Enthalpy-driven nuclease-like activity and mechanism of peptide-chlorambucil conjugates

Article abstract of DOI:10.1039/c4ob00123k

We report the results of attaching the anticancer drug chlorambucil (CLB) to two high-affinity DNA binding peptides: Met-Hyp-Arg-Lys-(Py) ₄-Lys-Arg-NH₂ (HyM-10) and Gln-Hyp-Arg-Lys-(Py) ₄-Lys-Arg-NH₂ (HyQ-10). These CLB-peptide conjugates cleave DNA very effectively and sequence-selectively without the use of chemicals, heat, or UV irradiation. Polyacrylamide gel electrophoresis identifies the sites where CLB-HyM-10 and CLB-HyQ-10 attack a complementary pair of 5′-³²P-labeled duplexes derived from pBR322 in the absence of piperidine or other chemical additives. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) has confirmed the preferential cleavage sites as well as a novel stepwise cleavage mechanism of sequence-selective DNA cleavage. Resembling restriction endonucleases, the CLB-peptide conjugates appear to be capable of producing double strand DNA breaks. Circular dichroism studies show that CLB-HyM-10 and CLB-HyQ-10 induce significant local conformational changes in DNA via the minor groove, possibly with dimeric binding stoichiometry. The energetic basis of DNA binding by these conjugates has been investigated by isothermal titration calorimetry, revealing that the binding of both the peptides and their CLB conjugates is overwhelmingly enthalpy-driven. The maintenance of a conserved negative binding free energy in DNA-conjugate interactions is a crucial feature of the universal enthalpy-entropy compensation phenomenon. The strongly enthalpy-driven binding of CLB-peptide conjugates to preferred loci in DNA furnishes the required proximity effect to generate the observed nuclease-like sequence-selective cleavage. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ob00123k

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Enthalpy-driven nuclease-like activity and
mechanism of peptide–chlorambucil conjugates†
Robin C. K. Yang,‡a Jonathan T. B. Huang,‡a Yu-Ling Chen,a Chia-Chun Hung,a
Mokai Liao,a Wen-Chen Yao,a Chiu-Heng Chen,a Chien-Chung Liou,a
Michael J. Waringb and Leung Sheh*a
Cite this: Org. Biomol. Chem., 2014,
12, 4890
We report the results of attaching the anticancer drug chlorambucil (CLB) to two high-affinity DNA
binding peptides: Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HyM-10) and Gln-Hyp-Arg-Lys-(Py)4-Lys-Arg-
NH2 (HyQ-10). These CLB–peptide conjugates cleave DNA very effectively and sequence-selectively
without the use of chemicals, heat, or UV irradiation. Polyacrylamide gel electrophoresis identifies the
sites where CLB–HyM-10 and CLB–HyQ-10 attack a complementary pair of 5’-32P-labeled duplexes
derived from pBR322 in the absence of piperidine or other chemical additives. Matrix-assisted laser de-
sorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) has confirmed the preferential
cleavage sites as well as a novel stepwise cleavage mechanism of sequence-selective DNA cleavage.
Resembling restriction endonucleases, the CLB–peptide conjugates appear to be capable of producing
double strand DNA breaks. Circular dichroism studies show that CLB–HyM-10 and CLB–HyQ-10 induce
significant local conformational changes in DNA via the minor groove, possibly with dimeric binding
stoichiometry. The energetic basis of DNA binding by these conjugates has been investigated by isother-
mal titration calorimetry, revealing that the binding of both the peptides and their CLB conjugates is over-
whelmingly enthalpy-driven. The maintenance of a conserved negative binding free energy in DNA–
conjugate interactions is a crucial feature of the universal enthalpy–entropy compensation phenomenon.
The strongly enthalpy-driven binding of CLB–peptide conjugates to preferred loci in DNA furnishes the
required proximity effect to generate the observed nuclease-like sequence-selective cleavage.
Received 17th January 2014,
Accepted 29th April 2014
DOI: 10.1039/c4ob00123k
www.rsc.org/obc
from the limitations imposed by chemical additives and/or UV
irradiation.
1. Introduction
The design and synthesis of sequence-selective chemical
Chlorambucil (CLB) is an anticancer drug widely used in
nucleases is an important topic in the application of chemistry clinical practice against chronic lymphocytic leukemia, non-
to biological and medical sciences. Most DNA cleavage agents Hodgkin’s lymphoma, and other types of cancer.6,7 CLB alkyl-
or chemical nucleases are metal complexes that mediate ates DNA preferentially at the guanine-N7 position. However,
strand cleavage via various reactive oxygen species.1–4 Others CLB and CLB–peptide conjugates do not generally cleave
which operate by photo-cleavage require strong UV irradiation DNA in the absence of additives like piperidine. In 1995, we
to generate free radicals.5 It has long been the aspiration of reported the synthesis of cyclic peptide–chlorambucil conju-
chemists and bioorganic researchers to design and synthesize gates endowed with DNA-alkylating activities that could cleave
chemical nucleases that are capable of cleaving nucleic acids DNA in the presence of piperidine.8 We also synthesized
with sequence specificity mimicking restriction enzymes, free dipeptide conjugates that were disubstituted with chlorambu-
cil and a 2,6-dimethoxyhydroquinone derivative, albeit requir-
ing ferrous ions or high temperatures (thermal cleavage) to
elicit DNA cleavage.9
Recently, we found that when chlorambucil is conjugated to
certain arginine-containing peptides, the conjugate is capable
of cleaving DNA at 37 °C or room temperature in the absence
of piperidine or other cleavage additives. Early studies com-
menced with the attachment of chlorambucil to a peptide with
weak DNA binding ability to afford the conjugate: CLB–
Asn-Arg-Arg-Ala-Asn-Ala-NH2 (CLB–NR6). Although the parent
aDepartment of Chemistry and Life Science Center, Tunghai Christian University,
Taichung, Taiwan 407, Republic of China. E-mail: Lsheh@thu.edu.tw;
Fax: +886-4-23590426; Tel: +886-4-23592674
bDepartment of Pharmacology, University of Cambridge, Tennis Court Road,
CB2 1PD England, UK
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c4ob00123k
‡These authors contributed equally to this work.
4890 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
peptide Asn-Arg-Arg-Ala-Asn-Ala-NH2 (NR6) did not prove posi- dues (Py) to XP(Hyp)RK peptides enhances their sequence-
tive in DNA footprinting studies,10 its CLB-conjugate did show specificity toward sequences containing consecutive arrays of
remarkable DNA cleavage in the absence of chemical additives A or T nucleotides.11,14–16
(Fig. 1A). In the search for CLB-conjugates with higher DNA
Here we report that peptides HyM-10 and HQ-10 bind satis-
sequence-selective cleavage capabilities, we went on to attach factorily to DNA in a sequence-selective manner as assessed
CLB to two designed peptides incorporating the XP(Hyp)RK by DNase I footprinting. Polyacrylamide gel electrophoresis is
motif: Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HyM-10) and Gln- ideal for exploring the sequence-selectivity of DNA cleavage by
Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HQ-10). Peptides incorporating the CLB–peptide conjugates CLB–HyM-10 and CLB–HyQ-10. In
such motifs possess excellent DNA binding capability in the addition, circular dichroism and isothermal titration calori-
sub-micromolar concentration range.11–16 Further attachment metry are employed to investigate induced DNA conformational
of one or more 4-amino-1-methylpyrrole-2-carboxylic acid resi- changes as well as the thermodynamic aspects of CLB–peptide
Fig. 1 Plots comparing the susceptibility of 158-mer (upper strand) and 135-mer (lower strand) of the pBR322 DNA fragment to cleavage induced
by CLB–peptide conjugates, or by DNase I (footprinting) in the presence of peptides: A, CLB–NR6; B, CLB–HyM-10; C, peptide HyM-10; D, CLB–
HyQ-10; E, peptide HyQ-10. The vertical dotted lines in C and E denote the positions of interstrand bidentate interactions between peptide moieties
and bases of DNA. F and G: representative cleavage of the pBR322 DNA fragment induced by CLB–HyM-10 and CLB–HyQ-10 at 900 nM and
1800 nM, respectively.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4891

View Article Online
Paper
Organic & Biomolecular Chemistry
conjugate–DNA interactions, respectively. Recall that the ener- paring the sequence-binding preferences from footprinting
getic basis of general DNA–peptide interactions has been dis- studies (Fig. 1C and 2) and the sequence cleavage preferences
cussed in detail recently.16 We have also used matrix-assisted (Fig. 1B and 2), it appears that there are correlations. On the
laser desorption ionization time-of-flight mass spectrometry upper strand, the sequence preferences of the peptide HyM-10
(MALDI-TOF-MS) to investigate preferential cleavage sites correspond to U84-90 and U104-115 whereas the corres-
as well as the mechanism of DNA cleavage by these ponding cleavage by CLB–HyM-10 occurs at U81-85, with
conjugates.
