Selective targeting of DNA for cleavage within DNA-histone assemblies by a spermine-[CpW(CO)3Ph]2 conjugate

Article abstract of DOI:10.1039/b504248h

In contrast to the histone-modifying action of other complexes of the type CpML_nR, the compound obtained by linking the phenyl rings of two CpW(CO)₃Ph moieties to the DNA-binding agent spermine selectively cleaves DNA in DNA-histone

Full text of DOI:10.1039/b504248h

View Article Online / Journal Homepage / Table of Contents for this issue
C O M M U N I C A T I O N
Selective targeting of DNA for cleavage within DNA–histone
assemblies by a spermine–[CpW(CO)3Ph]2 conjugate†
Thomas A. Shell and Debra L. Mohler*
Department of Chemistry, Emory University, 1515 Dickey Dr., Atlanta, Georgia, 30322, USA.
E-mail: dmohler@emory.edu; Fax: 404-727-6586; Tel: 404-727-6738
Received (Pittsburgh, PA, USA) 24th March 2005, Accepted 23rd June 2005
First published as an Advance Article on the web 20th July 2005
In contrast to the histone-modifying action of other com-
plexes of the type CpMLnR, the compound obtained by
linking the phenyl rings of two CpW(CO)3Ph moieties to
the DNA-binding agent spermine selectively cleaves DNA
in DNA–histone assemblies.
Due to their proven antibiotic and antitumor activity, molecules
that modify nucleic acids are important tools in chemistry,
biology, and medicine. Of particular note in the latter area are
compounds that cause the double-strand cleavage of DNA,1
since such lesions are more difficult than single-strand events
for the cellular machinery to repair.2 Molecules that modify
oligonucleotides are also useful for the study of primary, sec-
ondary, and tertiary structure of DNA3 and for the elucidation
of the binding modes of other molecules with nucleic acids.4 In
all of these applications, the design of systems that incorporate
triggering mechanisms5 for producing the active species provides
the potential ability to target specific substrates and structures.
For example, we have previously employed the photolysis of
organometallic species of the type CpM(CO)nR to generate
carbon-centered radicals that cause single- and/or double-
strand cleavage of plasmid DNA.6
Although radicals show potential as tools for the elucidation
of secondary and tertiary DNA structure, both carbon-centered
and hydroxy radicals are known to modify histone proteins,
causing either protein–DNA crosslinking7 or the dissociation
of protein–DNA assemblies.8 One approach to circumvent this
undesired reactivity is to attach DNA recognition elements to
a triggerable radical precursor, thus producing the radical in
close proximity to DNA. Therefore, we now report the synthesis
and DNA-cleaving behavior of compound 3, which is obtained
by linking the phenyl rings of two CpW(CO)3Ph moieties to
the DNA-binding agent spermine.9,10 Importantly, this molecule
selectively cleaves DNA in DNA–histone assemblies without
giving biomolecular dissociation.
Scheme 1 Photochemical production of carbon-centered radicals in
the photolysis of CpM(CO)nR complexes.
