Conformationally constrained analogues of bleomycin A5

Article abstract of DOI:10.1021/ja030057w

The bleomycin (BLM) group antitumor antibiotics are glycopeptide-derived natural products shown to cause sequence selective lesions in DNA. Prior studies have indicated that the linker region, composed of the methylvalerate and threonine residues, may be responsible for a conformational bend in the agent required for efficient DNA cleavage. We have synthesized a number of conformationally constrained methylvalerate analogues and incorporated them into deglycobleomycin A₅ congeners using our recently reported procedure for the solid phase construction of (deglyco)bleomycin and its analogues. These analogues were designed to probe the effects of conformational constraint of the native valerate moiety. Initial experiments indicated that the constrained molecules, none of which mimic the conformation proposed for the natural valerate linker, possessed DNA cleavage activity, albeit with potencies less than that of (deglyco)BLM and lacking sequence selectivity. Further experiments demonstrated that these analogues failed to produce alkali-labile lesions in DNA or sequence selective oxidative damage in RNA. However, two of the conformationally constrained deglycoBLM analogues were shown to mediate RNA cleavage in the absence of added Fe²⁺. The ability of the analogues to mediate the oxygenation of small molecules was also assayed, and it was shown that they were as competent in the transfer of oxygen to low molecular weight substrates as the parent compound.

Full text of DOI:10.1021/ja030057w

Published on Web 07/31/2003
Conformationally Constrained Analogues of Bleomycin A5
Michael J. Rishel, Craig J. Thomas, Zhi-Fu Tao, Corine Vialas,
Christopher J. Leitheiser, and Sidney M. Hecht*
Contribution from the Departments of Chemistry and Biology, UniVersity of Virginia,
CharlottesVille, Virginia 22904
Received January 28, 2003; E-mail: sidhecht@virginia.edu
Abstract: The bleomycin (BLM) group antitumor antibiotics are glycopeptide-derived natural products shown
to cause sequence selective lesions in DNA. Prior studies have indicated that the linker region, composed
of the methylvalerate and threonine residues, may be responsible for a conformational bend in the agent
required for efficient DNA cleavage. We have synthesized a number of conformationally constrained
methylvalerate analogues and incorporated them into deglycobleomycin A5 congeners using our recently
reported procedure for the solid phase construction of (deglyco)bleomycin and its analogues. These
analogues were designed to probe the effects of conformational constraint of the native valerate moiety.
Initial experiments indicated that the constrained molecules, none of which mimic the conformation proposed
for the natural valerate linker, possessed DNA cleavage activity, albeit with potencies less than that of
(deglyco)BLM and lacking sequence selectivity. Further experiments demonstrated that these analogues
failed to produce alkali-labile lesions in DNA or sequence selective oxidative damage in RNA. However,
two of the conformationally constrained deglycoBLM analogues were shown to mediate RNA cleavage in
the absence of added Fe2+. The ability of the analogues to mediate the oxygenation of small molecules
was also assayed, and it was shown that they were as competent in the transfer of oxygen to low molecular
weight substrates as the parent compound.
of the molecule and DNA.5 BLM can also mediate oxidative
RNA cleavage in the presence of Fe2+ and oxygen, and this
Introduction
The bleomycin (BLM) group antibiotics, represented by
bleomycin A5 (Figure 1), are a class of glycopeptide-derived
antitumor agents that have been applied clinically for the
treatment of a variety of malignancies, including those of the
testes and lymph nodes.1 The agents are composed of four
primary structural domains; a metal binding domain responsible
for the binding and activation of a metal ion cofactor, a nucleic
acid binding domain which contributes the necessary nucleic
acid binding affinity to the agent, a linker domain, and a
carbohydrate moiety (Figure 1).2 The bleomycins mediate
sequence selective DNA cleavage, which has been shown to
require a metal ion cofactor and O2.2 The observed sequence
selectivity of DNA cleavage by BLM has been suggested by
some to arise from selective interactions of the metal binding
domain with DNA,2-4 while others have asserted that sequence
selectivity is governed at the interface of the bithiazole portion
process is also highly selective with regard to the sites cleaved.6
Much information has been gathered regarding the structure7
and unique mechanism of action2 of the BLMs. This has
(4) (a) Kilkuskie, R. E.; Suguna, H.; Yellin, B.; Murugesan, N.; Hecht, S. M.
J. Am. Chem. Soc. 1985, 107, 260. (b) Shipley, J. B.; Hecht, S. M. Chem.
Res. Toxicol. 1988, 1, 25. (c) Carter, B. J.; Murty, V. S.; Reddy, K. S.;
Wang, S.-N.; Hecht, S. M. J. Biol. Chem. 1990, 265, 4193. (d) Carter, B.
J.; Reddy, K. S.; Hecht, S. M. Tetrahedron 1991, 47, 2463. (e) Hamamichi,
N.; Natrajan, A.; Hecht, S. M. J. Am. Chem. Soc. 1992, 114, 6278. (f)
Quada, J. C., Jr.; Levy, M. J.; Hecht, S. M. J. Am. Chem. Soc. 1993, 115,
12171. (g) Kane, S. A.; Natrajan, A.; Hecht, S. M. J. Biol. Chem. 1994,
269, 10899. (h) Boger, D. L.; Teramoto, S.; Honda, T.; Zhou, J. J. Am.
Chem. Soc. 1995, 117, 7338.
(5) (a) Sugiura, Y.; Suzuki, T. J. Biol. Chem. 1982, 257, 10544. (b) Kuwahara,
J.; Sugiura, Y. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 2459. (c) Dickerson,
R. E. In Mechanisms of DNA Damage and Repair: Implications for
Carcinogenesis and Risk Assessment in Basic Life Sciences; Sini, M. G.,
Grossman, L., Eds.; Plenum Press: New York, 1986; pp 245-255.
(6) (a) Magliozzo, R. S.; Peisach, J.; Cirolo, M. R. Mol. Pharmacol. 1989, 35,
428. (b) Carter, B. J.; de Vroom, E.; Long, E. C.; van der Marel, G. A.;
van Boom, J. H.; Hecht, S. M. Proc. Natl. Acad. Sci. U.S.A. 1990, 87,
9373. (c) Holmes, C. E.; Carter, B. J.; Hecht, S. M. Biochemistry 1993,
32, 4293. (d) Hecht, S. M. Bioconjugate Chem. 1994, 5, 513. (e) Holmes,
C. E.; Abraham, A. T.; Hecht, S. M.; Florentz, C.; Giege, R. Nucleic Acids
Res. 1996, 24, 3399. (f) Hecht, S. M. In The Many Faces of RNA; Eggleston,
D. S., Prescott, C. D., Pearson, N. D., Eds.; Academic Press: San Diego,
CA, 1998; pp 3-17. (g) Abraham, A. T.; Newton, D. L.; Lin, J.-J.; Rybak,
S.; Hecht, S. M. Chem. Biol. 2003, 10, 45.
(1) Bleomycin Chemotherapy; Sikic, B. I., Rozencweig, M., Carter, S. K., Eds.;
Academic Press: Orlando, FL, 1985.
(2) (a) Natrajan, A.; Hecht, S. M. In Molecular Aspects of Anticancer Drug-
DNA Interactions; Neidle, S., Waring, M. J., Eds.; MacMillan Press:
London, 1993; pp 197-242. (b) Kane, S. A.; Hecht, S. M. Prog. Nucleic
Acid Res. Mol. Biol. 1994, 49, 313. (c) Burger, R. M. Chem. ReV. 1998,
98, 1153. (d) Boger, D. L.; Cai, H. Angew. Chem., Int. Ed. 1999, 38, 448.
(3) (a) Guajardo, R. J.; Hudson, S. E.; Brown, S. J.; Mascharak, P. K. J. Am.
Chem. Soc. 1993, 115, 7971. (b) Guajardo, R. J.; Tan, J. D.; Mascharak,
P. K. Inorg. Chem. 1994, 33, 2838. (c) Loeb, K. E.; Zaleski, J. M.; Westre,
T. E.; Guajardo, R. J.; Mascharak, P. K.; Hedman, B.; Hodgson, K. O.;
Solomon, E. I. J. Am. Chem. Soc. 1995, 117, 4545. (d) Guajardo, R. J.;
Chavez, F.; Farinas, E. T.; Mascharak, P. K. J. Am. Chem. Soc. 1995, 117,
3883. (e) Farinas, E. T.; Tan, J. D.; Mascharak, P. K. Inorg. Chem. 1996,
35, 2637.
(7) (a) Takita, T.; Umezawa, Y.; Saito, S.; Morishima, H.; Naganawa, H.;
Umezawa, H.; Tsuchiya, T.; Miyake, T.; Kageyama, S.; Umezawa, S.;
Muraoka, Y.; Suzuki, M.; Otsuka, M.; Narita, M.; Kobayashi, S.; Ohno,
M. Tetrahedron Lett. 1982, 23, 521. (b) Aoyagi, Y.; Katano, K.; Suguna,
H.; Primeau, J.; Chang, L.-H.; Hecht, S. M. J. Am. Chem. Soc. 1982, 104,
5537. (c) Boger, D. L.; Colletti, S. L.; Honda, T.; Menezes, R. F. J. Am.
Chem. Soc. 1994, 116, 5607. (d) Katano, K.; An, H.; Aoyagi, Y.; Overhand,
M.; Sucheck, S. J.; Stevens, W. C., Jr.; Hess, C. D.; Zhou, X.; Hecht, S.
M. J. Am. Chem. Soc. 1998, 120, 11285.
9
10194
J. AM. CHEM. SOC. 2003, 125, 10194-10205
10.1021/ja030057w CCC: $25.00 © 2003 American Chemical Society

