Pyrrolobenzodiazepine dimers: novel sequence

Journal of Antimicrobial Chemotherapy (2005) 56, 513–518
doi:10.1093/jac/dki256
Advance Access publication 15 July 2005
Pyrrolobenzodiazepine dimers: novel sequence-selective,
DNA-interactive, cross-linking agents with activity against
Gram-positive bacteria
Tsveta Hadjivassileva, David E. Thurston and Peter W. Taylor*
School of Pharmacy, 29–39 Brunswick Square, London WC1N 1AX, UK
Received 31 March 2005; returned 17 June 2005; revised 21 June 2005; accepted 23 June 2005
Objectives: Pyrrolo[2,1-c][1,4]benzodiazepine (PBD) dimers are synthetic sequence-selective interstrand
DNA minor-groove cross-linking agents developed from anthramycins. We investigated the antibacterial
activity of three dimers, SJG-136, DRG-16 and ELB-21, which differ in the structure of the PBD monomeric
unit and the length of the linker region between the two identical PBD monomers.
Methods: MICs were determined against 38 methicillin-resistant Staphylococcus aureus (MRSA), 20
vancomycin-resistant enterococci (VRE), 12 isolates of Streptococcus pyogenes, 12 of Streptococcus
agalactiae, 12 of Listeria monocytogenes and 24 Gram-negative clinical isolates. Binding to doublestranded DNA was assessed by determination of the DNA melting temperature (Tm).
Results: MIC90 values for SJG-136 were 0.5 mg/L against MRSA, VRE and L. monocytogenes, 0.06 mg/L
against S. pyogenes and 0.03 mg/L against S. agalactiae; these were below the maximum tolerated dose of
the drug. MIC90s for DRG-16 were 0.125, >0.5, 0.125, 0.015 and <0.008 mg/L, respectively. The most potent
compound was ELB-21, with corresponding MIC90 values of 0.03, 0.06, 0.06, 0.015 and 0.015 mg/L. There was
little or no variation in sensitivity amongst isolates from any one species. All Gram-negative species
(Acinetobacter, Pseudomonas, Klebsiella, Proteus spp.) were not susceptible due to the barrier function
of the outer membrane. PBD dimers showed bactericidal activity against MRSA and VRE and there was a
significant post-antibiotic effect (1.5–3.5 h). Incubation of EMRSA-16 genomic DNA (50 mM) with 20 mM ELB21 resulted in a large increase in Tm suggesting that PBD dimers exert their antibacterial effect by crosslinking of the two DNA strands.
Conclusions: These data indicate that this novel class of antibacterial agents warrants further
investigation as potential antibiotics for the treatment of severe infections caused by Gram-positive
pathogens.
Keywords: DNA cross-linking agents, pyrrolobenzodiazepine dimers, Staphylococcus aureus,
vancomycin-resistant enterococci, MRSA
Introduction
Over the last two decades, drug-resistant Gram-positive pathogens,
particularly methicillin-resistant Staphylococcus aureus (MRSA),
multiresistant coagulase-negative staphylococci, penicillinresistant Streptococcus pneumoniae and vancomycin-resistant
enterococci (VRE), have emerged worldwide. Although panresistant Gram-positive isolates have yet to appear, MRSA and
VRE infections in particular are becoming more difficult to
treat.1 The recent introduction of novel antibacterial agents such
as quinupristin/dalfopristin (type B/type A streptogramins) and
linezolid (an oxazolidinone) has provided efficacious and safe
treatment options for MRSA and VRE infections, but there remains
an urgent need for new therapies, particularly as some VRE strains
are already showing decreased susceptibility to these agents.2,3 As
therapeutic options decrease, there has been recent interest in
the development of agents that exert their antibacterial effects
by binding selectively to specific sequences of duplex DNA.
For example, small molecules derived from polyheterocyclic antibiotics based on distamycin have emerged recently as candidates
for clinical development.4,5 These molecules are active against
Gram-positive isolates such as streptococci, staphylococci and
.............................................................................................................................................................................................................................................................................................................................................................................................................................
*Corresponding author. Tel/Fax: +44-20-7753-5867; E-mail: [email protected]
.............................................................................................................................................................................................................................................................................................................................................................................................................................