has been proposed that CpW(CO)2CH3 (3) reacts with another
molecule of starting material to produce the metal–metal
bonded species 4 and two methyl radicals. Furthermore, it has
been demonstrated that the 16-electron species CpW(CO)2CH3
(3) can coordinate a variety of ligands (e.g., L = PPh3, CH3CN,
THF, or H2O); and when CpW(CO)2(PPh3)CH3 (either purified
or produced in situ during the photolysis of 1 in the presence
of PPh3) is photolyzed, methyl radicals are formed. It is such
carbon-centered radicals that have been implicated as the active
species leading to DNA strand scission.6 The attachment of
spermine to the phenyl ring of CpW(CO)3Ph (instead of the Cp
moiety) is necessary to avoid diffusible radicals, which reduce
selectivity.14
The synthesis of 9 was straightforward (Scheme 2), begin-
ning with the preparation of the succinimide ester 7 from 4-
iodobenzoic acid. Subsequent treatment of this activated ester
Because spermine is fully protonated at physiological pH, it
is cationic and binds to anionic DNA via electrostatic forces;11
and this interaction is at most only slightly sequence-selective,12
although its exact manner is unclear. Compound 3 was expected
to bind and to cleave DNA with the same lack of specificity upon
photolysis at wavelengths greater than 300 nm, conditions which
are known to produce carbon-centered radicals in CpM(CO)nR
complexes, the general route for which is shown (Scheme 1).13
It is generally accepted that the primary photoprocess for
complexes of the formula 1, in which M = W or Fe and
R = CH3 or C6H5, involves loss of carbon monoxide (to give
3), which may be accompanied by homolysis of the metal–
methyl or metal–aryl bond to yield the metal-based radical 2
along with methyl or phenyl radical. However, radical formation
may occur by multiple pathways, as has been suggested for
the photolysis of CpW(CO)3CH3, the only complex whose
photochemistry has been extensively studied. In this case, it
† Electronic supplementary information (ESI) available: Full experimen-
tal details. See http://dx.doi.org/10.1039/b504248h
Scheme 2 Synthesis of 9.
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 3 0 9 1 – 3 0 9 3
3 0 9 1
©

View Article Online
with the zinc salt of CpW(CO)3 anion yielded the substituted
phenyl tungsten complex 8, which then reacted with spermine
to produce the desired compound 9.
double-strand events require the absorption of two photons (one
for each organometallic moiety, assuming a quantum yield of 1
for the production of each radical18) within the period of time
that 9 is associated with a particular stretch of DNA. While
ethidium bromide displacement assays19 show that 9 complexes
DNA with an apparent binding constant Kapp = 3.4 × 106 M−1,
apparently only high intensity irradiation is able to overcome
the low quantum efficiency of radical formation to produce two
radical centers within the residence time of 9 at a given site in
DNA, leading to the concurrent damaging of both strands of
the duplex.
Additional evidence for the lack of diffusion of the radicals
produced in the photolysis of 9 was provided by studies in which
radical scavengers were added to the reaction mixtures before
photolysis (Fig. 3). When 100 or 10 equivalents (lanes 3 and 4,
respectively) of cysteine, which can function as a general radical
trap,20 were present in the reaction mixture, similar amounts
of cleavage were observed as in a control reaction lane (lane
2). Likewise, 2,2,6,6-tetramethyl-1-piperdinyloxyl (TEMPO), a
stable nitroxide which traps carbon-centered radicals,21 but not
those on oxygen, also appeared not to suppress strand scission
by 9 (lanes 7 and 8 vs. lane 6). If the radical species responsible
for DNA cleavage were diffusible, both cysteine and TEMPO
would be expected to inhibit the formation of nicked DNA, as
was observed previously in the case of the unsubstituted complex
CpW(CO)3Ph.6
The DNA-cleaving activity of compound 9 was assessed
initially using a plasmid relaxation assay to monitor the
conversion of circular supercoiled DNA (form I) to relaxed
circular (form II) and linear DNA (form III). Thus, after
the photolysis of spermine derivative 9 through a Pyrex filter
in the presence of pBR322 DNA, a small amount of a 1%
SDS solution was added to the reaction mixtures, in order
to reduce the aggregration of DNA that is known to occur
with spermine itself.15 Analysis by agarose gel electrophoresis
(Fig. 1) showed a band corresponding to form III DNA
and resulting from nonrandom double-strand cleavage16 at
compound concentrations greater than 9 lM (lanes 4–6) or
0.30 molecules bp−1, a value lower than that reported for
light-induced double strand cleavage of DNA by the natural
enediyne dynemicin (0.75 molecules bp−1).17 Additionally, form
II DNA arising from single-strand cleavage occurred at ratios
as low as 0.038 molecules bp−1 (lane 7), representing a 39-fold
improvement (per radical center) in single-strand cleaving ability
over the simple complex CpW(CO)3CH3, which causes single-
strand scission at 1.5 molecules bp−1.6 One control experiment
demonstrated that the organometallic compound was necessary
for cleavage (lane 3); however, the other (lane 2) showed that high
intensity light was not required to effect strand scission. This
latter observation suggested that ambient light was sufficient to
activate 9, a surprising result, based on previous work with other
substituted or unsubstituted CpM(CO)nR complexes.
