Cisplatin Cytotoxicity Associated with Tetracycline Resistance Determinants in
Escherichia colil
By
Doriana Froim
B.S.-Biochemistry
Brandeis University, 1994
Submitted to the Biological Engineering Division in Partial Fulfillment of the Requirements for
the Degree of
Doctor of Philosophy in Molecular and Systems Toxicology and Pharmacology
at the
Massachusetts Institute of Technology
February 2005
C 2005 Massachusetts Institute of Technology
All rights reserved
Signature
ofAuthor:
...........
..
f..
.. ...
a ........................................
.....
Biological Engineering Division
December 21St,2004
Certified by:
Accepted
....................................................
Dr. John M. Essigmann
Professor of Toxicology and Chemistry
Thesis Supervisor
....................
by:......
Dr. Bevin P. Engelward
Chairman of the Thesis Committee
Associate Professor of Toxicology
Accepted by: .... ...
.................................................................................
Dr. Peter C. Dedon
Professor of Toxicology
Acceptedby:.....................................................................................................
Dr. Ram Sasisekharan
rMASaACHUS~
rrSINsrtEn-g
OF TECHNOLOGY
JAN 2 5 2005RIES
LIBRARIES ..
Professor of Biological Engineering
Cisplatin Cytotoxicity Associated with Tetracycline Resistance Determinants in
Escherichia coli
by
Doriana Froim
Submitted to the Biological Engineering Division on January 7,th 2005
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy in Molecular and Systems Toxicology and Pharmacology
Abstract
Tetracyclines, a broad-spectrum class of antibiotics, were discovered in the late 1940s,
and became widely used because of their important advantages: they are inexpensive, safe,
demonstrate good oral absorption, and are active against a broad range of bacterial pathogens.
Unfortunately, as with most antibiotics, the emergence of microbial resistance to tetracyclines has
become a serious problem. Today, most genera examined have tetracycline-resistant isolates,
although the percentage varies according to species and geographic location. Due to the
emergence of resistance, tetracyclines are no longer the antibiotics of choice in treatment of many
conditions, although they are still extensively used to treat a variety of bacterial infections.
Substantial research efforts have been directed towards reversing tetracycline resistance in
bacteria.
This work describes the development of a novel anti-bacterial treatment for diseases
caused by bacteria resistant to tetracycline. It was found that tetracycline-resistant bacteria
expressing the TnlO gene of tetracycline resistance, upon induction with tetracycline, became
extremely susceptible to destruction by the DNA-damaging anti-cancer drug cisplatin.
Tetracycline-resistant bacteria grown in tetracycline and subsequently treated with cisplatin in the
presence of tetracycline were killed about 105 -fold more effectively than wild-type bacteria and
tetracycline-resistant bacteria not exposed to tetracycline. This phenomenon was observed in
different strains of tetracycline-resistant E. coli. Other antibiotics tested with respective antibioticresistant bacteria did not produce the same effect of sensitization to cisplatin, suggesting a unique
relationship among cisplatin, tetracycline and the tetracycline resistance gene.
It was determined that levels of platinum DNA damage were higher in sensitized
tetracycline-resistant cells than in wild-type cells, although total cellular platinum levels in
sensitized tetracycline-resistant cells were not increased. At this time, the mechanism of increased
DNA damage formation and the mechanism underlying sensitization to cisplatin are still matters
of speculation. The experiments reported here, however, demonstrate that cells expressing the
genes of tetracycline resistance actually became primary targets for destruction by cisplatin.
Based on this study, it is suggested that the therapeutic power of the tetracyclines could be
restored and enhanced by using a complementary drug that, in combination with tetracycline,
would induce selective destruction of tetracycline-resistant bacteria.
Thesis Supervisor: John M. Essigmann
Title: Professor of Toxicology and Chemistry
2
Table of Contents.
Title Page....................................................................................................
Abstract .................................................................
2.........................
Table of Contents .........
3
. ...........................................................................
Chapter 1. Cisplatin: Lessons Learned from Bacteria .................................................
5
1.1. Discovery of Cisplatin Potential through Experiments in Bacteria .................... 5
1.2. Cisplatin Journey: from Bacteria to Clinic . ...............................................
6
1.3. Lessons Learned from Bacteria: Replication, Mutagenesis, and DNA Repair......8
1.4. Current Status of Cisplatin in Anti-Cancer Therapy ....................................
14
1.5. Structures of Platinum Compounds ......................................................
15
References........................................................................................
Chapter 2. Tetracycline Resistance in Bacteria .
......................................................
2.1. Tetracyclines in Clinical Practice ..........................................................
16
20
20
2.2. Tetracycline Resistance in Bacteria ......................................................
22
2.3. Current Research to Overcome Tetracycline Resistance..............................
25
2.4. Novel Approach to Tetracycline Resistance .............................................
26
2.5. Structures of Tetracycline and its Analogs.........
..........
..........................
28
References ..................................................................
29
Chapter 3. Discovery of Connection between Tetracycline Resistance and Cisplatin
Cytotoxicity in Escherichia coli ................................................................
36
Figures
.................................................................
38
Chapter 4. Tetracycline Resistance Determinant from Transposon TnlO Confers High
Sensitivity to Cisplatin in Escherichia coli upon Induction with Tetracycline..