common loci of 5′-AAA-3′ and 5′-GAAAA-3′, respectively. This
implies that the cleavage preference of the conjugate CLB–
HyM-10 on the upper strand is directed by preliminary
sequence-selective binding of the peptide portion of the conju-
gate to DNA. The correlation between binding and cleavage on
2. Results and discussion
Early DNA cleavage studies were carried out using the conju- the lower strand is much weaker; only one common locus is
gate CLB–Asn-Arg-Arg-Ala-Asn-Ala-NH2 (CLB–NR6). The design found at position L82-80, corresponding to the sequence
of the peptide portion NR-6 (Asn-Arg-Arg-Ala-Asn-Ala-NH2) was 5′-CAC-3′. The sequence-selective cleavage pattern seen in the
derived from residues 30–35 of tumor necrosis factor-α CLB–peptide conjugate CLB–Gln-Hyp-Arg-Lys-(Py)4-Lys-Arg-
(TNFα).17 Peptide NR-6 alone produced agarose gel retardation NH2 (CLB–HyQ-10) (Fig. 1D and 1E; ESI [I]†) appears to be
as well as some DNA nicking effects on the pBR322 plasmid. similar to that in conjugate CLB–HyM-10.
However, it did not produce any DNase I footprints on test
We envisage that the peptide moieties of the conjugate
DNA duplexes,10 suggesting that its binding to DNA must be molecules interact with the recognition motifs d(AAAA)–d-
quite weak. On the 135-mer lower strand of pBR322 the CLB– (TTTT) or d(AAA)–d(TTT) in the DNA minor groove via mono-
peptide conjugate CLB–NR6 produced efficient DNA cleavage dentate, interstrand bidentate,14,15,18 ionic and hydrophobic
in the absence of piperidine or other cleavage additives interactions. The position of DNA cleavage is expected to lie
(Fig. 1A). The cleavage by CLB–NR6 is not particularly close to the CLB moiety. However, the observed site of cleavage
sequence-selective at high concentrations but is more evidently is often rather wide, ranging over as many as 5–9 bases since
selective at lower concentrations (500–1000 nM), notably at two the conjugate may bind fairly selectively to quite a wide
loci: positions L89-87 and L114-112, both comprising the binding locus. Nevertheless, the peptide-directed sequence
sequence 5′-GGT-3′ (Fig. 1A).
recognition must surely underlie the proximity effect directing
Next, we conjugated chlorambucil to the N-terminal of the CLB moiety to the preferred cleavage site(s). It is also poss-
the peptide carboxamide Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 ible that the conjugate molecules may bind to recognition
(HyM-10) which shows strong DNA sequence-selective binding motifs in the duplex in dimeric form, either simultaneously or
to d(AAAA)–d(TTTT) or d(AAA)–d(TTT). The resulting conjugate as monomers at close intervals.
CLB–Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2
(CLB–HyM-10)
To investigate the stoichiometry and possible conformation-
induced significant DNA cleavage at four major sites on the al changes associated with the peptide–DNA interactions, a
lower strand at positions L80-78, L87-85, L92-88, L101-96 and 13-mer deoxyribonucleotide duplex d(TAGGAGAAAATAC)–d-
L115-111, corresponding to the sequences 5′-CAC-3′, 5′-TAT-3′, (GTATTTTCTCCTA) (U4A–L4T) corresponding to the recog-
5′-TGCGGTA-3′, 5′-ACGCA-3′, and 5′-GGTA-3′ (Fig. 1B, 2 and nition site in the pBR322 fragments at position 103–114 was
Table 1). Notably, on the 135-mer lower strand, the conjugate used as a substrate for CD studies. On raising the concen-
CLB–HyM-10 produces sequence-selective cleavages at position tration of the peptide HyM-10 (Fig. 3A) or HyQ-10 (ESI [II]†)
L80-78, corresponding to the sequence 5′-CAC-3′, position L88- added to the duplex U4A–L4T, a near-isoelliptic point around
86, corresponding to the sequence 5′-GTA-3′, and position 280 nm was identified, suggesting a predominantly two-com-
L114-112, corresponding to the sequence 5′-GGT-3′. It is strik- ponent binding process. This reflects the development of a
ing that the cleavage site around position L80-78 containing strong, broad dose-dependent CD enhancement around
the sequence 5′-CAC-3′ disappears at a conjugate concentration 330 nm, clearly visible in the difference spectrum (Fig. 3B) for
higher than 1000 nM (Fig. 2); this result is fully reproducible the conjugates CLB–HyM-10 and CLB–HyQ-10 as well.
when the experiment is repeated.
A number of previous studies19,20 have demonstrated that
A strong cleavage site also occurs around position L72-68 molecules which bind in the minor groove of DNA typically
but was truncated in the autoradiograph. On the other hand, induce the appearance of a positive CD band around 320 nm.
four major DNA cleavage sites are found at the upper strand The same seems to be the case for the peptides and CLB con-
positions: U76-80, U81-85, U98-106, and U107-110, corres- jugates studied here. Previous studies have also indicated that
ponding to the sequences 5′-CGGTG-3′, 5′-TGAAA-3′, at ligand/DNA ratios between 0.5 and 2.0 most likely one mole-
5′-GCGTAAGGA-3′, and 5′-GAAAA-3′ (Table 1). The most pro- cule of each peptide binds to loci such as d(AAAA)–d(TTTT) in
nounced cleavage sequences in the lower strand appear to be the minor groove, as suggested by a hint of a small plateau,
5′-GGTA-3′ and 5′-ATG-3′. However, it is notable that an while at ratios between 2 and 3 further small plateau regions
unusual cleavage sequence also occurs in the lower strand: can be tentatively discerned, suggesting that two molecules
5′-CAC-3′. On the other hand, the strongest cleavage sequences in can bind in a dimeric fashion to the minor groove. As the
the upper strand appear to be 5′-GGAG-3′ and 5′-AAAA-3′. Com- [ligand]/[duplex] ratio rises above 3 progressive increases in Δθ
4892 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 Autoradiographs showing cleavage of 158-mer (left panels) and 135-mer (right panels) of the pBR322 DNA fragment induced by DNase I
(footprinting) in the presence of peptide HyM-10 (upper panels), or induced by incubating with the conjugate CLB–HyM-10 in 5 mM sodium caco-
dylate buffer (pH 6.5) at 37 °C for 30 min (lower panels).
