binding in this cluster. The single mol eguiv of Pt 2+ which is found in the final product is probably bound to the thioether sulfur of met-1, which does not undergo airoxidation under these conditions. The reaction of Pt 2+ ions with the cysteine sulfurs must be much faster than the reaction with methionine sulfur as the Cd 2+ ions are lost and oxidation of the cysteines occurs even for the eguimolar mixture of K 2 PtCl 4 and CdyMT. The reaction of excess K^PtCl. with CdyMT under nitrogen to produce Pt 4 Cd 4 MT suggests that the & cluster region can accommodate 3 Pt 2+ ions, assuming that the remaining Pt 2+ is bound to met-1. Extending this line of reasoning, the arrangement of Pt 2+ ions in PtyMT may be that

PAGE 96

87 of two separate 3Pt-thiolate cluster regions with the seventh Pt 2+ ion bound at the methionine. The products from the anaerobic reaction of excess K 2 PtCl 4 with native MT were divided about equally between an oligomeric fraction and a monomeric fraction when chromatographed on a size exclusion column. The Stokes' radius of the monomeric products is 0.2 nm larger than that of original native MT, i.e. the same as that of PtyMT (ca. 1.8 nm) The oligomeric fraction was composed of a large dimer peak and progressively smaller trimer, tetramer, and higher oligomer peaks, which were pooled for analysis by ICP-AES (Table 4-1). The monomeric products show a 75% deficiency in Zn content but only a 35% deficiency in Cd content as compared to the native MT used as starting material. This result is consistent with displacement of metal ions occurring predominantly in the 6 cluster as it is known that Zn 2+ ions are localized in this region of the native protein. The metal contents of the monomeric fraction are similar to those obtained for the anaerobic reaction of CdyMT with K 2 PtCl 4 In both cases 4 mol equiv of Pt 2+ ions are gained and 3 mol equiv of the initially bound metal ions are lost. The metal contents of the oligomeric fraction are quite different than those found for the monomeric fraction. The total mol metal/mol MT is very high and although most of the Zn 2+ has been displaced, none of the Cd 2+ is lost. The high

PAGE 97

88 content of total bound metals suggests that extensive disruption of the usual metal cluster structure and conformation of MT has occurred in this fraction. Such a catastrophic structural change could certainly make the affected MT molecules susceptible to oligomerization via intermolecular S-Pt-S cross-linking. The relatively high concentrations of the reactants and the large excess of K 2 PtCl 4 employed may also promote the formation of oligomeric species. Examination of the data in Table 4-1 also reveals that while a net loss of Zn 2+ and, to a lesser extent, Cd occurs, essentially all of the Cu originally bound by native MT is retained in the products. It is also interesting that nearly all of this Cu ends up in the monomeric fraction of the products. It is known that Cu is more tightly bound by MT than Cd 2+ presumably as Cu(I) [41]. Studies by Winge and coworkers indicate that Cu + is preferentially bound by the B cluster region of MT [32] as originally suggested by Briggs and Armitage [171]. It is possible that those molecules of native MT in which the metal ions occupying the B cluster are all or nearly all Cu + ions are much less reactive with Pt(II) due to the presence of very tightly bound Cu(I). The results in Table 4-1 are consistent with this proposition but the Cu concentrations in the native MT employed are rather small relative to the experimental errors in the ICP-AES data.

PAGE 98

89 Reactions of Native Metallothioneins with Cisplatin Derivatives The analytical data obtained for the reactions of CdyMT and native MT with K 2 PtCl 4 strongly suggest that Pt(II) is preferentially displacing metal ions from the R clusters of these reactants. It is desirable, however, to obtain spectroscopic data from which precise structural information about the protein molecule can be derived. The use of 1 H NMR spectroscopy and the complete set of assignments available for native rat liver MT-2 [151] provide this opportunity. Changes in the resonance lines of individual amino acid residues of MT in the upfield region of the 1 H NMR spectrum of rat liver MT-2 ten minutes following the addition of excess cis-Pt(NH 3 ) 2 (N0 3 ) 2 are shown in Figure 4-1. This derivative was used instead of cisplatin because of the prohibitively low solubility of the latter. Note that the resonance lines of residues 8, 9, and 27, all of which lie in the & domain of MT (residues 1-29), are significantly broadened during the early stages of the reaction while the lines of residues in the a (residues 30-61) domain are relatively unaffected. This confirms that Pt(II) is preferentially displacing metal ions from the 6 cluster as originally suggested by the analytical data. The changes in the upfield region of the spectrum and a slight decrease in the intensity of the cysteine B-CH 2 line

PAGE 99

90 '/ a M b I | : i n I : i ; | i i i : I I I I I I I I ] I I I I p I I I j I I I | I I i q— 1.8 1.6 1.4 1.2 1.0 ??u B Figure 4-1. 300 MHz 1 H NMR spectra of (a) 0.33 mM native rat liver MT-2 in 20 mM phosphate buffer pH 7 and (b) the same sample 10 min after the addition of a 20-fold molar excess of cis-Pt(NH 3 ) 2 (N0 3 ) 2 • Selected resonance lines are labeled with the names of the amino acid residues from which they arise.

PAGE 100

91 /V \y\M VKjA 'I'"" I 1 Figure 4-2. 300 MHz ^-H NMR spectra of (a) 0.33 mM native rat liver MT-2 in 20 mM phosphate buffer pH 7 and (b) the same sample 12 hrs after the addition of a 20-fold molar excess of cis-Pt(NH 3 ) 2 (N0 3 ) 2

PAGE 101

92 (ca. 3.0 ppm) occur quite rapidly after mixing the reactants. There is also a change in the N-terminal methionine methyl peak (2.09 ppm) that occurs over a longer time period. Figure 4-2 shows the full spectrum before and 12 hours following addition of the Pt reagent. The 2.09 ppm line experiences a noticeable broadening over the 12 hrs. The methyl peak of the N-acetyl group attached to met-1 (2.03 ppm) also shows some broadening, but to a lesser extent than the met-1 methyl peak. A slight decrease in the ratio of signal-to-noise (S/N) occurs over 12 hrs (the spectra in Figures 4-1 and 4-2 were not altered by weighting of the FID). This overall degradation of spectral quality is due to the formation of oligomeric MT species and colloidal zinc hydroxides during the reaction. The methyl singlets at 2.03 and 2.09 ppm are both very liable to broaden due to overall loss of S/N because of their narrow line widths. The greater change which occurs to the 2.09 ppm line versus the 2.03 ppm line suggests, however, that Pt 2+ binding to met-1 may also be partly responsible for the broadening of this line. Figure 4-3 shows the elution profile of this reaction mixture on a size-exclusion column. The products were recovered as 50% monomers, 15% dimers, and 35% higher oligomers. The average Stokes' radius of the monomer ic fraction is similar to that of CdyMT and native MT (1.60 nm) Table 4-2 shows the analytical data obtained for

PAGE 102

93 05 CM CM < _20 Elution 40 Volume 60 80 ml_ Figure 4-3. Elution profile of the products from the reaction of native rat liver MT-2 with excess Pt(NH 3 ) 2 (N0 3 ) 2 fractionated on a Sephadex G-50 size-exclusion column, is the absorbance at 229 nm. 4 229