513
The Author 2005. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: [email protected]
Hadjivassileva et al.
enterococci and exert their bactericidal effects by binding within
the minor groove of DNA at A:T-rich sites commonly found in
bacterial promoters and origins of replication. Their in vitro
potencies are comparable to vancomycin and linezolid, they
show efficacy against MRSA in murine models of acute infection
and they display no genotoxicity.5
The pyrrolo[2,1-c][1,4]benzodiazepines (PBDs) are a family of
DNA-interactive antibiotics derived from various Streptomyces
species and examples include antitumour agents such as DC-81
and tomaymycin.6 They exert their biological activity through
covalent binding to the C2-NH2 of guanine within the minor groove
of DNA (Figure 1a). PBD monomers span three DNA base pairs
with a preference for Pu–G–Pu (where Pu = purine; G = guanine)
sequences and block transcription by inhibiting RNA polymerase
activity in a sequence-specific manner.7 The potency, DNAbinding affinity and sequence specificity of PBDs have been
markedly improved by linking two PBD units through their
C8/C800 positions to form a PBD dimer that can cross-link appropriately separated guanines on opposite DNA strands.8 PBD units
tethered through a three-carbon diether linkage are able to bind
selectively to regions of DNA that are 6–7 bp in length. For
example, the PBD dimer SJG-136 (Figure 1b) predominantly
binds covalently to embedded 50 -Pu-GATC-Py sequences by
cross-linking opposite-strand guanines separated by 2 bp, with little
or no disruption of DNA secondary structure.8,9 Sequence-selective
recognition can be further enhanced by extension of the linker
joining the two PBD monomers. Thus, use of a five-carbon diether
linker (Figure 1) facilitates selective binding to include an additional base pair and effects cross-linking of 50 -Pu-GATTC-Py and
50 -Pu-GAATC-Py sequences.8 In this study, we report the in vitro
antibacterial activity of the three PBD dimers SJG-136, DRG-16
and ELB-21.
Materials and methods
Materials
PBD dimers SJG-136, DRG-16 and ELB-21 (Figure 1) were designed
and synthesized according to previously described procedures.10,11
Paper discs impregnated with oxacillin, vancomycin, trimethoprim,
amikacin, gentamicin, tetracycline, rifampicin, ciprofloxacin,
erythromycin and clindamycin were used to determine bacterial
susceptibility to these antibiotics and were purchased from Oxoid
(Basingstoke, UK).
Bacterial strains
The international collection of 38 MRSA clinical isolates12 used in
this study was assembled at the Royal Free and University College
Medical School (RFUCMS), London. The 20 VRE, 12 Streptococcus
pyogenes, 12 Streptococcus agalactiae, 12 Listeria monocytogenes,
6 Klebsiella spp., 6 Pseudomonas aeruginosa, 6 Acinetobacter spp.,
and 6 Proteus mirabilis strains were isolated from clinical specimens
at the RFUCMS and were the gift of Professor Jeremy HamiltonMiller. Mycobacterium aurum, M. smegmatis, M. fortuitum,
M. abscessus, and M. phlei were obtained from previously described
sources.13
MIC assays
In vitro susceptibilities were determined by the NCCLS broth
microplate assay; 0.1 mL of dilutions of the test compound in
O
(a)
N
HN
H2N
..
2
O
N
N
DNA
H
11
A
H
N
H
N
N
R⬘
N
N
DNA
H
R⬘
B
N C
R⬙
R⬙
N
O
O
(b)
(CH2)n
H
N
11 10
8
O
O
8⬘
N
10⬘
11⬘
11a
H
11a⬘
N
R1
N
HN
OCH3
R1⬘
N
H3CO
O
O
R2
SJG-136: n = 1, R1/R1⬘ = R2/R2⬘ = H
DRG-16: n = 3, R1/R1⬘ = R2/R2⬘ = H
ELB-21: n = 3, R1/R1⬘ = CH3, R2/R2⬘ = H
R2⬘
Figure 1. PBDs bind covalently to duplex DNA: (a) mechanism of binding via N10-C11 imine/carbinolamine moiety to the C2-amino position of a guanine residue
within the minor groove of DNA. (b) Tethering of two identical PBD units through a C8/C8 diether linker improves DNA-binding efficiency.
514
Antibacterial pyrrolobenzodiazepine dimers
Mueller–Hinton broth (Oxoid) were added to 0.1 mL aliquots of
bacterial suspension to give a final bacterial count of 105 cfu/mL.