Conformationally Constrained Bleomycin Analogues
A R T I C L E S
Figure 1. Structures of bleomycin A5 (1a) and deglycobleomycin A5 (1b).
included studies of the linker domain.2d,4c,7c,8-12 A key finding
of the studies focusing on this domain is that the substantial
effect of the methylvalerate subunit on BLM cleavage efficiency
is of a conformational nature. An analysis of the linker region
of BLM has been presented, proposing that this portion of the
molecule is responsible for preorganizing BLM into a rigid,
compact conformation essential for efficient DNA binding and
cleavage (Figure 2).2d,11 Further, it has been suggested that the
insertion of appropriately constrained valerate analogues into
the linker region of bleomycin may accelerate rates of DNA
cleavage by lowering the entropic cost for DNA binding.12c We
have synthesized a series of novel conformationally constrained
methylvalerate analogues A-F13 (Figure 3) intended to probe
for alternative DNA bound conformations of the linker region
of deglycobleomycin A5. Although none of these constrained
analogues reproduce all aspects of the proposed bound confor-
mation, they serve as an initial query of such conformational
features. These analogues were incorporated into deglycoBLM
A5 using our methodology for the synthesis of (deglyco)BLM
analogues on a solid support.14,15 The analogues were assayed
for DNA cleavage efficiency and selectivity relative to authentic
deglycobleomycin A5. In addition, the ability of selected
analogues to mediate the oxygenation of styrene and the
degradation of an RNA were also studied.
(8) Chien, M.; Grollman, A. P.; Horwitz, S. B. Biochemistry 1977, 16, 3641.
(9) Umezawa, H.; Takita, T.; Saito, S.; Muraoka, Y.; Takahashi, K.; Ekimoto,
H.; Minamide, S.; Nishikawa, K.; Fukuoka, T.; Nakatani, T.; Fujii, A.;
Matsuda, A. In Bleomycin Chemotherapy; Sikic, B. I., Rozencweig, M.,
Carter, S. K., Eds.; Academic Press: Orlando, FL, 1985; pp 289-301.
(10) Owa, T.; Haupt, A.; Otsuka, M.; Kobayashi, S.; Tomioka, N.; Itai, A.;
Ohno, M. Tetrahedron 1992, 48, 1193.
(11) (a) Boger, D. L.; Colletti, S. L.; Teramoto, S.; Ramsey, T. M.; Zhou, J.
Bioorg. Med. Chem. 1995, 3, 1281. (b) Boger, D. L.; Ramsey, T. M.; Cai,
H.; Hoehn, S. T.; Stubbe, J. J. Am. Chem. Soc. 1998, 120, 9139. (c) Boger,
D. L.; Ramsey, T. M.; Cai, H.; Hoehn, S. T.; Stubbe, J. J. Am. Chem. Soc.
1998, 120, 9149.
(12) (a) Wu, W.; Vanderwall, D. E.; Stubbe, J.; Kozarich, J. W.; Turner, C. J.
J. Am. Chem. Soc. 1994, 116, 10843. (b) Wu, W.; Vanderwall, D. E.; Lui,
S. M.; Tang, X.-J.; Turner, C. J.; Kozarich, J. W.; Stubbe, J. J. Am. Chem.
Soc. 1996, 118, 1268. (c) Wu, W.; Vanderwall, D. E.; Turner, C. J.;
Kozarich, J. W.; Stubbe, J. J. Am. Chem. Soc. 1996, 118, 1281. (d)
Vanderwall, D. E.; Lui, S. M.; Wu, W.; Turner, C. J.; Kozarich, J. W.;
Stubbe, J. Chem. Biol. 1997, 4, 373. (e) Lui, S. M.; Vanderwall, D. E.;
Wu, W.; Tang, X.-J.; Turner, C. J.; Kozarich, J. W.; Stubbe, J. J. Am.
Chem. Soc. 1997, 119, 9603. (f) Wu, W.; Vanderwall, D. E.; Teramoto,
S.; Lui, S. M.; Hoehn, S.; Tang, X.-J.; Turner, C. J.; Boger, D. L.; Kozarich,
J. W.; Stubbe, J. J. Am. Chem. Soc. 1998, 120, 2239.
Results
Synthesis of Deglycobleomycin Analogues 2a-f. Incorpora-
tion of the constrained valerate analogues A-F (Supporting
Information, Schemes S1 and S2) into deglycobleomycin A5
(13) Rishel, M. J.; Hecht, S. M. Org. Lett. 2001, 3, 2867.
(14) (a) Leitheiser, C. J.; Rishel, M. J.; Wu, X.; Hecht, S. M. Org. Lett. 2000,
2, 3397. (b) Smith, K. L.; Tao, Z.-F.; Hashimoto, S.; Leitheiser, C. J.; Wu,
X.; Hecht, S. M. Org. Lett. 2002, 4, 1079. (c) Tao, Z.-F.; Leitheiser, C. J.;
Smith, K. L.; Hashimoto, S.; Hecht, S. M. Bioconjugate Chem. 2002, 13,
426.
(15) Thomas, C. J.; McCormick, M. M.; Vialas, C.; Tao, Z.-F.; Leitheiser, C.
J.; Rishel, M. J.; Wu, X.; Hecht, S. M. J. Am. Chem. Soc. 2002, 124, 3875.
9
J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003 10195

A R T I C L E S
Rishel et al.
11
Figure 2. Conformational analyses of the native methylvalerate moiety of deglycobleomycin A5 and valerate analogues A-F.
quantitative measurement of resin loading. Fmoc cleavage
analysis of an aliquot of dry resin treated with a solution of
20% piperidine in DMF indicated a loading of 0.24 mmol/g,
corresponding to a 78% yield for the production of the dipeptide
5 from the Dde-linked TentaGel starting material.
Extension to the tripeptide 7 was accomplished by removal
of the Fmoc group from 5 and subsequent coupling with
commercially available NR-Fmoc-(S)-threonine (6) in the pres-
ence of HBTU, HOBt, and Hunig’s base in DMF. Fmoc
cleavage analysis of the resin bound product 7 then afforded a
loading of 0.21 mmol/g, corresponding to a 90% yield for the
synthesis of tripeptide 7 from the dipeptide 5.
Figure 3. Conformationally constrained valerate analogues A-F.13
congeners was accomplished using our recently described
methodology for the solid-phase synthesis of (deglyco)bleomy-
cin A5.14,15 This technique has already proven quite efficient in
the synthesis of deglycobleomycin A5 and several analogues14,15
and is further highlighted here by our construction of de-
glycoBLMs 2a-f (Figure 4). The solid-phase synthesis of the
deglycobleomycin A5 analogues was accomplished using Tenta-
Gel resin (preloading of 0.45 mmol/g). Attachment of the
requisite deglycobleomycin A5 subunits to the solid support was
performed using a hydrazine-labile Dde linker of the type
described by Bycroft.16 The elaboration of intermediate 3
(Scheme 1) has been described previously.15
Synthesis of the tetrapeptides 8a-f was accomplished using
valerate analogues A-F (Figure 3), respectively and a synthetic
protocol identical to that used for the coupling of NR-Fmoc-
(S)-threonine (6). The use of A-F in the construction of the
tetrapeptides 8a-f gave the desired resin bound precursors in
varying yields. Fmoc analysis of the tetrapeptides 8b, 8c, and
8d indicated coupling yields of 91%, 88%, and 90%, respec-
tively. However, the tetrapeptide analogues 8a, 8e, and 8f were
formed in much lower yields of 45%, 48%, and 45%, respec-
tively.
The syntheses of the pentapeptide analogues 10a-f (Scheme
2) were effected by Fmoc removal from the corresponding
tetrapeptides 8a-f followed by the coupling of the tetrapeptides
with NR-Fmoc-Nim-trityl-â-hydroxyhistidine (9) via the agency
of HATU, HOAt, and Hunig’s base in DMF. Fmoc analysis of
an aliquot of each of the resin bound pentapeptides 10a-f
revealed resin loadings corresponding to yields of 93%, 82%,
56%, 50%, 86%, and 89% for the pentapeptides 10a-f,
respectively.
Synthesis of resin-bound dipeptide 5 (Scheme 1) was ac-
complished through the coupling of the resin-bound spermidine
moiety 3 and Fmoc bithiazole 4 in the presence of HBTU and
Hunig’s base in DMF. The addition of 4 permitted the initial
(16) (a) Bycroft, B. W.; Chan, W. C.; Hone, N. D.; Millington, S.; Nash, I. A.
J. Am. Chem. Soc. 1994, 116, 7415. (b) Chhabra, S. R.; Khan, A. N.;
Bycroft, B. W. Tetrahedron Lett. 1998, 39, 3585. (c) Chhabra, S. R.; Khan,
A. N.; Bycroft, B. W. Tetrahedron Lett. 2000, 41, 1095. (d) Chhabra, S.
R.; Khan, A. N.; Bycroft, B. W. Tetrahedron Lett. 2000, 41, 1099.
9
10196 J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003