PAGE 103

94 Table 4-2. ICP-AES of Products From the Reaction of Native Rat Liver MT-2 with cis-Pt(NH 3 ) 2 (N0 3 ) 2 Reactants (conditions) Metal Contents of Products Metal Mol/Mol MT* native MT-2 Cd

PAGE 104

95 the three fractions. The metal contents of the monomeric fraction are consistent with displacement of metal ions occurring exclusively in the zinc-rich ft cluster of native MT. Nearly 4 mol Cd/mol MT are retained in this fraction and 1 mol Pt/mol MT is bound. These data are nearly the same as those obtained for the aerobic reaction of CdyMT with excess K 2 PtCl 4 The metal contents of the dimeric fraction differ greatly from those of the monomeric fraction. Much less bound cadmium is present and the values for bound Pt and total bound metals are both much greater. The bound Cu contents of the monomeric and dimeric products are significantly higher than the Cu content in the starting material. Unfortunately the ICP-AES data obtained for the oligomeric fraction are of very poor quality and thus a mass balance cannot be obtained for Cu as was done for the anaerobic reaction of native MT with excess K 2 PtCl 4 The monomeric fraction was also tested for the presence of disulfide bonds by using the colorimetric reagent 2-nitro-5-thiosulfobenzoate (NTSB) [172]. Shown in Figure 4-4 is the cleavage of a disulfide by sulfite, followed by attack of NTSB on the resulting thiolate to yield the highly colored dianion of 2-nitro5-thiobenzoic acid (TNB 2). Two such assays on the monomeric fraction found ca. 3-4 mol disulf ide/mol MT. This is close to what would be expected for the oxidation of

PAGE 105

96 RSSR + SO
PAGE 106

97 cysteine side chains in the & cluster to cystine disulfides. It is possible that the NTSB undergoes an unwanted reaction with the metal-bound thiolates in the a cluster, thus giving a false analytical result. However, if these thiolates were not sufficiently protected from NTSB by the bound metal ions, one would expect a much larger production of TNB 2than actually found. The sulfur donor ligands, thiolates and thioethers, are known to have high trans-labilizing effects in Pt(II) complexes [49]. Thus, the binding of cis-diammine Pt(II) complexes to native MT should involve the loss of amine ligands via trans-labilization and the ultimate products would be the same as would be obtained from the binding of PtCl 4 2 ~ under the same conditions. The following experiments were designed to test whether or not such loss of amine occurs. The binding to native eguine kidney MT of Pt from the spin labeled complex, cis-Pt (ATMPO) 2 CI 2 (ATMPO = 4-amino2,2,6,6-tetramethylpiperidin-l-oxyl) [173], was followed by EPR spectroscopy. Free ATMPO ligand is readily distinguished from Pt(II) bound ATMPO by the difference in the EPR line shapes of this stable nitroxyl radical in the two states (Figure 4-5). Two broad bands occur between the central and low-field lines in the EPR spectrum of the Pt(II) complex due to electronic exchange between the two spin-labeled ligands located cis to one another.

PAGE 107

98 ATMPO = NH Figure 4-5. EPR line shapes of cis-Pt ( ATMPO ) 2 CI 2 and the ATMPO ligand in aqueous solutions.

PAGE 108

99 Figure 4-6 shows the distribution of metal ions in the size-exclusion chromatographic separation of the reaction mixture as determined by high-temperature-graphite-furnace atomic absorbance spectrometric (HTGF AAS) analysis of the fractions. The reaction mixture of native MT and cisPt(ATMPO) 2 Cl 2 (1:1 molar ratio Pt:MT) was kept at 37 C for 12 hours before separation. Nearly all the original MT was recovered in the prominent monomeric fraction. The Cd to Zn ratio in the most concentrated fraction (ca. 80 mL elution volume) is approximately the same as that of the starting material suggesting that no loss of Cd 2+ or Zn 2+ has occurred. The ratio of the sum of Cd and Zn concentrations to the Pt concentration in this fraction is 24.6 |j,M/3.3 p,M = 7.5. This ratio strongly suggests that a single Pt 2+ is bound for each MT. As apparently no displacement of Cd 2+ or Zn 2+ has occurred, this Pt 2+ is probably bound to the thioether sulfur of met-1. It is evident from the data in Figure 4-6 that a small portion of the Pt (ca. 24%) did not bind to MT and was recovered in the low molecular weight fractions (120-150 mL elution volume). The total Pt recovered, as determined by AAS of the fractions, was 82 nmol This accounts, within the limits of experimental error, for the 76 nmol of Pt complex used as starting material, also determined by AAS. The total MT recovered, as determined by AAS of Cd and Zn, was 55 nmol as compared to the 75 nmol of starting material,

PAGE 109

100 o • !— I cti d CD a o o o o o o pt a p a a p 2n A A A A A Cd. 50 100 Elution Volume 150 mL Figure 4-6. Elution profile on a Sephadex G-75 sizeexclusion column of the products from the reaction of 0.13 mM native equine kidney MT with 0.13 mM cis -Pt(ATMPO) 2 Clg in 3.6 mM Tris/HCl buffer pH 7.7. Metal concentrations of the 5 mL fractions were determined by HTGF AAS.

PAGE 110

101 determined gravimetrically This apparent loss of MT (ca. 27%) is probably not an actual loss of protein during the reaction but, rather, arises from an over-estimation of the amount of MT used as starting material. This experiment was done before the ICP-AES instrumentation was in place and simple gravimetry tends to give over-estimates of total protein due to water and salts trapped in lyophilized protein samples. The mass balance, then, also shows the binding of approximately 1 mol Pt/mol MT. Examination of EPR signals from both the protein fractions and the low MW fractions indicated that extensive loss of spin-labeled amine did indeed occur during the reaction. The most concentrated low MW fraction had a strong three-line EPR spectrum characteristic of the free ligand. Although about 75% of the available Pt was recovered in the protein-containing fractions, the most concentrated fraction in this region had only an extremely weak EPR signal which was visible above the background noise only at an amplification forty-times that used for the low MW fraction. A blank containing all reagents except MT was also kept at 37C for 12 hours and no loss of amine from the spin-labeled Pt(II) complex was observed. Figure 4-7 shows results from a similar experiment wherein the amount of MT was one fourth that used in the previous experiment (4:1 molar ratio Pt:MT). It is interesting that, with excess Pt complex present in the

PAGE 111

102 3 3. O • i— i cti d CD O a o o o o o o o pt, n d d Zn A A A A A Qd 50 100 150 Elution Volume / mL Figure 4-7. Elution profile on a Sephadex G-75 sizeexclusion column of the products from the reaction of 0.04 mM native equine kidney MT with 0.16 mM cis -Pt(ATMPCQ 2 cl 2 in 3.6 mM Tris/HCl buffer pH 7.7. Metal concentrations of the 5 mL fractions were determined by HTGF AAS.