suggest that further ligand molecules start to bind non-specifi- HyQ-10 and CLB–HyQ-10 is 2 : 1. It is likely that the peptide
cally – probably to the wide major groove. All these features are HyM-10 and its conjugate bind to the oligonucleotide
apparent in the CD behavior of the peptides and conjugates duplex U4A–L4T rather non-selectively at higher ligand con-
reported here (Fig. 3A–E; Fig. [II], ESI†). In this study, Job’s centrations, a common DNA-binding property shared by many
plots (Fig. 3E) show that the ligand/DNA ratio of the peptide of our synthetic peptides incorporating the XP(Hyp)RK
HyM-10 and its conjugate CLB–HyM-10 is 3 : 1 whereas that of motif.13–15
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4893

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Sequence-selective binding and cleavage by peptides and CLB–peptide conjugates on complementary 5’-32P-labeled upper (158-mer)
and 5’-32P-labeled lower (135-mer) DNA strandsa
Position of interstrand
bidentate interactions
Sequence-selective cleaved fragments
of CLB–peptide conjugates
Sequence-selective binding of peptides
Ka
nH
HyM-10
5′-AAATACCG-3′
(U84-90)
5′-GGAGAAAATACCG-3′
(U104-115)
5′-TTTCACAC-3′
(L85-78)
CLB–HyM-10
5′-CGGTGTGAAA-3′
(U76-85)
5′-GTAAGGAGAAAA-3′
(U100-111)
5′-TATG-3′ (L73-70)
5′-CACCG-3′ (L80-77)
5′-GTGCGGTA-3′
(L93-86)
3.5 × 106
3.4 × 106
5.0 × 107
1.0 × 107
1.1
2.4
1.9
2.1
84–85
105–111
5′-TTTTCTCCTTA-3′
(L111-101)
5′-GGTA-3′ (L115-112)
HyQ-10
5′-AAATACCG-3′
(U83-90)
5′-GAAAATACCG-3′
(U107-116)
5′-ATTTCAC-3′
(L86-80)
CLB–HyQ-10
5′-GGTGTGAA-3′
(U77-85)
5′TAAGGAGAAAA
(U101-111)
5′-GGCGCT-3′(U121-127)
3.8 × 106
5.6 × 106
2.9 × 106
5.7 × 106
2.7
3.6
3.7
1.8
83–86
107–112
5′-ATTTTCTCCTT-3′
(L112-102)
5′-GTGCGGTA-3′
(L93-86)
a Ka and nH are the apparent association constant and the Hill coefficient determined from concentration-dependent DNase I footprinting
studies, respectively. The binding site positions on the upper and lower strands are abbreviated as U and L, respectively.
Although the mechanism of DNA cleavage by established incubation with CLB–HyM-10 or CLB–HyQ-10 for 15–30 min at
alkylating agents is generally understood, much less is known room temperature or 37 °C, in the complete absence of special
about the complex mechanisms of action of designed agents. chemical additives. The CLB–peptide conjugates can also
To gain insight into the sequence-selective DNA cleavage cleave single strand oligonucleotides (data not shown) as well
mechanism of CLB–peptide conjugates we carried out MALDI- as double strand oligonucleotides. The cleavage of CLB–
TOF analyses of the cleavage fragments produced from a HyM-10 and CLB–HyQ-10 on the U4A–L4T duplex (Fig. 4A, B,
double-stranded oligonucleotide duplex d(TAGGAGAAAATAC)– E, F; Fig. [III], A, B, ESI†) is much less sequence-selective than
d(GTATTTTCTCCTA) (U4A–L4T), employing 3-hydroxy-picoli- that on the XU4A–XL4T duplex (Fig. 4C, D; Fig. [III], C,D,
nic acid/picolinic acid/ammonium citrate or trihydroxyaceto- ESI†). The major cleavage fragments most frequently detected
phenone/ammonium citrate as the matrix. Another (i.e. those observed at least twice in different duplex/CLB–
oligonucleotide duplex was also employed in MALDI-TOF ana- peptide conjugate reactions) are shown in Table 2, where the
lyses: d(AAGGAGAAAATAC)–d(GTATTTTCTCCTT) (XU4A–XL4T). preferential sites of cleavage and the corresponding cleavage
The sequence of this duplex was designed to verify the site of mechanisms are proposed. Two routes to hydrolysis of the
cleavage when two or more cleavage fragments have the same phosphodiester bond may be involved: depurination or depyri-
calculated mass.
midination. For DNA glycosylase, depurination involving an
MALDI-TOF analyses confirm that significant sequence- oxocarbocation intermediate that leads to subsequent hydro-
selective cleavage of the oligonucleotide duplexes occurs after lytic cleavage of the 5′- or 3′-phosphodiester bonds is well
4894 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 3 A: CD spectrum for the titration of the duplex U4A–L4T versus the peptide HyM-10 at peptide concentrations of 0.5, 1.0, 2.2, 2.4, 2.6, 2.8,
3.0, 3.2, 3.4, 4.0, 5.0, 7.0 μM at 37 °C. B: Corresponding CD difference spectra with the contribution of free duplex and peptide HyM-10 subtracted.
C: titration of U4A-L4T versus conjugate CLB–HyM-10 at conjugate concentrations of 0.5, 1.0, 1.5, 1.8, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 3.0, 3.5,
4.0, 5.0 μM at 37 °C. D: Corresponding difference spectra with the contribution of the free duplex and conjugate CLB–HyM-10 subtracted. E: Job’s
plots of Δθ·X versus mole fraction ligand at 322 nm (X = mole fraction ligand) for the titration of the duplex U4A–L4T versus different ligands. The
stoichiometry of ligands HyM-10, HyQ-10, CLB–HyM-10, and CLB–HyQ-10 binding to the duplex U4A–L4T is determined to be 3 : 1, 3 : 1, 2 : 1, and
2 : 1, respectively.
characterized.21 Since our CLB–peptide conjugates can induce recently been reported as a concerted bimolecular process, invol-
significant depurination and depyrimidination with concomi- ving nucleophilic attack of the 2-deoxyribosyl oxocarbenium
tant cleavage of the phosphodiester bond, a similar DNA clea- ion.22 In Table 2, the MALDI-TOF MS peaks at 998, 1948, 2337,
vage mechanism is proposed in Fig. [IV], ESI.† In addition, a and 2643 reveal the corresponding oligonucleotide fragments
mechanism of phosphodiester hydrolysis by thymidine phos- preserving purine or pyrimidine bases on the 5′- and 3′-ends.
phorylase via depyrimidination of thymine nucleotides has However, depurination and depyrimidination may occur on
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4895

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 4 Major cleavage fragments induced by conjugates CLB–HyM-10 (A) and CLB–HyQ-10 (panel B) on the oligonucleotide duplex U4A–L4T;
major cleavage fragments induced by conjugates CLB–HyM-10 (C) and CLB–HyQ-10 (D) on the oligonucleotide duplex XU4A–XL4T as detected by
MALDI-TOF mass spectroscopy; representative MALDI-TOF mass peaks of major cleavage fragments induced by conjugates CLB–HyM-10 (E) and
CLB–HyQ-10 (F) on the oligonucleotide duplex U4A–L4T. The sequences of the upper and lower strands of duplexes U4A–L4T and XU4A–XL4T
are shown in Table 2.
the 5′- or 3′-end of the adjacent fragment originally linking the supported by ionic and hydrophobic interactions, and thus the
cleaved fragment, causing hydrolysis of the corresponding CLB moiety is positioned close to certain DNA loci. After tight
phosphodiester bonds.
binding, the resulting phosphodiester bond cleavage must be
To interpret the sequence-selective cleavages we propose an dominated by proximity to the CLB moiety, which mediates
initial binding alignment of the peptide–CLB conjugate with depurination or depyrimidination of a base leading to sub-
the oligonucleotide duplex (Fig. 5A–D). The peptide moiety of sequent hydrolysis of the DNA backbone.