Conformationally Constrained Bleomycin Analogues
A R T I C L E S
Figure 4. Structures of the constrained deglycobleomycin A5 analogues 2a-2f.
Scheme 1
Production of resin-bound deglycoBLM A5 analogues 12a-f
was accomplished by a final Fmoc deprotection followed by
coupling of the pentapeptides to Boc pyrimidoblamic acid (11)17
utilizing the BOP reagent and Hunig’s base at 0 °C. The
coupling was performed overnight and in the absence of light.
Acid-mediated deprotection of the Boc and trityl groups was
accomplished through two successive 20-min exposures of the
functionalized resin to trifluoroacetic acid in the presence of
triisopropylsilane and dimethyl sulfide. Removal of the 2-nitro-
benzenesulfonyl (NBS) group was accomplished by treatment
(17) (a) Umezawa, Y.; Morishima, H.; Saito, S.; Takita, T.; Umezawa, H.;
Kobayashi, S.; Otsuka, M.; Narita, M.; Ohno, M. J. Am. Chem. Soc. 1980,
102, 6630. (b) Aoyagi, Y.; Chorghade, M. S.; Padmapryia, A. A.; Suguna,
H.; Hecht, S. M. J. Org. Chem. 1990, 55, 6291. (c) Boger, D. L.; Honda,
T.; Dang, Q. J. Am. Chem. Soc. 1994, 116, 5619.
9
J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003 10197

A R T I C L E S
Rishel et al.
Scheme 2
with a 1 M solution of sodium thiophenoxide in DMF over a
period of 2 h. Cleavage of the products 2a-f from the resin
was performed via treatment with a 2% solution of hy-
drazine in DMF. Precipitation of the product from CF3COOH
solution with diethyl ether, followed by HPLC purification,
afforded the desired deglycoBLM A5 analogues 2a-f as
colorless amorphous solids. The yields for the final coupling,
deprotection, and purification for each analogue were quanti-
fied by UV analysis and were found to be 22%, 18%, 9%,
35%, 37%, and 6% for deglycoBLM A5 analogues 2a-f,
respectively.
showed at least some ability to nick the circular DNA at a 10
µM concentration in the presence of Fe2+. A sample of 10 µM
deglycobleomycin A5 in the presence of 1.5 µM Fe2+ (lane 15)
quantitatively converted Form I (supercoiled) DNA to Form II
(relaxed circular) and Form III (linear duplex DNA). As is clear
from lane 1, the DNA employed for this study contained some
nicked species. The conformationally constrained deglycoBLM
analogues 2a-e, in the presence of a metal ion cofactor (lanes
4, 6, 8, 10, and 12), converted supercoiled DNA to relaxed
circular DNA to an extent approximately 15% greater than that
observed for untreated DNA. The unconstrained analogue 2f
(lane 14) was much more efficient than any of the constrained
analogues, relaxing all of the available supercoiled DNA and
producing a significant quantity (12%) of Form III DNA.
Deglycobleomycin A5 itself (lane 15) was only slightly more
DNA Degradation by Deglycobleomycin A5 Analogues
Modified in the Valerate Moiety. The synthetic deglycoBLM
analogues 2a-f were assayed for their ability to relax super-
coiled plasmid DNA (Figure 5). All of the analogues synthesized
Figure 5. Cleavage of supercoiled pBR322 DNA by valerate-modified deglycobleomycin A5 analogues. The incubation time was 30 min at 37 °C. Lane
1, DNA alone; lane 2, 1.5 µM Fe2+; lane 3, 10 µM 2a; lane 4, 10 µM 2a + 1.5 µM Fe2+; lane 5, 10 µM 2b; lane 6, 10 µM 2b + 1.5 µM Fe2+; lane 7, 10
µM 2c; lane 8, 10 µM 2c + 1.5 µM Fe2+; lane 9, 10 µM 2d; lane 10, 10 µM 2d + 1.5 µM Fe2+; lane 11, 10 µM 2e; lane 12, 10 µM 2e + 1.5 µM Fe2+
lane 13, 10 µM 2f; lane 14, 10 µM 2f + 1.5 µM Fe2+; lane 15, 10 µM deglycoBLM A5 + 1.5 µM Fe2+; lane 16, 10 µM deglycoBLM A5.
;
9
10198 J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003

Conformationally Constrained Bleomycin Analogues
A R T I C L E S
Figure 6. Cleavage of 5′-32P end-labeled 158-base-pair DNA duplex by
valerate-modified deglycobleomycin A5 analogues 2a-f. Lane 1, DNA
alone; lane 2, 10 µM Fe2+; lane 3, 10 µM 2a; lane 4, 10 µM 2a + 10 µM
Fe2+; lane 5, 10 µM 2b; lane 6, 10 µM 2b + 10 µM Fe2+; lane 7, 10 µM
2c; lane 8, 10 µM 2c + 10 µM Fe2+; lane 9, 10 µM 2d; lane 10, 10 µM 2d
+ 10 µM Fe2+; lane 11, 10 µM 2e; lane 12, 10 µM 2e + 10 µM Fe2+; lane
13, 10 µM 2f; lane 14, 10 µM 2f + 10 µM Fe2+; lane 15, 10 µM
deglycoBLM A5 + 10 µM Fe2+; lane 16, 10 µM deglycoBLM A5. The
band of intermediate mobility in lanes 1-13 and 16 was due to adventitious
nondenatured DNA duplex.
Figure 7. Assay for alkalai-labile lesions produced in a 5′-32P end-labeled
158-base-pair DNA duplex by valerate-modified deglycoBLM analogues
2a and 2c. Samples were initially incubated for 30 min at 37 °C and then
treated with 0.1 M piperidine for 20 min at 90 °C. Lane 1, 1 µM
deglycobleomycin A5 + 10 µM Fe2+; lane 2, DNA alone; lane 3, 10 µM
Fe2+; lane 4, 10 µM 2a; lane 5, 10 µM 2a + 10 µM Fe2+; lane 6, 10 µM
effective at relaxing the plasmid DNA substrate, producing 84%
Form II and 16% Form III DNA under the same conditions.
The constrained analogues were assayed for sequence selec-
tivity of DNA cleavage using a 5′-32P end-labeled 158-
nucleotide DNA oligonucleotide substrate (Figure 6). As shown,
the constrained analogues 2a-e cleaved the 158-base-pair
oligonucleotide weakly and at essentially every position, with
no clear sequence selectivity. The unconstrained analogue 2f,
on the other hand, cleaved DNA with a sequence selectivity
essentially identical to that of deglycobleomycin A5.
2c; lane 7, 10 µM 2c + 10 µM Fe2+
.
substrate to ascertain if any sequence selectivity could be
established for the constrained bleomycin analogues using these
types of substrates. As shown in Figure 8, in the presence of
Fe2+ no sequence selective oxidative RNA cleavage was
observed for any of the analogues, including 2f which had
exhibited sequence selective DNA cleavage (Figure 6, lane 14).
It may be noted, however, that RNA cleavage by BLM exhibits
a concentration optimum6b (cf Figure 8, lanes 12 and 13), and
it is conceivable that some of the deglycoBLM analogues 2 may
produce RNA lesions under narrowly defined conditions of
concentration.
Keck and Hecht et al.19 have shown previously that BLM
can effect the cleavage of tRNAPhe in the absence of Fe2+ by a
process that must involve phosphoryl transfer involving the 2′-
OH groups in the RNA substrate at the sites of cleavage. The
actual cleavage sites occurred at all pyrimidine-purine sites
not involving modified nucleosides; the most efficient cleavage
was at the five 5′-CA-3′ sequences.
In addition to frank DNA strand scission, BLM is also known
to create alkali-labile lesions;2 these have been shown to be more
prominent at low oxygen tension and in the presence of certain
modified DNA substrates.18 To determine whether the modified
deglycoBLM analogues 2a-2f produced a greater proportion
of such alkali-labile lesions, two arbitrarily chosen analogues
(2a and 2c) were also tested for their ability to produce sequence
selective alkali-labile lesions in a 5′-32P end-labeled 158-
nucleotide DNA oligonucleotide substrate (Figure 7). The
analogues tested showed no ability to produce sequence se-
lective lesions in DNA, analogous to those produced by
Fe(II)•deglycoBLM A5 (Figure 7, lane 1). While those tested
deglycoBLM analogues 2 mainly produced cleavage bands at
every position in the DNA substrate, it is interesting that some
of these bands were somewhat enhanced in relative intensity
following alkali treatment and were at sites different than those
produced by deglycoBLM A5.
As shown in Figure 8, deglycoBLM analogues 2b and 2e,
chosen to be representative of the constrained analogues,
effected cleavage of the 53-nucleotide RNA substrate employed
here in the absence of added Fe2+ but not in its presence,
suggesting the occurrence of “hydrolytic” cleavage analogous
to that observed previously for BLM itself.19 Interestingly, the
With the apparent lack of DNA cleavage sequence selectivity
by 2a-2e established, we turned our attention to an RNA
(18) Long, E. C.; Hecht, S. M.; van der Marel, G. A.; van Boom, J. H. J. Am.
Chem. Soc. 1990, 112, 5272.
(19) Keck, M. V.; Hecht, S. M. Biochemistry 1995, 34, 12029.
9
J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003 10199

