DNA and RNA damage by CuII-amikacin complexMałgorzata Je_zzowska-Bojczuk1, Wojciech Szczepanik1, Wojciech Les´niak1,*, Jerzy Ciesiołka2, Jan Wrzesin´ski2and Wojciech Bal3 1 Faculty of Ch
Trang 1DNA and RNA damage by Cu(II)-amikacin complex
Małgorzata Je_zzowska-Bojczuk1, Wojciech Szczepanik1, Wojciech Les´niak1,*, Jerzy Ciesiołka2,
Jan Wrzesin´ski2and Wojciech Bal3
1 Faculty of Chemistry, University of Wrocław, Wrocław, Poland; 2 Institute of Bioorganic Chemistry, Polish Academy of Sciences, Poznan´, Poland; 3 Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Warsaw, Poland
The oxidation-promoting reactivity of copper(II) complex of
aminoglycosidic antibiotic amikacin [Cu(II)-Ami] in the
presence of hydrogen peroxide, was studied at pH 7.4, using
2¢-deoxyguanosine (dG), pBR322 plasmid DNA and yeast
tRNAPheas target molecules The mixtures of complex with
H2O2were found to be efficient oxidants, converting dG to
its 8-oxo derivative, generating strand breaks in plasmid
DNA and multiple cleavages in tRNAPhe The complex
underwent autooxidation as well, with amikacin hydroper-oxides as likely major products This reactivity pattern was found to be due to a combination of metal-bound and free hydroxyl radicals
Keywords: copper(II) complexes; amikacin; 2¢-deoxyguano-sine oxidation; plasmid DNA damage; tRNA cleavage
Amikacin (Ami) is a semisynthetic aminoglycoside, a
derivative of kanamycin A, having the B1 amino group of
2-deoxystreptamine moiety modified by acylation with
4-amino-2-hydroxybutyric acid Previous studies
demon-strated that Ami is a strong chelator for Cu(II) ions [1,2]
Cu(II) is coordinated at B ring, which is in contrast with
unsubstituted aminoglycosides, where terminal aminosugar
rings are involved in the binding [3–11] In the presence of
H2O2the Cu(II)-Ami complex oxidizes 2¢-deoxyguanosine
to its 8-oxo derivative [1] Recently we presented the
mechanism of such H2O2 activation, which involves
generation of hydroxyl radicals [12]
Antimicrobial action of Ami and other aminoglycosides
is based on their interactions with ribosomal RNA [13] and
also with cytoplasmic membrane [14,15] Aminoglycosides
also have severe toxic side-effects, on kidney and inner ear
[16,17] Fe(II) complexation and generation of hydroxyl
radicals by another aminoglycoside, gentamicin has recently
been implicated in the mechanism of its toxicity [18,19]
Aminoglycosides themselves, including Ami, are
redox-inactive [12,20] While intracellular copper is under very
tight control in bacterial cells, exerted by specific chaperone
proteins [21], extracellular copper in the human blood serum
is coordinated in part by, a variety of low molecular mass
ligands [22,23] It is not controlled as tightly there, and is
elevated in some pathological conditions, including cancer
and inflammation [24] It is not unlikely that some copper
may be transferred to Ami, e.g during treatment of sepsis This physiological state also involves increased generation
of hydrogen peroxide and oxygen radical species [25] The ability of Cu(II) complexes of other aminoglycosides to cleave DNA and RNA was shown recently [11,26,27] Here
we present the interactions of Cu(II)-Ami with oxidation-susceptible biomolecules: 2¢-deoxyguanosine (dG), pBR322 plasmid DNA and yeast tRNAPhe in both presence and absence of hydrogen peroxide, as well as the complex autooxidation Some of these reactions may play a role in toxic effects of amikacin
E X P E R I M E N T A L P R O C E D U R E S Materials
Amikacin and Tris buffer were obtained from Fluka (Buchs, Switzerland); CuCl2, methanol, 2¢-deoxyguano-sine (dG), H2O2, Chelex 100 resin, EDTA, and sodium and potassium phosphates were purchased from Sigma Chemical Co (St Louis, MO) The reference sample of 7,8-dihydro-8-oxo-2¢-deoxyguanosine (8-oxo-dG) was syn-thesized and purified by HPLC according to a published procedure [28] Low EEO agarose was purchased from AppliChem (Darmstadt, Germany) Bromophenol blue and glycerol were obtained from POCH (Gliwice, Poland) The plasmid pBR322 was isolated by
J Zakrzewska-Czewin˜ska (Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland) Bacterial alkaline phosphatase and T4 polynucleotide kinase were from MBI Fermentas (Vilnius, Lithuania) [c-32P]ATP was purchased from ICN (Costa Mesa, CA, USA) Electrophoretic reagents: boric acid, acrylamide, bis-acrylamide and urea were obtained from Serva (Heidelberg, Germany)
Decomposition of dG and formation of 8-oxo-dG Solutions of dG (50 lM) in 50 mM sodium phosphate buffer, pH 7.4, were incubated in triplicate for 24 h at 25C
or 37C in the presence of combinations of Ami, Cu(II)
Correspondence to M Je_zzowska-Bojczuk, Faculty of Chemistry,
University of Wrocław, F Joliot-Curie 14, 50–383 Wrocław, Poland.
Fax: + 48 71 3282 348, Tel.: + 48 71 3757 281,
E-mail: MJB@wchuwr.chem.uni.wroc.pl
Abbreviations: Ami, amikacin; Cu(II)-Ami, cupric complex of
amikacin; dG, 2¢-deoxyguanosine; 8-oxo-dG,
7,8-dihydro-8-oxo-2¢-deoxyguanosine; ESI, electrospray ionization; MS/MS, tandem mass
spectrometry; D, dihydrouridine; Y, wybutosine.
*Present address: Kriesge Hearing Research Institute, Ann Arbor,
MI, USA.