Fig. 3 Effect of radical scavengers on the cleavage of pBR322 DNA
(30 lM bp−1 in 5% DMSO/10 mM Tris buffer, pH 8) by 9. Lanes 1
and 5, DNA alone; lanes 2 and 6, DNA and 9 (6.0 lM); lanes 3 and 4,
DNA, 9 (6.0 lM), and cysteine (600 and 60 lM, respectively) lanes 7
and 8, DNA, 9 (6.0 lM), and TEMPO (600 and 60 lM, respectively).
Samples in lanes 1 and 5 were incubated on the benchtop, and those in
lanes 2–4 and 6–8 were irradiated with Pyrex-filtered light from a 450 W
medium-pressure mercury arc lamp for 20 minutes.
Fig. 1 Cleavage of pBR322 DNA (30 lM bp−1 in 10% DMSO/10 mM
Tris buffer, pH 8) by 9. Lanes 1 and 3, DNA alone; lanes 2 and 4–7, DNA
and 9 (22, 14, 11, 9, and 7.2 lM, respectively). Samples in lanes 1 and 2
were incubated on the benchtop, and those in lanes 3–7 were irradiated
with Pyrex-filtered light from a 450 W medium pressure mercury arc
lamp for 20 minutes.
After the demonstration of the basic DNA-cleaving behavior
of 9, its capacity to damage DNA selectively within DNA/H1
assemblies was investigated. Thus, 9 was irradiated in the
presence of the supramolecular complex of pUC19 DNA with
calf thymus histone H1, and the results were analyzed by gel
electophoresis (Fig. 4). A band corresponding to the form III
DNA-H1 assembly was observed at compound concentrations
greater than 50 lM (lane 6) and single-strand cleavage of
complexed DNA22 occurred at concentrations of 9 greater than
3.1 lM (lane 10). Again, both the presence of 9 and high intensity
light were required to achieve significant strand scission (lanes 4
and 5, respectively), although ambient light caused some DNA
cleavage (lane 5). Importantly, in none of the reactions was
uncomplexed DNA observed, suggesting a much higher level
of DNA cleavage than histone modification.
To further probe the use of ordinary room light in inducing
DNA cleavage by 9, pBR322 DNA was incubated with varying
concentrations of the compound on the bench; and two control
samples were prepared and incubated in the dark (Fig. 2).
In these experiments, photolysis with room light produced
form II DNA at concentrations above 1.1 lM of 9 (lane 9);
and the linearization of DNA was observed at concentrations
higher than 7.2 lM (lane 8), presumably resulting from the
accumulation of single-strand damage.16 Control experiments
showed that both the presence of 9 (lane 3) and room light (lane
2) were required for strand scission.
Fig. 2 Cleavage of pBR322 DNA (30 lM bp−1 in 10% DMSO/10 mM
Tris buffer, pH 8) by 9 with activation by ambient light. Lanes 1 and 3,
DNA alone; lanes 2 and 4–14, DNA and 9 (22, 22, 18, 14, 11, 9, 7.2, 5.8,
4.6, 2.3, 1.15 and 0.58 lM, respectively). The reaction mixtures in lanes
1 and 2 were mixed and allowed to sit in the dark for 20 minutes; those
in all other lanes were exposed to ambient light for 20 minutes.