the conjugate first interacts with the recognition locus
The use of the XU4A–XL4T duplex unambiguously verifies
d(AAAA)–d(TTTT) in the DNA minor groove via monodentate the site of cleavage as well as confirms the validity of the
and interstrand bidentate hydrogen bonding interactions, assignment of the cleavage fragments. For example, the 5′-end
4896 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 Cleavage fragments most frequently detected in MALDI-TOF experiments with DNA duplexes and bound CLB–peptide conjugates
Mass
998
Cleavage fragments
Proposed cleavage mechanism
Depyrimidination of T17
1442
1598
1732
1948
Depurination of A5
Depurination of A16 and depyrimidination of T22
Depurination of G6
Depyrimidination of C21 and depurination of G14
2033
2337
Depyrimidination of T18 and T25
Depurination of G3 and depyrimidination of T11
2643
2944
Depyrimidination of T1 and depurination of A10
Depurination of G14 and depyrimidination of C24
For details of MALDI-TOF mass fragments, see Fig. 4A–F and ESI, Tables [Ia]–[Id].
nucleotide on the upper strand of U4A–L4T is T1 whereas that in a dimeric fashion to this truncated duplex locates a CLB
of XU4A–XL4T is A1 (Fig. 4A–D). The cleavage fragments moiety close to preferred sites and initiates cleavage of the
induced by the conjugate CLB–HyQ-10 are f1(mass 1442.2) and duplex releasing four or five single strand fragments. Depuri-
c3(mass 2177.6) from the upper strand of U4A–L4T, corres- nation of A5 or G6 in the upper strand and depyrimidination of
ponding to the fragments f1(mass 1451.2) and c3(mass 2186.9)
T22 and T20 in the lower strand are indicated by the production
from the upper strand of XU4A–XL4T (Fig. 4B and D), respect- of fragments from opposite strands: 5′-T1-A2-G3-G4-A5-dr-3′
ively, indicating that the site of phosphodiester bond cleavage (mass 1732.3), and cleaved fragments: 5′-T17-T18-T19-T20-C21-3′
by the CLB–peptides on different duplexes that differ by a (mass 1597.9) and 5′-T1-A2-G3-G4-dr-3′ (mass 1442.2) (Fig. 5A).
single nucleotide is the same.
The fragment 5′-T19-T20-C21-T22-C23-C24-dr-3′ (2033.1) is gener-
Fig. 5A–5D illustrate a novel “stepwise cleavage mechanism” ated by alignment of the CLB conjugate to the duplex near T18,
of the DNA duplex by the conjugate CLB–HyM-10 or CLB– causing depyrimidination of this nucleotide and subsequent
HyQ-10 on the U4A–L4T duplex. The proposed mechanisms hydrolysis of the 5′-T18-T19-3′ phosphodiester bond (Fig. 5A).
can be explained by successive association and dissociation Stepwise cleavage of the upper strand or lower strand initiated
processes directing the proximity of the CLB moiety to another by successive binding by a single molecule of this conjugate to
susceptible base, producing depurination or depyrimidination, a preferred cleavage site affords other major fragments: 5′-G4-
and generating smaller fragments after each cut. For the conju- A5-G6-A7-A8-A9-A10-3′ (2337.3), 5′-A2-G3-G4-A5-G6-A7-A8-A9-3′
gate CLB–HyM-10, different binding alignment of a single (2643.2), and 5′-G14-T15-A16-3′ (998.3).
CLB–HyM-10 molecule with the duplex would result in the
Similarly, binding alignment of a single molecule of the
removal of 5′-T25-A26-3′ by depyrimidination of T25 and also the conjugate CLB–HyQ-10 to the duplex U4A–L4T first removes
truncating of 5′-G14-T15-A16-3′ by depurination of A16 (Fig. 5A). the dinucleotide 5′-G14-T15-3′ or a G14 nucleotide from the
Next, binding of two molecules of the CLB–peptide conjugate lower strand, followed by dimeric binding of two molecules
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4897

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 5 (A) Proposed stepwise cleavage of the duplex U4A–L4T induced by the chlorambucil–peptide conjugate CLB–HyM-10 via an initial mono-
meric binding, followed by a dimeric binding mode, resulting in the generation of fragmental strands from opposite strands. (B) Proposed stepwise
cleavage of the duplex U4A–L4T induced by the conjugate CLB–HyQ-10 via a similar dimeric binding mode. Before DNA cleavage, the peptide moi-
eties of the conjugate first interact with the DNA minor groove via monodentate, interstrand bidentate, ionic and hydrophobic interactions (non-
covalent interactions not shown). (C) Proposed stepwise cleavage mechanism of the duplex XU4A–XL4T induced by the chlorambucil–peptide con-
jugate CLB–HyM-10. (D) Proposed stepwise cleavage mechanism of the duplex XU4A–XL4T induced by the conjugate CLB–HyQ-10. X, M, Q, Ph, Py
and dr represent residues of chlorambucil, methionine, glutamine, hydroxyproline, 4-amino-1-methylpyrrole-2-carboxylic acid, and deoxyribose,
respectively. Position of DNA cleavage is proposed to locate near the CLB moiety and is denoted by the red arrow sign . The underlined base indi-
cates where depurination or depyrimidination has occurred. The found mass of the cleaved fragments from MALDI-TOF experiments are as shown
in brackets.
4898 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
that initiates cleavage to afford fragments from opposite thermodynamic basis of DNA binding and cleavage by pep-
strands: 5′-T1-A2-G3-G4-A5-dr-3′ (mass 1732.7), 5′-T1-A2-G3-G4-dr- tides incorporating the XHypRK motif as well as the corres-
3′ (mass 1442.3), and 5′-T15-A16-T17-T18-T19-T20-3′ (mass 1948.6) ponding CLB conjugates, isothermal calorimetry titration
(Fig. 5B; Table [Ia] and [Ic], ESI†). Similar to the cleavage studies (ITC) were carried out (Fig. 6A and 6B). Plots of heat
mechanism of CLB–HyM-10, the fragment 5′-T19-T20-C21-T22- versus molar ratio were obtained by subtracting the heat of
C23-C24-dr-3′ (2033.9) is generated by depyrimidination of dilution for the addition of the peptide or conjugate to buffer.
nucleotide T18 and subsequent hydrolysis of the 5′-T18-T19-3′ In both CD and ITC experiments, the titrations were performed
phosphodiester bond. Stepwise cleavage of the upper strand or by adding a small aliquot of the CLB–peptide conjugate to
lower strand triggered by successive monomeric binding to a bulk DNA solution, so the corresponding detection of ellipti-
preferred site generates some other major fragments: 5′-G4-A3- city changes and enthalpy changes may be limited to relatively
G6-A7-A8-A9-A10-3′ (2337.5), 5′-A2-G3-G4-A5-G6-A7-A8-A9-3′ (2642.6), short periods of DNA-ligand binding, that is, DNA binding
5′-G14-T15-A16-3′ (998.3), 5′-A10-T11-3′ (712.4), 5′-T15-A16-T17-T18- must precede DNA cleavage. CD spectra of CLB–peptide conju-
T19- T20-C21-T22-C23-dr-3′ (2943.7), and 5′-A5-G6-A7-3′ (970.7).
gates titrated into DNA reveal significant concentration-depen-
Further support for the stepwise cleavage mechanism is pre- dent ellipticity increases around 330 nm, suggesting that the
sented by the cleavage of the duplex XU4A–XU4T by conju- CLB–peptide conjugates bind tightly to the intact minor
gates CLB–HyM-10 and CLB–HyQ-10 (Fig. 5C and D). In these groove of DNA. If the DNA is already cleaved, no corresponding
cleavage events the mechanism follows the general stepwise minor groove may be available for the concentration-depen-
cleavage pattern: an initial truncation of an upper or lower dent binding of CLB–peptide conjugates. Thus, we reason that
strand of the DNA duplex by a single molecule of the CLB– the ellipticity changes (CD) and enthalpy changes (ITC) reflect
peptide conjugate with repetitive binding to recognition sites. primary binding interactions of the CLB–peptide conjugate
Subsequently, dimeric binding of the conjugate CLB–HyM-10 with intact DNA rather than with nicked DNA, although we do
triggers the cleavage of 5′-A1-A2-G3-G4-A5-G6-A7-3′ (2186.9) and not preclude the possibility that some DNA was cleaved during
two fragmental pairs from opposite strands: 5′-A7-A8-A9-A10- ITC and CD experiments.