(Received 10 June 2002, revised 5 September 2002,
accepted 16 September 2002)
Trang 2(0 or 50 lM) and H2O2 (0.5 mM) At intervals, small
aliquots were removed from the incubating mixtures and
analyzed by HPLC without any additional pretreatment on
a HP 1100 system (Hewlett-Packard, Palo Alto, CA, USA)
with a Nucleosil 100 C18column (4.6 mm· 25 cm), using
UV detection at 254 nm The mobile phase was 50 mM
KH2PO4solution in 12% aqueous methanol Results were
quantified with the use of standard solutions containing
known amounts of dG and 8-oxo-dG
Mass spectrometry (MS)
The mass spectra were obtained with a Finnigan Mat TSQ
700 instrument (Thermo-Finnigan, Bremen, Germany),
using the ESI (electro-spray ionization) technique with N2
as carrier gas, and flow of 2 lLÆmin)1 Four kinds of
samples were investigated at pH 7.4, adjusted with acetic
acid and ammonium carbonate: free Ami, Ami-H2O2,
Cu(II)-Ami and Cu(II)-Ami-H2O2 All samples were
pre-pared as water solutions and incubated at 37C for 5, 20
and 40 min Then methanol was added to a final
concen-tration of 50%, in order to decrease surface tension during
sample evaporation Final sample concentrations were: Ami
and Cu(II), 0.5 mM; H2O2, 2.5 mM Further analysis of key
peaks was carried out using tandem MS (MS/MS)
experi-ments, using helium as a collision gas Its pressure was
adjusted so that the parent ion intensity was 45% of its
initial abundance
DNA strand break analysis
The ability of Cu(II)-Ami to induce strand breaks in
plasmid DNA in the absence and presence of H2O2 was
tested with the use of pBR322 plasmid Samples contained
15 lMDNA in 50 mM sodium phosphate buffer (pH 7.4)
and combinations of amikacin, CuCl2 and H2O2 The
concentrations of these reagents were as follows: Ami, 0 or
50 lM; Cu(II) 0 or 50 lM, H2O2, 0, 0.5, 5 or 50 lM After
1 h incu bations at 37C, reaction mixtures (20 lL) were
mixed with 4 lL of loading buffer (bromophenol blue in
30% glycerol) and loaded on 1% agarose gels, containing
ethidium bromide, in Tris/borate/EDTA buffer (90 mM
Tris/borate, 20 mM EDTA, pH 8.0) Gel electrophoresis
was carried out at constant voltage of 100 V (4 VÆcm)1), for
20 min As control for double strand breaks, reference
plasmid samples were linearized with XhoI endonuclease
The procedure used in kinetic experiments was similar, with
the use of Ami and H2O2, Cu (II) and H2O2and Cu(II)-Ami
and H2O2(each 50 lM), incubated at 37C for time periods
varied between 5 and 95 min, typically with 10 min
intervals The gels were photographed and processed with
a Digital Imaging System (Syngen Biotech, Wrocław,
Poland)
Isolation of yeast tRNAPhe
Yeast tRNAPhe of specific phenylalanine acceptance
1200–1400 pmol per A260 unit was prepared from crude
baker’s yeast tRNA by standard column chromatography
procedures including benzoylated DEAE-cellulose and
Sepharose-4B (Pharmacia, Sweden) Final purification was
carried out by HPLC on TSK-gel DEAE 2SW column
(Toyo Soda, Japan)
Labeling of tRNAPhe Yeast tRNAPhewas dephosphorylated with bacterial alka-line phosphatase and subsequently 5¢-end-labeled with [c-32P]ATP and polynucleotide kinase The tRNA was purified on a denaturing 12.5% polyacrylamide gel, located
by autoradiography, excised and eluted from the gel with the 0.3M potassium acetate buffer, pH 5.1, containing
1 mM EDTA and 0.1% SDS The eluted tRNAPhe was precipitated with ethanol, dissolved in water and stored at )20 C before u se
TRNAPhecleavage analysis Prior to the reaction, the 32P labeled tRNAPhe was supplemented with carrier tRNAPhe to the final RNA concentration of 1 lM and subjected to denaturation/ renaturation procedure by heating the samples at 65C for 2 min and slow cooling to 25C Cleavage reactions induced by Ami and Cu(II)-Ami complexes (1–100 lM) were performed in 10 mM sodium phosphate buffer at
pH 7.4, in the absence or presence of H2O2(50 or 100 lM) Further details of the reaction conditions are specified
in figure legends All reactions were stopped by mixing with equal volume of 8M urea/dyes/20 mM EDTA solu-tion and loaded on a 15% polyacrylamide, 7M urea gel Electrophoresis was carried out at 1500 V for 3 h, followed by autoradiography at)80 C with an intensifying screen
In order to assign the cleavage sites, products of cleavage reaction were run along with the products of alkaline degradation and limited T1nuclease digestion of the same tRNAPhe The alkaline hydrolysis ladder was generated by incubation of 32P labeled tRNAPhe with 5 volumes of formamide in boiling water for 10 min Partial T1nuclease digestion was performed in denaturing conditions (50 mM sodium citrate, pH 4.5, 7 M urea) with 0.1 unit of the enzyme The reaction mixture was incubated for 10 min at
50C
R E S U L T S Mass spectrometry of the Cu(II)-Ami-H2O2system Figure 1 presents the molecule of Cu(II)-amikacin complex with the possible places of breakage indicated, and Fig 2 provides typical mass spectra for the antibiotic and its complex both in presence and absence of hydrogen perox-ide According to the data presented in Table 1, the solution containing Ami alone can be characterized by signals at m/z
of 587 and 294 (predominant), corresponding to single- and double-charged Ami ions, respectively Several peaks of low intensity were also observed, resulting from dimerization and fragmentation of the Ami molecule during the ioniza-tion process No differences were observed between the mass-spectra of the samples of Ami and Ami + H2O2 Detection of the signals at m/z of 647 and 324 in the Cu(II) + Ami samples (Cu(II)-Ami complex ions, carrying charges of +1 and +2, respectively) confirmed the presence
of 1 : 1 complexes in solution at pH 7.4 New peaks were also seen in the high m/z region, at m/z of 1232 and 1294 They can be assigned to single-charged molecules of Cu(II)-Ami and Cu(II) Ami, respectively Also, two additional
Trang 3peaks at m/z 710 and 355 appeared in the spectrum,