A R T I C L E S
Rishel et al.
acetaldehyde and styrene oxide in the presence of Fe3+ and
hydrogen peroxide. The results from this experiment are shown
in Table 1. The data clearly demonstrate that analogue 2d,
chosen to be respresentative of the constrained analogues, was
as efficient as bleomycin and more efficient than deglycoBLM
A5 in mediating the conversion of styrene to phenylacetaldehyde
(13) and styrene oxide (14).
Discussion
Oxidative damage of DNA by bleomycin can be viewed as
a two-step process.21 Bleomycin first coordinates a metal
cofactor and oxygen and undergoes activation to afford a species
capable of oxidative DNA degradation. The drug then binds to
DNA at a preferred cleavage site, leading to abstraction of the
C-4′ H from the deoxyribose moiety, which is apparently the
rate-limiting event in the oxidative strand cleavage of a nucleic
acid substrate.22 It is generally accepted that interactions between
the activated metal binding domain and the minor groove of
the bound DNA duplex are of primary importance in maintain-
ing the sequence selectivity observed for Fe•BLM.3a,4c,23
Although analogues of the BLM metal binding domain have
been shown to mediate sequence selective cleavage of DNA,3
it has been demonstrated that the actual metal binding subunit
of deglycobleomycin alone was unable to produce significant
sequence selective DNA cleavage in the presence of Fe2+ 4a,h
.
The C-terminus of the agent, although not the primary
determinant of DNA sequence selectivity, has been shown to
be responsible for the majority of the observed DNA binding
affinity of bleomycin.8 Further, studies of the photoinduced
cleavage of DNA by chlorobithiazoles and structurally related
deglycobleomycin analogues incorporating chlorinated bithiazole
moieties have suggested that the bithiazole moiety itself may
in fact exhibit some sequence selective DNA binding.4f,24 This
selectivity, however, was found not to be the same as that
associated with oxidatively induced Fe•BLM lesions. This may
explain why certain GpY sequences are cleaved more efficiently
by bleomycin than others, as those sites best suited to accom-
modate the preferences of both the metal binding domain and
the bithiazole moiety may be cleaved with the greatest ef-
ficiency.
Figure 8. Cleavage of a 53-nt RNA by valerate-modified deglycobleomycin
A5 analogues 2b, 2e, and 2f. Lane 1, RNA alone; lane 2, 100 µM Fe2+
;
lane 3, 100 µM 2b; lane 4, 10 µM 2b + 10 µM Fe2+; lane 5, 100 µM 2b
+ 100 µM Fe2+; lane 6, 100 µM 2e; lane 7, 10 µM 2e + 10 µM Fe2+; lane
8, 100 µM 2e + 100 µM Fe2+; lane 9, 100 µM 2f; lane 10, 10 µM 2f + 10
µM Fe2+; lane 11, 100 µM 2f + 100 µM Fe2+; lane 12, 100 µM
deglycoBLM A5 + 100 µM Fe2+; lane 13, 10 µM bleomycin A5 + 10 mM
Fe2+. The position of cleavage (arrow) obtained with deglycoBLM A5 in
lane 13 was inferred by analogy with the cleavage of Bacillus subtilis
tRNAHis precursor transcript.6b
The linker domain, while not implicated directly in sequence
selective DNA binding or cleavage, has been demonstrated to
be important for maintaining the efficiency with which bleo-
mycin cleaves DNA.11 Although alterations to this region have
been shown to have a significant impact on DNA cleavage
efficiency, they have not been reported to result in a change in
DNA sequence selectivity. In short, the function of the linker
domain has been suggested to be one of a conformational nature,
directing the agent into a conformation amenable to the efficient,
sequence selective cleavage of DNA. Consistent with this
analysis, it has been suggested that the appropriate conforma-
efficiency of this process was significantly better for constrained
deglycoBLM analogues 2b and 2e than for unconstrained
analogue 2f.
Oxidation of Styrene by Deglycobleomycin A5 Analogues.
In addition to their ability to mediate oxidative damage of DNA
and RNA substrates, bleomycin and some of its analogues have
also been shown to be capable of supporting the oxygenation
of a number of low molecular weight substrates including cis-
stilbene and styrene and also the hydroxylation of aromatic
substrates such as naphthalene and anisole.4a,4e,20 To determine
whether the constrained analogues 2 were bleomycin-like in their
capacity to deliver oxygen to small molecules, we tested one
of the analogues for its ability to convert styrene to phenyl-
(21) (a) Hecht, S. M. Acc. Chem. Res. 1986, 19, 383. (b) van Atta, R. B.; Long,
E. C.; Hecht, S. M.; van der Marel, G. A.; van Boom, J. H. J. Am. Chem.
Soc. 1989, 111, 2722. (c) Natrajan, A.; Hecht, S. M.; van der Marel, G.
A.; van Boom, J. H. J. Am. Chem. Soc. 1990, 112, 3997.
(22) Kozarich, J. W.; Worth, L., Jr.; Frank, B. L.; Christner, D. F.; Vanderwall,
D. E.; Stubbe, J. Science 1989, 245, 1396.
(23) Sugiyama, H.; Kilkuskie, R. E.; Chang, L.-H.; Ma, L.-T.; Hecht, S. M.;
van der Marel, G. A.; van Boom, J. H. J. Am. Chem. Soc. 1986, 108, 3852.
(24) (a) Zuber, G.; Quada, J. C., Jr.; Hecht, S. M. J. Am. Chem. Soc. 1998, 120,
9368. (b) Quada, J. C., Jr.; Zuber, G.; Hecht, S. M. Pure Appl. Chem.
1998, 70, 307. (c) Quada, J. C., Jr.; Boturyn, D.; Hecht, S. M. Bioorg.
Med. Chem. 2001, 9, 2303.
(20) (a) Murugesan, N.; Ehrenfeld, G. M.; Hecht, S. M. J. Biol. Chem. 1982,
257, 8600. (b) Ehrenfeld, G. M.; Murugesan, N.; Hecht, S. M. Inorg. Chem.
1984, 23, 1496. (c) Murugesan, N.; Hecht, S. M. J. Am. Chem. Soc. 1985,
107, 493. (d) Heimbrook, D. C.; Mulholland, R. L., Jr.; Hecht, S. M. J.
Am. Chem. Soc. 1986, 108, 7839. (e) Ehrenfeld, G. M.; Shipley, J. B.;
Heimbrook, D. C.; Sugiyama, H.; Long, E. C.; van Boom, J. H.; van der
Marel, G. A.; Oppenheimer, N. J.; Hecht, S. M. Biochemistry 1987, 26,
931. (f) Heimbrook, D. C.; Carr, S. A.; Mentzer, M. A.; Long, E. C.; Hecht,
S. M. Inorg. Chem. 1987, 26, 3835.
9
10200 J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003