.......... 44
4.1. Abstract .................................................................
44
4.3.
4.4.
4.5.
4.6.
4.7.
4.8.
47
49
52
57
58
72
4.2.Introduction
......... . .......................................................45
Materials and Methods .................................................................
Results .................................................................
Discussion.................................................................
Acknowledgements .................................................................
Figures .................................................................
References .................................................................
Chapter 5. Future work .................................................................
References
.................................................................
3
78
82
Appendices
........................................
4
Appendix I. Analysis of the "Transcription Factor Hijacking Hypothesis" in
vivo .................................................................................................
84
Appendix II. Analysis of DNA Damage Following Treatment of Estrogen ReceptorPositive and Estrogen Receptor-Negative Breast Cancer Cells by Experimental Drug
E27a .
...........................
87
Appendix 11I. Involvement of Base Excision Repair Proteins in the Processing of
Cisplatin Lesions in Escherichia coli..............................
90
References
................................
93
Biographical Note and Acknowledgements .........
4
...........
...........
.. 95
Chapter 1. Cisplatin: Lessons Learned from Bacteria.
1.1. Discovery of Cisplatin Potential through Experiments in Bacteria.
This review will seek to bring into focus the importance of the work done in bacteria,
which helped elucidate many important aspects underlying the activity of the anti-tumor
drug cisplatin. While a lot of work is being done with cisplatin in mammalian cells, many
hypotheses and discoveries have originated from work in bacteria - and the work in
bacteria often proved critical later.
Although the cisplatin molecule had long been known to chemists (Lippard, 1982), its
biological activity and therapeutic potential were not realized until 40 years ago. The true
discovery of cisplatin the world owes to bacteria. It was in February of 1965 that the first
paper appeared in Nature describing a new exciting phenomenon that later had a lot of
impact on the anti-cancer chemotherapy field (Rosenberg et al., 1965).
In the 1965 paper, Rosenberg and coworkers reported the unusual and unexpected
results of their study of the possible effects of an electric field on growth processes in
bacteria. Experimental conditions were chosen so as to eliminate electrolysis effects, and
platinum electrodes were chosen because of supposed chemical inertness of platinum.
Voltage was applied for the first two hours of the experiment. Once the current was turned
on in the chamber containing E. coli in the culture medium, turbidity of the culture began
to decrease after 1 hour. Within 1-2 hours, bacteria ceased to divide and began to elongate.
Within a few hours after the start of the experiment, all bacteria were in the form of long
filaments; bacteria continued to form filaments for 1-2 hours after the current was turned
off. Thereafter, cell division started anew, and after 8 hours the culture density returned to
its previous levels.
Rosenberg and colleagues methodically eliminated the possibility that any of the
physical and chemical agents known to cause filamentous growth in bacteria were involved
in the observed phenomena. Filamentous growth in bacteria occurs when bacteria continue
to grow and form daughter cells, while cell division is inhibited, and long strings of
unseparated daughter cells form. Among relevant agents causing filamentous growth in
bacteria are certain chemicals, near ultra-violet irradiation, osmotic pressure changes,
temperature changes, transfer to unaccustomed medium, and magnesium deficiency or
excess. Through a variety of tests, Rosenberg and coworkers excluded ultra-violet light,
temperature, pH and magnesium concentration as potentially involved in this effect of the
electrical field. They suspected that new chemical products were forming in the culture
medium as a result of electrolysis, and that these new products were the causative agents of
the filamentous growth. To test this hypothesis, authors passed electric current through the
culture medium in the chamber not containing bacteria, and then transferred this medium
into another chamber inoculated with bacteria. This test conclusively showed that new
long-lived chemical species were created in the culture medium by electric current in the
first chamber, and that these new newly formed compounds were responsible for the effect
of bacterial elongation in the second chamber.
5
Further tests presented in this study (Rosenberg et al., 1965) proved that formation of
the causative agent of filamentous growth required oxygen, and that electrolysis indeed
generated an oxidizing agent. The time course of oxidizing agent formation in the
electrolyzed medium was strikingly similar to the time course of the elongation process,
implicating this new substance as a causative agent. Each component of the culture
medium was individually tested, and the authors determined that oxidizing agent formation
was dependent on the chlorides in the culture medium. They suggested that soluble
platinum salts were forming as a result of chloride attack on the platinum electrodes. A
solution of (NH4 )2 PtCl6 produced an exact duplication of the results with the electrolyzed
medium, verifying that platinum salt was indeed an active agent.
Rosenberg and coworkers tested a number of group VIIIb compounds to determine
which metal ions were most effective in inducing filamentous growth. They confirmed that
various platinum salts were indeed inhibiting cell division, and rhodium was as effective as
platinum in that regard; other metal salts caused either no effect, or bacterial death in these
experiments. The importance of various metallic oxidation states, various ligands and their
spatial orientation could not be specified at that time.
This fundamental study presented, along with an unexpected discovery, many
important questions: what is the mechanism of the described phenomenon? Where in the
bacterial cell does action take place? Is there a connection between this phenomenon and
mechanism of action of other agents causing filamentous growth? And, would cell division
be similarly inhibited in other bacteria - and in other, non-bacterial cells?