suggesting the formation of a Cu(II)2Ami complex The
signal at m/z of 385 belongs to a complex decomposition
product
The samples consisting of Cu(II), amikacin and H2O2
gave signal-rich mass-spectra There were no qualitative
differences between the samples at various incubation times
Novel, high intensity signals at m/z values of 617, 600, 308
and 300 indicate the formation of Ami-oxygen adducts
(single and double-protonated Ami-O and Ami-O2
mole-cules) Another set of weaker signals, at m/z values of 679,
662, 340 and 331 likely represents Cu(II) complexes of these
adducts (see Table 1) Also, many low intensity peaks
appeared in the m/z region of 439–512 In order to assign
them and to gain better understanding of the oxidation
products, the MS/MS experiments were performed for
peaks at m/z 586 (Ami), 600 (Ami + O) and 617
(Ami + 2O) The results are presented in Table 2 Only
the peaks with m/z higher than c 400 were resolved to a
degree allowing assignments
Oxidation of 2¢-deoxyguanosine
We previously found that Cu(II)-Ami has a particular
ability to oxidize dG to 8-oxo-dG [1] The kinetics of this
process at 37C was investigated here The reactions of
decomposition of dG and formation of 8-oxo-dG are
illustrated on Fig 3A and B, respectively The shapes of
kinetic curves indicate that both these reactions are in the
second-order mode at times longer than 10 min Still, the
kinetic constants of first-order reactions could be calculated
from the first 10 min, to be 1.836 ± 0.039· 10)3min)1for
dG decomposition and 2.084 ± 0.061· 10)1min)1 for
8-oxo-dG formation For comparison with previous
experi-ments [1], data at 24 h were also recorded for 25 and 37C,
and were found to fall within the same ranges (data not
shown) The efficacy of dG conversion into 8-oxo-dG was
50% for 25 C and 25% for 37 C
Interaction with pBR322 plasmid Figure 4 presents the strand break assays of pBR322, performed by means of agarose gel electrophoresis The plasmid can be detected in one of three forms: supercoiled (I), nicked/relaxed (II) or linear (III) Part A of Fig 4 shows DNA cleavage by Cu(II)-Ami complex both in the presence and absence of H2O2(50 lM), part B – in two different, lower concentrations of H2O2: 5 and 0.5 lM In the first experiment (part A) the samples, dissolved in 50 mM phosphate buffer of pH 7.4, contained combinations of
50 lMCu(II), Ami and H2O2 At these conditions 55% of Cu(II) was present as CuHAmi, 35% as CuAmi, and the remaining 10% as Cu(II) aqua ion [1]
Cu(II) alone (lane 7) and H2O2alone (lane 3) increased the amount of form (II), but only marginally under those conditions The presence of Ami alone (lane 8), with Cu(II) (lane 9) or with H2O2(lane 5) resulted in a higher conversion
to form (II) The combination of Cu(II) and H2O2(lane 4) was more active, yielding mostly form (II) and also a trace
of linear form (III), while the mixture of all three compo-nents (the Cu(II)-Ami complex in the presence of H2O2 (lane 6) destroyed form (I) completely in these conditions, yielding 2/3 of form (II) and 1/3 of form (III)
Another effect, seen specifically for samples containing Ami, was a faint smear between the well and the band of form II It had the highest intensity for Cu(II)-free samples, both with and without H2O2(lanes 5 and 8)
Having established the requirement of H2O2presence for complex-mediated DNA cleavage, we checked its reactivity
at H2O2 concentrations 10 and hundred times lower than the initial 50 lM (Fig 4 part B) The effects of lower concentrations of H2O2alone (lanes 3 and 7) were at the background level The activity of the mixture of Ami and
H2O2was maintained at the constant level, independently of
H2O2concentration The activity of the Cu(II) and H2O2 combination was found to be roughly proportional to the logarithm of H2O2 concentration The same can be estimated for the Cu(II)-Ami-H2O2system, but at a higher level of activity (lanes 6, 10)
Figure 5 present the kinetics of pBR322 cleavage with Ami + H2O2 (part A), Cu(II) + H2O2 (part B), and Ami + Cu(II) + H2O2 (part C) for incubation times between 5 and 95 min As clearly seen, Ami + H2O2did not produce form (III), and the ratio of forms (I) and (II) remained constant throughout the experiment In the presence of Cu(II) and the complex, the degradation process proceeded from decomposition of form (I) (almost complete
at 90 min in case of Cu (II) aqu a ion and 15 min for the complex), through the relaxed form II, to the linear form III In the Cu(II) + H2O2system, the amount of form II increased during the duration of the experiment Only for the complex in the presence of hydrogen peroxide a further degradation of both forms II and III was seen, to a smear of short DNA fragments at times longer than 75 min
Cleavage of yeast tRNAPhe induced by the Cu(II)-Ami-H2O2system
Figure 6 shows the results of the strand-break assay of yeast tRNAPhe in the presence of Ami, Cu(II)-Ami and the Cu(II)-Ami-H2O2 system, performed by means of poly-acrylamide gel electrophoresis Strikingly, in the presence of
Fig 1 The schematic drawing of the Cu(II)-amikacin complexat
pH 7.4, with the sites of breakage indicated by ESI-MS and MS/MS
experiments Arrows indicate the positions of breakages occurring in:
ligand molecule, in the absence of Cu(II), ; Cu(II)-Ami complex,
dotted arrow ; in both cases, fi
Trang 4Ami or Cu(II)-Ami highly specific cleavages occurred in the
anticodon loop, at Y37 (Fig 7) These cleavages were
already observed with 5-fold molar excess of the antibiotic
or its copper(II) complex over tRNAPhe (lanes 4 and 9, Fig 6) The cleavage intensities increased with the concen-tration of Ami and Cu(II)-Ami gradually, up to their
Fig 2 Typical ESI-MS spectra of amikacin (A), its Cu(II) complexin the absence (B) and presence (C) of H 2 O 2 , as well as the MS/MS experiment on the 587 m/z peak of amikacin (D).