PAGE 112

103 reaction mixture, all of the Zn 2+ originally present in the MT is liberated. The Zn 2+ lost from MT precipitated as a gelatinous hydroxide and was not recovered in the low MW fraction. The ratio of Cd to Pt in the most concentrated MT-containing fraction is 3.6 |j.M/0.7 nM = 5.1, nearly the same ratio that was obtained in the 1:1 Pt:MT experiment (16.4 nM/3.3 M-M = 5.0). This suggests that, again, no Cd 2+ was displaced and that a single mol Pt/mol MT was bound. The ratio of free Pt recovered versus bound Pt recovered (ca. 3.4) also supports this conclusion. The EPR spectra observed for the fractions from the 4 : 1 Pt:MT experiment were the same as those obtained in the 1:1 Pt:MT experiment. A strong threeline EPR spectrum, characteristic of the free amine ligand, was again found for the most concentrated low MW fraction and the most concentrated MT-containing fraction had only a very weak signal Ismail and Sadler [167] have also observed translabilization of amine ligands in reactions of cis diamminedichloro complexes of Pt(II) with N-acetylmethionine, glycine-methionine dipeptide, and RNase A. Release of both coordinated NH 3 and ethylenediamine was observed via 15 N NMR. The reactivity of Pt(ATMPO) 2 C1 2 w ^ th native MT is similar to that of K 2 PtCl 4 with CdyMT except that no displacement of metal ions from the MT occurred in the 1:1

PAGE 113

104 Pt(ATMPO) 2 Cl 2 :MT reaction mixture despite reactant concentrations about twenty times greater than in the 1 : 1 K 2 PtCl 4 :CdyMT mixture and the higher temperature. The lower reactivity of Pt(ATMPO) 2 ci 2 may be due to the bulky ATMPO ligands preventing the complex from penetrating into the metal clusters of the MT molecule. The difference in charge between the neutral spin-labeled complex and the tetrachloroplatinate ion may also be important in the absence of extensive hydrolysis. Conclusions The experimental results described in Chapter 3 provide evidence that, despite the square planar geometry of Pt(II), thiolato-bridged Pt 2+ clusters exist in PtyMT that are similar to the clusters in other metal MT adducts. Experiments described in this chapter provide information about the distribution of Pt 2+ ions along the protein chain and the reactivity of cisplatin, and analogous Pt anticancer drugs, with MT in vitro These limited data do not provide a final all-encompassing model for the various Pt(II) metallothioneins, but the evidence can be assembled into a partial model. An interpretation of the binding experiments is diagrammed in Figure 4-8. Note that native MT has been represented as an idealized Cd 4 Zn 3 MT for the sake of

PAGE 114

105 fi metl K 2 PtCI 4 in N-SCH 3 K 2 PtCI 4 in N-, K 2 PtCI 4 in air Figure 4-8. Possible products from the reactions of Note that the K 2 PtCl 4 with Cd ? MT and native MT. distribution of Cd and Zn in the native MT shown above is only one of many possible distributions.

PAGE 115

106 simplicity. Native MT is actually heterogenous with respect to the distribution of Cd 2+ and Zn 2+ ions and usually contains smaller amounts of other metal ions such as Cu + Fe 2+ and Pb 2+ The interpretation in Figure 4-8 is drawn from the following observations: (1) Displacement of metal ions by Pt(II) occurs exclusively in the B cluster region, as evidenced by the extensive loss of Zn 2+ from native MT upon reaction with excess K 2 PtCl 4 cis-Pt(NH 3 ) 2 (N0 3 ) 2 and Pt (ATMP0) 2 C1 2 It is known that Zn 2+ is concentrated in this region of native MT, whereas Cd 2+ is concentrated in the a cluster [165], It has also been shown that the 6 region of fully metallated MT will react with strong electrophiles such as DTNB and iodoacetamide while the a region is kinetically inaccessible to these reagents [37]. This supposition is also consistent with the retention of four Cd 2+ ions in the monomeric products of the reactions of K 2 PtCl 4 with Cd y MT and with native MT and the reaction of cis-Pt(NH 3 ) 2 (N0 3 ) 2 with native MT. (2) One mole eguiv of Pt(II) binds to the N-terminal methionine residue as evidenced by the binding of a single mol eguiv of Pt 2+ to air-oxidized MT from the aerobic reaction of K 2 PtCl 4 with Cd 7 MT and from the aerobic reactions of cis-Pt(NH 3 ) 2 (N0 3 ) 2 and Pt(ATMPO) 2 Cl 2 with native MT. The cysteine thiolates are susceptible to airoxidation, whereas the thioether side chain of methionine is

PAGE 116

107 not air-oxidizable. This conclusion is further supported by the decreased binding of Pt 2+ to apoMT containing methionine sulfoxide in place of methionine, and the disappearance of the 2.09 ppm methyl singlet of methionine in the 1 H NMR spectrum of PtyMT via coupling to 195 Pt. The binding of Pt(II) to the methionine of MT is not surprising because of the well known propensity of Pt(II) complexes to bind to exposed methionine residues in proteins [47, 50]. (3) If displacement of metal ions occurs exclusively in the R cluster, and one of the four mol eguiv of Pt 2+ ions found in the monomeric products of the anaerobic reactions of K 2 PtCl 4 with Cd-MT and native MT is bound to methionine, then three mol eguiv of Pt 2+ are bound in the 6 cluster region. It is likely that the same 3-Pt 6 cluster in the mixed metal Cd 4 Pt 4 MT adduct is also present in PtyMT A hypothetical 3-Pt & cluster is depicted in Figure 4-9b. The Pt-S bond distances in Pt(II) metallothioneins should be about the same as Zn-S bond distances of MT bound Zn 2+ ions (ca. 0.23 nm) based on crystallographic data for small thiolate containing Pt(II) complexes [174, 175]. Pt 2+ thiolate clusters are not, however, expected to be isomorphic with tetrahedral Cd 2+ and Zn 2+ thiolate clusters as significant distortion of sguare-planar geometry around Pt(II) in these adducts is not likely to be energetically feasible. The S-Pt-S and Pt-S-Pt bond angles in Figure 4-9b are an idealized 90 and 110 respectively and all bond

PAGE 117

108 Figure 4-9. Shown are (a) the 3-metal 6 cluster of native MT and (b) a hypothetical 3-Pt 6 cluster. Larger circles represent metal centers; smaller circles are thiolate sulfur positions.

PAGE 118

109 angles are 110 in the idealized 3-Zn 6 cluster (Figure 4-9a) This particular structural model was chosen because it is easily incorporated into the crystallographic conformation found for native MT (Figure 1-2). We have found, through the use of molecular models, that only very slight changes in this conformation of the protein are needed to insert the proposed 3-Pt cluster in place of the 3-Zn cluster. The three metals and the three bridging sulfurs in the two clusters can be almost exactly superimposed upon one another, thus the only significant difference between the two structures is in the orientations of the terminal sulfurs. The approximate differences in the orientations of these terminal ligands in going from the 3-Zn to the 3-Pt cluster are 90 about the S 4 rotation-reflection axes bisecting the plane of each Zn 2+ and its two terminal ligands. Despite the different orientations of the terminal sulfurs, this 3-Pt cluster can be inserted into the B region of the crystallographic MT structure, without significantly changing the conformation of the polypeptide backbone, by rotations about C a -C R bonds of the nine cysteines in this region. A better fit may be obtained by permitting changes in the Pt-S-Pt bond angles from the idealized 110. According to a recent review by Dance [176], bonds angles