MIC determinations for mycobacterial strains were performed as previously described.13 PBD dimers were dissolved in methanol prior to
dilution in broth; appropriate concentrations of methanol were incorporated into controls to ensure that the low concentrations of solvent
used had no effect on the growth of test strains.
5 computer package (MicroCal Inc, Northampton, MA, USA) for
analysis and the melting temperature Tm determined at the midpoint
of the normalized melting profiles using a published procedure.16 PBD
dimers (20 mM) were added to the DNA and incubated at 37 C for 1 h
prior to determination of the melting profile.
Kinetics of bacterial killing
In vitro antibacterial activity
Mid-logarithmic or stationary phase bacteria in Mueller–Hinton
broth (106 cfu/mL) were exposed to appropriate concentrations
(MIC; MIC ·2; MIC ·4) of PBD dimers and viable counts performed
at various time points following incubation with agitation (200 orbits
per min on an orbital shaker) at 37 C. Experiments were carried out in
duplicate.
We determined the antibacterial activity of three PBD dimers
(SJG-136, DRG-16 and ELB-21) with differing potencies and
sequence selectivities against 94 Gram-positive isolates, 30 Gramnegative isolates and five atypical mycobacteria (Table 1). All
Gram-negative isolates were insensitive to all three dimers, with
MIC90 values greater than 16 mg/mL; strains could be sensitized by
perturbation of the outer membrane with 1 mM EDTA, indicating
that this structure precluded the agents from entering the cell. In
contrast, all Gram-positive isolates were susceptible to the three
dimers with MIC90 values below 1 mg/L in all cases. The most
potent compound was ELB-21, with MIC90 values of 0.03 mg/L for
MRSA and 0.06 mg/L for VRE. Overall, DRG-16 proved more
potent than SJG-136 and all three compounds showed good activity
against mycobacteria. There was little or no variation in susceptibility to each compound among isolates from any one species;
this is exemplified in Table 1, which summarizes MIC data for
the 38 MRSA isolates. Disc diffusion assays using the antibiotics
listed in the Materials and methods section provided no evidence
of cross-resistance between the three PBD dimers and other antibiotics (data not shown).
Electron microscopy
Negatively stained images were obtained as described previously.14
Post-antibiotic effect (PAE)
Mid-logarithmic phase bacteria growing in Mueller–Hinton broth
were incubated at 37 C in the presence of various concentrations of
the PBD dimer for 1 h. The agent was removed by centrifugation
and the cells washed with pre-warmed PBS. The pellet was
suspended in pre-warmed Mueller–Hinton broth, incubated at 37 C
and bacterial numbers determined over a 24 h period by viable counting. The PAE was calculated by the standard equation T–C, where T
is the time taken for the viable count of the compound-treated
culture to increase 10-fold after removal of the agent and C is the
time taken for the untreated control to increase 10-fold under the
same conditions.
Melting curves of duplex DNA
Genomic DNA was prepared from MRSA strains by guanidium
thiocyanate extraction according to Pitcher et al.15 The DNA was
washed with ethanol and dissolved in 10 mM Na2HPO4 containing
1 mM EDTA (pH 7.0) to give a concentration of 50 mM. Melting
curves were determined spectrophotometrically (l260) in a heated
chamber with a controlled increase in temperature of 1 C/min
over the range 40–98 C. Optical data were imported into the Origin
Results
Bactericidal activity of PBD dimers
The capacity of ELB-21 to kill EMRSA-16 and VRE was determined. As can be seen from Figure 2, ELB-21 exhibited potent
bactericidal activity against logarithmic phase EMRSA-16 cells.
Stationary phase cells were also efficiently killed by ELB-21 and
the curves obtained were similar to those for logarithmic phase
cells (data not shown). The re-growth that occurred after 24 h
was not due to the emergence of resistant genotypes. Logarithmic
phase VRE cells were highly susceptible to killing by ELB-21,
with a five-log reduction in viability after 4 h incubation at the
Table 1. MIC values (mg/L) of PBD dimers against various bacterial species
MIC90
Bacterial isolates
MRSA (range)
VRE
S. pyogenes
S. agalactiae
L. monocytogenes
Citrobacter spp.
Klebsiella spp.
Proteus spp.
P. aeruginosa
Mycobacterium spp.