Subsequent studies showed that certain group VIIIb transition metal compounds could
inhibit cell division in E. coli, causing filamentous growth (Rosenberg et al., 1967). Gramnegative bacilli were the most sensitive to this effect; some Gram-positive bacilli showed
slight elongation, but only at near-toxic levels of the metal; and none of the cocci tested
showed any apparent effect, even at relatively high concentrations of platinum. Cell
division in the platinum-induced filaments could be initiated by removal of platinum salts.
Only rhodium salts produced significant elongation comparable to the most active
platinum salt [(NH4)2PtCl6 ], but only at concentrations of metal much higher than were
required with the platinum salt. Results with these rhodium salts were less clear-cut,
though, and presented more difficulty in interpretation.
1.2. Cisplatin Journey: from Bacteria to Clinic.
Experiments with animal tumors quickly followed experiments in bacteria. A number
of platinum (II) and platinum (VI) compounds - cisplatin among them - were tested for
anti-tumor activity in mice and preliminary results were published as early as 1969.
(Rosenberg et al., 1969). Platinum compounds were found to inhibit sarcoma 180 and
leukemia L1210 in mice, and results suggested cisplatin as a potent anti-leukemic agent.
These observations were soon confirmed with the demonstration that platinum compounds
could cause complete regression of large sarcoma 180 tumors in mice with 63-100%
6
success (Rosenberg and Van Camp, 1970). This made platinum compounds the first
chemotherapeutic agents able to accomplish such regression of large tumors. Further
experiments with Dunning ascitic leukemia and intramuscular Walker 256 carcinosarcoma
in rats demonstrated that cisplatin was capable of inhibition of tumor development in these
neoplasms as well; treatment of both tumors even during more advanced stages of
development resulted in pronounced regression (Kociba et al., 1970). Cisplatin was also
shown to increase survival in mice with virus-induced reticulum cell sarcoma (Talley,
1970), and to be highly effective in promoting regression of rat mammary carcinoma
(Welsh, 1971), an experimental system closely resembling human breast cancer.
A number of other animal studies with cisplatin had been undertaken, and some of the
findings, as reviewed and summarized by Rosenberg (1973, 1985), were the following:
the drug exhibited marked, rather than marginal anti-tumor activity; it was a broadspectrum drug, active against drug-resistant as well as drug sensitive tumors; it was active
against slow growing as well as fast growing tumors, and against disseminated as well as
solid tumors; it exhibited no animal specificity, and caused regression of transplantable,
virally induced and chemically induced tumors.
It was not long before platinum compounds made their way into the clinic. Phase I
clinical trials of cisplatin, conducted on terminally ill cancer patients to determine
appropriate dose levels and schedules which would permit acceptable toxicities, were
concluded in mid-1972. Incidentally, tumor remissions were reported in 10-25% of
terminal patients, for 28 different tumor types, suggesting that cisplatin could also be a
broad-spectrum anti-tumor drug in humans (Rosenberg, 1973). Importantly, among the
tumors that had shown response to cisplatin there were both drug-sensitive and drugresistant tumor types, and many of these tumors were no longer responsive to the classic
chemotherapeutic agents.
Remarkably, the tumors most responsive to the drug were found to be testicular
tumors. Higby and coworkers reported tumor regressions in 9 out of 11 patients with
various types of testicular cancers in a phase I clinical trial for cisplatin (Higby et al., 1973
and 1974a). Although phase I and II clinical trials suggested that cisplatin might be useful
against tumors such as lymphosarcoma, Hodgkin's disease, endometrial carcinoma,
fibrosarcoma, squamous cell carcinoma, and renal and breast carcinoma (Higby et al.,
1974a), the remarkable responsiveness of testicular cancers to cisplatin suggested that
testicular tumors were differentially more sensitive to this drug.
Only five years after its approval by FDA in 1979, cisplatin became one of the most
important agents in clinical oncology. It achieved spectacular success in the treatment of
testicular cancer, showed major activity in ovarian cancer, and also became important in
the treatment of bladder cancer, cervical and endometrial carcinoma, lung cancer, head and
neck cancers, both Hodgkin's and non-Hodgkin's lymphoma, and esophageal cancer
(Loehrer and Einhorn, 1984).
In the 1960s, the standard treatment for advanced disseminated testicular cancer was
actinomycin D, with or without methotrexate and chlorambucil (Williams et al., 1984).
7
Treatment afforded an objective response rate of 40-50% and a complete remission rate of
10-20%. Around one half of complete responders never relapsed (5-10% cure rate), and if
recurrence occurred, it was within 2 years. Later, a synergistic regimen of vinblastinebleomycin achieved a 25% long-term disease-free survival (Einhorn, 2002). After the
striking activity of cisplatin was recognized, these numbers changed dramatically. In 1984,
80% of patients with disseminated disease were projected to achieve disease-free status
with cisplatin-based chemotherapy. The relapse rate was expected to be 10%, and around
70% would be long survivors. Maximum benefit from initial chemotherapy would be
attained in 9-12 weeks, and maintenance therapy to prevent relapse was not required for
patients with complete response.
Initial combination chemotherapy with cisplatin included cisplatin, vinblastin and
bleomycin. Later studies demonstrated that a bleomycin-etoposide-cisplatin regimen had
less toxicity and a higher cure rate, and since 1984 it has been standard chemotherapy for
disseminated testicular cancer, affording 90% cure rates (Einhorn, 2002).