Trang 550-fold molar excess (lanes 6, 7 and 11, 12, Fig 6) This may
reflect a saturation effect of a tight antibiotic binding site
Remarkably, the intensities of cleavages induced by free
Ami were slightly higher than those induced by its Cu(II)
complex
Cleavages generated in the presence of Cu(II)-Ami-H2O2
system were more numerous Two cleavages with similar
intensities were observed at Y37 and A36 in the anticodon
loop (lanes 13–19, Fig 6) Moreover, several cleavages
appeared also in theD-arm and their intensities increased
with H2O2concentration The most prominent cleavage site
was localized at D17 The cleavages were weaker when the
Cu(II)-Ami-H2O2 system was prepared one hour before
incubation with tRNAPhe(lanes 20 and 21, Fig 6) In this
case, three major cleavages were observed, at Y37, A36 and
D17 It has to be noted, however, that the cleavage at D17
and in the D and anticodon loops occurred to some extent
already in the presence of 10 lMCu(II) ions and 50 lMH2O2
(data not shown) No cleavage was observed, however, at
Y37 in the absence of Ami or its Cu(II)-complex
Interest-ingly, in another RNA molecule, the 3¢ product of the
genomic delta ribozyme, Ami and its Cu(II) complex
induced no specific cleavages Only upon the addition of
HO several weak, nonspecific cleavages were found
D I S C U S S I O N The initial target of oxidations promoted by Cu(II)-Ami is the complexed aminoglycoside molecule (Fig 1) Two kinds
of reactions were seen in the presence of Cu(II) and H2O2 Fragmentations occurred predominantly at glycosidic bonds between the rings and in the amide bond, linking the aglycon to ring B (see arrows at Fig 1 and spectra at Fig 2) Single and double oxygen additions were also seen (Table 1) The fragmentation experiments performed on the
MS peaks of these adducts indicated their heterogeneity The majority of secondary peaks are identical to those found for unmodified Ami (Table 2) This suggests that labile forms, perhaps hydroperoxides, constitute the major-ity of adducts, rather than products of hydroxylation But even those two peaks which could be assigned to stable oxygen adducts (at m/z 528 and 436, Table 2) originate from two different molecules – the former from an adduct
on one of the terminal rings and the latter from an adduct
on the aglycon chain Such a lack of specificity is charac-teristic for hydroxyl radical-like agents, and in this case can
be ascribed to the reactivity of an activated oxygen atom, bound at the copper atom coordinated to the Ami molecule All the peaks with Cu(II)/Ami ratios different than 1 : 1 are
Table 1 Summary of the observed ESI-MS ions (m/z) at pH 7.4 Concentrations: Cu (II), Ami, 0.5 m M and H 2 O 2 , 2.5 m M w, weak, i, intermediate,
s, strong signal; agl, aglycon chain; pep., aglycon chain broken at peptide bond; CO – NH, Ca, aglycon chain broken between CO and Ca.
408 w Ami – ring A or C with O (glycosidic bond) + H +
315w Ami – ring A or C – agl Ca– 2OH + H+
177w ring A or C with O (glycosidic bond) + H+
385 i Cu(II)-Ami – ring A or C – agl pep + H +
324 i Cu(II)-Ami + 2H + and all the peaks seen for Ami
and all the peaks seen for Ami and Cu(II)-Ami
Trang 6due to the decomposition of the Cu2Ami2dimer, which is
formed during reduction of droplet volume in the process of
injection into the MS instrument (see [12] for the studies of
Cu2Ami2dimer formation)
8-Oxo-dG is the first relatively stable product of dG
oxidation, used as a simple chemical model for introductory
assessment of promutagenic/procarcinogenic properties of
the agents studied [29–31] Further oxidation products of
dG are easily formed in vitro, because 8-oxo-dG has a lower
oxidation potential than dG [32–34] The main final
products of oxidative dG destruction are more hydrophilic than dG and absorb UV weaker than either dG or
8-oxo-dG [35] Therefore, they cannot be quantified in the same HPLC assay, and the loss of dG should be used instead, as
an overall measure of reactivity
The susceptibility of 8-oxo-dG to further oxidation usually leads to a low steady-state abundance of 8-oxo-dG
in model experiments in vitro Thus, the yield of as much as 27% of 8-oxo-dG vs total dG loss at 37C and as mu ch as 53% at 25C, indicates a dG-specific, and thus a relatively mild oxidizing agent Our study of the mechanism of H2O2 activation by Cu(II)-Ami suggests the presence of a copper-bound oxygen radical [12]
DNA cleavage is another type of genotoxic activity The initial experiment (Fig 4) indicated that the ability of uncomplexed Ami to generate form (II) of the pBR322 plasmid equaled that of the complex at 0 or 0.5 lMH2O2, while Cu(II) alone in these conditions did not have such activity The lack of effects of H2O2suggested a nonredox process, especially as Ami itself is redox-inactive [12] Form (II) is the relaxed form of the plasmid, devoid of super-helicity, therefore we speculated whether its formation was due to single-strand nicking or superhelix unwinding Further experiments, aimed at finding out which of these options is true, demonstrated the lack of effect of incubation time and the lack of concentration effect of Ami between
5 lMand 25 mM, and differences in the extent of formation
of form II between experiments (compare lanes 5 and 8 in Fig 4A and lanes 5 and 9 in Fig 4B with those in Fig 5A), while the cleavage results for Cu(II) and Cu(II)-Ami were reproducible This suggests the occasional presence of an
Fig 3 The experimental curves of the kinetic of dG decomposition (A)
and 8-oxo-dG (B) formation caused by Cu(II)-Ami complexes in the
presence of hydrogen peroxide Solution contained sodium phosphate
buffer, 50 m M ; Ami and Cu (II), 50 l M ; H 2 O 2 , 0.5 m M ; dG, 50 l M
Incubation at 37 C.
