Nickel-dependent oxidative cross-linking of a protein

Article abstract of DOI:10.1021/tx960170i

A model protein, ribonuclease A (bovine pancreas), was examined for its ability to coordinate Ni²⁺ and promote selective oxidation. In the presence of a peracid such as monopersulfate, HSO₅^-, nickel induced the monomeric RNase A to form dimers, trimers, tetramers, and higher oligomers without producing fragmentation of the polypeptide backbone. Co²⁺ and to a lesser extent Cu²⁺ exhibited similar activity. The nickel-dependent reaction appeared to result from a specific association between the protein and Ni²⁺ that allowed for transient and in situ oxidation of the bound nickel to yield intermolecular tyrosine-tyrosine cross-links. Macrocylic nickel complexes that had previously been shown to promote guanine oxidation were unable to mimic the activity of the free metal salt. Amino acid analysis of the protein dimer confirmed the expected consumption of one tyrosine per polypeptide and formation of dityrosine. The presence of excess tyrosine efficiently inhibited formation of the protein dimer and produced instead a ribonuclease- tyrosine cross-link. In contrast, high concentrations of the hydroxyl radical quenching agent mannitol only partially inhibited ribonuclease dimerization. The polypeptide-mediated activation of nickel and its subsequent reactivity mimic a process that could contribute to the adverse effects of nickel in vivo.

Full text of DOI:10.1021/tx960170i

302
Chem. Res. Toxicol. 1997, 10, 302-309
Nick el-Dep en d en t Oxid a tive Cr oss-Lin k in g of a P r otein †
Gurpreet Gill, Angelika A. Richter-Rusli, Madhushree Ghosh,‡
Cynthia J . Burrows,§ and Steven E. Rokita*,‡
Department of Chemistry, State University of New York at Stony Brook,
Stony Brook, New York 11794
Received October 3, 1996X
A model protein, ribonuclease A (bovine pancreas), was examined for its ability to coordinate
Ni2+ and promote selective oxidation. In the presence of a peracid such as monopersulfate,
HSO5-, nickel induced the monomeric RNase A to form dimers, trimers, tetramers, and higher
oligomers without producing fragmentation of the polypeptide backbone. Co2+ and to a lesser
extent Cu2+ exhibited similar activity. The nickel-dependent reaction appeared to result from
a specific association between the protein and Ni2+ that allowed for transient and in situ
oxidation of the bound nickel to yield intermolecular tyrosine-tyrosine cross-links. Macrocylic
nickel complexes that had previously been shown to promote guanine oxidation were unable
to mimic the activity of the free metal salt. Amino acid analysis of the protein dimer confirmed
the expected consumption of one tyrosine per polypeptide and formation of dityrosine. The
presence of excess tyrosine efficiently inhibited formation of the protein dimer and produced
instead a ribonuclease-tyrosine cross-link. In contrast, high concentrations of the hydroxyl
radical quenching agent mannitol only partially inhibited ribonuclease dimerization. The
polypeptide-mediated activation of nickel and its subsequent reactivity mimic a process that
could contribute to the adverse effects of nickel in vivo.
The chemical and biological properties of nickel are
highly dependent on its surrounding ligands. Both the
coordination geometry and oxidation-reduction potential
are significantly altered when simple salts forming Ni-
(H2O)62+ in aqueous solution are chelated by strong donor
ligands. Crystalline, amorphous, and metallic nickel
have been shown to produce tumors in animals and may
share a common mechanism for this involving intracel-
lular Ni(II) despite differences in their initial exposure,
uptake, and solubilization (for review, see refs 1-3). This
transition metal ion stongly associates with amino acids,
peptides, proteins, and nucleic acids. Such chelation can
be extremely stable and exhibit dissociation constants
with the tripeptides, GGH, GGG, and DAH, or serum
albumin on the order of 10-10 M-1 (4, 5). Some deleteri-
ous effects of nickel may result solely from coordination,
but most are likely related to the oxidative activity of
these complexes.
The ability to control the redox chemistry of nickel by
appropriate ligand design has also allowed for develop-
ment of numerous reagents applicable in vitro. Our
laboratories have characterized nucleic acid structure
with a series of macrocyclic nickel complexes including
(2,12-dimethyl-3,7,11,17-tetraazabicyclo[11.3.1]heptadeca-
1(17),2,11,13,15-pentaenato) nickel(II) (NiCR)1 and 1,4,8,-
11-tetraazacyclotetradecane nickel(II) (Nicyclam) that
oxidize guanine residues according to the accessibility of
the guanine N7 atom (13, 14). Use of an alternative
ligand promoted covalent bond formation between the
nickel complex and the same accessible guanines (15).
These examples all relied on a peracid as the terminal
Under physiological conditions, simple nickel salts
exhibit little intrinsic oxidation or reduction chemistry
particularly in contrast to iron, copper, and cobalt (2, 6).
However, after nickel binds to the backbone of peptides
•-
and proteins, it becomes quite easily oxidized by O2, O2
,
and H2O2 (7, 8). GGH-Ni(II), glutathione-Ni(II), and
other oligopeptide complexes of Ni(II) have been shown
to promote lipid and protein oxidation in the presence of
peroxides to a much greater extent than either free Ni2+
or ligand alone (9, 10). Similarly, GGGG-Ni(II) induced
polymerization of histones (11), and histidine-Ni facili-
tated oxidation of deoxyguanosine (12).
† This work is based in part on the M.S. thesis of A.R.-R. (August,
1991) and the Ph.D. dissertation of G.G. (December, 1993).
‡ Current address: Department of Chemistry and Biochemistry,
University of Maryland, College Park, MD 20742.
Abbreviations: CAPS, 3-cyclohexylamino-1-propanesulfonate; 2D-
1
COSY, two-dimensional correlated spectroscopy; MCPBA, meta-chlo-
roperbenzoate; MMPP, magnesium monoperoxyphthalate; NiCR, (2,12-
dimethyl-3,7,11,17-tetraazabicyclo[11.3.1]heptadeca-1(17),2,11,13,15-
pentaenato)nickel(II); PVDF, poly(vinylidene difluoride); TSP,
3-(trimethylsilyl)propionate; tren, tris(2-aminoethyl)amine nickel(II);
cyclam, 1,4,8,11-tetraazacyclotetradecane.
§ Current address: Department of Chemistry, University of Utah,
Salt Lake City, UT 84112.
X Abstract published in Advance ACS Abstracts, February 15, 1997.