PAGE 119

110 for M-S-M bonds in metal-thiolate complexes generally vary between about 90-120. Refinement of the crystal lographic structure in Figure 1-2 from 230 to 190 pm resolution has revealed that the M-S-M bond angles in native MT actually vary from 80-130 [9]. In addition, the net forces of hydrogen-bonding and other weak interactions in the polypeptide backbone of MT probably cause slight distortions from D 4h about the Pt(II) centers. Low molecular weight Pt(II)-thiolate complexes containing the structural unit in Figure 4-9b are not known, and this is not surprising as Pt(II) reacts with simple thiols to form [Pt(n~SR) 2 ] polymers. Two Pt 3 S 3 rings do occur in a hexanuclear complex of Pt(II) with 2-amino-ethanethiol, recently isolated by Gibson and Lippard [175]. In addition, the thiolato bridged trimers, Pd 3 (n-SC 2 H 5 ) 3 (S 2 CSC 2 H 5 ) 3 [177] and Pd 3 (SCH 2 CH 2 SCH 2 CH 2 S) 3 [178], both adopt structures of the type shown in Figure 4-9b. If the Pt 3 (Cys) g 3 ~ in Figure 4-9b occurs in Pt ? MT for the nine cysteines in the 6 region of MT, and a single Pt 2+ ion in PtyMT is bound to methionine, that leaves three Pt 2+ ions and eleven cysteines in the MT a region to be accounted for. Binding of 3 Pt 2+ ions in this region involving all 11 cysteines would reguire the presence of a (cys) 3 Pt(p,cys)Pt(Cys) 3 3 ~ unit and a Pt(cys) 4 2 ~ unit. The residual reactivity of approximately 2 mol cys/mol MT with DTNB for

PAGE 120

Ill 3PtyMT (Figure 3-6) suggests that a Pt 3 (cys) g J ~ cluster might also occur in the a region of PtyMT leaving two unbound cysteines in this region of the protein. Materials and Methods Reaction of K 2 PtCl 4 with Cd y MT in Air 1.1 mL of 16.3 m-M K 2 PtC14 was added to 2.0 mL of 8.56 M.M equine kidney Cd y MT in 3.5 mM phosphate buffer, pH 7.2 (1:1 molar ratio Pt:Cd y MT). After standing for 20 hrs, the reaction mixture was diluted to 20 mL and then concentrated by ultrafiltration. The resulting solution and ultraf iltrate blank were assayed for S, Cd ,and Pt by ICPAES. A similar experiment was conducted with a reaction mixture containing 1.5 mL of the above CdyMT solution and 60 |j.L of a 2.2 mM solution of K 2 PtCl 4 (10:1 molar ratio Pt:Cd 7 MT) Reaction of K 2 PtCl 4 with Cd y MT under Nitrogen 1.65 mL of 10.5 \iK equine kidney Cd ? MT in 3.5 mM phosphate buffer, pH 7.2, was mixed with 0.85 mL of 163 nM K 2 PtCl 4 (8:1 molar ratio Pt:Cd7MT) in the glove-box. After 24 hrs the mixture was removed from the glove-box, diluted to 12 mL, concentrated by ultrafiltration, and assayed for S, Cd, and Pt by ICP-AES.

PAGE 121

112 Reaction of K 2 PtCl 4 with Native MT under Nitrogen 1 mL of 28 |j,M native equine kidney MT was mixed with 10 [xL of 0.1 M K 2 PtCl 4 and allowed to stand for 24 hrs. The products were fractionated on a Sephadex G-50 column. Monomer fractions were pooled and concentrated by ultrafiltration as were fractions consisting of dimers and higher oligomers, and the two resulting solutions assayed for Cd, Zn, Cu, Pt, and S by ICP-AES. A solution of the native MT used in this experiment was also assayed for Cd, Zn, Cu, and S. Preparation of cis-Ptf ATMP01 2 C1 2 1.00 g K 2 PtCl 4 and 0.41 g of 4-amino-2 2 ,6 6tetramethylpiperidin-1-oxyl (ATMPO) were dissolved in 25 mL of water and stirred overnight under a slow stream of nitrogen. The brown powder that precipitated, was recrystallized from HC1 solution to yield a bright yellow powder. Analytical Calculated: C, 35.52; H, 6.29; N, 9.21. Found: C, 34.25; H, 6.30; N, 8.67. Preparation of cis-Pt(NH 3 ) 2 fNQ 3 ;> 2 0.250 g of cis-Pt(NH 3 ) 2 C1 2 was dissolved in 10 mL of water with 0.282 g of AgN0 3 and stirred overnight. The AgCl precipitate was removed by passage through a filter paper followed by a 0.2 |j,m filter. The water was removed in the

PAGE 122

113 lyophilizer and the resulting pale yellow residue used without further purification. Reaction of cis-Pt(NH 3 X 2 (N0 3 J. 2 with Native MT 100 jiL of a 50 mM cis-Pt (NH 3 ) 2 (N0 3 ) 2 solution in D20 was added to an NMR tube containing 600 n-L of a 0.325 mM solution of native rat liver MT-2 buffered at pH 7.2 with 20 mM phosphate. 1 H NMR spectra of the reaction mixture at 25C were recorded at 10 min intervals for about 2 hrs. After standing 24 hrs the reaction mixture was fractionated on a Sephadex G-50 column. Fractions were pooled, concentrated by ultrafiltration, and assayed for Cd, Zn, Cu, Pt, and S. A sample of the native MT-2 used in this experiment was assayed for Cd, Zn, Cu, and S. NTSB Disulfide Assay A 2-nitro-5-thiosulfobenzoate (NTSB) assay solution containing 0.25 mM NTSB and 0.1 M Na 2 S0 3 was prepared according to a literature method [172]. Two assays were performed by mixing 0.200 mL of the above solution (pH 9) with 0.500 mL of a 11.7 p.M solution of the monomeric products from the reaction of native MT with cis Pt(NH 3 ) 2 (N0 3 ) 2 and allowing them to react in the dark until the absorbance at 412 nm reached a maximum. The extent of reaction was calculated using the molar absorptivity of the

PAGE 123

114 resulting dianion of 2-nitro-5-thiobenzoic acid, 14150 M -1 cm -1 at 412 nm [168]. Reaction of cis-Ptf ATMP0U C1 2 with Native MT 460 M.L of 165 |J.M cis-Pt (ATMPO) 2 C1 2 was mixed with 100 HL of 750 m-M native rat liver MT-1 (1:1 molar ratio Pt:MT) in 20 mM Tris/HCl buffer, pH 7.7. After standing for 15 hrs in a 35C water bath the mixture was fractionated on a Sephadex G-75 column. Fractions were assayed for Cd, Zn, and Pt by high-temperature graphite-furnace atomic absorption spectrometry on a Perkin-Elmer 303 AA spectrometer equipped with a Perkin-Elmer HGA-2000 graphitefurnace. EPR spectra were recorded on a Bruker ER200D Spectrometer for the major protein containing fraction and Pt containing low molecular weight fraction using a quartz flat-cell. The same procedure was performed for a mixture Of 460 |XL of 165 M-M cis-Pt ( ATMPO ) 2 CI 2 with 25 M-L of 750 p.M native rat liver MT-1 (4:1 molar ratio Pt:MT).