Fig. 4 Cleavage of pUC19 DNA (102 lM bp−1 DNA and 0.23 mg
mL−1 histone H1 in 10% DMSO/10 mM Tris buffer, pH 8) by 9. Lanes 1
and 2, DNA alone; lanes 3 and 4, DNA and histone; lanes 5–11, DNA,
histone, and 9 (50, 50, 25, 13, 6.3, 3.1, and 1.5 lM, respectively). Mixtures
in lanes 2, 4 and 6–11 were irradiated with Pyrex-filtered light from a
450 W medium pressure mercury arc lamp for 20 minutes.22
Thus, ambient and high intensity irradiation show similar
efficiencies in causing single-strand cleavage; but in contrast to
the previous experiments, none of the ambient light activated
cleavage mixtures contained all three forms of DNA, indicating
that true double-strand scission does not occur under these
conditions. The most likely cause for this phenomenon is that
In conclusion, the spermine–[CpW(CO)3Ph]2 conjugate 3 has
demonstrated the ability to cause both double- and single-strand
breaks at low compound : DNA ratios in purified DNA. While
3 0 9 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 3 0 9 1 – 3 0 9 3

View Article Online
high intensity light is necessary for true double strand cleavage,
and ambient light is sufficient for cutting only one strand.
Additionally, 9 gives a high degree of nucleic acid cleavage
in DNA–histone assemblies without causing dissociation of
the biomolecular complex. Thus, the attachment of a suitably
charged DNA recognition element to a cleaving agent provides
a method to cleave DNA selectively within DNA–histone
assemblies by exploiting the differences in charge between the
two biomolecules.
8 D. L. Mohler and M. P. Maddox, III, unpublished work.
9 For a review on the chemistry and biology of polyamines and their
conjugates: G. Karigiannis and D. Papaioannou, Eur. J. Org. Chem.,
2000, 1841.
10 For lead references to other DNA-cleaving agents with attached
polyamines: I. Suzuki, A. Shigenaga, H. Nemoto and M. Shibuya,
Tetrahedron Lett., 2004, 45, 1955.
11 For lead references: C. W. Tabor and H. Tabor, Annu. Rev. Biochem.,
1984, 53, 749.
12 For example, see: N. Schmid and J. P. Behr, Biochemistry, 1991, 30,
4357.
13 B. Klein, R. J. Kazlauskas and M. S. Wrighton, Organometallics,
1982, 1, 1338; A. S. Goldman and D. R. Tyler, Inorg. Chem., 1986, 25,
706; D. R. Tyler, Inorg. Chem., 1981, 20, 2257–2261; R. J. Kazlauskas
and M. S. Wrighton, J. Am. Chem. Soc., 1982, 104, 6005; A. S.
Goldman and D. R. Tyler, J. Am. Chem. Soc., 1986, 108, 89; E.
Samuel, M. D. Rausch, T. E. Gismondi, E. A. Mintz and C. Giannoti,
J. Organomet. Chem., 1979, 172, 309; G. K. Yang, K. S. Peters and
V. Va i d a , J. Am. Chem. Soc., 1986, 108, 2511; R. G. Severson and A.
Wojcicki, J. Organomet. Chem., 1978, 157, 173; M. W. Virrels, George,
F. P. A. Johnson, J. P. Turner and J. R. Westwell, Organometallics,
1995, 14, 5203.
14 A. L. Hurley, M. P. Maddox, III, T. L. Scott, M. R. Flood and D. L.
Mohler, Org. Lett., 2001, 3, 2761.
15 E. Raspaud, I. Chaperon, A. Leforestier and F. Livolant, Biophys. J.,
1999, 77, 1547; J. S. Choi, D. K. Joo, C. H. Kim, K. Kim and J. S.
Park, J. Am. Chem. Soc., 2000, 122, 474.
16 The observation of all three forms of DNA in one cleavage reaction
mixture is indicative of a true double-strand cleavage event (as
opposed to the linearization of DNA via the random accumulation
of single-strand damage): L. F. Povirk, W. Wu¨bker, W. Ko¨hnlein and
F. Hutchinson, Nucleic Acids Res., 1977, 4, 3573.
17 T. Shiraki and Y. Sugiura, Biochemistry, 1990, 29, 9795.
18 While no quantum yields for CpW(CO)3Ph have been determined,
the data reported for CpW(CO)3Me suggests that radical production
in these systems may not be very efficient: the quantum yield for the
disappearance of CpW(CO)3Me has been measured as 0.4; however,
the value for the production of [CpW(CO)3]2, which is thought
to be coupled to methyl radical formation, is much lower (0.04)
(see ref. 13).