Conformationally Constrained Bleomycin Analogues
A R T I C L E S
Table 1. Styrene Oxidation by BLM, DeglycoBLM A5, and DeglycoBLM Analogue 2da,b
compound
compound
compound
compound
compound
compound
13
14
13
14
13
14
oxidant
40 min
80 min
120 min
bleomycin
deglycoBLM A5
deglycoBLM 2d
4.8 ( 1.3
2.9 ( 0.8
5.0 ( 1.5
0.6 ( 0.2
0.5 ( 0.2
1.1 ( 0.3
2.9 ( 0.8
2.7 ( 0.8
4.8 ( 1.3
0.2 ( 0.06
0.2 ( 0.05
0.7 ( 0.2
2.9 ( 0.8
2.1 ( 0.6
2.9 ( 0.7
0.09 ( 0.03
0.05 ( 0.02
0.3 ( 0.08
a Reaction conditions, 0.73 mM Fe3+, 73 mM styrene, 1.6 mM ethyl benzoate (internal standard), 0.72 mM BLM, deglycobleomycin A5 or 2d, and 15
mM H2O2. The reactions were maintained at 4 °C for the times indicated. b Results are in mmol/L; reported values are the average of three experiments.
tional constraint within the linker domain could lower the
entropic cost for DNA binding and thus facilitate cleavage.12c
Boger et al.11c reported an insightful, quantitative conforma-
tional analysis of the linker region in which they stressed the
importance of the valerate C4 methyl substituent in permitting
the proper orientation of the metal binding domain with respect
to the bound DNA substrate. Further, these authors identified a
swivel point within the valerate allowing for ca. 120° of rotation
between C1 and C2. This swivel point was suggested to allow
the bleomycins “access to several related DNA bound confor-
mations” depending on the microstructure of the DNA complex.
Based on this analysis we prepared the valerate analogues
shown in Figures 2 and 3 and the respective deglycoBLMs
(Figure 4). These analogues were chosen for their ability to
provide an initial query of requisite conformations within the
methylvalerate moiety and also because they were reasonably
accessible synthetically. While none of these analogues closely
resembles the conformation proposed to facilitate the formation
of a compact linker region associated with efficient DNA
cleavage, the existence of swivel points that may provide BLM
access to related DNA bound conformations suggested that
alternate BLM conformations might conceivably also produce
DNA damage.
We utilized our solid phase synthesis strategy14,15 to elaborate
the deglycobleomycin A5 analogues (2a-f) in which the valerate
moiety was replaced with the valerate analogues A-F, respec-
tively. It has been proposed that the construction of appropriately
constrained analogues should facilitate accelerated rates of DNA
cleavage, as the energy for DNA binding should be lowered
significantly.12c It was of interest to us to see whether any of
the constrained analogues 2a-e was capable of eliciting the
desired effect even though none of them embody all of the
conformational constraints of the natural valerate linker.
As demonstrated in Figure 5, the constrained analogues 2a-e
exhibited greatly reduced abilities to relax supercoiled plasmid
DNA to produce circular (Form II) DNA. Further, in no case
was any linear duplex (Form III) DNA observed. On the other
hand, the unconstrained analogue 2f, prepared as a control,
relaxed supercoiled plasmid DNA to produce both circular and
linear duplex DNA with an efficiency quite comparable to that
of deglycobleomycin A5 itself. To ensure that the modifications
to deglycobleomycin A5 had not affected its ability to complex
with a metal ion cofactor and activate bound oxygen, the
analogue 2d was tested for its ability to oxygenate styrene to
afford phenylacetaldehyde (13) and styrene oxide (14).20 The
analogue performed similarly to authentic BLM and deglyco-
bleomycin A5 (Table 1), indicating that the ability of this
analogue to form an activated metal complex had not been
impaired.
The sequence selectivity of DNA cleavage by analogues 2a-f
was assayed using a 5′-32P end-labeled linear DNA duplex
(Figure 6). The results showed that the constrained analogues
2a-e cleaved the linear DNA substrate at essentially every
position, while the unconstrained analogue 2f maintained the
sequence selectivity of the parent deglycoBLM. Analogues 2a
and 2c were studied further to see whether they formed alkali-
labile lesions in the DNA substrate. As shown in Figure 7, none
of these analogues produced alkali-labile lesions at the same
sites as deglycoBLM A5, but there were sites at which some
enhancement of cleavage may have occurred.
There is accumulating evidence that RNA may also be a
therapeutically important target for BLM.6 Accordingly, it
seemed important to characterize RNA cleavage by the con-
formationally constrained deglycobleomycin analogues. In fact,
an interesting result was obtained when analogues 2b and 2e
were tested for their ability to cleave RNA. The substrate
employed was a synthetic RNA identical in sequence with the
core region of the B. subtilis tRNAHis precursor, the latter of
which undergoes efficient cleavage by Fe•BLM predominantly
at a single site.6b The 53-nt RNA has also been shown to
undergo efficient cleavage by Fe•BLM at what is presumably
the analogous position.15 While neither 2b nor 2e effected
cleavage of this RNA under oxidative conditions, the metal free
analogues both effected RNA cleavage at a number of sites via
a “hydrolytic” process, as has been noted previously for BLM
itself.19 Further, nonoxidative RNA cleavage by 2b and 2e was
considerably more efficient than that mediated by unconstrained
analogue 2f. The observation that analogues 2b and 2e were
able to produce sequence selective RNA cleavage in the absence
of a metal cofactor indicates that the binding mode for this
process is different than that required for oxidative RNA
cleavage.
At least two possibilities are consistent with the behavior
observed for deglycoBLM analogues 2a-2e. The first is that
deviations from an optimal folded conformation of the linker
domain may greatly affect the orientation of the metal binding
domain with respect to its DNA substrate. The specific structural
constraints employed in the present study thus precluded the
adoption of a conformation necessary for efficient DNA binding
and cleavage. The second possibility is that some conformational
flexibility per se within the linker domain may be necessary
for efficient sequence selective binding and cleavage of DNA
9
J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003 10201