PAGE 124

CHAPTER 5 GENERAL CONCLUSIONS The importance of MT as a major in vivo binding site for platinum anticancer drugs is well established [53, 57], but the molecular nature of in vivo Pt(II) MT adducts is not known. Concentrations of MT in the cytosol of mammalian cells are on the micromolar level. The experiment in Chapter 4 involving an eguimolar mixture of cis Pt(ATMPO)Cl 2 and native MT at approximately physiological temperature and pH suggests that in vivo binding of Pt drugs to MT may be confined to the met-1 residue. Zn 2+ is readily displaced from MT by Pt 2+ however, and higher doses of Pt drugs could lead to binding of Pt(II) to cysteines in the 6 cluster region of MT or the formation of cystine disulfides in the presence of dioxygen and other oxidants. MT concentrations in cells can be increased over 100-fold by induction of MT biosynthesis [179] and the elevated levels of MT could also lead to more extensive binding of Pt(II) from Pt drugs to MT. The Pt-induced oligomerization of MT found for the reactions of K 2 PtCl 4 and cis -Pt f NH 3 ) 2 ( N0 3 ) 2 with native MT should be taken into account when analyzing the distribution of Pt and other metals in size-exclusion chromatographic 115

PAGE 125

116 profiles of liver and kidney extracts from Pt treated animals. Such oligomeric MT might very well contribute to the metal contents of high-molecular-weight-fractions attributed to higher MW proteins. Sulfur assays might be a more valid way of identifying the MT-containing fractions than hydrodynamic properties. It is somewhat surprising that Pt(II) compounds show little or no ability to induce the biosynthesis of MT [57, 58] as induction is most prominent for heavy metal ions with toxicities and type b acceptor properties similar to Pt(II) [180]. The genetic mechanism of MT induction by metal ions is not well understood yet and it is possible that the stereochemistry of the metal ion is important, and in this respect Pt(II) is considerably different from the other metal ions that have been studied. The experimental results presented in Chapter 4 clearly show that Pt(II) from cisplatin derivatives displaces Zn 2+ from native MT leaving the more strongly bound Cd 2+ and Cu + intact. Depletion of available cisplatin by MT in tumor cells could diminish the effectiveness of cancer treatment with this drug [61, 181]. If MT is important in the regulation of Zn levels the binding of Pt(II) to the 6 cluster region of MT could also have deleterious effects on the patient by disrupting Zn homeostasis. On the other hand, seguestering of toxic Pt(II) metabolites by MT could be very beneficial in protecting the

PAGE 126

117 patient from the inactivation of enzymes and other more critically important biomolecules. Naganuma and coworkers [64] have reported that the toxic effects of cisplatin in rats are substantially prevented, following induction of MT biosynthesis by Bi(N0 3 ) 3 without compromising the antitumor activity of cisplatin. Bi(N0 3 ) 3 and other agents may selectively induce biosynthesis of MT in healthy tissue but not in the tumor cells. It is likely that MT plays an important role in the sequestering of Cd 2+ and other toxic metal ions, and a better understanding of the interactions of MT with metal ions could lead to new treatments for heavy metal poisoning.

PAGE 127

REFERENCES 1. Kagi, J. H. R. ; Nordberg, M. Eds. Metallothionein ; Birkhauser-Verlag: Basel, 1979. 2. Vasak, M. ; Kagi, J. H. R. Metal Ions in Biological Systems ; Sigel, H. Ed.; Marcel Dekker: New York 1983, 15, 213. 3. Vasak, M. J. Mol Catalysis 1984, 23, 293. 4. Kagi, J. H. R. ; Kojima, Y., Eds. Metallothionein-2 ; Birkhauser-Verlag: Basel, 1988. 5. Margoshes, M.; Vallee, B. L. J. Am. Chem. Soc. 1957, 235 4813. 6. Kojima, Y.; Berger, C; Kagi, J. H. R. Metallothionein ; Kagi, J. H. R. ; Nordberg, M. Eds.; Birkhauser: Basel, 1979, p. 153. 7. Winge, D. R. ; Nielson, K. B.; Zeikus, R. D.; Gray, W. R. J. Biol. Chem. 1984, 259, 11419. 8. Furey, W. F.; Robbins A. H. ; Clancy, L. L. ; Winge, D. R. ; Wang, B. C; Stout, C. D. Science 1986, 231 704. 9. Collett, S. A.; Stout, C. D. Reel. Trav. Chim. Pays-Bas: Proceedings of the Third International Conference on Bioinorganic Chemistry f Noordwi jkerhout Netherlands 1 ; van Koten, G.; Reedijk, J., Eds.; Elsevier: Amsterdam, 1987, 106 182. 10. Kagi, J. H. R. 8th Int. Conar. Biochem. Abstr. 1970, Switzerland, pl30. 11. Mehra, R. K.; Bremner, I. Biochem. J. 1984, 219, 539. 12. Szymanska, J. A.; Zelazowski A. J.; Stillman, M. J. Biochem. Biophys. Res. Comm. 1983, 115 167. 118

PAGE 128

119 13. Day, F. A.; Funk, A. E.; Brady, F. O. Chem. -Biol. Interactions 1984, 50, 159. 14. Funk, A. E.; Day, F. A.; Brady, F. O. Comp. Biochem. Physiol. 1987, 86C, 1. 15. Ikebuchi, H.; Teshima, R. ; Suzuki, K. ; Sawada, J.; Terao, T.; Yamane, Y. Biochem. Biophys. Res. Comm. 1986, 136, 535. 16. Schmitz, G.; Minkel, D. T.; Gingrich, D.; Shaw, C. F. Ill J. Inorg. Biochem. 1980, 12, 293. 17. Szymanska, J. A.; Stillman, M. J. Biochem. Biophys. Res Comm 1982, 108 919. 18. Zelazowski, A. J.; Garvey, J. S.; Hoeschele, J. D. Arch. Biochem. Biophys. 1984, 229 246. 19. Otvos, J. D. ; Engeseth, H. R. ; Wehrli, S. Biochemistry 1985, 24, 6735. 20. Nettesheim, D. G.; Engeseth, H. R.; Otvos, J. D. Biochemistry 1985, 24., 6744. 21. Stillman, M. J.; Cai W. ; Zelazowski, A. J. J. Biol. Chem. 1987, 262, 4538. 22. Stillman, M. J.; Zelazowski, A. J. J. Biol. Chem. 1988, 263 6128. 23. Vasak, M. ; Worgotter, E.; Wagner, G. ; Kagi J. H. R.; Wuthrich, K. J. Mol. Biol. 1987, 196, 711. 24. Szymanska, J. A.; Law, A. Y. C; Stillman, M. J.; Zelazowski, A. J. Inorg. Chim. Acta 1983, 79, 123. 25. Johnson, B. A.; Armitage, I. M. Inorg. Chem. 1987, 26, 3139. 26. Laib, J. E.; Shaw, C. F. Ill; Petering, D. H. ; Eidsness, M. K. ; Elder, R. C; Garvey, J. S. Biochemistry 1985, 24, 1977. 27. Bongers, J.; Bell, J. U. ; Richardson, D. E. J. Inorg. Biochem. 1988 3_4, 55. 28. Waalkes, M. P.; Harvey, M. J.; Klaassen, C. D. Toxicology Letters 1984, 20, 33. 29. Vasak, M. ; Kagi, J. H. R. ; Holmquist, B. ; Vallee, B. L. Biochemistry 1981, 20, 6659.