1.3. Lessons Learned from Bacteria: Replication, Mutagenesis, and DNA Repair.
Early on, inhibition of DNA synthesis by cisplatin was demonstrated in human cells in
vitro (Harder and Rosenberg, 1970). The rationale for initiating this work was again
derived from observations in bacteria: since UV and X-irradiation, as well as alkylating
agents, were capable of inducing filamentous growth in E. coli, and all could cause damage
to DNA, it was predicted that platinum compounds could also react with DNA and inhibit
DNA synthesis (Roberts and Thomson, 1979; Pinto and Lippard, 1985; Howle and Gale,
1970a). Harder and Rosenberg found that, at a low dose, only DNA synthesis was impaired
by cisplatin, whereas a higher cisplatin dose resulted in suppression of RNA and protein
synthesis as well. Authors suggested that DNA synthesis was the primary target of
platinum compounds, and that inhibition of RNA and protein synthesis was a secondary
effect at higher platinum concentrations. Simultaneously, profound and extended
suppression of DNA synthesis by cisplatin was demonstrated in mice by another group
(Howle and Gale, 1970b). These authors observed that although RNA and protein
synthesis were also impaired initially by cisplatin treatment, the rates of RNA and proteins
synthesis returned to normal levels relatively quickly, while inhibition of DNA synthesis
was far more persistent.
Preferential inhibition of DNA synthesis by platinum compounds was soon
demonstrated
in bacteria
as well
(Shimizu
and Rosenberg,
1973; Beck,
1973).
Subsequently, direct interactions of cisplatin with DNA were studied extensively, leading
to qualitative and quantitative characterization of DNA adducts and their respective
importance in cisplatin toxicity. Remarkably, the distribution of cisplatin adducts was
essentially the same in bacterial and mammalian cells as in DNA treated in vitro
(Fichtinger-Schepman et al., 1986; Pinto and Lippard, 1985; Eastman, 1987).
Since interactions of other drugs with DNA frequently lead to mutagenesis (Walker,
1984), cisplatin was also tested for mutation induction in bacteria. Evidence was presented
8
(Beck and Brubaker, 1975) that demonstrated that cisplatin is indeed an efficient bacterial
mutagen. Although no direct evidence regarding the mechanism of cisplatin-induced
mutagenesis was obtained at that time, the observations indicated that cisplatin can induce
base substitutions, at least some of which are transitions.
A number of studies in bacteria have been undertaken since to elucidate the nature and
significance of mutation induction by cisplatin (Yarema et al., 1994; Burnouf et al., 1987).
Considering that cisplatin is a carcinogen in laboratory animals, and that cisplatin is
suspected of inducing secondary tumors in cancer patients receiving cisplatin-based
therapy, it was important to evaluate the relative toxic and mutagenic potential of
individual cisplatin lesions. The emergence of secondary tumors due to the drug
mutagenicity is a serious factor to consider in development of new platinum-based
therapies, and therefore it is important to identify the adduct that has the highest toxicity
combined with the lowest mutagenicity.
A study by Yarema and colleagues (1995) provided detailed analysis of toxicity and
mutagenicity in E. coli associated individually with each of the three major cisplatin
adducts:
1,2-d(GpG),
1,2-d(ApG) and 1,3-d(GpNpG).
In order to study the relative
contributions of each of these adducts towards toxicity and mutagenicity, without the
influence of other adducts, each adduct was incorporated site-specifically within a viral
genome and introduced into E. coli cells. The authors found that in SOS induced E. coli,
GG and AG adducts gave rise predominantly to G to T and A to T transversions,
respectively, which were targeted to the 5' modified base; A to G transitions were also
detected for the AG adduct. The less abundant AG adduct was found to be 4-5 times more
mutagenic than GG adduct, with mutation frequencies of 6% and 1.4%, respectively. The
GTG adduct was found in this study to be not more mutagenic than the unmodified DNA
sequence. The GG adduct was found to be most toxic lesion; although the AG adduct was
less toxic than GG adduct, the difference in toxicity disappeared upon induction of SOS in
E. coli; the GTG lesion showed intermediate toxicity, which was not affected by SOS
induction. In summary, this study identified the GG adduct as an ideal adduct to look for in
future platinum-based drug candidates since it displayed the highest ratio of toxicity to
mutagenicity, and the AG adduct as the one formation of which it would be desirable to
minimize. It must be noted that the results of this work are in good agreement with a
number of other studies (Bradley et al., 1993; Burnouf et al., 1990; Brandsma et al., 1996).
However, this is the only systematic study that was undertaken to analyze and compare the
toxicity and mutagenicity of each of the major cisplatin adducts under the same
experimental conditions.