Fig 4 Agarose gel electrophoresis of pBR322 plasmid cleavage by Cu(II)-Ami complex The samples, incubated for 1 h at 37 C, were ran
on a 1% agarose gel, containing ethidium bromide for 1 h at 4 VÆcm)1
in Tris/borate/EDTA buffer (A) Lane 1, plasmid; lane 2, plasmid linearized with XhoI endonuclease; lane 3, plasmid + 50 l M H 2 O 2 ; lane 4, plasmid + 50 l M CuCl 2 + 50 l M H 2 O 2 ; lane 5, plasmid +
50 l M Ami + 50 l M H 2 O 2 ; lane 6, plasmid + 50 l M complex +
50 l M H 2 O 2 ; lane 7, plasmid + 50 l M CuCl 2 ; lane 8, plasmid + 50 l M
Ami; lane 9, plasmid + 50 l M complex (B) Lane 1, plasmid; lane 2, plasmid linearized with XhoI endonuclease; lane 3, plasmid + 5 l M
H 2 O 2 ; lane 4, plasmid + 50 l M CuCl 2 + 5 l M H 2 O 2 ; lane 5, plasmid + 50 l M Ami + 5 l M H 2 O 2 ; lane 6, plasmid + 50 l M complex +
5 l M H 2 O 2 ; lane 7, plasmid + 0.5 l M H 2 O 2 ; lane 8, plasmid + 50 l M
CuCl 2 + 0.5 l M H 2 O 2 ; lane 9, plasmid + 50 l M Ami + 0.5 l M
H 2 O 2 ; lane 10, plasmid + 50 l M complex + 0.5 l M H 2 O 2
Table 2 Results of MS/MS experiments w; weak, i-intermediate,
s-strong signal agl.; aglycon chain pep.; aglycon chain broken at
peptide bond, CO – NH; C a ; aglycon chain broken between CO and
C a ; C b ; aglycon chain broken between C a and C b ; C c ; aglycon chain
broken between Cband Cc.
Amikacin (Ami + H + , m/z 587)
453 i Ami – agl pep – 2 NH 2 + H +
Cu(II)-amikacin-oxygen adduct (Ami + O + H + , m/z 600)
453 i Ami- agl pep – 2 NH 2 + H +
Cu(II)-amikacin-dioxygen adduct (Ami + 2O + H + , m/z 617)
Trang 7unknown DNA-cleaving impurity in our samples, rather
than a specific effect of uncomplexed Ami towards plasmid
DNA
The Cu(II)-Ami complex was a superior DNA cleavage
agent at 5 and 50 lM H2O2 (Fig 4) Only plasmid
linearization, but not further degradation to short DNA
fragments, was seen in this initial experiment To obtain
further information, the reaction kinetics was followed As
shown in Fig 5A, the Ami-H2O2system was not able to
convert the plasmid to form (III), even at long incubation
times, and the ratio of forms (I) and (II) remained constant
In contrast, the extension of incubation times for the
Cu(II)-Ami-H2O2system clearly demonstrated the gradual
degra-dation of superhelical DNA to its linear form, and further,
to a continuum of short DNA fragments (Fig 5C) The
relative persistence of form (II) and the stepwise character of
plasmid DNA degradation through all the forms suggest
that the DNA fragmentation is a result of accumulation of
random single strand breaks in the plasmid, rather than the
immediate formation of double strand breaks which was
proposed previously for Cu(II) aqua ion interacting with
DNA [36,37], but which would lead to the direct formation
of form III from form I In fact, our control mixture of
Cu(II) ions and H2O2 (Fig 5B), also produced double
strand scission in a stepwise fashion, degrading form I to
form II and then to form III, but much slower than the
complex Altogether, our results indicate that Ami strongly
activates Cu(II) ions to cleave plasmid DNA on a redox mechanism As proposed previously, Ami complexation makes the both Cu(I)/Cu(II) and Cu(II)/Cu(III) redox pairs available, while the DNA-interacting Cu(II) aqua ion can only access the Cu(I)/Cu(II) redox pair [12] This additional redox mechanism is clearly more efficient in exerting DNA strand breaks It should be noted that the metal-bound rather than free oxygen radicals are the cleaving species in both systems [12,36]
Aminoglycoside antibiotics are known to interact with a variety of RNA targets including ribosomal RNAs [38], group I introns [39] and ribozymes [40] In order to evaluate the ability of Cu(II)-Ami-H2O2system to promote cleavage
of RNA we have chosen yeast tRNAPhe as a model substrate Ami and Cu(II)-Ami complex induced highly specific cleavage in the anticodon loop, at the hipermodified base Y37 (wybutine) (Figs 6 and 7) Also other amino-glycosides, as kanamycin A, neomycin B and a synthetic neomycin-neomycin dimer have been shown to induce fragmentation of tRNAPhe at Y37 and additionally at
m7G46 in the variable loop region [41] Although no specific chemical mechanism was determined, two different types of
Fig 5 Kinetics of cleavage of pBR322 plasmid (15 l M per DNA bp) in
50 m M sodium phosphate buffer, pH 7.4, with 50 l M H 2 O 2 , followed by
agarose gel electrophoresis, in the presence of Ami alone (A), Cu(II) ions
alone (B), and Cu(II)-Ami complex(C) The samples were incubated at
37 C and then ran for 1 h on a 1% agarose gel, containing ethidium
bromide, at 4 V cm)1in Tris/borate/EDTA buffer (A) Lane 1,
plas-mid; lane 2, plasmid linearized with XhoI endonuclease; lanes 3–12,
plasmid + 50 l M Ami + 50 l M H 2 O 2 incubated for 5, 15, 25, 35, 45,
55, 65, 75, 85, 95 min, respectively (B) Lane 1, plasmid; lane 2, plasmid
linearized with XhoI endonuclease; lanes 3–12, plasmid + 50 l M
Cu(II) + 50 l M H 2 O 2 incubated for 5, 15, 25, 35, 45, 55, 65, 75, 85,
95 min, respectively (C) Lane 1, plasmid; lane 2, plasmid linearized
with XhoI endonuclease; lanes 3–12, plasmid + 50 l M
Cu(II) + 50 l M Ami + 50 l M H 2 O 2 incubated during 5, 15, 25, 35,
45, 55, 65, 75, 85, 95 min, respectively.