A R T I C L E S
Rishel et al.
and RNA substrates. Accordingly, the conformational constraints
imposed in analogues 2a-2e may have precluded efficient
cleavage. In this context, it may be mentioned that a parallel
study employed analogues of deglycoBLM having conforma-
tionally constrained replacements for the methylvalerate moiety
based on the model of Zn(II)•deglycoBLM bound to DNA;25,26
no sequence selective DNA cleavage was obtained.27 The failure
of several conformationally constrained deglycoBLM analogues
to mediate sequence selective DNA cleavage raises the pos-
sibility that the compact valerate conformation adopted by
(deglyco)BLM itself11 may be uniquely suited to orient the
molecule in a conformation conducive to selective DNA
cleavage. In principle, the lack of sequence selective DNA and
RNA cleavage by analogues 2a-2e might also reflect unfavor-
able steric interactions with substrate not present for deglyco-
bleomycin itself. However, this seems unlikely, based both on
the sequence selective cleavage mediated by analogue 2f and
also by the recent finding that deglycobleomycin analogues
having C4 substituents of increased size within the methyl-
valerate moiety actually exhibited enhanced cleavage of super-
coiled plasmid DNA.28 Clearly, it may be possible to decide
among the foregoing possibilities by preparing constrained
analogues that more closely mimic the proposed preferred
conformation.
Finally, it may be noted that the assays employed in the
present study measured the cleavage of DNA and RNA, an event
that is believed to be preceded by selective binding of these
substrates. While 2a-2e failed to mediate sequence selective
oxidative cleavage of the single DNA and RNA substrates tested
in this initial study, it seems possible that these analogues may
still exhibit DNA and RNA binding. This possibility is
underscored by the fact that H atom abstraction from DNA is
the rate-limiting step in DNA cleavage even for BLM itself.22
Accordingly, these analogues could exhibit productive binding
and cleavage of non-B-form DNA structures such as DNA
triplexes29 and bulges,30 not to mention RNA structures not
cleaved efficiently by BLM itself.6c In this context, it is
interesting to note that McLean et al.31 found that Co•BLM and
Fe•BLM, which have the same valerate moiety, nonetheless
exhibited different patterns of DNA cleavage; each cleaved DNA
at some positions not cleaved by the other metallobleomycin.
This argues that not all activated BLMs exhibit the same unique
DNA binding pattern and conversely that one may anticipate
that individual BLM analogues could exhibit a particular facility
for the cleavage of a structurally unique DNA or RNA substrate.
Also worthy of comment are the implications of the present
study for models of BLM binding and cleavage published
previously. Assuming that the failure of analogues 2a-2e to
mediate sequence selective DNA cleavage is not due to their
lack of flexibility, the present results support the thesis that
stabilization of a rather narrowly defined conformation of the
methylvalerate moiety is essential for sequence selective DNA
cleavage. However, it is perhaps somewhat surprising that
conformational alterations around the proposed swivel points11b,c
did not result in DNA cleavage, if only at altered positions.
A related issue is the proposal12b,c,32 that both Fe(II)•BLMs
and Co(III)•BLMs, which were both used to develop the binding
model being tested in this study, interact with and cleave their
nucleic acid substrates in the same fashion. There is compelling
evidence that this is not the case. DNA cleavage by Fe•BLM is
a dark reaction, while cleavage by Co(III)•BLM requires light.33
The affinity of Co(III)•BLM for its DNA substrate is 10-fold
greater than that of Fe(II)•BLM.33b The specificities of these
two metallobleomycins were compared by examining the
relative extents of cleavage of each DNA strand in a DNA
duplex; neither the selectivity nor the extent of cleavage by the
metallobleomycins at specific sites along the DNA backbone
was comparable.31 Specifically, Fe(II)•BLM exhibited charac-
teristic GT and GC selectivity, while Co(III)•BLM showed a
preponderance of cleavage at GA and at every GT sequence.
The cleavage pattern for Co(III)•BLM was sequence selective
but different than that of Fe(II)•BLM. Co(III)•BLM has also
been reported to produce less double-strand cleavage than
Fe(II)•BLM.33b Additionally, it was suggested that the actual
mechanism of cleavage by the two metallobleomycins is
different, an assertion that is consistent with differences in the
nature of the degradation products formed,34 as well as differ-
ences in the isotope effects associated with DNA degrada-
tion.22,35 Further, each metalloBLM cleaved DNA at sites not
cleaved by the other metalloBLM.31 The data argues strongly
that the two metallobleomycins utilize different chemistry for
DNA cleavage and exhibit different sequence selectivity for
DNA cleavage and that caution should be exercised when
making generalizations regarding commonalities in the behavior
of different metallobleomycin species.
Experimental Section
General Methods. Mass determination was accomplished by
electrospray ionization on a Finnigan 3200 Quadrupole mass spec-
trometer. HPLC purifications were performed on a Varian Associates
HPLC using an Altech Alltima C18 reversed phase column (250 × 10
mm, 5 µm). TentaGel resin, HBTU, HOBt, BOP, and NR-Fmoc-(S)-
threonine were purchased from Novabiochem. Microspin G-25 columns
were obtained from Qiagen. DMF was purchased from Acros Organics.
All other reagents were purchased from Aldrich Chemical Co. The 53-
nucleotide RNA substrate was purchased from Dharmacon Research.
All solvents were of analytical grade. Tetrahydrofuran was distilled
from potassium metal and benzophenone ketyl prior to use. Methanol
was dried over 3 Å molecular sieves for 48 h prior to use. All synthetic
transformations were carried out under dry argon or nitrogen. Dry resin
was swollen in CH2Cl2 and then in the appropriate reaction solvent
prior to use.
Yield determination for reactions performed on solid support was
accomplished by UV quantification of the dibenzylfulvene-piperidine
adduct formed upon treatment of the resin with piperidine.36 The molar
(25) Sucheck, S. J.; Ellena, J. F.; Hecht, S. M. J. Am. Chem. Soc. 1998, 120,
7450.
(26) Sucheck, S. J., Ph.D. Thesis, University of Virginia, 1998.
(27) Cagir, A.; Tao, Z.-F.; Leitheiser, C. J.; Sucheck, S. J.; Hecht, S. M. Bioorg.
Med. Chem., submitted.
(28) Leitheiser, C. J.; Smith, K. L.; Rishel, M. J.; Hashimoto, S.; Konishi, K.;
Thomas, C. J.; Li, C.; McCormick, M. M.; Hecht, S. M. J. Am. Chem.
Soc. 2003, 125, 8218.
(29) Kane, S. A.; Hecht, S. M.; Sun, J.-S.; Garestier, T.; He´le`ne, C. Biochemistry
1995, 34, 16715.
(30) Williams, L. D.; Goldberg, I. H. Biochemistry 1988, 27, 3004.
(31) McLean, M. J.; Dar, A.; Waring, M. J. J. Mol. Recog. 1989, 1, 184.
(32) (a) Lehmann, T. E. J. Biol. Inorg. Chem. 2002, 7, 305. (b) Zhao, C.; Xia,
C.; Qunkai, M.; Fo¨rsterling, H.; DeRose, E.; Antholine, W. E.; Subczynski,
W. K.; Petering, D. H. J. Inorg. Biochem. 2002, 91, 259.
(33) (a) Chang, C.-H.; Meares, C. F. Biochemistry 1982, 21, 6332. (b) Chang,
C.-H.; Meares, C. F. Biochemistry 1984, 23, 2268.
(34) Saito, I.; Morii, T.; Sugiyama, H.; Matsuura, T.; Meares, C. F.; Hecht, S.
M. J. Am. Chem. Soc. 1989, 111, 2307.
(35) Worth, L., Jr.; Frank, B. L.; Christner, D. F.; Absalon, M. J.; Stubbe, J.;
Kozarich, J. W. Biochemistry 1993, 32, 2601.
(36) (a) Fields, G. B.; Noble, R. L. Int. J. Peptide Protein Res. 1990, 35, 161,
(b) Gordeev, M. F.; Luehr, G. W.; Hui, H. C.; Gordon, E. M.; Patel, D. V.
Tetrahedron 1998, 54, 15879.
9
10202 J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003