Acknowledgements
We acknowledge the National Science Foundation and Emory
University for support of this work, as well as shared instrumen-
tation at Emory provided by grants from the NIH and NSF.
Notes and references
1 For lead references: S. V. Kovalenko and I. Alabugin, Chem.
Commun., 2005, DOI: 10.1039/b417012; A. L. Hurley and D. L.
Mohler, Org. Lett., 2000, 2, 2745; J. L. Epstein, X. Zhang, G. A.
Doss, J. M. Liesch, B. Krishnan, J. Stubbe and J. W. Kozarich, J. Am.
Chem. Soc., 1997, 119, 6731.
2 L. F. Povirk, Molecular Aspects of Anti-Cancer Drug Action, ed.
S. Neidle and M. Waring, Verlag-Chemie, Weinheim, 1983.
3 For example, see: G. B. Schuster, Acc. Chem. Res., 2000, 33, 253.
4 For examples, see: C.-H. Yang, W.-F. Chen, M.-C. Jong, B.-J. Jong,
J.-C. Chang, M. J. Waring, L. Ma and L. Sheh, J. Am. Chem. Soc.,
2004, 126, 8104; L. M. Ottinger and T. D. Tullius, J. Am. Chem.
Soc., 2000, 122, 5901; S. R. Rajski, S. Kumar, R. J. Roberts and J. K.
Barton, J. Am. Chem. Soc., 1999, 121, 5615; P. J. Carter, C.-C. Cheng
and H. H. Thorp, J. Am. Chem. Soc., 1998, 120, 632; R. W. Frazee,
J. A. Taylor and T. D. Tullius, J. Mol. Biol., 2002, 323, 665; J. W.
Trauger and P. B. Dervan, Methods Enzymol., 2001, 340, 450.
5 For lead references, see: M. E. Maier, Synlett, 1995, 13; B. Armitage,
Chem. Rev., 1998, 98, 1171; H. Shimakoshi, T. Kaieda, T. Matsuo, H.
Sato and Y. Hisaeda, Tetrahedron Lett., 2003, 44, 5197; P. A. Wender,
S. M. Touami, C. Alayrac and U. C. Philipp, J. Am. Chem. Soc., 1996,
118, 6522; D. P. Arya and D. J. Jebaratnam, Tetrahedron Lett., 1995,
36, 4369.
6 D. L. Mohler, D. R. Dain, A. D. Kerekes, W. R. Nadler and T. L.
Scott, Bioorg. Med. Chem. Lett., 1998, 8, 871; D. L. Mohler, E. K.
Barnhardt and A. L. Hurley, J. Org. Chem., 2002, 67, 4982.
7 E. Gajewski, A. F. Fuciarelli and M. Dizdaroglu, Int. J. Radiat. Biol.,
1988, 54, 445; H. Schuessler and E. Jung, Int. J. Radiat. Biol., 1989,
56, 423; C. Luxford, R. T. Dean and M. Davies, Chem. Res. Toxicol.,
2000, 13, 665.
19 A. R. Morgan, J. S. Lee, D. E. Pulleyblank, N. L. Murray and D. H.
Evans, Nucleic Acids Res., 1979, 7, 547.
20 P. Huston, J. H. Espenson and A. Bakac, Inorg. Chem., 1992, 31,
720.
21 T. J. Connolly, M. V. Baldov´ı, N. Mohtat and J. C. Scaiano,
Tetrahedron Lett., 1996, 37, 4919.
22 The assignment of bands in this gel was confirmed by comparison to
markers and by the trends in the intensity of these bands with respect
to concentration of 9; see ESI†.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 3 0 9 1 – 3 0 9 3
3 0 9 3