Fig 6 Specificity of cleavages in yeast tRNAPhe, induced by Ami and its Cu(II) complex 5¢-end-labeled tRNA Phe was mixed with unlabeled tRNA Phe to obtain final concentration of 1 l M Lane 1, untreated tRNAPhe, lane 2, tRNAPhe+ 50 l M Cu(II) ions; lanes 3–7, tRNAPhe + 1, 5, 20, 50, 100 l M amikacin, respectively; lanes 8–12, tRNA Phe +
1, 5, 20, 50, 100 l M complex, respectively; lane 13, tRNA Phe + 20 l M
complex + 50 l M H 2 O 2 ; lane 14, tRNAPhe+ 50 l M complex +
50 l M H 2 O 2 ; lane 15, tRNAPhe+ 100 l M complex + 50 l M H 2 O 2 ; lane 16, tRNA Phe + 50 l M complex + 10 l M H 2 O 2 ; lane 17, tRNA Phe
+ 50 l M complex + 20 l M H 2 O 2 ; lane 18, tRNAPhe+ 50 l M com-plex + 50 l M H 2 O 2 ; lane 19, tRNAPhe+ 50 l M complex + 100 l M
H 2 O 2 ; lane 20, tRNA Phe + 50 l M complex + 50 l M H 2 O 2 , prepared
1 h before incubation with RNA; lane 21, tRNAPhe+ 50 l M complex + 100 l M H 2 O 2 prepared 1 h before incubation with RNA; L, formamide ladder.
Trang 8mechanisms could explain the observed cleavages One
possibility is that Ami or Cu(II)-Ami complex binds to the
RNA molecule and provides the basic functional group
responsible for deprotonation of the critical 2¢-hydroxyl
group and subsequent cleavage of the RNA chain via
transesterification mechanism A second possible
mechan-ism involves the binding of the antibiotic to the RNA such
that the resulting complex adopts a conformation more
conducive to polynucleotide chain scission at these specific
sites The cleavages observed in the presence of Ami or
Cu(II)-Ami were, however, exceptionally specific and
effi-cient This clearly suggests that Ami and its copper(II)
complex bind tightly to the anticodon loop of tRNAPhe, in
the vicinity of Y37, which might be directly involved in the
binding, especially as no specific cleavages were detected for
the genomic delta ribozyme lacking modified bases in the
presence of Ami or Cu(II)-Ami complex (our unpublished
data) Hydrated Mg2+ion is bound directly to wybutine in
the crystal structure of yeast tRNAPhe [42,43] and the
presence of a tight metal-ion-binding site in the anticodon
loop was also confirmed in solution by means of Eu3+
-induced cleavage method [44] Specific cleavages at U33 and
A36 were observed and they disappeared after removing
wybutine Interestingly, enhancement of wybutine
fluores-cence was observed after addition of neomycin B or
kanamycin A to tRNAPhe and the effect was similar to
that one caused by magnesium ions [41] Occupation of the
same binding site for metal ions and aminoglycoside
antibiotics is often observed in RNA and, for example,
displacement of metal ions by antibiotics inhibits catalytic
properties of ribozymes [39,40] The decrease of tRNA
cleavage upon the addition of Cu(II) simply reflects the lowered affinity of Ami to the nucleic acid in the complex, due to electrostatic and/or conformational reasons, same as with DNA Different cleavage patterns of yeast tRNAPhe obtained in the presence of Cu(II)-Ami and the
Cu(II)-Ami-H2O2system can be explained by different mechanisms of these reactions The addition of H2O2 resulted in the appearance of additional cleavages in several new positions, near the putative complex binding site, but also in the D-arm The increase of H2O2 concentration resulted in multiple cleavages and a gradual loss of specificity (Fig 6) This pattern suggests that, in addition to copper-bound active oxygen, also free radicals are generated as side-reaction and confer further damage These results are in excellent agreement with the mechanism of H2O2activation
by Cu(II)-Ami [12]
One can speculate that the binding of Cu(II)-Ami to wybutine is enhanced by the formation of a coordination bond between Ami-complexed Cu(II), which is coordina-tively unsaturated at pH 7.4 [1,12], and the secondary amine nitrogen present in the wybutine residue
C O N C L U S I O N S The processes of dG, DNA and RNA oxidation clearly demonstrate that Cu(II)-Ami is a potentially dangerous genotoxic agent Cu(II)-coordinated Ami retains the ability
to bind to nucleic acids and cleave phosphodiester bonds The complex activates H2O2, producing both metal bound and diffusible radical species, capable of conferring various kinds of oxidative damage to nucleic acids and their components Ami hydroperoxides, formation of which was suggested by MS experiments, may also be involved
in these reactions Concentrations of amikacin used in all experiments were comparable to its therapeutic ones, which are in the range of 50–70 lMin blood serum [45–48] The issue of intracellular copper bioavailability is currently debated [49–51] However, there is a mobile pool of copper
in blood serum, which may be additionally increased by deleterious oxidative events [52] The story of bleomycin demonstrates that an antibiotic can sequester copper and introduce it into cell [53] Therefore, our data may provide elements for the yet unknown mechanisms of toxicity of aminoglycoside antibiotics
A C K N O W L E D G E M E N T S
This work was supported by the Polish State Committee for Scientific Research (KBN), grant no 3 T09A 06818.
R E F E R E N C E S
1 Je_zzowska-Bojczuk, M & Bal, W (1998) Co-ordination of copper (II) by amikacin Complexation equilibria in solution and oxygen activation by resulting complexes J Chem Soc Dalton Trans 153–159.
2 Gaggelli, E., Gaggelli, N., Maccotta, A., Valensin, G., Marini, D.,
Di Cocco, M.E., Manetti, C & Delfini, M (1999) Delineation of conformational and structural features of the amikacin-Cu (II) complex in water solution by 13C-NMR spectroscopy Spectro-chim Acta Part A 55, 205–210.