T11-A12-3′ (1960.3) and 5′-T15-A16-T17-T18-T19-T20-3′ (1948.5). The
The ITC results indicate that either the peptide or the CLB-
fragment 5′-T17-T18-T19-T20-C21-3′ (1597.9) was produced by trun- conjugate prefers to bind in a dimeric fashion to d(AAAA)–
cating 5′-T15-A16-3′ from the lower strand followed by depyrimidi- d(TTTT) sites. In Breslauer’s pioneering study of the binding
nation of T22 to cleave the 5′-C21-T22-3′ phosphodiester bond of netropsin, distamycin, ethidium and daunorubicin to the
(Fig. 5C). By a similar mechanism, the conjugate CLB–HyQ-10 duplex poly[dA–dT]·[dA–dT], the ΔH values lie within the range
induces the generation of fragmental pairs 5′-A1-A2-G3-G4-A5-G6- −8.9 to −18.5 kcal mol−1 23
.
In comparison, the negative
A7-3′ (2186.6) and 5′-T19-T20-C21-T22-C23-C24-dr-3′ (2033.9), and enthalpy changes for binding of the peptide and its CLB conju-
the fragment 5′-G4-A5-G6-A7-A8-A9-3′ (2024.8) (Fig. 5D). gate incorporating the XHypRK motif into DNA are signifi-
Some evidence of generation of double strand DNA breaks cantly greater at their preferred DNA binding sequences (ΔH
is demonstrated by cleavage experiments on the XU4A–XL4T lies within the range −25.4 to −74.4 kcal mol−1; Table 3).
duplex (Fig. 5C and D). Typical fragmental pairs from opposite Thus, the ITC results indicate that binding of the peptides
strands of the duplex XU4A–XL4T are produced by the conju- and CLB–peptide conjugates to DNA is overwhelmingly
gate CLB–HyM-10: 5′-A7-A8-A9-A10-T11-A12-3′ (1960.3) and 5′-T15- enthalpy-driven. However, the opposing entropic effects for all
A16-T17-T18-T19-T20-3′ (1948.3) (Table [Ib], ESI†). Similarly, frag- of the peptides and CLB–peptide conjugates are very unfavor-
mental pairs from opposite strands of the duplex XU4A–XL4T able: TΔS falls within the range −17.8 to −65.9 kcal mol−1
are generated by the conjugate CLB–HyQ-10: 5′-A1-A2-G3-G4-A5- (Table 3).
G6-A7-3′ (2186.6) and 5′-T19-T20-C21-T22-C23-C24-dr-3′ (2033.9)
The unusually strong exothermic enthalpy change could be
(Table [Id], ESI†). We reason that the CLB–peptide conjugates are due to the formation of ionic interactions between the side
capable of generating double strand DNA breaks when the CLB chains of basic residues like Arg and Lys and the DNA phos-
moiety is repeatedly located near the preferred cleavage points.
phates as well as the interstrand bidentate interactions
We suggest that the stepwise cleavage mechanism points to between the side chains of Met, Arg, and Lys of peptides and
a general feature: the truncation of an upper or lower strand the O2 of thymine and N3 of adenine, as will be considered
can be initiated by binding of a single molecule of the CLB– below. First it should be noted that the peptides and their
peptide conjugate. Next, monomeric binding of the conjugate corresponding chlorambucil conjugates clearly exhibit the
may produce more fragments whereas dimeric binding of the enthalpy–entropy compensation (EEC) phenomenon that has
conjugate to the DNA duplex may generate fragmental pairs been described before (Fig. 6B). Fig. 6C shows a plot of ΔH
from opposite strands, producing double strand breaks in the versus TΔS for four peptides and four CLB–peptide conjugates.
meantime. Stepwise cleavage of the duplex DNA can often be The linearity of ΔH versus TΔS for these eight ligands is
achieved by binding of a single molecule of the CLB–peptide impressive: slope = 1.0155; y-intercept = −7.9 kcal mol−1. A
conjugate repeatedly to various recognition sites, yielding 3-dimensional figure of ΔH, TΔS and ΔG can be plotted to
different fragments.
emphasize the linear enthalpy–entropy–free energy relation-
The energetic basis of general DNA–peptide interactions ship that underlines the EEC phenomenon in DNA–small
was discussed in detail recently.16 To gain insight into the ligand interactions (Fig. 6D). A near-straight line is suspended
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4899

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 6 (A) ITC curves for titration of decapeptides HyM-10, HyQ-10, CLB–HyM-10 and CLB–HyQ-10 into the U4A–L4T duplex at 25 °C. For each
experiment the top panel represents the raw heat of binding generated with successive additions of the peptide, and in the bottom panel the inte-
grated heat is plotted versus peptide/DNA molar ratio. Data acquisition and analysis were performed using the nonlinear least-squares fitting algo-
rithm software (Microcal Origin 7.1). (B) Plot of enthalpy (ΔH) versus entropy (TΔS) from ITC of HyM-10, HyQ-10, CLB–HyM-10 and CLB–HyQ-10
added to the U4A–L4T duplex at 25 °C (298 K). (C) Plot of enthalpy (ΔH) versus entropy (TΔS) from ITC of HyM-10 [1], HyQ-10 [2], HyH-10 [3],
HyS-10 [4], CLB–HyM-10 [1a], CLB–HyQ-10 [2a], CLB–HyH-10 [3a], CLB–HyS-10 [4a] added to the U4A–L4T duplex at 25 °C (298 K). (D) Three-
dimensional plot of ΔH, TΔS, ΔG from titration of peptides and CLB–peptide conjugates into the U4A–L4T duplex. The short vertical projection
plane (in green) corresponds to ΔG magnitudes preserved around −7.9 kcal mol−1. The figure represents a view of the ‘cube’ from below.
in thermodynamic space, with ΔG values maintained within Ionic interactions between the positively charged Arg and Lys
an almost constant range around −7.9 kcal mol−1
.
side chains of the peptides or conjugates and internucleotide
In a recent study16 as well as the work reported here, we phosphates probably restrict the conformational mobility of
have put forward the view that strong exothermic enthalpy is the DNA duplex and may account for the dramatic decrease in
needed to overcome the very unfavorable entropy required to total entropy. According to the law of conservation of energy,
attain efficient binding to DNA. In many cases, when a any loss in entropic energy must transform to exothermic
protein24 or a peptide binds to DNA, cations from the DNA are energy, which in turn causes an increase in enthalpy. We envi-
released to bulk solution, and the increase in entropy from sage that high exothermic energy is required for allosteric
solvent reorganization ΔS°H (hydrophobicity) and ΔSi°ons (poly- adjustment of DNA conformation within the binding loci by
electrolyte) favors DNA–ligand binding. However, the sum of means of hydrogen bonding and ionic interactions between
entropy change ΔS°rt (rotation and translation), ΔSv°ib the DNA duplex and the peptide,16 augmenting the favorable
(vibration) and ΔS°conf (configuration) is typically negative and energy components ΔGh, ΣΔGp, ΔGconf.