No. of strains
SJG-136
DRG-16
ELB-21
38
20
12
12
12
6
6
6
12
5
0.5 (0.125–0.5)
0.5
0.06
0.03
0.5
>32
>32
>32
>32
2
0.125 (0.015–0.125)
0.5
0.015
<0.008
0.125
>16
>16
>16
>16
1
0.03 (0.008–0.06)
0.06
0.015
0.015
0.06
16 (0.5)a
16 (2)
16 (1)
16 (1)
0.5
a
MICs obtained for Gram-negative bacteria in the presence of 1 mM EDTA in parentheses.
515
Hadjivassileva et al.
12
(a)
Log10 cfu/mL
10
8
6
4
2
0
0
2
4
6
8
10
12
14
Time (h)
16
18
20
22
24
0
2
4
6
8
10
12
14
Time (h)
16
18
20
22
24
10
(b)
Log10 cfu/mL
8
6
4
2
0
Figure 2. In vitro killing kinetics of ELB-21 against (a) logarithmic phase EMRSA-16 and (b) a logarithmic phase VRE isolate. The MICs for EMRSA-16 and VRE
were 0.015 and 0.06 mg/L, respectively. Open squares, control; filled squares, MIC; filled triangles, twice the MIC; open circles, four times the MIC. Values shown are
the averages of duplicate experiments and error bars indicate the range of values.
(a)
(b)
Figure 3. Sections of glutaraldehyde-fixed EMRSA-16 stained with uranyl acetate and Reynold’s lead citrate. (a) Untreated cells and (b) cells incubated with ELB-21
at the MIC for 4 h.
MIC. Incubation of EMRSA-16 with ELB-21 at the MIC for 4 h
resulted in a marked degree of morphological heterogeneity and
drug-induced lysis (Figure 3), suggesting that ELB-21 inhibits the
transcription of a multiplicity of key genes.
PAE
In vitro PAEs for EMRSA-16 and VRE were determined against
ELB-21 over a range of concentrations. At the MIC, the PAEs
for the two strains were 1.5 and 3.5 h, respectively. The PAE
516
Antibacterial pyrrolobenzodiazepine dimers
for VRE could not be determined at concentrations above the MIC
due to rapid bacterial killing at these concentrations. PAEs against
MRSA-16 at twice and four times the MIC were 1.5 and 3.5 h,
respectively.
DNA melting
Covalent cross-linking of the two strands of the DNA duplex will
increase the temperature required to denature the complex;17 melting of the complex results in a large increase in absorbance that can
be monitored spectrophotometrically. Addition of 20 mM of either
DRG-16 or ELB-21 to genomic EMRSA-16 DNA had a profound
effect on the Tm: DRG-16 raised the Tm by, typically, 28.07 C
(from 66.70 to 94.77 C) and ELB-21 by 30.81 C (error –0.1 C).
Discussion
The large majority of currently useful antibacterial agents have
been derived from natural products produced by microorganisms
forced to compete for nutrients in hostile habitats such as soil. The
majority of targets for clinically useful natural and semi-synthetic
antibiotics are enzymes or structural proteins involved in essential
pathways of macromolecule synthesis and most of these have been
extensively exploited for antibiotic development. Naturally occurring antibacterial agents such as distamycin, anthramycin and chromomycin that inhibit gene transcription by binding selectively to
specific sequences of duplex DNA have been known for many
years18 but it is only recently that efforts have been made to develop
such agents as antibacterial drugs. Even though the simplest DNA
binding antibiotics show selectivity for various sequence motifs,18
they are generally cytotoxic and appear unattractive as candidates
for the treatment of infection in humans. However, the increasing
prevalence of multidrug resistance, associated, in particular, with
nosocomial pathogens has created circumstances under which such
molecules may have therapeutic utility.
PBD dimers are attracting interest as potential therapeutics for
the treatment of various cancers due to their capacity to recognize
and cross-link DNA in a sequence-selective manner. In particular,
the PBD dimer SJG-136, which produces interstrand cross-links
at embedded Pu-GATC-Py target sites within duplex DNA,8,9 is
currently in Phase I clinical development.19 This compound, comprised of two PBD units tethered through a three-carbon diether
linkage (Figure 1), possesses potent bactericidal activity against the
five species of Gram-positive bacteria used in this study (Table 1)
and, in particular, all MRSA and VRE isolates proved to be highly
susceptible. Searching the EMRSA-16 genome20 using the Artemis
genome viewer (www.sanger.ac.uk/Software/Artemis) indicated
that the preferred SJG-136 binding sequence AGATCT8,9 occurs
162 times within coding sequences, including pseudo-genes, and
109 times within non-coding regions of the EMRSA-16 genome;
cross-linking at these sites is highly likely to account for its
observed in vitro potency.