S0893-228x(96)00170-1 CCC: $14.00 © 1997 American Chemical Society

Protein Cross-Linking Induced by a Simple Nickel Salt
Chem. Res. Toxicol., Vol. 10, No. 3, 1997 303
â-mercaptoethanol, 150 mM Tris (pH 6.8), and 0.1% bromophe-
nol blue). Samples were then heated to 100 °C for 5 min and
analyzed by SDS-polyacrylamide gel electrophoresis using a
Laemmili discontinuous system with 5% stacking and 15%
resolving gels (34). Proteins were stained by Coomassie Blue
R-250. Reactions were quantified by densitometric analysis
using an Enprotech scanner and software (Natick, MA).
oxidant, but related derivatives have been investigated
that activate O2 for DNA modification (16). Others have
demonstrated the use of a bleomycin-nickel complex for
similar identification of guanine (17, 18). Nickel-peptide
complexes have additionally been adapted for sequence
and structure specific probes of DNA (19-22). In each
case, the nickel-based reagents benefit from the forma-
tion of a non-diffusible oxidant.
RNase was alternatively cross-linked by horseradish peroxi-
dase (35) under the same conditions described above except 0.08
µM peroxidase was used in place of the nickel salt, and 800 µM
H2O2 was used in place of the KHSO5. The reaction was
quenched, denatured, and analyzed as described above.
The redox chemistry of nickel-peptide complexes has
concurrently been applied to mapping protein-drug (23)
and protein-protein interactions (24). Likewise, other
transition metal complexes have been employed to char-
acterize protein structure in analogy to their use in the
field of nucleic acids. For example, the nonspecific
cleavage of a protein by EDTA-Fe(II) (25) could be
localized by tethering this complex to a protein (26-29)
or biotin (30). A comparable study made use of the native
recognition of tetracycline-Fe(II) for specific protein
cleavage (31). Sequence and structure specific hydrolysis
of proteins may also soon be possible as a consequence
of the exciting observation that Pd(II) could selectively
hydrolyze dipeptides containing N-acetyl-His-X (32).
This latter approach may ultimately allow for polypeptide
fragmentation at accessible sites adjacent to histidine
residues with a simple metal complex thereby eliminat-
ing the need for specialized ligands or tethers. We
pursued a similar strategy for protein oxidation based
on the intrinsic ability of free Ni2+ to bind a peptide
backbone spontaneously (5) and create a metal-peptide
site that would be sensitive to oxidants (8). Flexible loop
and coil regions were the expected targets since Ni(II)
would most easily be accommodated in these regions.
Initial studies described below demonstrate that nickel
readily bound to the monomeric protein RNase A and led
to its oxidative cross-linking to form multimeric products.
An alternative process, protein fragmentation, was not
observed and therefore not competitive with amino acid
side chain oxidation.
Am in o Acid An a lysis of RNa se A a n d Its Der iva tives
a fter Sep a r a tion by Gel Electr op h or esis. Protein fractions
were transferred from polyacrylamide gels to poly(vinylidene
difluoride) (PVDF) membranes (Bio-Rad) by electroblotting. The
regions of the membrane containing protein were excised and
submitted for amino acid analysis (Picotag, Millipore Corp.,
Milford, MA) to the Center for Analysis and Synthesis of
Macromolecules at Stony Brook. A standard of 3,3′-dityrosine
was synthesized according to the procedure below.
Syn th esis a n d Mod el F or m a tion of 3,3′-Dityr osin e.
Tyrosine (13 mM), Ni(OAc)2 (26 mM) and 6 mM KHSO5 were
combined in 10 mM borate buffer (pH 10) and mixed at 22 °C
for 30 min. Reaction was followed by removing and filtering
(Acrodisc, Fisher Scientific) small aliquots and then measuring
the characteristic fluorescence of this dityrosine isomer (36). The
desired product was obtained after isocratic elution (50 mM
triethylammonium acetate, pH 6) through reverse phase HPLC
(C-18 column) in an unoptimized yield of 11% based on the
limiting reagent, KHSO5. 1H-NMR and 2D-COSY NMR (AMX-
600) of the product are consistent with 3,3′-dimerization. 1H-
NMR (0.1 M TSP, pH 11) δ 2.78 (dd, J ) 15, 5 Hz, 1H, Hâ1′), 2.8
(dd, J ) 13, 5 Hz, 1H, Hâ1), 3.02 (dd, J ) 13, 5 Hz, 1H, Hâ2′),
3.14 (dd, J ) 13, 5 Hz, 1H, Hâ2), 3.5 (dd, J ) 7.8, 5 Hz, 1H, HR′),
3.56 (dd, J ) 7.8, 5 Hz, 1H, HR), 6.84 (d, J ) 8.1 Hz, 1H, H-6′),
7.09 (d, J ) 8.1 Hz, 1H, H-6), 7.13 (d, J ) 8.3 Hz, 1H, H-5′),
7.27 (d, J ) 8.2 Hz, 1H, H-5), 7.3 (s, 1H, H-2′), 7.4 (s, 1H, H-2).
Resu lts
Ni(II) P r om oted Oxid a tive Cr oss-Lin k in g of Mon -
om er ic RNa se A in th e P r esen ce of KHSO5. Incuba-
tion of RNase A (13.7 kDa) in the presence of Ni2+ and
HSO5- generated high molecular weight products corre-
sponding to RNase dimer, trimer, and higher oligomers
(Figure 1A, lane 2). Both Ni2+ and HSO5- were required
for this process, and addition of a chelator such as EDTA
(4 mM) completely inhibited cross-linking. No competing
fragmention of RNase was evident under any conditions,
and therefore, oxidative modification of the polypeptide
backbone was unlikely under these conditions. The
dimerization process was mimicked by the action of
horseradish peroxidase and H2O2 (Figure 1A, lane 3),
which is known to couple proteins via formation of
dityrosine (35, 36). Consistent with this model process,
the nickel-dependent reaction was similarly controlled
by pH, and an optimal yield was observed at pH 10.
Accordingly, all further experiments were performed at
this pH using borate buffer. Although borate may
selectively complex dityrosyl residues (37), this buffer was
not necessary for RNase dimerization and could be
replaced with 3-cyclohexylamino-1-propanesulfonate
(CAPS, 10 mM).
Ma ter ia ls a n d Meth od s
Ca u tion . Perchlorate salts of metal complexes with organic
ligands are potentially explosive and should be handled carefully
to avoid mechanical agitation. These salts should be prepared
in small quantities only.
Ma ter ia ls. Ribonuclease A from bovine pancreas Type III-A
was purchased from Sigma (St. Louis, MO) and used directly
in all experiments. Molecular weight standards for electro-
phoresis under denaturing conditions were obtained from Bio-
Rad (Hercules, CA) and contained Escherichia coli â-galactosi-
dase (97 kDa), bovine serum albumin (66 kDa), egg white
ovalbumin (42 kDa), bovine carbonic anhydrase (31 kDa),
soybean trypsin inhibitor (21 kDa), and lysozyme (14 kDa). All
buffer solutions were made with distilled, deionized, and filtered
water (Nanopure, Sybron/Barnstead, Dubuque, IA). Reagents
and buffers were purchased from Fisher Scientific (Pittsburgh,
PA), Aldrich (Milwaukee, WI), and Sigma in the highest purity
available and used without further purification. Metal salts had
a grade of Certified (American Chemical Society). Diaqua [tris-
(2-aminoethyl)amine nickel(II)] acetate (Ni(II)-tren) and [1,4,8,-
11-tetraazacyclotetradecane nickel(II)] perchlorate (Ni(II)-
cyclam) were synthesized by Dr. J ames Muller (University of
Utah) according to published procedures (33).