Conformationally Constrained Bleomycin Analogues
A R T I C L E S
absorptivities of 5540 M-1 at 290 nm and 7300 M-1 at 300 nm were
used to calculate the loading from a known weight of dry resin.
and subsequent wash procedure was repeated 3 times, and the resin
was then washed successively with three 20-mL portions of DMF, three
20-mL portions of CH2Cl2, and three 20-mL portions of DMF. An
aliquot of the resin was removed, and a qualitative Kaiser free amine
test was performed, indicating the presence of a free amine. The resin
was suspended in a solution of 110 mg (0.3 mmol) of the Fmoc
protected, constrained methylvalerate analogue A,13 106 mg (0.3 mmol)
of HBTU, 37 mg (0.3 mmol) of HOBt, and 95 µL (70 mg, 0.5 mmol)
of Hunig’s base in 2 mL of DMF. The reaction mixture was stirred for
30 min, and the solvent was then removed by filtration. The resin was
washed with three 20-mL portions of DMF, three 20-mL portions of
CH2Cl2, and three 20-mL portions of methanol. An aliquot of the resin
was removed, and a qualitative Kaiser free amine test was performed,
indicating the absence of a free amine. An aliquot of the putative
tetrapeptide 8a was cleaved from the resin using a solution of 2%
hydrazine in DMF. After removal of the solvent under diminished
pressure, the residue was dissolved in CF3COOH and the solution was
treated with diethyl ether to effect precipitation of the cleaved
product: mass spectrum (electrospray), m/z 824.4 (M + H)+, theoretical
m/z 825.0 (M + H)+. An aliquot of the resin was removed for Fmoc
cleavage analysis and revealed a loading of 0.08 mmol/g (corresponding
to a 45% yield from the tripeptide). The resin was dried under
diminished pressure.
Resin-Bound Dipeptide 5. To 500 mg of swollen Nbs and Boc
protected spermidine resin (3) was added a 10% solution of triiso-
propylsilane in 2 mL of CH2Cl2 and 1 mL of CF3COOH. The reaction
mixture was stirred for 1 h, and then the solvent was removed by
filtration and the entire process was repeated. The resin was washed
with a 10% solution of triisopropylsilane in CH2Cl2. The solvent was
removed by filtration and the resin was washed successively with three
20-mL portions of CH2Cl2, three 20-mL portions of DMF, and three
20-mL portions of a 10% solution of Hunig’s base in DMF. An aliquot
of the resin was removed, and a qualitative Kaiser free amine test37
was performed, indicating the presence of a free amine. The NBS-
protected spermidine resin 3 was suspended in a solution of 421 mg
(0.9 mmol) of Fmoc protected bithiazole 4, 341 mg (0.9 mmol) of
HBTU, and 240 µL (170 mg, 1.4 mmol) of Hunig’s base in 2 mL of
DMF. The reaction mixture was stirred for 30 min, and then the solvent
was removed by filtration and the resin was washed with three 20-mL
portions of DMF, three 20-mL portions of CH2Cl2, and three 20-mL
portions of methanol. An aliquot of the resin was removed, and a
qualitative Kaiser free amine test was performed, indicating the absence
of a free amine. An aliquot of the putative dipeptide 5 was cleaved
from the resin with a solution of 2% hydrazine in DMF. After removal
of the solvent under diminished pressure, the residue was dissolved in
CF3COOH and the solution was treated with diethyl ether to effect
precipitation of the cleaved product: mass spectrum (electrospray), m/z
568.3 (M + H)+, theoretical m/z 568.7 (M + H)+. An aliquot of the
resin was subjected to Fmoc cleavage analysis, which indicated a
loading of 0.24 mmol/g (corresponding to a 78% overall yield over
the first six steps). The resin was dried under diminished pressure.
Resin-Bound Pentapeptide 10a. To 200 mg (0.08 mmol/g) of
swollen tetrapeptide resin 8a was added 3 mL of a 20% solution of
piperidine in DMF. The reaction mixture was shaken for 10 min, and
the solvent was then removed by filtration and the resin was washed
with 10 mL of a 20% solution of piperidine in DMF. The piperidine
deblocking and subsequent wash procedure was repeated 3 times, and
the resin was then washed successively with three 20-mL portions of
DMF, three 20-mL portions of CH2Cl2, and three 20-mL portions of
DMF. An aliquot of the resin was removed, and a qualitative Kaiser
free amine test was performed, indicating the presence of a free amine.
The resin was suspended in a solution of 51 mg (0.08 mmol) of NR-
Fmoc-Nim-trityl-â-hydroxyhistidine (9),15 30 mg (0.08 mmol) of HATU,
11 mg (0.08 mmol) of HOAt, and 28 µL (21 mg, 0.16 mmol) of Hunig’s
base in 2 mL of DMF. The reaction mixture was stirred for 30 min,
and the solvent was then removed by filtration. The resin was washed
successively with three 20-mL portions of DMF, three 20-mL portions
of CH2Cl2, and three 20-mL portions of methanol. An aliquot of the
resin was removed, and a qualitative Kaiser free amine test was
performed, indicating the absence of a free amine. An aliquot of the
putative pentapeptide 10a was cleaved from the resin with a solution
of 2% hydrazine in DMF. After removal of the solvent under diminished
pressure, the residue was dissolved in CF3COOH and the solution was
treated with diethyl ether to effect precipitation of the cleaved
product: mass spectrum (electrospray), m/z 977.4 (M + H)+, theoretical
m/z 978.1 (M + H)+. Fmoc cleavage analysis on an aliquot of the resin
revealed a loading of 0.067 mmol/g (corresponding to an overall 93%
yield from the tetrapeptide). The resin was dried under diminished
pressure.
Resin-Bound Tripeptide 7. To 480 mg (0.24 mmol/g) of swollen
dipeptide resin 5 was added 3 mL of a 20% solution of piperidine in
DMF. The reaction mixture was shaken for 10 min, and the solvent
was then removed by filtration and the resin was washed with 10 mL
of a 20% solution of piperidine in DMF. The piperidine deblocking
and subsequent wash procedure was repeated 3 times, and the resin
was washed successively with three 20-mL portions of DMF, three
20-mL portions of CH2Cl2, and three 20-mL portions of DMF. An
aliquot of the resin was removed, and a qualitative Kaiser free amine
test was performed, indicating the presence of a free amine. The resin
was suspended in a solution of 230 mg (0.7 mmol) of NR-Fmoc-(S)-
threonine (6), 255 mg (0.7 mmol) of HBTU, 91 mg (0.7 mmol) of
HOBt, and 205 µL (152 mg, 1.2 mmol) of Hunig’s base in 2 mL of
DMF. The reaction mixture was stirred for 30 min, and the solvent
was then removed by filtration. The resin was washed with three 20-
mL portions of DMF, three 20-mL portions of CH2Cl2, and three 20-
mL portions of methanol. An aliquot of the resin was removed, and a
qualitative Kaiser free amine test was performed on this aliquot,
indicating the absence of a free amine. An aliquot of the putative
tripeptide 7 was cleaved from the resin with a solution of 2% hydrazine
in DMF. After removal of the solvent under diminished pressure, the
residue was dissolved in CF3COOH and the solution was treated with
diethyl ether to effect precipitation of the cleaved product: mass
spectrum (electrospray), m/z 669.4 (M+H)+, theoretical m/z 669.8 (M
+ H)+. An aliquot of the resin was removed for Fmoc cleavage analysis
and revealed a loading of 0.21 mmol/g (corresponding to an overall
90% yield from the dipeptide). The resin was dried under diminished
pressure.
Deglycobleomycin Analogue 2a. To 150 mg (0.067 mmol/g) of
swollen pentapeptide resin 10a was added 3 mL of a 20% solution of
piperidine in DMF. The reaction mixture was shaken for 10 min, and
the solvent was then removed by filtration and the resin was washed
with 10 mL of a 20% solution of piperidine in DMF. The piperidine
deblocking and subsequent wash procedure was repeated 3 times, and
the resin was then washed successively with three 20-mL portions of
DMF, three 20-mL portions of CH2Cl2, and three 20-mL portions of
DMF. An aliquot of the resin was removed, and a qualitative Kaiser
free amine test was performed, indicating the presence of a free amine.
The resin was suspended in a solution of 11 mg (0.03 mmol) of Boc
pyrimidoblamic acid (11),17 31 mg (0.07 mmol) of BOP reagent, and
20 µL (15 mg, 0.1 mmol) of Hunig’s base in 1 mL of DMF at 0 °C.
The reaction mixture was stirred for 12 h, and the solvent was then
removed by filtration. The resin was washed successively with three
Resin-Bound Tetrapeptide 8a. To 220 mg (0.18 mmol/g) of swollen
tripeptide resin 7 was added 3 mL of a 20% solution of piperidine in
DMF. The reaction mixture was shaken for 10 min, and the solvent
was then removed by filtration and the resin was washed with 10 mL
of a 20% solution of piperidine in DMF. The piperidine deblocking
(37) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Anal. Biochem.
1970, 34, 595.
9
J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003 10203