Mismatch repair of DNA damage (MMR) was also identified through work in bacteria
as one of the mechanisms involved in cellular responses to cisplatin adducts (Lin et al.,
1999). In 1985, Fram et al. carried out experiments to determine whether cisplatin
produced adducts susceptible to mismatch repair. Cytotoxicity of cisplatin was assessed in
mismatch repair-deficient dam-, dammut and wild-type (dam+) E. coli strains after 2-hour
exposure to cisplatin. Toxicity in methylation-deficient mutants (dam-) was markedly
higher than in the wild-type (dam+)cells. However, introduction of an additional mutation
in either mutS or mutL genes abolished the increased sensitivity to cisplatin, and double
mutants (dammutf) were as sensitive to cisplatin as wild-type cells. The authors further
9
sought to find out whether excision of cisplatin adducts was also different in wild-type and
mismatch-repair deficient strains. Although the total platinum content after exposure to
cisplatin was similar in all strains, the excision of cisplatin adducts was slower in
methylation-deficient (dam) compared to wild-type cells over the course of a 6-hr recovery
period. Therefore, higher cisplatin cytotoxicity in dam mutants was accompanied by less
efficient excision of cisplatin adducts compared to the wild-type cells, although the
mechanism mediating inefficient repair remained unclear.
This study by Fram et al. (1985) demonstrated that mismatch repair was indeed
important in cellular responses to cisplatin. The hypothesis was put forward at that time
that, in the absence of a strand-discrimination signal in a methylation-deficient
background, mismatch repair proteins may initiate futile cycles of abortive repair opposite
cisplatin adducts (Karran and Marinus, 1982). Additional mutation in mismatch-repair
genes would then render cells unable to engage in abortive mismatch repair cycles, and
therefore afford protection against mismatch-repair mediated cisplatin toxicity.
A recent study (Zdraveski et al., 2002) examined in vitro the binding of the bacterial
mismatch repair protein MutS to cisplatin adducts, as well as to the adducts formed by two
cisplatin analogs with a DACH (diamminocyclohexane) ligand - oxaliplatin and
Pt(DACH)C12. These analogs present promising therapeutic agents since, remarkably, they
do not elicit resistance in mismatch-repair deficient cells. Oxaliplatin and Pt(DACH)C12
form DNA adducts differing from those of cisplatin by their bulky, non-polar DACH
ligand, which could present a different recognition substrate for mismatch repair proteins.
Recognition and binding of mismatch repair proteins to DNA-platinum adducts in cells
could possibly cause a range of consequences, such as abortive cycles of mismatch repair
or inhibition of replicative or recombinational bypass of such adducts. Therefore it was
important to examine these interactions directly and to establish any possible links between
cellular responses and recognition of platinum adducts by mismatch repair proteins.
Binding reactions between purified E. coli MutS and DNA globally modified with cisplatin
or DACH compounds were studied by electrophoretic mobility shift assay. Interestingly,
MutS recognized both types of adducts, but cisplatin adducts were recognized with 2-fold
higher affinity than DACH adducts, perhaps because of the differences in adduct geometry.
E. coli mutants deficient in mismatch repair were analyzed in this study (Zdraveski et
al., 2002) for sensitivity to treatment with cisplatin. Similar to the results described above
(Fram et al., 1985), methylation-deficient mutants showed high sensitivity to cisplatin,
which was abrogated by additional mutations in either the mutS or mutL mismatch repair
genes. Survival of these mutants was also examined following treatment with oxaliplatin
and Pt(DACH)C12, and a similar pattern of toxicity was observed with these compounds.
However, dam mutants were more sensitive to cisplatin than to DACH compounds,
possibly mirroring the higher affinity of the mismatch repair protein MutS to cisplatin than
to DACH compounds. These observations together support the hypothesis that differential
mismatch repair-mediated cellular responses to the two compounds reflect differential
recognition of the respective adducts by mismatch-repair proteins.
10
All recombination-deficient mutants analyzed by Zdraveski et al. (2002) showed
striking sensitivity to Pt(DACH)C12, comparable to sensitivity to cisplatin. Therefore, both
cisplatin and DACH compounds require recombinational repair for cellular survival,
indicating that both types of adducts likely present replication blocks in E. coli.
While the study by Zdraveski et al. (2002) indicated that E. coli MutS did indeed
recognize cisplatin lesions, the precise nature of the lesions involved in recognition by
mismatch-repair proteins remained to be investigated. The subsequent study by Fourrier
and colleagues (2003) addressed this question. Since replicative bypass of cisplatin lesions
frequently leads to misincorporation of a mismatched base opposite the platinated base, the
authors investigated MutS binding not only to cisplatin lesions per se, but also to cisplatin
lesions in the context of a mismatch - the so-called compound lesions. Since mismatches
are natural substrates for MutS protein, the presence of cisplatin lesion in the context of a
mismatch could affect recognition of the mismatch by MutS, either enhancing or
weakening it.
Four cisplatin crosslinks were investigated by Fourrier et al. (2003) in competition
experiments using an electrophoretic mobility shift assay. Among the four crosslinks (1,2d(GpG), 1,2-d(ApG), 1,3-d(GpCpG), and inter-strand crosslink), only the 1,2-d(GpG)
intrastrand crosslink was recognized by E. coli MutS with 1.5-fold higher affinity relative
to the homoduplex DNA substrate; all other cisplatin cross-links were recognized by MutS
less well than homoduplex DNA. In comparison, the heteroduplex GG/CT containing a
G/T mismatch (which is a natural substrate of MutS) was recognized with 47-fold higher
affinity than homoduplex DNA.
Interesting results were obtained in this study with cisplatin cross-links placed in
the context of a mismatch (compound lesions). When mismatched T was placed opposite
the 3'G in the GG/CT substrate, the presence of a cisplatin adduct enhanced the binding of
MutS about 3-fold over the unplatinated mismatched substrate. However, when the
mismatched T was placed opposite 5'G in the GG/TC substrate, the presence of a cisplatin
crosslink reduced the affinity of MutS about 4-fold relative to an unplatinated mismatch.