Sta n d a r d Con d ition s for Oxid a tive Cr oss-Lin k in g of
RNa se A. Reaction mixtures (50 µL) of RNase (400 µM) and
Ni(OAc)2 (50 µM) were pre-incubated in 10 mM sodium borate
(pH 10) for 30 min under ambient temperature. Cross-linking
was initiated by addition of KHSO5 (800 µM) and quenched after
30 min by adding an equal volume of a standard denaturing
buffer for electrophoresis (4% SDS, 25% glycerol, 140 mM
Under standard conditions, the oxidant HSO5- was the
limiting reagent and reaction was complete prior to
quenching (30 min). Use of additional HSO5- (>800 µM)
resulted in further consumption and extensive oligomer-
ization of the monomeric enzyme to yield high molecular
weight products. Nickel was not consumed in this

304 Chem. Res. Toxicol., Vol. 10, No. 3, 1997
Gill et al.
F igu r e 1. Oxidative cross-linking of RNase A in the presence
of Ni(OAc)2 and KHSO5. (A) RNase A (lane 1) was alternatively
treated with Ni(OAc)2/KHSO5 (lane 2) or horseradish peroxidase/
H2O2 (lane 3) under standard reaction conditions described in
Materials and Methods. RNase A was also treated with KHSO5
in the absence of nickel (lane 4). Samples were analyzed by
SDS-polyacrylamide gel electrophoresis, stained with Coo-
massie Blue, and compared to molecular weight markers (lane
5). (B) RNase (400 µM, lane 1) was incubated with Ni(OAc)2
(50 µM) in 10 mM borate buffer pH 10 (lane 2) or 10 mM
potassium phosphate buffer pH 7 (lane 3) for 30 min, dialyzed
against water (3 × 5 h), and readjusted to pH 10 with 10 mM
borate. Reaction was then initiated with KHSO5 (800 µM) and
analyzed under standard conditions.
F igu r e 2. Metal dependence of RNase A oxidative cross-
linking. RNase A (400 µM) and various metal salts (50 µM) were
combined under standard conditions prior to addition of KHSO5
(800 µM). Samples were incubated for 30 min under ambient
conditions and then analyzed by SDS-polyacrylamide gel
electrophoresis. Lane 1, Ni(OAc)2; lane 2, Ni(II)tren and 4 mM
tren, the free macrocyclic ligand; lane 3, Ni(II)cyclam and 4 mM
cyclam, the free macrocyclic ligand; lane 4, Co(OAc)2; lane 5,
Cu(OAc)2; lane 6, Fe(NH4)2(SO4)2; lane 7, Mn(OAc)2; lane 8,
PdCl2; lane 9, Mg(OH)2; lane 10, Zn(NO3)2.
Initial studies had suggested that a limited number of
macrocyclic Ni(II) complexes forming square planar and
octahedral geometries were also capable of dimerizing
RNase albeit in much lower yields than Ni(OAc)2. How-
ever, no such activity was detected when additional free
ligand (4 mM) was added to minimize dissociation of
nickel from the macrocyclic complex (Figure 2, lanes 2
and 3). After the Kodadek laboratory reported the use
of GGH-Ni for cross-linking multimeric proteins (24),
we tested this nickel peptide complex with RNase under
their published conditions [pH 7, 10 µM protein, 100 µM
GGH-Ni, and 100 µM magnesium monoperoxyphthalate
(MMPP), a peracid analogous to HSO5-]. Interestingly,
this also led to formation of RNase dimer, trimer, and
tetramer (data not shown). The mechanism of this latter
reaction may differ from that described here since none
of the nickel species presented in Figure 2 exhibited the
activity of GGH-Ni at pH 7.
Oxid a n t Dep en d en ce of RNa se Cr oss-Lin k in g.
Most biochemical applications of nickel peptide and
macrocyclic complexes have used the peracids HSO5- and
MMPP as terminal oxidants. Such reagents were also
most effective in supporting nickel-dependent cross-
linking of RNase. The strongest oxidants, HSO5-, MMPP,
and meta-chloroperbenzoate (MCPBA), provided the high-
est yield of oligomeric RNase (Figure 3). Persulfate,
S2O82-, also supported efficient cross-linking (Figure 3,
lane 2). In contrast, peroxides were inactive (Figure 3,
lanes 6-8).
process and appeared to act catalytically. Although it
bound tightly to RNase (see below), it remained kineti-
cally labile enough to dissociate from cross-linked protein
and reassociate with the native, monomeric protein.
Quantifying the reaction illustrated in Figure 1B (lane
3) suggested that less than 50 µM Ni(OAc)2 promoted
consumption of approximately 250 µM RNase, the result
of more than 2.5 turnovers.
Str on g Associa tion betw een Nick el a n d RNa se A.
If an association between Ni2+ and RNase were required
for reaction as predicted, then lack of reaction at pH 7
might have been a function of metal binding or protein
oxidation. RNase affinity for Ni2+ was qualitatively
examined by incubating this protein with Ni(OAc)2 at pH
7 (10 mM potassium phosphate) and pH 10 (10 mM
sodium borate), dialyzing against water, and then as-
saying for bound Ni(II) by its oxidation of RNase after
-
subsequent addition of HSO5 at pH 10. Sufficient Ni-
(II) resisted dialysis at pH 7 to promote protein cross-
linking equivalent to that of samples dialyzed at pH 10
(Figure 1B). This suggests that Ni(II) binds strongly to
RNase in a manner independent of pH 7-10. As
expected, alternative dialysis in the presence of EDTA
generated RNase samples that yielded no dimerization
-
after addition of HSO5
.
Meta l Dep en d en ce of RNa se Cr oss-Lin k in g.
A
series of common transition metals were surveyed for an
equivalent ability to induce protein cross-linking in the
presence of HSO5-. This was necessary in part to verify
that a contaminating metal was not responsible for the
activity ascribed to nickel. Even trace metals had the
potential to dominate reaction since Ni2+ is not often
associated with oxidation chemistry under aqueous con-
ditions. In contrast to many expectations, Ni(OAc)2 was
one of the two most active metal salts tested for RNase
cross-linking (Figure 2). Cobalt and copper were also
competent, but only cobalt exhibited an efficiency similar
to nickel.
Qu en ch in g by F r ee Ra d ica l Tr a p p in g Agen ts.