3 Hanessian, S & Patil, G (1978) Carbon-NMR studies of the initial binding of Cu (II) to aminoglycoside antibiotics – a useful structural and functional probe Tetrahedron Lett 12, 1031–1034.
Fig 7 The secondary structure of the yeast tRNAPhewith cleavage sites
induced by Ami and its Cu(II) complex Arrows, Ami and Cu(II)-Ami;
lines (higher intensity cleavage) and dotted lines (lower intensity
clea-vage), Cu(II) aqua ions and Cu(II)-Ami in the presence of H 2 O 2
Trang 94 Hanessian, S & Patil, G (1978) Aminoglycoside antibiotics – a
method for selective N-acylation based on the temporary
protec-tion of amino alcohol funcprotec-tions as copper chelates Tetrahedron
Lett 12, 1035–1038.
5 Nagabhushan, T.L., Cooper, A.B., Turner, W.N., Tsai, H.,
McCombie, S., Mallams, A.K., Rane, D., Wright, J.J., Reichert,
P., Boxler, D.L & Weinstein, J (1978) Interaction of vicinal and
nonvicinal amino-hydroxy group pairs in
aminoglycoside-ami-nocyclitol antibiotics with transition metal cations Selective
N-protection J Am Chem Soc 100, 5253–5254.
6 Abu-El-Wafa, S.M., El-Ries, M.A., Abou-Attica, F.M & Issa,
R.M (1989) Coordination chemical studies of some polymeric
transition metal complexes with neomycin and their biological
activity uses Indirect determination of neomycin by atomic
absorption spectroscopy Anal Lett 22, 2703–2716.
7 Je_zzowska-Bojczuk, M., Bal, W & Kozłowski, H (1998)
Kana-mycin revisited: a combined potentiometric and spectroscopic
study of copper (II) binding to kanamycin B Inorg Chim Acta
275/276, 541–545.
8 Je_zzowska-Bojczuk, M., Karaczyn, A & Kozłowski, H (1998)
Copper (II) binding to tobramycin Potentiometric and
spectro-scopic studies Carbohydr Res 313, 265–269.
9 Je_zzowska-Bojczuk, M., Karaczyn, A & Bal, W (1998) Binding of
Cu(II) ions to geneticin, a gentamicin analog J Inorg Biochem.
71, 129–134.
10 Grapsas, I., Massova, I & Mobashery, S (1998)1H-NMR
ana-lysis of copper-aminoglycoside complexes in solution and its
im-plication for regioselective modification of multifunctional
aminoglycoside antibiotics Tetrahedron 54, 7705–7720.
11 Sreedhara, A., Freed, J.D & Cowan, J.A (2000) Efficient deoxy
ribonucleases Greater than 50-million-fold rate enhancement in
enzyme-like DNA cleavage J Am Chem Soc 122, 8814–8824.
12 Je_zzowska-Bojczuk, M., Les´niak, W., Bal, W., Kozłowski, H.,
Gatner, K., Jezierski, A., Sobczak, J., Mangani, S &
Meyer-Klaucke, W (2001) Molecular mechanism of hydrogen peroxide
conversion and activation by Cu(II)-amikacin complexes Chem.
Res Toxicol 14, 1353–1362.
13 Tally, F.P., Louie, T.J., Weinstein, W.M., Bartlett, J.G &
Gorbach, S.L (1975) Amikacin therapy for severe gram-negative
sepsis Emphasis on infections with gentamicin-resistant
organ-isms Ann Intern Med 83, 484–488.
14 Davis, B.D (1987) Mechanism of bactericidal action of
amino-glycosides Microbiol Rev 51, 341–350.
15 Busse, H.J., Wostmann, C & Bakker, E.P (1992) The bactericidal
action of streptomycin: membrane permeabilization caused by the
insertion of mistranslated proteins into the cytoplasmic membrane
of Escherichia coli and subsequent caging of the antibiotic inside
the cells due to degradation of these proteins J General Bacteriol.
138, 551–561.
16 Swan, S.K (1997) Aminoglycoside nephrotoxicity Semin
Neph-rol 17, 27–33.
17 Humes, H.D (1999) Insights into ototoxicity Analogies to
nephrotoxicity Ann N Y Acad Sci 884, 15–18.
18 Forge, A & Schacht, J (2000) Aminoglycoside antibiotics Audiol.
Neurootol 5, 3–22.
19 Priuska, E.M., Clark-Baldwin, K., Pecoraro, V.L & Schacht, J.
(1998) NMR studies of iron-gentamycin complexes and the
implications for aminoglycoside toxicity Inorg Chim Acta 273,
85–91.
20 Guo, L.H., Hill, H.A.O., Hopper, D.J., Lawrance, G.A &
Sanghera, G.S (1990) Direct voltammetry of the Chromatium
vinosum enzyme, sulfide: cytochrome c oxidoreductase
(flavocy-tochrome c552) J Biol Chem 265, 1958–1963.
21 Rae, T.D., Schmidt, P.J., Pufahl, R.A., Culotta, V.C &
O’Halloran, T.V (1999) Undetectable intracellular free copper:
the requirement of a copper chaperone for superoxide dismutase.
Science 284, 805–808.
22 Theophanides, T & Anastassopoulou, J (2002) Copper and carcinogenesis Crit Rev Oncol./Hematol 42, 57–64.
23 Bar-Or, D., Rael, L.T., Lau , E.P., Rao, N.K., Thomas, G.W., Winkler, J.V., Yukl, R.L., Kingston, R.G & Curtis, C.G (2001)
An analog of the human albumin N-terminus (Asp-Ala-His-Lys) prevents formation of copper-induced reactive oxygen species Biochem Biophys Res Commun 284, 856–862.
24 Linder, M.C (1991) Biochemistry of Copper Plenum Press, New York.
25 Halliwell, B & Gutteridge, J.M.C (1999) Free Radicals in Biology and Medicine, 3rd edn Oxford University Press, Oxford.
26 Sreedhara, A & Cowan, J.A (1998) Efficient catalytic cleavage of DNA mediated by metalloaminoglycosides Chem Commun 1737–1738.