Loadman and coworkers have shown that after intraperitoneal
administration of SJG-136 to NCR/Nu mice at the MTD (0.2 mg/
kg), SJG-136 shows a favourable pharmacokinetic profile:19 high
peak plasma concentrations of Cmax = 336 nM at 30 min after
dosing were observed. A calculated terminal t1/2 of 0.98 h and
AUC of 0.34 mM·h resulted in a clearance rate of 17.7 mL/min
per kg. Furthermore, the dimer appeared to bind only moderately
to plasma proteins. These data indicate that a therapeutic index for
BPD dimers as antibacterial agents is achievable, at least in the case
of SJG-136, as plasma drug concentrations are more than 10-fold
higher than those required for antibacterial efficacy in vitro. These
authors define the term ‘maximum tolerated dose’ that we used in
the context of this study.
The PBD units within DRG-16 and ELB-21 are tethered through
a five-carbon diether linker, rather than three as is the case
for SJG-136 (Figure 1), and both are thought to recognize and
bind to Pu-GAATC-Py sequences;8 these compounds, particularly
ELB-21, are more potent than SJG-136 (Table 1). In contrast to
distamycin-based polyheterocyclic antibiotics that, like vancomycin, are bactericidal against staphylococci but bacteriostatic
against enterococci,4 PBD dimers are bactericidal against both
MRSA and VRE with a significant PAE. Modelling of the interstrand DNA minor-groove cross-linked adducts formed with
PBD dimers indicates that there is little or no disruption of
DNA secondary structure including base-mispairing and that
induced distortion effects are absent.10 As these compounds are
well accommodated within the minor groove such that little of
either PBD unit remains exposed beyond the periphery of the
host DNA duplex, the complexes are resistant to repair enzymes
that are reliant on tracing distortion or helical perturbation in
DNA. Both halves of the dimer behave in an essentially identical
fashion without disruption of DNA integrity, and the flexible
regions linking the two subunits make favourable non-bonding
contacts with the floor and walls of the minor groove to enhance
overall binding.10 It is likely that the failure of repair surveillance
mechanisms to detect sites of cross-linking accounts in large part
for the bactericidal nature of these molecules as interstrand crosslinks, if not repaired, will interfere with the process of DNA replication and result in cell death.
In conclusion, the preliminary data presented here indicate that
these compounds exert their antibacterial effects by cross-linking
duplex DNA at selected sites on the bacterial genome and these
observations are currently being extended using a variety of
techniques to quantify the extent and selective nature of the
cross-linking process. The compounds appear to cross the bacterial
cytoplasmic membrane efficiently but are unable to traverse the
outer membrane of Gram-negative bacteria. The molecular weights
of these relatively hydrophilic compounds fall within the 500–
600 Da range and it is not therefore surprising that they are
excluded from passage through the protein channels of this barrier.21 There may, however, be scope for synergy with compounds
that disrupt the integrity of the outer membrane.
The basis of the increase in antibacterial potency of PBD dimers
resulting from extension of the linker region is at present unclear
and we are investigating whether these changes are accompanied
by increases in cytotoxicity or genotoxicity. The potency of this
new class of agents against Gram-positive bacteria is high and there
appears to be an opportunity for the design of sequence-selective
dimers that have a high affinity for motifs within the DNA of key
Gram-positive pathogens.
Acknowledgements
We thank Tony Reszka for valuable help with DNA melting
experiments and Spirogen Ltd. for financial support. P. W. T. thanks
the EPSRC for a QUOTA studentship award.
References
1. Levy SB, Marshall B. Antibacterial resistance worldwide: causes,
challenges and responses. Nature Med 2004; 10: S122–9.
517
Hadjivassileva et al.
2. Baysallar M, Kilic A, Aydogan H et al. Linezolid and quinupristin/
dalfopristin resistance in vancomycin-resistant enterococci and
methicillin-resistant Staphylococcus aureus prior to clinical use in Turkey.
Int J Antimicrob Agents 2004; 23: 510–2.
3. Raad II, Hanna HA, Hachem RY et al. Clinical-use-associated
decrease in susceptibility of vancomycin-resistant Enterococcus faecium
to linezolid: a comparison with quinupristin–dalfopristin. Antimicrob
Agents Chemother 2004; 48: 3583–5.