Once Ni(II) becomes redox active through coordination
to a peptide backbone, a variety of oxidative processes
may be stimulated. In the presence of H2O2, GGH-Ni
promoted formation of the diffusible hydroxyl radical that
led to oxidative degradation of tryptophan and formation
of dityrosine in bovine serum albumin and lactate dehy-
drogenase (10). This reaction was inhibited by standard
radical quenching agents including ethanol, butanol, and
dimethyl sulfoxide. To examine the possible role of
hydroxyl radical or other activated oxygen species in the

Protein Cross-Linking Induced by a Simple Nickel Salt
Chem. Res. Toxicol., Vol. 10, No. 3, 1997 305
F igu r e 5. Effect of free amino acids on the oxidative cross-
linking of RNase A. Ni(OAc)2, oxidant, and RNase were reacted
under standard conditions (lane 1) and in the added presence
of the indicated amino acids (800 µM).
F igu r e 3. Oxidant dependence of RNase A cross-linking. The
nature of the oxidant required for the nickel-mediated reaction
of RNase was examined by substituting KHSO5 (lane 1) with
equimolar concentrations (800 µM) of K2S2O8 (lane 2), MCPBA
(lane 3), MMPP (lane 4), peracetate (lane 5), hydrogen peroxide
(lane 6), hydrogen peroxide/ascorbate (lane 7), and tert-butyl
hydroperoxide (lane 8).
F igu r e 6. Effect of tyrosine on the oxidative cross-linking of
RNase A. Ni(OAc)2, oxidant, and RNase were reacted under
standard conditions (lane 1) and in the added presence of
tyrosine (lanes 2-6).
amino acid was added to compete with RNase in the
nickel-dependent reaction. Tyrosine completely sup-
pressed RNase dimerization (Figure 5, lane 2) when
added in a 2-fold excess over protein. Tryptophan and,
to a much lesser extent, histidine also inhibited reaction
under these conditions, whereas lysine, phenylalanine,
serine, glycine, and arginine had no effect on protein
oligomerization (Figure 5, lanes 3-9). These latter
results suggest that quenching was not simply due to the
ability of R-amino acids to bind Ni2+ and perhaps
sequester it from RNase.
F igu r e 4. Effect of hydroxyl radical scavengers on the oxidative
cross-linking of RNase A. Ni(OAc)2, oxidant, and RNase were
reacted under standard conditions (lane 1) and in the added
presence of mannitol or methanol.
modification of RNase, the nickel-dependent oxidation
was repeated in the presence of 1-100 mM mannitol and
10-1000 mM methanol (Figure 4). These radical scav-
engers had little effect on protein dimerization, and only
partial inhibition was observed for samples containing
the highest concentration of mannitol. Therefore, the
reactive intermediates generated by the bound nickel
remain either highly localized or relatively unaffected by
hydrogen atom donors.
Qu en ch in g by F r ee Am in o Acid s. Accessible ty-
rosines on the surface of RNase were obvious candidates
for intermolecular coupling, and additional targets of
oxidation were histidine, lysine, and arginine (38-40).
Tryptophan might also have been expected to react, but
this residue is not present in RNase A. To identify which
residues could be responsible for cross-linking, each
Protein oligomerization was extremely sensitive to free
tyrosine. Inhibition of this reaction was evident after
addition of tyrosine at concentrations equimolar to Ni-
(OAc)2 (0.05 mM, Figure 6, lane 3), and quenching was
almost complete (82%) when the tyrosine concentration
was increased to 0.15 mM (Figure 6, lane 5) but well
-
below the concentration of RNase (0.4 mM) and HSO5
(0.8 mM).
Am in o Acid An a lysis of RNa se A a n d Its P r od u cts
After Oxid a tion . Amino acid analysis of RNase was
used to ascertain which amino acids were subject to the

306 Chem. Res. Toxicol., Vol. 10, No. 3, 1997
Gill et al.
Ta ble 1. Am in o Acid An a lysis of RNa se A a n d Its
Der iva tives Isola ted a fter Tr ea tm en t w ith Ni(OAc)2 a n d
KHSO5
RNase. Chromatographic analysis detected only the
presence of 3,3′-dityrosine and unmodified starting mate-
rial. Under these conditions, tyrosine coupling was
promoted by Ni(OAc)2 and the nickel complexes and
strongly inhibited by mannitol (100 mM). In contrast to
protein dimerization, this model reaction likely depended
on the generation of a diffusible radical.
amino acid residues per polypeptidea
monomer
tyrosine-coupled
monomerc
residue (theoretical) monomerb dimer
Dityr
Tyr
-
6
e0.2
6.1
0.6
5.1
0.6
5.3
Discu ssion
Ala
Arg
Asx
Gly
Glx
His
Ile
12
4
15
3
12
4
3
11.9
4.1
15.4
3.4
12.4
4.0
2.4
11.7
4.3
14.9
3.5
12.4
3.7
2.3
11.5
4.2
14.9
3.2
11.9
3.8
2.3
Ap p lica tion of a Sim p le Nick el Sa lt. Much of
nickel’s biological activity is controlled by the endogenous
ligands that surround solubilized Ni2+ (1, 2, 6). Similarly,
much of nickel’s utility as a biochemical reagent relies
on proper ligand design since coordination alone does not
guarantee an oxidation chemistry compatible with aque-
ous conditions. Peptides and proteins provide one of the
most significant classes of ligands for consideration both
in vitro and in vivo. Nickel has a strong propensity to
bind and deprotonate an amide backbone which allows
for subsequent reaction with a variety of oxidants.
Selective modification of proteins may therefore be
anticipated from coordination to compliant peptide se-
quences. Related processes are likely responsible for
stimulating oxidative stress and aging associated with
exposure to nickel (11, 43-45).
The present study demonstrates the ability of a simple
nickel salt to bind to a monomeric protein, RNase A, and
promote its oligomerization in the presence of a peracid,
HSO5-. Individual dimer, trimer and tetramers of RNase
were evident from electrophoretic analysis of protein
oxidation (Figure 1A). Little heterogeneity was apparent
in these products, and no protein fragmentation was
discernible. These observations presented the first in-
dication that the nickel-dependent oxidation of RNase
was therefore selective in its chemistry.
While the use of RNase at high concentration (400 µM)
is convenient for the present investigation, the cross-
linking remained detectable for reaction mixtures diluted
50-fold [1 µM Ni(OAc)2, 8 µM RNase, 16 µM HSO5-].
These alternative conditions more closely mimicked those
used to cross-link a series of proteins by an oxidative
process mediated by GGH-Ni (24). In contrast to the
activity of this nickel tripeptide complex, cross-linking
induced by the RNase-Ni complex was not observed at
pH 7. Nickel appeared to bind spontaneously to RNase
at both pH 7 and 10 (Figure 1B), but protein coupling
required the higher pH. The partial deprotonation of
Y73, Y76, Y92, and Y115 under these conditions (46, 47)
would assist oxidation by lowering the one-electron
potential of tyrosine (48).