PAGE 129

120 30. Vasak, M. J. Am. Chem. Soc. 1980, 102 3953. 31. Good, M.; Vasak, M. Biochemistry 1986, 25, 8353. 32. Nielson, K. B.; Winge, D. R. J. Biol. Chem. 1984, 259, 4941. 33. Byrd, J.; Winge, D. R. Arch. Biochem. Biophys. 1986, 250 233. 34. George, G. N. ; Winge, D. ; Stout, C. D.; Cramer, S. P. J. Inorq. Biochem. 1986, 27, 213. 35. Nielson, K. B.; Winge, D. R. J. Biol. Chem. 1983, 258, 13063. 36. Avdeef, A.; Zelazowski, A. J.; Garvey, J. S. Inorq. Chem. 1985, 24, 1928. 37. Bernhard, W. R. ; Vasak, M. ; Kagi J. H. R. Biochemistry 1986, 2_5, 1975. 38. Willner, H.; Vasak M. ; Kagi, J. H. R. Biochemistry 1987, 26, 6287. 39. Vasak, M.; Hawkes, G. E. ; Nicholson, J. K. ; Sadler, P. J. Biochemistry 1985, 24, 740. 40. Bernhard, W. R.; Good, M. ; Vasak, M. ; Kagi, J. H. R. Inorq. Chim. Acta 1983, 7_9, 154. 41. Nielson, K. B.; Atkin, C. L. ; Winge, D. R. J. Biol. Chem. 1985, 260, 5342. 42. Thomson, A. J.; Williams, R. J. P.; Reslova, S. Structure and Bonding Springer-Verlag: New York, 1972, 11, 1. 43. Cisplatin; Current Status and New Developments Prestayko, A. W. ; Crooke, S. T.; Carter, S. K.; Alder, N. A., Eds.; Academic Press: New York, 1980. 44. Rosenberg, B.; Van Camp, L. ; Trosko, J. E., Mansour, V. H. Nature 1969, 222, 385. 45. Pinto, A. L. ; Lippard, S. J. Biochem. Biophys. Acta 1985, 780, 167. 46. Howe-Grant, M. E.; Lippard, S. J. Metal Ions in Biological Systems ; Sigel, H. Ed.; Marcel Dekker: New York, 1980, 11, 63.

PAGE 130

121 47. Melius, P.; Friedman, M. E. Inorg. Perspect. Biol. Med. 1978, 1, 1. 48. Von Hoff, D. D.; Schilsky, R.; Reichert, C. M. ; Reddick, R. L. ; Rozencweig, M. ; Young, R. C; Muggia, F. M. Cancer Treat. Rep. 1979, 63 1527. 49. Hartley, F. R. The Chemistry of Platinum and Palladium ; John Wiley and Sons: New York, 1973. 50. Blundell, T. L. ; Johnson, L. N. Protein Crystal lography ; Academic Press: New York, 1976, Chpt. 8. 51. Hoeschele, J. D. ; Butler, T.A. ; Roberts, J. A. Inorganic Chemistry in Biology and Medicine ; Martell, A. E. Ed.; Am. Chem. Soc. : Washington D. C, 1980, p. 181. 52. Robins, A. B.; Leach, M. 0. Cancer Treat. Rep. 1983, 67, 245. 53. Litterst, C. L. Agents and Actions 1984, 15_, 520. 54. Sharma, R. P.; Edwards, I. R. Biochem. Pharmacol. 1983, 32, 2665. 55. Choie, D. D. ; del Campo, A. A.; Guarino, A. M. Tox. Appl. Pharmacol. 1980, 55, 245. 56. Sharma, R. P. Pharmacol. Res. Comm. 1985, 17., 197. 57. Zelazowski, A. J.; Garvey, J. S. ; Hoeschele, J. D. Arch. Biochem. Biophys. 1984, 229 246. 58. Kinsler, S.; Bell, J. U. Biochemistry Int. 1985, 10, 847. 59. Bakka, A.; Endresen, L. Johnsen, A. B. S., Edminson, P. D.; Rugstad, H. E. Toxicol. Appl. Pharmacol 1981, 61, 215. 60. Endresen, L.; Schjerven, L. ; Rugstad, H. E. Acta Pharmacol et Toxicol 1984, 55, 183. 61. Kelly, S. L. ; Basu, A.; Teicher, B. A. ; Hacker, M. P.; Hamer, D. H.; Lazo, J. S. Science 1988, 241 1813. 62. Naganuma, A.; Satoh, M. ; Koyama, Y.; Imura, N. Toxicol Lett. 1985, 24, 203.

PAGE 131

122 63. Satoh, M. ; Naganuma, A.; Imura, N. Cancer Chemother. Pharmacol 1988, 21, 176. 64. Naganuma, A.; Satoh, M. Imura, N. Cancer Res. 1987, 47, 983. 65. Buhler, R. H. 0.; Kagi J. H. R. Metallothionein ; Kagi, J. H. R. ; Nordberg, M. Eds.; Birkhauser: Basel, 1979, p. 211. 66. Wetlaufer, D. B. Adv. Protein Chem. 1962, 12, 296. 67. Bongers, J.; Walton, C. D.; Richardson, D. E.; Bell, J. U. Anal Chem. in press. 68. Heine, D. R. ; Babis, J. S.; Denton, M. B. Appl Spectrosc. 1980, 34./ 595. 69. Nygaard, D. D.; Leighty, D. A. Appl Spectrosc 1985, 39, 968. 70. Suzuki, K. T. ; Kobayashi E.; Sunaga H.; Shimojo, N. Anal. Lett. 1986, 19, 863. 71. Morita, M. ; Uehiro, T. ; Fuwa, K. Anal. Chim. Acta 1984, 166, 283. 72. Sunaga, H. ; Kobayashi, E. ; Shimojo, N. ; Suzuki, K. T. Anal.Biochem. 1987, 160, 160. 73. Kojima, Y. ; Berger, C; Vallee, B. L.; Kagi, J. H. R. Proc. Natl. Acad. Sci 1976, 73, 3413. 74. Vasak, M. ; Kagi, J. H. R. ; Hill, H. A. 0. Biochemistry 1981, 20, 2852. 75. Margoliash, E. ; Smith, E. L. Nature 1961, 192 1125. 76. Myer, Y. P.; MacDonald, L. H. ; Verma, B. C. ; Pande, A. Biochemistry 1980, 19, 199. 77. Butt, W. D.; Keilin, D. Proc. Roy. Soc. 1962, B156 429. 78 Margoliash, E. ; Frohwirt, N. Biochem. J. 1959, 71 570. 79. Kirschenbaum, D. M. Handbook of Biochemistry and MolecularBiolocry 3rd ed.; Fasman, G. D., Ed.; CRC Press: Cleveland, 1975; 2, p. 383.