opposes binding, resulting in a decrease in total entropy.24 Williams’ equation,25 and facilitating positively cooperative
, and ΔGvdw in
4900 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 3 Comparison of thermodynamic parameters from titration of peptides, chlorambucil–peptide conjugates, and drugs versus DNA duplexes
Entry
Peptide/conjugate (mM)
[DNA] (mM)
ΔH (kcal mol−1
)
ΔS (cal mol−1
)
ΔG (kcal mol−1
)
TΔS (kcal mol−1
)
Ka (M−1
)
1
2
3
4
1a
2a
3a
4a
HyM-10 (0.15)
HyQ-10 (0.15)
HyH-10 (0.05)
0.0033
0.01
−50.5
−74.4
−51.2
−31.2
−25.4
−46.0
−30.4
−26.6
−11.2
−18.5
−140
−221
−142
−74.5
−56.3
−126
−75
−8.8
−8.6
−8.8
−9.0
−8.6
−8.3
−8.0
−7.8
−12.7
−12.6
−41.7
−65.9
−42.3
−22.2
−17.8
−37.5
−22.4
−18.8
—
2.9 × 106
1.9 × 106
2.9 × 106
4.2 × 106
2 × 106
1.2 × 106
7.7 × 105
5.3 × 105
—
0.0033
0.006
0.005
0.007
0.005
0.005
—
HyS-10 (0.15)
CLB–HyM-10 (0.15)
CLB–HyQ-10 (0.15)
CLB–HyH-10 (0.15)
CLB–HyS-10 (0.15)
Netropsin
−63.1
+5
Distamycin A
—
−20
—
—
Peptides HyM-10, HyQ-10, HyH-10, HyS-10 and the corresponding CLB–peptide conjugates were titrated against the oligonucleotide duplex (U4A–
L4T) employing isothermal titration calorimetry. Netropsin and distamycin A were titrated versus the copolymer duplex poly[d(A–T)]·poly[d(A–T)]
using batch calorimetry.22 The amino acid sequence of the peptides are: Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HyM-10), Gln-Hyp-Arg-Lys-(Py)4-
Lys-Arg-NH2 (HyQ-10),
His-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HyH-10),
Ser-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HyS-10).
Conjugation
of
chlorambucil to the N-terminal of these peptides afforded the corresponding CLB–peptide conjugates.
binding of the approaching peptide or CLB–peptide molecules. these peptides and conjugates is evidently required for a
These considerations formalize a situation in which the process of energy compensation, needed to sustain an accepta-
strongly negative binding enthalpy for these peptides and CLB- ble negative value of net binding free energy. We believe that
conjugates is required for the compensation of energy, sustain- this strongly enthalpy-driven binding of the peptide conjugate
ing an acceptable negative value of net binding free energy ΔG to DNA leads to the proximity effect that positions the CLB
mandatory for DNA–ligand interactions.
moiety close to the preferred cleavage site, manifested in the
observed nuclease-like sequence-selective cleavage.
The highly effective cleavage of DNA by the CLB–peptide
conjugates that we have described, in the absence of UV
irradiation, chemical additives or thermal assistance, mimics
3. Conclusion
This study shows that chlorambucil–peptide conjugates elicit the action of natural nucleases. To our knowledge, this new
significant sequence-selective cleavage of DNA without the series of CLB–peptide conjugates is one of the most effective
chemical additives or thermal assistance that are usually sequence-selective chemical nucleases reported. As such, the
required for chemical nucleases. Gel electrophoresis experi- conjugates and their congeners merit further investigation.
ments reveal that although the cleavage sites are mostly adjacent
to A and G nucleotides, some sequence-selective cleavages also
occur on the 5′- or 3′-end of pyrimidines such as 5′-CAC-3′,
5′-CGGTG-3′ and 5′-TGAA-3′. MALDI-TOF studies of cleavage
fragments from duplexes U4A–L4T and XU4A–XL4T indicate
4. Experimental
4.1. Chemicals and biochemicals
preferential cutting at comparable sites. Notably, MALDI-TOF All the protected amino acid derivatives were purchased from
studies of conjugates CLB–HyM-10 and CLB–HyQ-10 versus the AnaSpec, Inc. (San Jose, CA) and all peptides are synthesized
two DNA duplexes, a total of four reactions, all point to a novel in our laboratory. All other analytical reagents were purchased
stepwise scissioning mechanism. Generation of a particular from Acros, Tedia or Sigma. Melting points were determined
sequence-selective cleavage fragment can often be interpreted on Mel-Temp apparatus (Cambridge, Mass) and are uncor-
as resulting from binding of the CLB–peptide conjugate to its rected. Optical rotations were determined on a Rudolph
cognate recognition site on DNA so as to place the CLB moiety Autopol II instrument. Radiolabeled nucleoside triphosphates
close to a susceptible base, causing depurination or depyrimi- [γ-32P]dATP were obtained from NEN Life Science Products at a
dination, and inducing the hydrolysis of a phosphodiester specific activity of 6000 Ci mmol−1. Taq polymerase, T4 poly-
bond, generating smaller fragments after each cut. The general nucleotide kinase, and DNase I were purchased from Promega.
features of the stepwise scissioning mechanism are: following All of the biochemicals were used according to the supplier’s
the initial truncating of an upper or lower strand of the DNA recommended protocol in the activity buffer provided. Other
duplex by a single molecule of the CLB–peptide conjugate, chemicals were analytical grade reagents, and all solutions
monomeric binding of the conjugate may produce more frag- were prepared using doubly deionized, Millipore filtered water.
ments whereas dimeric binding of the conjugate to the DNA Semi-preparative and analytical HPLC (Vydac reversed-phase
duplex may generate fragmental pairs from opposite strands, columns, TP201; column 1, 1 × 25 cm; column 2, 0.4 × 25 cm)
producing double strand breaks in the meantime.
were performed using a Hitachi L-7100 pump equipped with a
It also appears that the binding of CLB–peptide conjugates gradient elution device and a Soma S-3702 variable wavelength
to the recognition locus on DNA is overwhelmingly enthalpy- UV detector which is connected to a PC computer installed
driven. The strongly negative binding enthalpy measured for with the Hitachi HPLC analytical software. Mass spectra were
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4901

View Article Online
Paper
Organic & Biomolecular Chemistry
recorded with a Finnigan/Thermo Quest MAT 95XL instrument strand 5′-32P-labeled) were prepared by PCR amplification in a
operating in the electrospray ionization (ESI) mode in National thermal cycler (ABI model 9700) as previously described. The
Chung-Hsing University.
concentrations of DNA as determined by UV spectroscopy were
around 10−7 M. Reactions were conducted in a total volume of
10 μL. Briefly, radiolabeled DNA (2 μL) was mixed with varying
4.2. Chemical methods
Synthesis of peptides and chlorambucil–peptide conjugates concentrations of 2 μL of CLB–HyM-10 or CLB–HyQ-10 dis-
Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (HyM-10). This peptide solved in 5 mM sodium cacodylate buffer (pH 6.5) and the
was synthesized using the solid phase methodology by manual reaction was allowed to proceed at 37 °C for 30 min, cooled on
operation of a Protein Technology PS3 peptide synthesizer. ice, ethanol-precipitated, and lyophilized. Samples were heated
The first Fmoc-protected amino acid was coupled to the Nova at 90 °C for 4 min prior to electrophoresis.