P r otein Oligom er iza tion Is F or m ed by In ter m o-
lecu la r Cou p lin g of Tyr osin e Resid u es. Formation
of 3,3′-dityrosine is a common product of protein oxidation
and was first to be considered in the oligomerization of
RNase. Dityrosine has been detected in a wide variety
of proteins subjected to oxidation by hydroxyl radical,
H2O2, peroxidase, H2Cl6Ir, and ultraviolet light (36). This
amino acid derivative also occurs naturally in a number
of connective tissue proteins (38) and serves as a deter-
minant of protein aging (39). Already RNase had been
shown to dimerize via tyrosine coupling when exposed
to γ-irradiation (49). In addition, heteroprotein cross-
linking by GGH-Ni was similarly thought to involve
intermolecular tyrosine-tyrosine bonds (24).
Leu
Lys
Phe
Pro
Ser
Thr
Val
2
10
3
2.5
9.5
3.1
4.6
13.6
9.6
9.1
2.4
8.7
3.0
4.7
13.7
9.7
9.1
2.5
9.0
3.0
4.4
12.8
8.6
9.0
4
15
10
9
a
RNase A contains four Met and eight Cys that were not
measured in these analyses. All standard deviations were less than
b
5% of the average from three independent determinations. This
species represents the monomeric protein isolated after reaction
by gel electrophoresis. c This derivative was isolated as the
monomer after standard nickel-dependent oxidation in the added
presence of 16 mM tyrosine.
nickel-dependent oxidation. Both monomeric and dimer-
ic protein were blotted from standard polyacrylamide
gels, subjected to exhaustive hydrolysis (6 N HCl, 24 h,
110 °C) and quantified using standard chromatography.
These procedures were also repeated for the monomeric
protein oxidized in the presence of added tyrosine. The
average composition from three independent determina-
tions is summarized in Table 1. To facilitate comparison
between samples, stoichiometries are listed per single
polypeptide rather than per mole since the RNase dimer
contains two polypeptides.
Analysis of the monomeric enzyme recovered after
oxidation indicates that few, if any, residues were modi-
fied other than those participating in cross-linking. The
ratio of amino acids for this protein was close to the
theoretical value. Only the concentrations of isoleucine,
lysine, threonine, and particularly serine were lower than
expected although still typical of past analyses of RNase
(for example, see ref 41). The covalent dimer generated
one less tyrosine per polypeptide strand and produced a
new material that co-eluted with 3,3′-dityrosine synthe-
sized as described below. The slight excess of dityrosine
detected over the expected 0.5 residue/polypeptide (or 1
residue/dimer) may be attributed to the e0.2 residue
background observed for the monomeric protein recov-
ered after reaction. Analysis of a similar monomeric
-
derivative isolated after treatment with Ni(OAc)2, HSO5
,
and a large excess of free tyrosine also indicated the loss
of a tyrosyl residue. However, a substoichiometric yield
of 3,3′-dityrosine was detected which suggested alterna-
tive products such as an O-coupled isodityrosyl derivative
might have formed (42). Finally, the characteristic
fluorescence excitation (310 nm) and emission (410 nm)
maxima of 3,3′-dityrosyl residues (36) was used to confirm
its presence in the covalent dimer of RNase and its
absence in the products of control reactions containing
-
RNase alone, RNase and Ni(OAc)2, or RNase and HSO5
.
Syn th esis of 3,3′-Dityr osin e As a Sta n d a r d for
Am in o Acid An a lysis. Tyrosine was oxidatively dimer-
ized under conditions related to those used to modify
The first experimental evidence implicating tyrosine
as the site of oxidition induced by nickel was obtained
from a series of competition studies. The oligomerization

Protein Cross-Linking Induced by a Simple Nickel Salt
Chem. Res. Toxicol., Vol. 10, No. 3, 1997 307
Sch em e 1
of RNase was not inhibited by addition of excess amino
acids that might coordinate to Ni2+ through the common
R-amino acid groups (glycine, serine, phenylalanine, etc.,
Figure 5) or cationic side chains (arginine, lysine). In
contrast, easily oxidized amino acids such as tyrosine and
tryptophan effectively quenched protein dimerization
(Figure 5).2 Reaction was most sensitive to tyrosine due
to its ability to substitute for RNase as the target for
intermolecular coupling to form a protein-tyrosine cross-
link (see below). Participation of tyrosine was subse-
quently confirmed by amino acid analysis. Only one
tyrosyl residue per polypeptide was oxidized in the
dimeric protein generated by nickel and HSO5-, and no
other residues were significantly modified when compar-
ing the dimeric vs. monomeric RNase recovered after
oxidation (Table 1). Accordingly, tyrosyl modification
appeared to dominate intermolecular cross-linking of
RNase. Formation of the 3,3′-dityrosyl residue was
indicated by both amino acid and fluorometric analysis.
Meta l, Oxid a n t, a n d Mech a n ism . Redox active
metals ions including Fe2+, Cu2+, and Mn2+ all promote
oxidative modification of proteins under a variety of
conditions (40), but these simple ions exhibited little
-
ability to mediate the oxidation of RNase by HSO5
(Figure 2). In contrast, the normally unreactive ion Ni2+
demonstrated efficient oxidation of RNase while a nickel
complex [Ni(cyclam)]2+ previously known to induce DNA
oxidation remained unreactive (50). Such observations
are consistent with the ability of the simple nickel ion to
bind an accessible region of a polypeptide and promote a
unique and localized oxidation. Oligomerization of mon-
omeric RNase by intermolecular reaction of tyrosyl
residues was only partially affected by high concentra-
tions of mannitol (Figure 4) and much more sensitive to
free tyrosine concentration. Consequently, a diffusible
hydroxyl radical is not a plausible intermediate in the
nickel-dependent reaction described in this report (10,
51). Oxidants typically used to generate hydroxyl radical
were also unreactive with RNase-Ni2+ (Figure 3). In-
stead, this protein-nickel complex may act by promoting
the transient formation of a Ni(III) intermediate that
abstracts an electron from a neighboring tyrosine in
analogy to a mechanism previously proposed for cross-
linking between a phenolate containing complex of nickel
(salen) and DNA (14, 15). The peracids and persulfate
are all capable of oxidizing the proposed complex from
RNase-Ni(II) to RNase-Ni(III) (52, 53). In the absence
of readily oxidized side chains, the Ni(III) intermediate
might alternatively abstract a hydrogen from the polypep-
tide backbone ultimately promoting protein fragmenta-
tion, a process not observed with RNase (see Scheme 1).
The low activity of Cu2+ (Figure 2) may result from a
mechanism equivalent to, but less effective than, that of
nickel. Copper(II) has a similar high affinity for binding
peptides and proteins, and the resulting complexes
stabilize the related formation of Cu(III) (54). In con-
trast, the Co2+ reaction may not require prior binding to
the protein. Only Co2+ readily generates a diffusible
examine the Co2+-dependent reaction of RNase A.