PAGE 132

123 80. Webster, G. C. Biochim. Biophys. Acta 1970, 207 371. 81 Mayer, M. M. ; Miller, J. A. Anal. Biochem. 1970, 36, 91. 82. Dautrevaux, M. ; Boulanger, Y.; Han, K. ; Biserte, G. Eur. J. Biochem. 1969, 11, 267. 83. Antonini, E.; Brunori, M. Hemoglobin and Myoglobin in their Reactions with Ligands ; American Elsevier: New York, 1971, p. 44. 84. Romberg, R. W. ; Kassner, R. J. Biochemistry 1979, 18 5387. 85. Antonini, E. Physiol. Rev. 1965, 45, 123. 86. Rossi-Fanelli, A.; Antonini, E.; Povoledo, D. Symposium on Protein Structure ; Neuberger A Ed ; Camelot Press Ltd.: London, 1958; p. 144. 87. Schecter, A. N. ; Epstein, C. J. J. Mol Biol. 1968, 35, 567. 88. Smyth, D. G.; Stein, W. H. ; Moore, S. J. Biol. Chem. 1963, 238 227. 89. Meadows, D. H. ; Jardetzky, O. Proc. Natl. Acad. Sci 1968, 61, 406. 90. Sage, H. J.; Singer, S. J. Biochemistry 1962, 1, 305. 91. Blumenfeld, O. O. ; Levy, M. Arch. Biochem. Biophys. 1958, 76, 97. 92. Harrington, W. F. ; Schellman, J. A. Comptes Rend Lab. Carlsberg Ser. Chim. 1956, 30, 21. 93. Kirschenbaum, D.M. Anal. Biochem. 1975, 68, 465-484. 94. Krauz, L. M. ; Becker, R. R. J. Biol. Chem. 1968, 243 4606. 95. Scott, R. A.; Scheraga, H. A. J. Am. Chem. Soc. 1963, 85, 3866. 96. Tanford, C. ; Havenstein, J. D. ; Rands, D. G. J. Am. Chem. Soc. 1955, 77, 6409.

PAGE 133

124 97. Mikes, 0.; Holeysovsky, V.; Tomasek, V.; Sorm, F. Biochim. Biophys. Res. Comm. 1966, 24, 346. 98. Maekawa, K.; Leiner, I. E. Arch. Biochem. Biophys. 1960, 91, 108. 99. Kirschenbaum, D.M. Anal. Biochem. 1977, 80/ 193. 100. Meloun, B.; Fric, I.; Sorm, F. Eur. J. Biochem. 1968, 4, 112. 101. Green, M. N. ; Neurath, H. J. Biol. Chem. 1953, 204 379. 102. Mihalyi, E.; Harrington, W. F. Biochim. Biophys. Acta 1959, 36, 447. 103. D'Albis, A. Biochim. Biophys. Acta 1970, 200 34. 104. Jacquot-Armand, Y ; Hill, M. FEBS Lett. 1970, 11, 249. 105. Gentou, C; Yon, J.; Filitti-Wurmser S. Bull. Soc. Chim. Biol. 1960, 50, 2003. 106. Green, N. M. Biochem. J. 1957, 66/ 407. 107. Buck, F. F.; Vithayathil, A. J.; Bier, M. ; Nord, F.F. Arch. Biochem. Biophys. 1962, 9_Z, 417. 108. Schroeder, D. D.; Shaw, E. Arch. Biochem. Biophys. 1971, 142, 340. 109. Robinson, N. C. ; Neurath, H. ; Walsh, K. A. Biochemistry 1973, 12, 414. 110. Berrens, L. ; Bleumink, E. Int. Arch. Allergy 1965, 28, 150. 111. Liu, W. ; Trzeciak, H. ; Schussler, H. ; Meienhoffer, J. Biochemistry 1971, 10, 2849. 112. Abita, J. P.; Delaage, M. ; Lazdunski, M. ; Savrda, J. Eur. J. Biochem. 1969, 8,314. 113. Blow, D. M.; Birktoft, J. J.; Hartley, B. S. Nature 1969, 221, 337. 114. Eisenberg, M. A.; Schwert, G. W. J. Gen. Physiol. 1951, 34, 583. 115. Nichol, J. C. J. Biol. Chem. 1968, 243, 4065.

PAGE 134

125 116. Brandts, J.; Lumry, R. J Phys Chem 1963, 67 1484. 117. Wilcox, P. E.; Cohen, E.; Tan, W. J. Biol. Chem. 1957, 228, 999. 118. Chervenka, C. H. Biochim. Biophys. Acta 1959, H, 85. 119. Jackson, L. M. ; Becker, R. R. Biochemistry 1970, 9, 2294. 120. Smillie, L. B. ; Enenkel A. G.; Kay, C. M. J. Biol. Chem. 1966, 241, 2097. 121. Wu, F. C; Laskowski, M. J. Biol. Chem. 1955, 213, 609. 122. Guy, O.; Gratecos D. ; Rovery, M. ; Desnvelle, D. Biochim. Biophys. Acta 1966, 115 404. 12 3. Raval, D. N. ; Schellman, J. A. Biochim Biophys Acta 1965, 107 463. 124. Myers, S. A.; Tracy, D. H. Spectrochim. Acta 1983, 38B, 1227. 125. Boumans, P. W. J. M. Inductively Coupled Plasma Emission Spectrometry Part 1: Methodology Instrumentation, and Performance ; Boumans P. W. J. M. Ed.; Wiley-Interscience: New York, 1987, p. 85. 126. Christian, G. D. ; Ruzicka, J. Spectrochim. Acta 1987, 42B, 157. 127. Lawrence, K. E.; Rice, G. W. ; Fassel V. A. Anal. Chem. 1984, 56, 289. 128. Koropchak, J. A.; Winn, D. H. Anal. Chem. 1986, 58 2558. 129. Lee, J.; Pritchard, M. W. Spectrochim. Acta Biochem. Biotechnol. 1982, 36B, 591. 130. Sobocinski, P. Z.; Canterbury, W. J. Jr.; Mapes, C. A.; Dinterman, D. E. Am. J. Physiol. 1978, 234., E399. 131. Wallace, G. F.; Pirc, V. V.; Ediger, R. D. Can. J. Spectrosc. 1982, 27, 46. 132. McGuinness, E. T. J. Chem. Ed. 1973, 50, 826.