Thus, depending on the position of the mismatched T in the compound lesion, the cisplatin
adduct can either stimulate or impair recognition of the mismatch by MutS. However,
MutS bound with similar affinity to the mismatched AG/TT substrates either with or
without a cisplatin cross-link present. This suggested that cisplatin cross-links don't
interfere with AG/TT mismatch recognition by MutS, but rather 1,2-(dApG) cisplatin
cross-links may be recognized by MutS when a mismatch is present.
Other compound lesions containing all other possible mispairs opposite the 5' or
the 3' of the 1,2-d(GpG) and 1,2-d(ApG) cross-links were also analyzed by Fourrier et al.
(2003) in order to identify which ones could be specifically recognized by MutS. All five
of the compound lesions with the 1,2-d(GpG) adduct, and three of the compound lesions
with the 1,2-d(ApG) adduct were specifically recognized by MutS; two compound lesions
containing A/C and A/G mismatches were not recognized by MutS. Interestingly, the
presence of the 1,2-d(GpG) cross-link opposite the mismatch enhanced recognition of the
mismatched heteroduplex by MutS, whereas compound lesions with 1,2-(ApG) cross-links
11
were less well recognized compared to unplatinated mismatched substrates. In summary,
it can be concluded that a wide range of cisplatin compound lesions, including those
principally formed during replicative bypass, are good substrates for E. coli MutS, and
may well be the critical lesions mediating MMR involvement in cellular responses to
cisplatin. Whether or not the cisplatin lesions, or cisplatin compound lesions, can actually
engage MMR activity or other downstream events following recognition by MutS remains
to be determined.
Involvement of other repair systems in cellular responses to cisplatin has also been
revealed through a number of studies in bacteria. It was noted early on, thanks to research
in bacteria, that DNA repair deficiency has a detrimental effect on the ability of cells to
deal with cisplatin assault. In 1973, Beck and Brubacker reported that E. coli mutants
lacking certain DNA repair functions were significantly less viable than wild-type cells
after cisplatin exposure. This early study implicated both nucleotide excision repair (NER)
and recombination as mechanisms essential for protection against cisplatin damage.
Mutants deficient in both pathways were significantly more sensitive to cisplatin than
mutants defective in either nucleotide excision repair, or recombination alone, suggesting
that the two mechanisms might be independent of one another. Later study by Popoff and
colleagues (1987) suggested that while uvrB gene function was essential in E. coli for
repair of plasmid DNA damaged with cisplatin, the functional recA product seemed to be
of secondary importance for repair of such damage. Further analysis (Beck et al., 1985) of
bacterial strains defective in individual nucleotide repair components showed that all of the
NER mutants tested (uvrA, uvrB and uvrC mutants, which are all blocked in the first step
of the excision repair pathway) were exceptionally sensitive to cisplatin, and that
proficiency in excision repair was required for survival upon cisplatin damage.
In vitro experiments with the bacterial nucleotide-excision complex UvrABC (Beck et
al., 1985) provided further support for the role of NER in protection against cisplatin
lesions. Purified uvr gene products were used to reconstitute the UvrABC nuclease, and its
incision activity was analyzed on platinum-treated DNA. All of the three protein products
of the uvr genes were required to cut the cisplatin-damaged plasmid in vitro. In a
subsequent study, Husain et al. (1985) also observed that nucleotide excision repair was
the major mechanism of repair of platinum adducts in E. coli; however, the recombinationdependent pathway also contributed significantly to survival. The authors also
demonstrated the ability of UvrABC excinuclease to remove platinum adducts from the
plasmid in vitro.
Other early studies in bacteria also highlighted the importance of nucleotide
excision repair and recombination in cellular responses to cisplatin (Konishi et al., 1981;
Beck et al., 1975). Alazard and colleagues (1982) had also noted that nucleotide excision
and recombination deficient mutants were more sensitive to cisplatin than their wild-type
counterparts, and that the double uvrArecA mutant was twice as sensitive as a single recA
mutant. These authors reported that inhibition of DNA synthesis was correlated with
sensitivities of different repair-deficient and wild-type strains to cisplatin. They suggested
that the toxic effect of cisplatin was mediated by inhibition of DNA synthesis by cisplatin
lesions, and that the greater inhibition of DNA synthesis in repair-deficient E. coli and
12
their greater sensitivity to cisplatin compared to wild-type parental strains was due to the
presence of unrepaired DNA lesions.
In concurrent work, Alazard and Germanier (1982) reported that cisplatin treatment
produced single strand breaks and gaps in the DNA of excision-deficient E. coli, and that
post-treatment incubation led to rejoining of the DNA; the process of DNA conversion to
normal size fragments was abolished in a recombination-deficient E. coli mutant. These
observations, along with others, underscored the importance of recombinational repair in
E. coli responses to cisplatin exposure.
Recent studies have provided further evidence for the role of recombinational
repair in cellular responses to cisplatin in E. coli. A systematic study was carried out by
Zdraveski et al. (2000), which dissected the involvement of recombinational pathways in
the responses of E. coli to cisplatin damage. This detailed genetic study analyzed responses
to cisplatin of a series of E. coli mutants deficient in the major pathways of recombination.