The terminal oxidant of this investigation, HSO5-, was
initially chosen due to its ease of handling and successful
application to the field of nucleic acids (14, 58). However,
the cobalt reaction highlights the additional relevance
of this oxidant in biology. Cobalt generates sulfate
radical anion from sulfite, SO3-, and O2 via intermediate
formation of HSO5 (59, 60). Therefore, Co2+ mediates
-
both the production and consumption of this peracid. Our
study then describes a process and type of protein
modification that could result from natural exposure to
cobalt and the common food preservative sulfite. Simi-
-
larly, a wide variety of nickel complexes utilize HSO5
as a terminal oxidant for biopolymer modification, and
certain of these complexes will also generate HSO5- from
sulfite and O2 (61). The nickel-dependent reaction of this
report consequently mimicks one of the processes that
may contribute to the toxicity and carcinogenicity of this
metal.
Associa tion betw een Nick el a n d RNa se. An im-
portant remaining question is where and how does Ni2+
bind RNase to promote oxidation. This is the subject of
continuing investigations which are aided by numerous
peptide studies (see for example refs 8, 62, and 63).
Nickel may coordinate to RNase through backbone and
side chain interactions (5) although the proposed forma-
tion of Ni(III) would require stabilization afforded by the
peptide-like interactions of a deprotonated amide back-
bone (8). RNase has previously been shown to bind Cu2+
and Zn2+ at multiple sites, but simultaneous ligation
expected from both active site histidines (H12 and H119)
was not detected (64, 65). Interactions with Cu2+ and
Mn2+ have since been characterized by 13C-NMR studies
of RNase (66, 67). These again indicate variable associa-
tion with H12, H119, and also H105 and no direct ligation
to tyrosine residues. Possible coordination to the polypep-
tide backbone was not examined then, and visual inspec-
-
sulfate radical anion from HSO5 (55, 56). This radical
has previously been shown to produce dityrosine and
other oxidized products when generated in the presence
of albumin (57). Quenching studies have now begun to
2
Cysteine thiolates are expected to be additional targets of oxida-
tion, but their disulfide derivatives would be much less reactive. RNase
contains six cysteine residues that form three disulfide bonds. Free
cysteine was not examined in the competition studies since quenching
could result from its avid binding to nickel, its facile oxidation, or both.

308 Chem. Res. Toxicol., Vol. 10, No. 3, 1997
Gill et al.
(5) Sigel, H., and Martin, R. B. (1982) Coordinating properties of the
amide bond. Stability and structure of metal ion complexes of
peptides and related ligands. Chem. Rev. 82, 385-426.
(6) Hausinger, R. P. (1993) Biochemistry of Nickel, Plenum Press,
New York, NY.
tion of the RNase A crystal structure does not reveal any
obvious sites for this other than the N-terminus. As an
alternative, nickel might coordinate to an accessible and
disordered region of the RNase backbone, but no such
regions have been identified (47, 68-70).
(7) Bossu, F. P., Chellappa, K. L., and Margerum, D. W. (1977)
Ligand effects on the thermodynamic stabilization of copper(III)-
peptide complexes. J . Am. Chem. Soc. 99, 2195-2203.
(8) Bossu, F. P., Paniago, E. B., Margerum, D. W., and Kirksey, S.
T., J r. (1978) Trivalent nickel catalysis of the autoxidation of
nickel(II) tetraglycine. Inorg. Chem. 17, 1034-1042.
(9) Shi, X., Dalal, N. S., and Kasprzak, K. S. (1992) Generation of
free radicals from lipid hydroperoxides by Ni2+ in the presence
of oligopeptides. Arch. Biochem. Biophys. 299, 154-162.
(10) Torreilles, J ., and Guerin, M-C. (1990) Nickel(II) as a temporary
catalyst for hydroxyl radical generation. FEBS Lett. 272, 58-60.
(11) Kasprzak, K. S., and Bare, R. M. (1989) In vitro polymerization
of histones by carcinogenic nickel compounds. Carcinogenesis 10,
621-624.
(12) Datta, A. K., Misra, M., North, S. L., and Kasprzak, K. S. (1992)
Enhancement by nickel(II) and L-histidine of 2′-deoxyguanosine
oxidation with hydrogen peroxide. Carcinogenesis 13, 283-287.
(13) Chen, X., Woodson, S. A., Burrows, C. J ., and Rokita, S. E. (1993)
A highly sensitive probe for guanine N7 in folded structures of
RNA: Application to tRNAphe and Tetrahymena group I intron.
Biochemistry 32, 7610-7616.
(14) Burrows, C. J ., and Rokita, S. E. (1994) Probing guanine structure
in nucleic acid folding using nickel complexes. Acc. Chem. Res.
27, 295-301.
(15) Muller, J . G., Paikoff, S. J ., Rokita, S. E., and Burrows, C. J . (1994)
DNA modification promoted by water-soluble nickel(II) salen com-
plexes: A switch to DNA alkylation. J . Inorg. Biochem. 54, 199-
206.
(16) Cheng, C.-C., Rokita, S. E., and Burrows, C. J . (1993) Nickel-
(III)-promoted DNA scission using ambient dioxygen. Angew.
Chem., Int. Ed. Engl. 32, 277-278.
(17) Guan, L. L., Totsuka, R., Kuwahara, J ., Otsuka, M., and Sugiura,
Y. (1993) Cleavage of Yeast tRNAphe with Ni(III) and Co(III)
complexes of bleomycin. Biochem. Biophys. Res. Commun. 191,
1338-1346.
(18) Guan, L. L., Kuwahara, J ., and Sugiura, Y. (1993) Guanine-
specific binding by bleomycin-nickel(III) complex and its reactiv-
ity for guanine-quartet telomeric DNA. Biochemistry 32, 6141-
6145.
(19) Mack, D. P., and Dervan, P. B. (1990) Nickel-mediated sequence-
specific oxidative cleavage of DNA by a designed metalloprotein.
J . Am. Chem. Soc. 112, 4604-4606.
(20) Mack, D. P., and Dervan, P. B. (1992) Sequence-specific oxidative
cleavage of DNA by a designed metalloprotein, Ni(II)-GGH-
(Hin139-190). Biochemistry 31, 9399-9405.
(21) Shullenberger, D. F., Eason, P. D., and Long, E. C. (1993) Design
and synthesis of a versatile DNA-cleaving metallopeptide struc-
tural domain. J . Am. Chem. Soc. 115, 11038-11039.
(22) Liang, Q., Eason, P. D., and Long, E. C. (1995) Metallopeptide-
DNA interactions: Site-selectivity based on amino acid composi-
tion and chirality. J . Am. Chem. Soc. 117, 9625-9631.