27 Sreedhara, A., Patwardhan, A & Cowan, J.A (1999) Novel reagents for targeted cleavage of RNA sequences: towards a new family of inorganic pharmaceuticals Chem Commun 1147–1148.
28 Białkowski, K & Kasprzak, K.S (1998) A novel assay of 8-oxo-2¢-deoxyguanosine 5¢-triphosphate pyrophosphohydrolase (8-oxo-dGTPase) activity in cultured cells and its use for evaluation of cadmium (II) inhibition of this activity Nucleic Acids Res 26, 3194–3201.
29 Kasprzak, K.S (1995) Possible role of oxidative damage in metal-induced carcinogenesis Cancer Invest 13, 411–430.
30 Floyd, R.A (1990) The role of 8-hydroxyguanosine in carcino-genesis Carcinogenesis 11, 1447–1450.
31 Tan, X., Grollman, A.P & Shibutani, S (1999) Comparison of the mutagenic properties of 8-oxo-7,8-dihydro-2¢-deoxyadenosine and 8-oxo-7,8-dihydro-2¢-deoxyguanosine DNA lesions in mamma-lian cells Carcinogenesis 11, 1447–1450.
32 Burrows, C.J & Muller, J.G (1998) Oxidative nucleobase mod-ifications leading to strand scission Chem Rev 98, 1109–1151.
33 Steenken, S & Jovanovic, S (1997) How easily oxidizable is DNA? One-electron reduction potentials of adenosine and gua-nosine radicals in aqueous solution J Am Chem Soc 119, 617– 618.
34 Yanagawa, H., Ogawa, Y & Ueno, M (1992) Redox ribonu-cleosides Isolation and characterization of 5-ydroxyuridine, 8-hydroxyguanosine & 8-hydroxyadenosine from Torula yeast RNA J Biol Chem 267, 13320–13326.
35 Raoul, S & Cadet, J (1996) Photosensitized reaction of 8-oxo-7,8-dihydro-2¢-deoxyguanosine: identification of 1-(2-deoxy-b-D-erythro-pentofuranosyl) cyanuric acid as the major singlet oxygen oxidation product J Am Chem Soc 118, 1892–1898.
36 Yamamoto, K & Kawanishi, S (1989) Hydroxyl free radical is not the main active species in site-specific DNA damage induced
by copper (II) ion and hydrogen peroxide J Biol Chem 264, 15435–15440.
37 Oikawa, S & Kawanishi, S (1996) Site-specific DNA damage induced by NADH in the presence of copper (II): role of active oxygen species Biochemistry 35, 4584–4590.
38 Schroeder, R., Waldsich, C & Wank, H (2000) Modulation
of RNA function by aminoglycoside antibiotics EMBO J 19, 1–9.
39 Wallis, M.G & Schroeder, R (1997) The binding of antibiotics to RNA Prog Biophys Mol Biol 67, 141–154.
40 Walter, F., Vicens, Q & Westhof, E (1999) Aminoglycoside– RNA interactions Curr Opin Chem Biol 3, 694–704.
41 Kirk, S.R & Tor, Y (1999) tRNA (Phe) binds aminoglycoside antibiotics Bioorg Med Chem 7, 1979–1991.
42 Holbrook, S.R., Sussman, J.L., Warrant, R.W., Church, G.M & Kim, S.H (1977) RNA–ligand interactions (I) Magnesium binding sites in yeast tRNAPhe Nucleic Acids Res 4, 2811–2820.
43 Quigley, G.J., Teeter, M.M & Rich, A (1978) Structural analysis
of spermine and magnesium ion binding to yeast phenylalanine transfer RNA Proc Natl Acad Sci USA 75, 64–68.
Trang 1044 Ciesiołka, J., Marciniec, T & Krzy_zzosiak, W (1989) Probing the
environment of lanthanide binding sites in yeast tRNA (Phe)
by specific metal-ion-promoted cleavages Eur J Biochem 182,
445–450.
45 Belfayol, L., Talon, P., Eveillard, M., Alet, P & Fauvelle, F.
(1996) Pharmacokinetics of once-daily amikacin in pediatric
patients Clin Microbiol Infect 2, 186–191.
46 Giamarellou, H., Yiallouros, K., Petrikkos, G., Moschovakis, E.,
Vavouraki, E., Voutsinas, D & Sfikatis, P (1991) Comparative
kinetics and efficacy of amikacin administrated once or twice
daily in the treatment of the systemic gram-negative infections.
J Antimicrob Chemother 27, 73–79.
47 Kafetzis, D.A., Sinaniotis, C.A., Papadatos, C.J & Kosmidis, J.
(1979) Pharmacokinetics of amikacin in infants and pre-school
children Acta Pediatr Scan 68 (3), 419–422.
48 Viscoli, C., Dudley, M., Ferrea, G., Boni, L., Castagnola, E.,
Barretta, M.A., Lanino, E., Loy, A., Moroni, C & Somenzi, M.
(1991) Serum concentrations and safety of single daily dosing of
amikacin in children undergoing bone marrow transplantation.
J Antimicrob Chemother 27, 113–120.
49 Barbou ti, A., Dou lias, P.-T., Zhu , B.-Z., Frei, B & Galaris, D (2001) Intracellular iron, but not copper, plays critical role in hydrogen peroxide – induced DNA damage Free Rad Biol Med.
31, 490–498.
50 Bar-Or, D & Winkler, J.V (2002) Copper is involved in hydrogen – peroxide – induced DNA damage Free Rad Biol Med 32, 197–198.
51 Galaris, D & Zhu, B.-Z (2002) On the role of iron and copper ions in hydrogen peroxide – induced cellular DNA damage Free Rad Biol Med 32, 198–199.
52 Yim, M.B., Chock, P.B & Stadtman, E.R (1993) Enzyme func-tion of copper, zinc superoxide dismutase as a free radical generator J Biol Chem 268, 4099–4105.
53 Umezawa, H (1983) Roles of amines in the action of antibiotics – bleomycin, spergualin & aminoglycosides Adv Polyamine Res 4, 1–15.