The study found that recombination-deficient mutants were strikingly sensitive to cisplatin,
and that both daughter-strand gap and double-strand break recombination pathways were
critical for survival upon cisplatin treatment. Therefore, this study by Zdraveski et al.
(2000) confirmed that, as suggested by early observations in E. coli, recombination was
essential for repair of DNA breaks and gaps produced by cisplatin. Moreover, this study
confirmed and extended previous observations that recombinational repair plays a role in
countering cisplatin damage that is as important as nucleotide excision repair. Most
recombination-deficient mutants were as sensitive to cisplatin as the uvrA mutant. Mutants
deficient in both pathways showed much higher sensitivity to cisplatin than single mutants,
which indicated that recombination and NER pathways are independent of one another.
Extreme sensitivity of recombination-deficient mutants to cisplatin, which implies
that recombination is important for survival upon cisplatin exposure, also suggested that
cisplatin might be inducing high levels of recombination in surviving cells. Indeed,
cisplatin proved to be potently recombinogenic in E. coli as compared to other DNAdamaging compounds (Zdraveski et al., 2000). A subsequent study by Nowosielska et al.
(2004) confirmed that cisplatin caused recombination in E. coli, and determined the
specific genetic requirements for cisplatin-induced recombination.
One of the most important conclusions from the studies of cisplatin and
recombination repair in bacteria is that there might be a strong connection between
recombination and specific susceptibility of testicular tumors to cisplatin. Testicular
tumors mostly derive from germ cells, which are unique in that they undergo meiotic
recombination during cell division. Meiotic recombination is a highly regulated event, and
cisplatin-induced recombination may be disruptive to such cells, forcing them to enter
apoptosis. Therefore, it is possible that the unique sensitivity of germ cell tumors to the
drug may be explained, at least in part, by their dependence on recombination, which may
be deregulated by cisplatin exposure.
13
1.4. Current Status of Cisplatin in Anti-Cancer Therapy.
Although testicular tumors set the most spectacular example of cisplatin success
(Einhorn, 2002), they are indeed followed by a long list of other tumors against which
cisplatin is currently being used. Cisplatin and structurally similar platinum compounds are
employed as a first-line chemotherapy against lung, ovarian, cervical, bladder, head and
neck, esophageal, gastric, colorectal and pancreatic cancers. They also may be used as a
second- or third-line treatment against melanoma, cancers of the breast, prostate, and brain
as well as other tumors (Boulikas and Vougiouka, 2004).
In some tumor types, the therapeutic efficacy is equivalent for cisplatin and
carboplatin; however, in some, cisplatin seems to be more effective. In certain tumors,
such as esophageal and gastric cancers, carboplatin, unlike cisplatin, is inactive (Lokich,
2001). On the other hand, carboplatin is consistently better than cisplatin with regard to
non-hematologic toxicity and convenience of administration. Still, in most tumors, the
therapeutic index defined by the efficacy versus toxicity profile appears to favor cisplatin
over carboplatin. Cisplatin, unlike carboplatin, does not contribute to hematologic toxicity
and therefore permits the full dose of other potentially myelosuppressive agents to be
administered in multi-drug regimens. The only cancer for which comparable therapeutic
efficacy has been definitively established is ovarian cancer, and since carboplatin is better
tolerated, it has assumed a dominant role in the treatment of ovarian cancer. Carboplatin is
also used in the treatment of lung cancer, and may have potential in advanced endometrial
cancer and in pediatric patients with such cancers as Wilms tumor (nephroblastoma),
hepatocellular carcinoma, and retinoblastoma (Boulikas and Vougiouka, 2004).
Another promising analog of cisplatin is oxaliplatin, which is indicated as a firstline treatment of advanced colorectal cancer patients (Boulikas and Vougiouka, 2004). It is
also active in platinum-resistant ovarian cancer (Lokich, 2001). Importantly, this drug is
not cross-resistant with some platinum-resistant experimental tumors, and it therefore may
expand the spectrum of tumors against which platinum drugs are active. The toxicity
profile of oxaliplatin is closer to cisplatin than carboplatin.
Thousands of analogues have been synthesized in the quest to improve the
therapeutic index of cisplatin, expand its anti-tumor spectrum, and overcome the problem
of cisplatin resistance. However, just over a dozen of compounds have reached clinical
trials, and most of these have not shown any advantage over cisplatin (Weiss and
Christian, 1993). Many years after its remarkable discovery in bacterial experiments, and
after decades dedicated to the development of more successful analogs, cisplatin remains at
the forefront of the cancer chemotherapy.
14
1.5. Structures of Platinum Compounds.
Cisplatin
NH3
Carboplatin
NH3
Oxaliplatin
Pt(DACH)C12
12
Transplatin
[(en)PtCI 2 ]
CI
NH 3
Pt
CI
15
NH 3
References:
Alazard, R.J., and M. Germanier. 1982. Post replication repair in an excision defective
strain of Escherichia coli following treatment with cis-dichlorodiammineplatinum(II).
Biochem. Biophys. Res. Comm. 104(2):693-700.
Alazard R., M. Germanier and N.P Johnson. 1982. Mechanism of toxicity of platinum
(II) compounds in repair-deficient strains of Escherichia coli. Mutat. Res. 93:327-337.