(23) Cuenoud, B., Tarasow, T. M., and Schepartz, A. (1992) A new
strategy for directed protein cleavage. Tetrahedron Lett. 33, 895-
898.
(24) Brown, K. C., Yang, S.-H., and Kodadek, T. (1995) Highly specific
oxidative cross-linking of proteins mediated by a nickel-peptide
complex. Biochemistry 34, 4733-4739.
(25) Heyduk, E., and Heyduk, T. (1994) Mapping protein domains
involved in macromolecular interactions: a novel protein foot-
printing approach. Biochemistry 33, 9643-9650.
RNase A contains six tyrosine residues of which four
are accessible to chemical modification at pH 10 (68). Y73,
Y76, and Y115 cluster at one surface and Y92 resides on
a distant surface. Any of these may participate in
dimerization, but for steric considerations, further oli-
gomerization to trimers, tetramers, etc. would likely
involve Y92 and only one of the three neighboring
residues Y73, Y76, and Y115. Tyrosine 73 seems least
likely of the three to undergo intermolecular reaction
since it appears least exposed on the protein surface.
However, the phenolic ring of this residue is directly
adjacent to that of Y115, and intramolecular reaction
might have been expected. Amino acid analysis indicates
that this process was disfavored, and little dityrosine was
detected in the monomeric protein recovered after oxida-
tion (Table 1).
Con clu sion
Although simple nickel salts are not often associated
with oxidation under aqueous conditions, their peptide
complexes can be quite reactive with a range of oxidants.
Such complexes are thought to facilitate the cellular
distribution of nickel as well as promote protein and
nucleic acid degradation that can induce allergic, toxic
and carcinogenic responses. RNase A served as a model
for examining the ability of Ni2+ to promote localized and
selective oxidation of a protein which, in this example,
resulted in polypeptide oligomerization through tyrosine
coupling. Radical trapping agents provided little protec-
tion against protein cross-linking in the presence of Ni2+
,
and diffusible Ni(II) macrocyclic complexes used previ-
ously for DNA oxidation did not mimic the activity of free
nickel. Protein oxidation is likely mediated by transient
formation of a Ni(III) intermediate that abstracts an
electron from neighboring residues. The ultimate reac-
tion induced by this process may then depend on the
available electron donors which in turn contribute to the
deleterious effects of nickel exposure.
Ack n ow led gm en t. This work was supported by
grants from the National Institutes of Health (NIH) and
the American Cancer Society. A.R.-R. was supported by
a fellowship from the German-American Fulbright Com-
mission and the German National Scholarship Founda-
tion. Amino acid analysis was performed by the Center
for the Analysis and Synthesis of Macromolecules at
Stony Brook that is supported in part by NIH and the
New York State Science and Technology Foundation.
Acquisition of the AMX-600 NMR spectrometer was made
possible by grants from the NIH, NSF, and the New York
State Science and Technology Foundation.
(26) Rana, T. M., and Meares, C. F. (1990) Specific cleavage of a protein
by an attached iron chelate. J . Am. Chem. Soc. 112, 2457-2458.
(27) Rana, T. M., and Meares, C. F. (1991) Transfer of oxygen from
an artificial protease to peptide carbon during proteolysis. Proc.
Natl. Acad. Sci. U.S.A. 88, 10578-10582.
(28) Schepartz, A., and Cuenoud, B. (1990) Site-specific cleavage of
the protein calmodulin using a trifluoperazine-based affinity
reagent. J . Am. Chem. Soc. 112, 3247-3249.
(29) Platis, I. E., Erma´cora, M. R., and Fox, R. O. (1993) Oxidative
polypeptide cleavage mediated by EDTA-Fe covalently linked to
cysteine residues. Biochemistry 32, 12761-12767.
(30) Hoyer, D., Cho, H., and Schultz, P. G. (1990) A new strategy for
selective protein cleavage. J . Am. Chem. Soc. 112, 3249-3250.
(31) Ettner, N., Metzger, J . W., Lederer, T., Hulmers, J . D., Kisker,
C., Hinrichs, W., Ellestad, G. A., and Hillen, W. (1995) Proximity
mapping of the Tet repressor-tetracycline-Fe2+ complex by
hydrogen peroxide-mediated protein cleavage. Biochemistry 34,
22-31.
Refer en ces
(1) Costa, M. (1991) Molecular mechanisms of nickel carcinogenesis.
Annu. Rev. Pharmacol. Toxicol. 31, 321-327.
(2) Kasprzak, K. S. (1991) The role of oxidative damage in metal
carcinogenicity. Chem. Res. Toxicol. 4, 604-615.
(3) Wetterhahn, K. E., and Dudek, E. J . (1996) Environmental metal
carcinogens: genotoxicity and altered gene expression by direct
metal-mediated and indirect oxidative pathways. New J . Chem.
20, 199-203.
(4) Glennon, J . D., and Sarkar, B. (1982) Nickel(II) transport in
human blood serum. Biochem. J . 203, 15-23.
(32) Parac, T. N., and Kostic´, N. M. (1996) New selectivity and
turnover in peptide hydrolysis by metal complexes. A palladium-

Protein Cross-Linking Induced by a Simple Nickel Salt
Chem. Res. Toxicol., Vol. 10, No. 3, 1997 309
(II) aqua complex catalyzes cleavage of peptides next to the
histidine residue. J . Am. Chem. Soc. 118, 51-58.
(33) Barefield, E. K. (1973) A new synthesis of 1,4,8,11-tetraazacy-
clotetradecane (cyclam) via the Ni(II) complex. Inorg. Chem. 11,
2273-2274.
(34) Scha¨gger, H., and von J agow, G. (1987) Tricine-sodium dodecyl
sulfate-polyacrylamide gel electrophoresis for the separation of
proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368-
379.
(35) Aeschhach, R., Amado`, R., and Neukom, H. (1976) Formation of
dityrosine cross-links in proteins by oxidation of tyrosine. Bio-
chim. Biophys. Acta 439, 292-301.
(36) Giulivi, C., and Davies, K. J . A. (1994) Dityrosine: a marker for
oxidatively modified proteins and selective proteolysis. Methods
Enzymol. 223, 363-371.
(37) Malencik, D. A., and Anderson, S. R. (1991) Fluorometric char-
acterization of dityrosine: Complex formation with boric acid and
borate ion. Biophys. Biochem. Res. Commun. 178, 60-67.
(38) Amado`, R., Aeschbach, R., and Neukom, H. (1984) Dityrosine: in
vitro production and characterization. Methods Enzymol. 107,
377-388.
(39) Stadtman, E. R. (1992) Protein oxidation and aging. Science 257,
1220-1224.
(40) Stadtman, E. R. (1993) Oxidation of free amino acids and amino
acid residues in proteins by radiolysis and by metal-catalyzed
reactions. Annu. Rev. Biochem. 62, 797-821.