Rink amide AM resin using PyBOP/NMM in DMF. All of the
4.5. MALDI-TOF mass spectrometry
Nα-Fmoc-protected amino acids (in 4 equivalent ratio excess to
the resin) were coupled in a stepwise fashion using PyBOP/ Measurements were performed with a Microflex MALDI mass
NMM in DMF after deprotection of the N-Fmoc group by piper- spectrometer (Bruker Daltonik GmbH, Leipzig, Germany).
idine. The side chains of Arg and Lys were protected by the 20 μL 40 μM U4A–L4T (duplex) in 5 mM sodium cacodylate
Pmc and Boc groups, respectively. After coupling the last N- buffer, pH 6.5 and 20 μL 4 μM CLB–peptide conjugate in the
terminal Fmoc-amino acid residue, the resin was treated with same buffer were mixed with a pipette and vortexed for 3
the cleavage reagent (0.75 g phenol, 10 mL TFA, 0.5 mL thio- seconds, and centrifuged at 10 000 rpm for 10 seconds. The
anisole, 0.25 mL EDT) for 1.5 h, and then lyophilized. The mixture was incubated at 37 °C for 30 min. To this was added
resin was washed with dry ether (2 × 30 mL), filtered, and then 1 mL 99% alcohol and 5 μL NaOAc (3M, pH 5.2), and the
washed with 5% acetic acid (200 mL). The combined filtrate mixture was centrifuged at 10 000 rpm for 10 seconds. The
was lyophilized and the product was purified by semi-prepara- product was stored at −20 °C overnight, and then centrifuged
tive reversed-phase HPLC (column 1) using gradient elution. at 4 °C at 14 000 rpm for 30 min. The ethanol was removed
Eluent A: 5% MeCN, 95% H2O, 0.1% TFA; eluent B, 95% and 1 mL ethanol was again added and the mixture was centri-
MeCN, 5% H2O, 0.1% TFA. A linear gradient was achieved by fuged at 4 °C at 14 000 rpm for 10 min. This treatment was
increasing the MeCN content from eluent A to eluent B in repeated twice and finally the ethanol was removed using a
30 minutes. tR (column 2), 15.09 min, m.p. 148–151 °C, [α]2D7 speedvac (800 rpm) at 25 °C for 5 min. 1 mL 0.5 M trihydroxy-
−25.56 (c 0.30, MeOH–H2O, 1 : 1); ESIMS requires: 1318.56, acetophenone (THAP) in methanol was prepared. The matrix
found: 1319.0.
was prepared by adding 40 μL 0.5 M diammonium hydrogen
CLB–Met-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (CLB–HyM-10). The citrate (H2O) to 100 μL 0.5 M trihydroxyacetophenone (THAP)
Fmoc group of Fmoc-Met-Hyp-Arg(Pmc)-Lys(Boc)-(Py)4-Lys in methanol. 2.5 μL of the matrix was introduced in the plate
(Boc)-Arg(Pmc)-resin was removed with 20% piperidine in and allowed to dry for 3 min in air. To this matrix layer was
DMF and chlorambucil and PyBOP/NMM (in 4 equivalent ratio added 1 μL of a mixture of the fresh matrix and the reacted
excess to the resin) were added. The reaction was allowed to DNA sample (1 : 1). After 3 min the sample was subjected to
react for about 3 h, and the resin was treated with the cleavage MALDI-TOF analysis with 45% laser power, 400 ns delay time,
reagent (0.75 g phenol, 10 mL TFA, 0.5 mL thioanisole, 100 shots, using a linear and positive ion mode.
0.25 mL EDT) for 1.5 h, and then lyophilized and purified as
4.6. Circular dichroism (CD) studies
165–169 °C, [α]2D4 −32.14 (c 0.9, H2O); ESIMS calcd 1604.75, CD spectra were measured at 37 °C with a Jasco J-815 instru-
for peptide HyM-10. tR (column 2), 18.91 min, m.p.
found: 1605.0.
ment in the Institute of Chemistry, Academia Sinica. The
CLB–Gln-Hyp-Arg-Lys-(Py)4-Lys-Arg-NH2 (CLB–HyQ-10). The duplex DNA was adjusted to 1.0 μM in 5 mM sodium caco-
Fmoc group of Fmoc-Gln-Hyp-Arg(Pmc)-Lys(Boc)-(Py)4-Lys dylate buffer (pH 6.5) and peptides, dissolved in the same
(Boc)-Arg(Pmc)-resin was removed with 20% piperidine in buffer, were added to maintain final concentrations of 0.2, 1.0,
DMF and chlorambucil and PyBOP/NMM (in 4 equivalent ratio 2.2, 2.4, 2.6, 2.8, 3.0, 3.2, 3.4, 4.0, and 5.0 μM. Spectra were
excess to the resin) were added. The reaction was allowed to recorded after 60 min incubation at 37 °C.
react for about 3 h, then worked up and purified as for CLB–
4.7. Isothermal titration calorimetry
HyM-10. tR (column 2), 17.32 min, m.p. 163–166 °C, [α]D24
−23.08 (c 0.87, H2O); ESIMS calcd 1599.8, found: 1601.0.
ITC experiments were performed using a Microcal VP-ITC
(MicroCal, Northampton, MA) with a reaction cell volume of
2 mL for oligonucleotide duplex (U4A–L4T) at 25 °C or other
4.3. DNase I footprinting
Full protocols for footprinting experiments have been pub- temperatures as indicated in Table 3 and all samples were
lished previously.11,14,15
degassed under vacuum for 5 min. The DNA solutions at
appropriate concentrations in 5 mM sodium cacodylate, pH
6.5 (Table 3) were placed in the calorimeter cell and the
4.4. DNA cleavage induced by CLB–peptide conjugates
Radiolabeled pBR322 DNA fragments: 158-mer DNA duplex peptide at an appropriate concentration in 5 mM sodium caco-
(upper strand 5′-32P-labeled) and 135-mer DNA duplex (lower dylate, pH 6.5 (273 μL) was placed in the titration syringe and
4902 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
injected into aliquots of 7 μL with 220 s intervals between the
individual injections and stirring at 304 rpm, for a total of 39
injections. Data acquisition and analysis were performed using
the nonlinear least-squares fitting algorithm software (single
site binding model, Microcal Origin 7.1 software). The Ka
values were obtained from the ITC instrument by computer fit
of the ITC isotherms and the corresponding ΔG values were
calculated from the equation:
with chlorambucil as primary therapy for chronic lympho-
cytic leukemia, N. Engl. J. Med., 2000, 343, 1750–1757.
8 L. Sheh, H. W. Chang, C. W. Ong, S. L. Chen, C. Bailly,
R. C. M. Linssen and M. J. Waring, Synthesis, DNA binding,
and sequence specificity of DNA alkylation by some novel
cyclic peptide-chlorambucil conjugates, Anti-Cancer Drug Des.,
1995, 10, 373–388.
9 A. Minnock, L. S. Lin, J. Morgan, S. D. G. Crow,
M. J. Waring and L. Sheh, Sequence-specific DNA cleavage
by dipeptides disubstituted with chlorambucil and 2,6-
dimethoxyhydroquinone-3-mercaptoacetic acid, Bioconju-
gate Chem., 2001, 12, 870–882.
ΔG ¼ ꢀRTꢁln Ka
Acknowledgements
10 C. Bailly, Personal communications.
11 J. C. Chang, C. H. Yang, P. J. Chou, W. H. Yang, I. C. Chou,
C. T. Lu, P. H. Lin, R. C. W. Hou, K. C. G. Jeng, C. C. Cheng
and L. Sheh, DNA sequence-specific recognition of pep-
tides incorporating the HPRK and polyamide motifs,
Bioorg. Med. Chem., 2004, 12, 53–61.
12 C. H. Yang, W. F. Chen, M. C. Jong, B. J. Jong, J. C. Chang,
M. J. Waring, L. Ma and L. Sheh, Semiquinone footprint-
ing, J. Am. Chem. Soc., 2004, 126, 8104–8105.