A R T I C L E S
Rishel et al.
20-mL portions of DMF, three 20-mL portions of CH2Cl2, and three
20-mL portions of methanol. The resin was dried under diminished
pressure giving putative resin-bound hexapeptide 12a. To 80 mg of
swollen resin 12a was added 0.1 mL of triisopropylsilane and 0.1 mL
of dimethyl sulfide. The reaction mixture was agitated for 30 s, and
then 1 mL of CF3COOH was added. The reaction mixture was shaken
for 3 h, and the solvent was then removed by filtration. The resin was
washed successively with three 20-mL portions of DMF, three 20-mL
portions of CH2Cl2, and three 20-mL portions of DMF. The resin was
suspended in 1.5 mL of a 1.0 M solution of sodium thiophenoxide in
DMF. The solution was shaken for 30 min, and the solvent was then
removed by filtration. The sodium thiophenoxide treatment was
repeated, and the solvent was once again removed by filtration. The
resin was washed successively with three 20-mL portions of DMF,
three 20-mL portions of CH2Cl2, and three 20-mL portions of DMF.
The product was cleaved from the resin by treatment with 10 mL of a
2% hydrazine solution in DMF. After removal of the solvent under
diminished pressure, the residue was dissolved in CF3COOH and the
product was precipitated by the addition of diethyl ether. The precipitate
was then dissolved in 0.1% aqueous CF3COOH and lyophilized,
yielding a colorless, solid product. Purification of the crude product
was effected by C18 reversed phase HPLC (250 × 10 mm, 5µ). Elution
with a linear gradient of 0.1% CF3COOH containing increasing amounts
of acetonitrile (13%f22% acetonitrile over a period of 25 min, at a
flow rate of 4 mL/min; tr 13.2 min) provided deglycobleomycin A5
analogue 2a as a colorless solid following lyophilization: yield 1.51
mg (26% for pyrimidoblamic acid coupling, deprotection, cleavage from
applied to a 1% agarose gel containing 0.5 µg/mL ethidium bromide.
Horizontal gel electrophoresis was carried out in 9 mM Tris-borate
buffer, pH 8.3, containing 320 µM disodium EDTA, at 156 W for 2 h.
The DNA bands were visualized under UV light.
Preparation of a 5′-32P End-Labeled DNA Restriction Fragment.
Plasmid pBR322 (25 µg) was incubated with 100 units of restriction
endonuclease HindIII in a 100 µL (total volume) reaction mixture
containing 5 mM NaCl, 1 mM Tris-HCl, pH 7.9, 1 mM MgCl2, and
0.1 mM dithiothreitol (DTT). The digestion reaction was carried out
at 37 °C for 3 h. The DNA was recovered by ethanol precipitation.
The linearized DNA was dephosphorylated with 1 unit of calf intestinal
alkaline phosphatase in a reaction mixture (200 µL total volume)
containing 10 mM NaCl, 5 mM Tris-HCl, pH 7.9, 1 mM MgCl2, and
0.1 mM DTT. The reaction mixture was incubated at 37 °C for 1 h.
The enzyme was inactivated by heating at 75 °C for 10 min in 5 mM
EDTA. The DNA was then recovered by ethanol precipitation following
phenol extraction. The dephosphorylated DNA was 5′-32P-end-labeled
by incubation with 20 units of T4 polynucleotide kinase in a reaction
mixture (50 µL total volume) containing 7 mM Tris-HCl, pH 7.6, 1
mM MgCl2, 0.5 mM DTT, and 0.5 mCi of [γ-32P]ATP. The reaction
mixture was incubated at 37 °C for 1 h. The enzyme was inactivated
by heating at 65 °C for 20 min. Unreacted [γ-32P]ATP was removed
by the use of a microspin G-25 column and the DNA was recovered
by ethanol precipitation. The 5′-32P end-labeled DNA was digested with
100 units of restriction endonuclease EcoRV in a reaction mixture (150
µL total volume) containing 10 mM NaCl, 5 mM Tris-HCl, pH 7.9, 1
mM MgCl2 and 0.1 mM DTT. The reaction medium was incubated at
37 °C for 3 h, and the enzyme was then inactivated by heating at 80
°C for 20 min. The sample was desalted using a microspin G-25 column.
The sample was applied to an 8% native polyacrylamide gel after
addition of 50 µL of loading dye (50% glycerol (w/v), 25 mM EDTA,
0.25% bromophenol blue). Electrophoresis was carried out at 10 W
for 3.5 h. The DNA was visualized by audioradiography, and the band
of interest was excised from the gel. The 158-base-pair 5′-32P end-
labeled DNA duplex was eluted into 2 M LiClO4 at 37 °C for 12 h
and finally recovered by ethanol precipitation.
1
the resin and HPLC purification); H NMR (D2O) δ 0.94 (d, 3H, J )
9.5 Hz), 1.46 (m, 4H), 1.59 (m, 4H), 1.75 (s, 3H), 1.87 (m, 3H), 2.42
(d, 1H, J ) 5.5 Hz), 2.45 (m, 1H), 2.56 (m, 1H), 2.83 (m, 4H), 2.94
(m, 4H), 3.05 (m, 2H), 3.37 (m, 3H), 3.46 (m, 1H), 3.78 (m, 2H), 3.92
(m, 1H), 3.98 (m, 2H), 4.05 (m, 1H), 4.72 (m, 3H), 5.08 (d, 2H, J )
12.5 Hz), 7.32 (s, 1H), 7.83 (s, 1H), 8.04 (s, 1H), and 8.55 (s, 1H);
mass spectrum (electrospray), m/z 1099.5 (M + H)+, theoretical m/z
1100.3 (M + H)+, HRMS (FAB), m/z 1099.5040 (M + H)+
(C46H71N18O10S2 requires 1099.5042).
Oxidation of Styrene by Fe(III)•BLM and Analogues. A reaction
mixture containing 0.4 mg (0.28 µmol) of bleomycin and 0.14 mg (0.28
µmol) of ferric perchlorate in 220 µL of water was treated with 100
µL of a methanol solution that was 277 mM in styrene and 6.21 mM
in ethyl benzoate. The resulting solution was cooled to 4 °C, and 62.7
µL of an aqueous 91.5 mM H2O2 stock solution was added to bring
the final concentration of the reaction mixture to 15 mM in H2O2; this
resulted in a total reaction volume of 383 µL. Aliquots (127 µL each)
were taken from the reaction mixture at 40-min intervals. Each aliquot
was diluted with 300 µL of H2O and then extracted with three 100-µL
portions of CHCl3. The combined organic extract was dried over
anhydrous MgSO4, filtered, and analyzed by gas chromatography.
Injections (10 µL, splitless mode) were applied to a 5% phenylsiloxane
capillary column (0.25 mm ID × 30 m). The products were detected
using a flame ionization detector. The following temperature program
was employed at a carrier gas (helium) flow rate of 1.0 mL/min: 60
°C for 22 min; 60 f 85 °C at 10 °C/ min; then 85 °C for 15 min.
Under these conditions, the observed retention times (calculated from
the elution of an injected air peak) were as follows: styrene, 4.6 min;
phenylacetaldehyde, 15.4 min; styrene oxide, 18.8 min; ethyl benzoate,
28.9 min.
Cleavage of 5′-32P End-Labeled DNA Duplex by Deglycobleo-
mycin Analogues. Reactions were carried out in 20 µL (total volume)
of 10 mM sodium cacodylate buffer, pH 7.0, containing 32P end-labeled
DNA (∼3 × 104 cpm) and the appropriate concentrations of BLM and
Fe2+. Reaction mixtures were incubated at 4 °C for 30 min and
lyophilized. The samples were dissolved in 5 µL of loading dye (80%
formamide, 2 mM EDTA, 1% (w/v) xylene cyanol and 1% (w/v)
bromophenol blue), heated at 90 °C for 10 min, and then chilled on
ice. The solutions were finally applied to a 10% denaturing polyacryl-
amide gel (7 M urea). Electrophoresis was carried out at 50 W for 2 h.
The gel was analyzed using a phosphoimager (Molecular Dynamics).
The bands were correlated with those produced according to a Maxam-
Gilbert A+G sequencing protocol.38
Alkali-Induced Cleavage of 5′-32P End-Labeled DNA Duplex by
Deglycobleomycin Analogues. Reactions were carried out in 20 µL
(total volume) of 10 mM sodium cacodylate, pH 7.0, containing 32P
end-labeled DNA (∼3 × 104 cpm) and the appropriate concentrations
of BLM and Fe2+. Reaction mixtures were incubated at 37 °C for 30
min, treated with 0.1 M piperidine for 20 min at 90 °C, and then
lyophilized. The samples were dissolved in 5 µL of loading dye (80%
formamide, 2 mM EDTA, 1% (w/v) xylene cyanol and 1% (w/v)
bromophenol blue), heated at 90 °C for 10 min, and then chilled on
ice. The solutions were applied to a 10% denaturing polyacrylamide
gel (7 M urea). Electrophoresis was carried out at 50 W for 2 h. The
gel was analyzed using a phosphorimager (Molecular Dynamics). The
bands were correlated with those produced according to a Maxam-
Gilbert A+G sequencing protocol.38
Relaxation of Supercoiled DNA by Modified Deglycobleomycin
A5 Analogues. Reactions were carried out in 25 µL (total volume) of
10 mM sodium cacodylate buffer, pH 7.0, containing 300 ng of pBR322
plasmid DNA and the appropriate concentrations of BLM and Fe2+
.
The Fe2+ solutions were freshly prepared from FeII(NH4)2(SO4)2•6H2O
for all experiments. Reaction mixtures were incubated at 37 °C for 30
min. The reactions were quenched by the addition of 5 µL of loading
dye (30% glycerol containing 0.125% (w/v) bromophenol blue) and
(38) Maxam, A. M.; Gilbert, W. Methods Enzymol. 1980, 65, 499.
9
10204 J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003

Conformationally Constrained Bleomycin Analogues
A R T I C L E S
Cleavage of a 53-Nucleotide 5′-32P End-Labeled RNA by Deglyco-
bleomycin A5 Analogues. The 53-nt RNA was radiolabeled in
essentially the same manner as described.39 RNA cleavage reactions
were carried out in 10 µL (total volume) of 10 mM sodium phosphate,
pH 7.0, containing 5′-32P end-labeled RNA (5 × 104 cpm) and the
appropriate concentrations of reagents as indicated in the legend to
Figure 8. Each reaction mixture was incubated at 23 °C for 30 min
and then frozen in dry ice, lyophilized, and dissolved in a gel loading
solution (80% formamide, 2 mM EDTA, 0.5% (w/v) xylene cyanol,
and 0.5% (w/v) bromophenol blue). The resulting solution was heated
at 70 °C for 3 min, chilled on ice, and then applied to a 20% denaturing
polyacrylamide gel. The gel was analyzed by the use of a phosphor-
imager.
Acknowledgment. This work was supported by NIH Re-
search Grants CA76297 and CA77284 awarded by the National
Cancer Institute. High resolution mass spectral data were
obtained at the Michigan State University Mass Spectrometry
Facility which is supported, in part, by a grant (DRR-00480)
from the Biotechnology Research Technology Program, Na-
tional Center for Research Resources, National Institutes of
Health.
Supporting Information Available: Synthetic procedures for
valerate analogues A-F, 8b-8f, 10b-10f, and 2b-2f. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(39) Holmes, C. E.; Duff, R. J.; van der Marel, G. A.; van Boom, J.; Hecht, S.
M. Bioorg. Med. Chem. 1997, 5, 1235.
JA030057W
9
J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003 10205