(41) Marfey, P. S., Uziel, M., and Little, J . (1965) Reaction of bovine
pancreatic ribonuclease A with 1,5-difluoro-2,4-dinitrobenzene.
J . Biol. Chem. 240, 3270-3275.
(42) Fry, S. C. (1984) Isodityrosine, a diphenyl ether cross-link in plant
cell wall glycoprotein: identification, assay and chemical synthe-
sis. Methods Enzymol. 107, 388-397.
(43) Lee, J . E., Ciccarelli, R. B., and J ennette, K. W. (1982) Solubili-
zation of the carcinogen nickel subsulfide and its interaction with
deoxyribonucleic acid and protein. Biochemistry 21, 771-778.
(44) Misra, M., Olinski, R., Dizdaroglu, M., and Kasprzak, K. S. (1993)
Enhancement by L-histidine of nickel(II)-induced DNA-protein
cross-linking and oxidative DNA base damage in the rat kidney.
Chem. Res. Toxicol. 6, 33-37.
(52) Minisci, F., Citterio, A., and Giordano, C. (1983) Electron-transfer
processes: Peroxydisulfate, a useful and versatile reagent in
organic chemistry. Acc. Chem. Res. 16, 27-32.
(53) Muller, J . G., Chen, X., Dadiz, A. C., Rokita, S. E., and Burrows,
C. J . (1993) Macrocyclic nickel complexes in DNA recognition and
oxidation. Pure Appl. Chem. 65, 545-550.
(54) Bossu, F. P., and Margerum, D. W. (1977) Electrode potentials
of nickel(III,II)-peptide complexes. Inorg. Chem. 16, 1210-1214.
(55) Ball, D. L., and Edwards, J . O. (1958) The catalysis of the
decomposition of Caro’s Acid. J . Phys. Chem. 62, 343-345.
(56) Muller, J . G., Zheng, P., Rokita, S. E., and Burrows, C. J . (1996)
DNA modification promoted by [Co(H2O)6]Cl2: Probing temper-
ature-dependent conformations. J . Am. Chem. Soc. 118, 2320-
232.
(57) Ito, K., and Kawanishi, S. (1991) Site-specific fragmentation and
modification of albumin by sulfite in the presence of metal ions
or peroxidase/H2O2: Role of sulfate radical. Biochem. Biophys.
Res. Commun. 176, 1306-1312.
(58) Meunier, B. (1992) Potassium monopersulfate: J ust another
primary oxidant or a highly versatile oxygen atom donor in
metalloporphyrin-mediated oxygenation and oxidation reactions.
New J . Chem. 16, 203-211.
(59) Coichev, N., and van Eldik, R. (1991) Kinetics and mechanism of
the sulfite-induced autoxidation of cobalt(II) in aqueous azide
medium. Inorg. Chem. 30, 2375-2380.
(60) Brandt, C., and van Eldik, R. (1995) Transition metal-catalyzed
oxidation of sulfur(IV) oxides. Atmospheric-relevant processes and
mechanisms. Chem. Rev. 95, 119-190.
(61) Muller, J . G., Hickerson, R. P., Perez, R. J ., and Burrows, C. J .
(1997) DNA damage from sulfite autoxidation catalyzed by a
nickel(II) peptide. J . Am. Chem. Soc. 119, 1501-1506.
(62) Predki, P. F., Harford, C., Brar, P., and Sarkar, B. (1992) Further
characterization of the N-terminal copper(II)- and nickel(II)-
binding motifs of proteins. Biochem. J . 287, 211-215.
(63) Bal, W., Lukszo, J ., Bojczuk-J ezowska, M., and Kasprzak, K. S.
(1995) Interactions of nickel(II) with histones. Stability and
solution structure of complexes with CH3CO-Cys-Ala-Ile-His-NH2,
a putative metal binding sequence of histone H3. Chem. Res.
Toxicol. 8, 683-692.
(45) Zhuang, Z., Huang, X., and Costa, M. (1994) Protein oxidation
and amino acid-DNA crosslinking by nickel compounds in intact
cultured cells. Toxicol. Appl. Pharm. 126, 319-325.
(46) Egan, W., Shindo, H., and Cohen, J . S. (1978) On the tyrosine
residues of ribonuclease A. J . Biol. Chem. 253, 16-17.
(47) Lenstra, J . A., Bolscher, B. G. J . M., Stob, S., Beintema, J . J .,
and Kaptein, R. (1979) The aromatic residues of bovine pancreatic
ribonuclease studied by 1H nuclear magnetic resonance. Eur. J .
Biochem. 98, 385-397.
(64) Breslow, E., and Girottie, A. W. (1966) The interaction of
ribonuclease with metal ions, I. J . Biol. Chem. 241, 5651-5660.
(65) Girotti, A. W., and Breslow, E. (1968) The interaction of ribonu-
clease with metal ions, II. J . Biol. Chem. 243, 216-218.
(66) Walters, D. E., and Allerhand, A. (1980) Tautomeric states of the
histidine residues of bovine pancreatic ribonuclease A. J . Biol.
Chem. 255, 6200-6204.
(67) Howarth, O. W., and Lian, L. Y. (1984) Ribonuclease A: Carbon-
13 nuclear magnetic resonance assignments, binding sites, and
conformational flexibility. Biochemistry 23, 3515-3521.
(68) Blackburn, P., and Moore, S. (1982) Pancreatic ribonuclease.
Enzymes 15, 317-433.
(48) Butler, J ., Land, E. J ., Pru¨tz, W. A., and Swallow, A. J . (1986)
Reversibility of charge transfer between tryptophan and tyrosine.
J . Chem. Soc., Chem. Commun., 348-349.
(49) Boguta, G., and Dancewicz, A. M. (1983) Radiolytic and enzymatic
dimerization of tyrosyl residues in insulin, ribonuclease, papain
and collagen. Int. J . Radiat. Biol. 43, 249-265.
(50) Muller, J . G., Chen, X., Dadiz, A. C., Rokita, S. E., and Burrows,
C. J . (1992) Ligand effects associated with the intrinsic selectivity
of DNA oxidation promoted by nickel(II) macrocyclic complexes.
J . Am. Chem. Soc. 114, 6407-6411.
(69) Sosnick, T. R., and Trewhella, J . (1992) Denatured states of
ribonuclease A have compact dimensions and residual secondary
structure. Biochemistry 31, 8329-8335.
(70) Tilton, R. F., J r., Dewan, J . C., and Petsko, G. A. (1992) Effects
of temperature on protein structure and dynamics: X-ray crystal-
lographic studies of the protein ribonuclease A at nine different
temperatures from 98 to 320 K. Biochemistry 31, 2469-2481.
(51) Davies, K. J ., and Delsignore, M. E. (1987) Protein damage and
degradation by oxygen radicals. J . Biol. Chem. 262, 9908-9913.
TX960170I