4. Ge Y, Difuntorum S, Touami S et al. In vitro antimicrobial activity
of GSQ1530, a new heteroaromatic polycyclic compound. Antimicrob
Agents Chemother 2002; 46: 3168–74.
5. Gross M, Bürli R, Jones P et al. Pharmacology of novel heteroaromatic polycycle antibacterials. Antimicrob Agents Chemother 2003;
47: 3448–57.
6. Thurston DE. Advances in the study of pyrrolo[2,1-c][1,4]-benzodiazepine antitumour antibiotics. In: Waring MJ, Neidle S, eds. Molecular
Aspects of Anticancer Drug–DNA Interactions. London: Macmillan, 1993;
54–88.
7. Puvvada MS, Forrow SA, Hartley JA et al. Inhibition of bacteriophage T7 RNA polymerase in vitro transcription by DNA-binding
pyrrolo[2,1-c][1,4]-benzodiazepines. Biochemistry 1997; 36: 2478–84.
8. Smellie M, Bose DS, Thompson AS et al. Sequence-selective
recognition of duplex DNA through covalent interstrand cross-linking:
kinetic and molecular modelling studies with pyrrolobenzodiazepine
dimers. Biochemistry 2003; 42: 8232–9.
9. Martin C, Ellis T, McGurk CJ et al. Sequence-selective interaction
of the minor-groove interstrand cross-linking agent SJG-136 with naked
and cellular DNA: footprinting and enzyme inhibition studies. Biochemistry
2005; 44: 4135–47.
10. Gregson SJ, Howard PW, Hartley JA et al. Design, synthesis, and
evaluation of a novel pyrrolobenzodiazepine DNA-interactive agent with
highly efficient cross-linking ability and potent cytotoxicity. J Med Chem
2001; 44: 737–48.
11. Gregson SJ, Howard PW, Gullick DR et al. Linker length modulates
DNA cross-linking reactivity and cytotoxic potency of C8/C80 ether-linked
C2-exo-unsaturated pyrrolo[2,1-c][1,4]benzodiazepine (PBD) dimers.
J Med Chem 2004; 47: 1161–74.
12. Stapleton PD, Shah S, Anderson JC et al. Modulation of b-lactam
resistance in Staphylococcus aureus by catechins and gallates. Int J
Antimicrob Agents 2004; 23: 462–7.
13. Seidel V, Taylor PW. In vitro activity of extracts and constituents
of Pelargonium against rapidly growing mycobacteria. Int J Antimicrob
Agents 2004; 23: 613–9.
14. Stapleton PD, Shah S, Hamilton-Miller JMT et al. AntiStaphylococcus aureus activity and oxacillin resistance modulating
capacity of 3-O-acyl-catechins. Int J Antimicrob Agents 2004; 24:
374–80.
15. Pitcher DG, Saunders NA, Owen RJ. Rapid extraction of bacterial
genomic DNA with guanidium thiocyanate. Lett Appl Microbiol 1989; 8:
151–6.
16. Jones GB, Davey CL, Jenkins TC et al. The non-covalent interaction
of pyrrolo[2,1-c][1,4]benzodiazepine-5,11-diones with DNA. Anti-Cancer
Drug Des 1990; 5: 249–64.
17. James PL, Le Strat L, Ellervik U et al. Effects of a hairpin polyamide
on DNA melting: comparison with distamycin and Hoechst 33258.
Biophys Chem 2004; 111: 205–12.
18. Dervan PB. Molecular recognition of DNA by small molecules.
Bioorg Med Chem 2001; 9: 2215–35.
19. Wilkinson GP, Taylor JP, Shnyder S et al. Preliminary pharmacokinetic and bioanalytical studies of SJG-136 (NSC 694501), a sequenceselective pyrrolobenzodiazepine dimer DNA-cross-linking agent. Invest
New Drugs 2004; 22: 231–40.
20. Holden MT, Feil EJ, Lindsay LA et al. Complete genomes of
two clinical Staphylococcus aureus strains: evidence for the rapid
evolution of virulence and drug resistance. Proc Natl Acad Sci USA
2004; 101: 9786–91.
21. Nikaido H. Molecular basis of bacterial outer membrane permeability. Microbiol Mol Biol Rev 2003; 67: 593–656.
518