We thank professors L.S. Kan, D.K. Chang, S.T. Chen, S.H. Wu
and Y.T. Tau, Academia Sinica, for helpful advice and access to
ITC and CD facilities. We also thank professors Lin Ma, C.C.
Cheng, Y.J. Chen and C.W. Ong for encouragement and
helpful advice. This work was supported by grant NSC97-2113-
M029-005 and funds from Tunghai Christian University and
the Jeng Ching Ho Foundation.
13 C. H. Yang, K. C. G. Jeng, W. H. Yang, Y. L. Chen,
C. C. Hung, J. W. Lin, S. T. Chen, S. Richardson,
C. R. H. Martin, M. J. Waring and L. Sheh, Unusually
strong positive cooperativity in binding of peptides to
latent membrane protein-1 DNA fragments of the Epstein-
Barr viral gene, ChemBioChem, 2006, 7, 1187–1196.
References
1 D. S. Sigman, A. Mazumder and D. M. Perrin, Chemical
nuclease, Chem. Rev., 1993, 93, 2295–2316; J. Gallagher,
C.-H. B. Chen, C. Q. Pan, D. M. Perrin, Y. M. Cho and
D. S. Sigman, Optimizing the targeted chemical nuclease 14 K. L. Kao, J. C. T. Huang, C. K. Yang, K. C. G. Jeng,
activity of 1,10-phenanthroline-copper by ligand modifi-
cation, Bioconjugate Chem., 1996, 7, 413–420.
J. C. Chang, W. C. Yao, S. C. Hsien, M. J. Waring,
M. H. Chen, L. Ma and L. Sheh, Detection of multiple
network-based allosteric interactions between peptides and
arrays of DNA binding sites, Bioorg. Med. Chem., 2010, 18,
366–376; J. T. B. Huang, R. C. K. Yang, W. K. Hung,
M. J. Waring and L. Sheh, in Molecular Recognition, ed.
J. A. McEvoy, Nova Science Publishers Inc., New York, ISBN
978-1-61122-734-5, 2011, open access available.
2 P. B. Dervan, Design of sequence-specific DNA-binding
molecules, Science, 1986, 232, 464–471; P. G. Schultz and
P. B. Dervan, Distamycin and penta-N-methylpyrrolecarbox-
amide binding sites on native DNA. A comparison of
methidiumpropyl-EDTA-Fe(II) footprinting and DNA
affinity cleaving, J. Biomol. Struct. Dyn., 1984, 1, 1133–1147.
3 S. Routier, H. Vezin, E. Lamour, J. L. Bernier, J. P. Catteau 15 J. T. B. Huang, Y. C. Chen, J. C. Chang, K. C. G. Jeng,
and C. Bailly, DNA cleavage by hydroxylsalicylidene-ethyle-
nediamine-iron complex, Nucleic Acids Res., 1999, 27, 4160–
4166.
K. L. Kao, C. K. Yang, L. S. Kan, M. T. Wey, M. J. Waring,
C. S. Chen, W. J. Chien and L. Sheh, Novel DNA-peptide
interaction networks, Bioorg. Med. Chem., 2010, 18, 2575–
2585.
4 C. C. Cheng, W. C. Huang-Fu, K. C. Hung, P. J. Chen,
W. J. Wang and Y. J. Chen, Mechanistic aspects of CoII- 16 R. C. K. Yang, J. T. B. Huang, S. C. Chien, R. Huang,
(HAPP)(TFA)2 in DNA bulge-specific recognition, Nucleic
Acids Res., 2003, 31, 2227–2233.
5 Q. Qian Li, W. R. Browne and G. Roelfes, DNA cleavage
activity of Fe(II)N4Py under photo irradiation in the pres-
ence of 1,8-Naphthalimide and 9-Aminoacridine: Unex-
K. C. G. Jeng, Y. C. Chen, M. Liao, J. R. Wu, W. K. Hung,
C. C. Hung, Y. L. Chen, M. J. Waring and L. Sheh, Energetic
studies on DNA-peptide interaction in relation to the
enthalpy-entropy compensation paradox, Org. Biomol.
Chem., 2013, 11, 48–61.
pected effects of reactive oxygen species scavengers, Inorg. 17 M. J. Eck and S. Sprang, The structure of tumor necrosis
Chem., 2011, 50, 8318–8325.
6 C. H. Takimoto and E. Calvo, in Principles of oncologic phar-
factor-α at 2.6 Angstrom resolution, J. Biol. Chem., 1989,
264, 17595–17605 and reference quoted therein.
macotherapy in cancer management:
a multidisciplinary 18 N. M. Luscombe, R. A. Laskowski and J. M. Thornton,
approach. ed. R. Pazdur, L. D. Wagman, K. A. Camphausen
and W. J. Hoskins, 11th edn, 2008.
7 K. R. Rai, B. L. Peterson, F. R. Appelbaum, J. Kolitz,
Amino acid-base interactions: a three dimensional analysis
of protein-DNA interactions at an atomic level, Nucleic
Acids Res., 2001, 29, 2860–2874.
L. Elias, L. Shepherd, J. Hines, G. A. Threatte, R. A. Larson, 19 K. L. Buchmueller, A. M. Staples, C. M. Howard,
B. D. Cheson and C. A. Schiffer, Fludarabine compared
S. M. Horick, P. B. Uthe, N. Le Minh, K. K. Cox, B. Nguyen,
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4890–4904 | 4903

View Article Online
Paper
Organic & Biomolecular Chemistry
K. A. O. Pacheco, W. D. Wilson and M. Lee, Extending the
language of DNA molecular recognition by polyamides:
thymidine by human thymidine phosphorylase, Biochemis-
try, 2011, 50, 1412–1420.
unexpected influence of imidazole and pyrrole arrange- 23 K. J. Breslauer, D. P. Rometa, W. Y. Chou, R. Ferrante,
ment on binding affinity and specificity, J. Am. Chem. Soc.,
2005, 127, 742–750.
20 T. Brown, H. Mackay, M. Turlinton, A. Sutterfield, T. Smith,
J. Curry, D. Zaunczkowski, J. G. Snyder and L. A. Marky,
Enthalpy-entropy compensations in drug-DNA binding
studies, Proc. Natl. Acad. Sci. U. S. A., 1987, 84, 8922–8926.
A. Sielaff, L. Westrate, C. Bruce, J. Kluza, C. O’Hare, 24 S. B. Dixit, D. Q. Andrews and D. L. Beveridge, Induced fit
B. Nguyen, W. D. Wilson, J. A. Hartley and M. Lee, Modify-
ing the N-terminus of polyamides: PyImPyIm has improved
sequence specificity over f-ImPyIm, Bioorg. Med. Chem.,
2008, 16, 5266–5276.
and the entropy of structural adaptation in the complexa-
tion of CAP and λ-repressor with cognate DNA sequence,
Biophys. J., 2005, 88, 3147–3157.
25 D. H. Williams, E. Stephens, D. P. O’Brien and M. Zhou,
Understanding noncovalent interactions: ligand binding
energy and catalytic efficiency from ligand-induced
reductions in motions within receptors and enzymes,
Angew. Chem., Int. Ed., 2004, 43, 6596–6616, and references
cited therein.
21 T. L. Sheppard, P. Ordoukhanian and G. F. Joyce, A DNA
enzyme with N-glycosylase activity, Proc. Natl. Acad.
Sci. U. S. A., 2000, 97, 7802–7807.
22 P. A. Schwartz, M. J. Verticatt and V. L. Schramm, Tran-
sition state analysis of the arsenolytic depyrimidination of
4904 | Org. Biomol. Chem., 2014, 12, 4890–4904
This journal is © The Royal Society of Chemistry 2014