PAGE 135

126 133. Schechter, Y. J. Biol. Chem. 1986, 261, 66. 134. Kagi, J. H. R. ; Hiramelhoch, S. R. ; Whanger, P. D. ; Bethune,J. L. ; Vallee, B. L. J. Biol. Chem. 1961, 236, 2435. 135. Galdes, A.; Hill, H. A. 0.; Kagi, J. H. R.; Vasak, M. ; Bremner, I.; Young, B. W. Metallothionein ; Kagi, J. H. R. ; Nordberg, M. Eds.; Birkhauser: Basel, 1979, p. 241. 136. Takahashi, H. ; Nara, Y.; Yoshida, T.; Tuzimura, K.; Meguro, H. Agric. Biol. Chem. 1981, 45, 79. 137. Vasak, M.; Kagi, J. H. R. Proc. Natl. Acad. Sci. USA 1981, 78, 6709. 138. Gray, H. B. Transition Metal Chemistry ; 1965, 1, 239. 139. Martin, D. S. Jr.; Rush, R. M. ; Peters, T. J. Inorg. Chem. 1976, 15, 669. 140. Day, P.; Smith, M. J.; Williams, R. J. P. J. Chem. Soc. A. 1968, 668. 141. Mason, W. R. Ill; Gray, H. B. J. Am. Chem. Soc. 1968, 90, 5722. 142. Olsson, L. Inorg. Chem. 1986, 25, 1697. 143. Kroening, R. F.; Rush, R. F.; Martin, D. S.; Clardy, J. C. Inorg. Chem. 1974, 13, 1366. 144. Jorgensen, Proa. Inora. Chem. 1970, 12, 101. 145. Kozelka, J.; Luthi H. ; Dubler, E. ; Kunz R. W. Inorg. Chim. Acta 1984, 8.6, 155. 146. Rawn, D. J. Biochemistry ; Harper & Row: New York, 1983, Chapter 4. 147. Rupp, H. ; Weser, U. Biochim. Biophys Acta 1978, 533 209. 148. Pande, J.; Pande, C; Gilg, D.; Vasak, M. Callender, R.; Kagi, J. H. R. Biochemistry 1986, 25, 5526. 149. Otvos, J. D.; Armitage, I. M. J. Am. Chem. Soc. 1979, 101, 7734.

PAGE 136

127 150. Wagner, G. ; Neuhaus D. ; Worgotter, E.; Vasak, M. ; Kagi, J. H. R. ; Wuthrich, K. Eur. J. Biochem. 1986, 157 275. 151. Worgotter, E.; Wagner, G.; Vasak, M. ; Kagi, J. H. R.; Wuthrich, K. Eur. J. Biochem. 1987, 167 457. 152. Braun, W. ; Wagner, G.; Worgotter, E.; Vasak, M. ; Kagi, J. H. R.; Wuthrich, K. J. Mol. Biol. 1986, 187 125. 153. Wagner, G. ; Neuhaus, D.; Worgotter, E.; Vasak, M. ; Kagi, J. H. R.; Wuthrich, K. J. Mol. Biol. 1986, 187 131. 154. Vasak, M. ; Galdes, A.; Hill, H. A. 0.; Kagi, J H R ; Bremner I ; Young B W Biochemistry 1980, 19, 416. 155. Vasak, M. ; Berger, C. ; Kagi, J. H. R. FEBS Lett. 1984, 168, 174. 156. Pande, J.; Vasak, M. ; Kagi, J. H. R. Biochemistry 1985, 24, 6717. 157. Vasak, M. ; McClelland, C. E. ; Hill, H. A. 0., Kagi, J. H. R. Experientia 1985, 41, 30. 158. Mitsumori, F.; Tohyama, C. J. Biochem. 1984, 96, 533. 159. Campbell, I. D. ; Dobson, C. M. ; Williams, R. J. P.; Wright, P. E. FEBS Lett. 1975, 57, 96. 160. Galdes, A.; Vasak, M. ; Hill, H. A. 0.; Kagi, J. H. R. FEBS Lett. 1978, 92, 17. 161. Ismail, I. M. ; Kerrison, S. J. S.; Sadler, P. J. Polyhedron 1982, 1, 57. 162. Gummin, D. D. ; Ratilla, E. M. A.; Kostic, N. M. Inora. Chero. 1986, 25, 2429. 163. Boswell, A. P.; Moore, G. R. ; Williams, R. J. P. Biochem J. 1982, 201 523. 164. Bell, J. D.; Norman, R. E.; Sadler, P. J. J. Inorg. Biochem. 1987, 3_1, 241.

PAGE 137

128 165. Otvos, J. D.; Olafson, R. W. ; Armitage, I. M. Biochemical Structure Determination by NMR ; Sykes, B. D.; Glickson, J.; Bothner-By, A. A., Eds.; Marcel Dekker: New York, 1981, p. 65. 166. Hunt, C. T. ; Boulanger, Y.; Fesik, S. W. ; Armitage, I. M. Env. Health Persp. 1984, 54, 135. 167. Ismail, I. M. ; Sadler, P. J. Platinum Gold, and Other Metal Chemotheraputic Agents ; A. C. S. Symp. Ser. 1983, 209, 171. 168. Riddles, P. W. ; Blakeley, R. L. ; Zerner, B. Anal. Biochem. 1979, 94./ 75. 169. Carr, H. Y.; Purcell, E. M. Phys. Rev. 1954, 94, 630. 170. Templeton, D. M. ; Dean, P. A. W. ; Cherian, M. G. Biochem. J. 1986, 234, 685. 171. Briggs, R. W. ; Armitage, I. M. J. Biol. Chem. 1982, 257 1259. 172. Thannhauser, T. W. ; Konishi, Y.; Scheraga, H. A. Methods in Enzymology ; Jakoby, W. B. ; Griffith, O. W. Eds.; Academic Press: Orlando, 1987, 143 115. 173. Mathew, A.; Bergguist, B. ; Zimbrick, J. J Chem Soc Chem Comm 1979, 222. 174. Bird, P. H.; Siriwardane, U.; Rabin, D. L. ; Shaver, A. Can. J. Chem. 1982, 60, 2075. 175. Gibson, D. ; Lippard, S. J. Inorg. Chem. 1986, 25 219. 176. Dance, I. G. Polyhedron 1986, 5, 1037. 177. Fackler, J. P. Jr.; Zegarski W. J. J Am Chem Soc. 1973, 95, 8566. 178. McPartlin, E. M. ; Stephenson, N. C. Acta Cryst. 1969, B25, 1659. 179. Karin, M. Cell 1985, 41, 9. 180. Jones, M. M. ; Meredith, M. J.; Dodson, M. L. ; Topping, R. J.; Baralt, E. Inorg. Chim. Acta 1988, 153 87.

PAGE 138

129 181. Andrews, P. A.; Murphy, M. P.; Howell, S. B. Cancer Chemother. Pharmacol. 1987, 19, 149.

PAGE 139

BIOGRAPHICAL SKETCH Jacob Bongers was born in Milwaukee on February 6, 1959, to Donald and June Bongers. He spent most of his childhood in Green Bay, Wisconsin, where he graduated from South West High School. Jacob attended Macalester College in Saint Paul, Minnesota, for one year. After spending a semester away from school to complete training as a medic for the army national guard, he enrolled at the University of Wisconsin in Green Bay. After receiving his B.S. degree in 1982, he began work towards a doctoral degree in chemistry at the University of Florida, under the supervision of Dr. David E. Richardson. On July 23, 1988, Jacob married Cyndy Walton of London Grove, Pennsylvania, who is also a doctoral candidate in chemistry. 130

PAGE 140

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the deqree of Doctor of Philosophy. \^AcKJ,< David E. Richardson, Chairman Associate Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Paul W. Chun Professor of Biochemistry and lolecular Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. eft^JbktjAtK George E. Ryschkewitsch Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. R. Carl Stoufer'" 77 Associate Professor of Chemistry

PAGE 141

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. j UJULA — U_\jJ$L^~> William Weltner Jr, Professor of Chemistry ^~ ) This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the College of Liberal Arts and Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. May, 1989 Dean, Graduate School

PAGE 142

UNIVERSITY OF FLORIDA 3 1262 08553 4583