Beck, D.J., and R.R. Brubaker. 1973. Effect of cis-platinum(II)diamminodichloride on
wild-type and deoxyribonucleic acid repair-deficient mutants of Escherichia coli. J.
Bacteriol. 116(3):1247-1252.
Beck D.J., and R.R. Brubaker. 1975. Mutagenic properties of
plantinum(II)diammino-dichloride in Escherichia coli. Mutat. Res. Feb;27(2): 181-9.
cis-
Beck, D.J., S. Popoff, A. Sancar, and W.D. Rupp. 1985. Reactions of the UvrABC
excision nuclease with DNA damaged by diamminedichloroplatinum(II). Nucleic Acids
Research. 13(20):7395-7412.
Beck, D.J., D. Vazquez, and R. Cook. 1975. Effect of cis-platinum(II)
diamminodichloride on wild type and a recA mutant of Escherichia coli. J. Clin. Hematol.
Oncol. 7(2):726-740.
Boulikas, T., and M. Vougiouka. 2004. Recent clinical trials using cisplatin, carboplatin
and their combination chemotherapy drugs. Oncology Reports. 11:559-595.
Bradley, L.J.N., K.J. Yarema, S.J. Lippard, and J.M. Essigmann. 1993. Mutagenicity
and genotoxicity of the major DNA adduct of the antitumor drug cisdiamminedichloroplatinum(II). Biochemistry. 32(3):982-988.
Brandsma, J.A., M. de Ruijter, R. Visse, D. van Meerten, M. van der Kaaden, J.G.
Moggs, P. van de Putte. 1996. The in vitro more efficiently repaired cisplatin adduct cisPt.GG is in vivo a more mutagenic lesion than the relatively slowly repaired cis-Pt.GCG
adduct. Mutat. Res. 362:29-40.
Burnouf, D., M. Duane, and R.P.P. Fuchs. 1987. Spectrum of cisplatin-induced
mutations in Escherichia coli. Proc. Natl. Acad. Sci. U S A. 84(11):3758-3762.
Burnouf, D., C. Gauthier, J.C. Chottard, and R.P.P. Fuchs. 1990. Single d(ApG)/cisdiamminedichloroplatinum(II) adduct-induced mutagenesis in Escherichia coli. Proc. Natl.
Acad. Sci. USA. 87:6087-6091.
16
Eastman, A. 1987. The formation, isolation and characterization of DNA adducts
produced by anticancer platinum complexes. Pharmac. Ther. 34:155-166.
Einhorn, L.H. 2002. Curing metastatic testicular cancer. Proc. Natl. Acad. Sci. USA.
99(7):4592-4595.
Fichtinger-Schepman, A.M.J., P.H.M. Lohman, and F. Berends. 1986. Interactions of
the antitumour drug cisplatin with DNA in vitro and in vivo. IARC Sci. Publ. 78:83-99.
Fourrier, L., P. Brooks, and J.M. Malinge. 2003. Binding discrimination of MutS to a
set of lesions and compound lesions (base damage and mismatch) reveals its potential role
as a cisplatin-damaged DNA sensing protein. J. Biol. Chem. 278(23):21267-21275.
Fram, RJ., P.S. Cusick, J.M. Wilson, and M.G. Marinus. 1985. Mismatch repair of cisdiamminedichloroplatinum(II)-induced DNA damage. Molecular Pharmacology. 28:51-55.
Harder H.C., and B. Rosenberg. 1970. Inhibitory effects of antitumor platinum
compounds on DNA, RNA and protein synthesis in mammalian cells in vitro. Int. J.
Cancer. 6:207-216.
Higby, D.J., H.J. Wallace, D.J. Albert, and J.F Holland. 1974a.
Diaminodichloroplatinum: a phase I study showing responses in testicular and other
tumors. Cancer. 33:1219-1225.
Higby, DJ.,
H.J. Wallace, DJ.
Albert, and J.F Holland. 1974b.
Diamminodichloroplatinum in the chemotherapy of testicular tumors. J Urol. 112:100-104.
Higby, DJ., H.J. Wallace, J.F. Holland. 1973. cis-Diamminedichloroplatinum (NSC119875): a phase I study. Cancer Chemotherapy Reports. 57(4):459-463.
Howle, J.A., and G.R. Gale. 1970a. cis-Dichlorodiammineplatinum(II): cytological
changes induced in Escherichia coli. J. Bacteriol. 103(1):259-260.
Howle, J.A., and G.R. Gale. 1970b. cis-Dichlorodiamine-platinum
(II). Persistent and
selective inhibition of deoxyribonucleic acid synthesis in vivo. Biochem. Pharmacol.
19:2757-2762.
Husain, I., S.G. Chaney, and A. Sancar. 1985. Repair of cis-platinum-DNA adducts by
ABC excinuclease in vivo and in vitro. J. Bacteriol. 163(3):817-823.
Karran P., and M. Bignami. 1996. Drug-related killings: a case of mistaken identity.
Chem. Biol. 3(11):875-9.
Karran P., and M.G. Marinus. 1982. Mismatch correction 0 6-methylguanine residues in
E. coli DNA. Nature. 296:868-869.
17
Kociba, RJ., S.D. Sleight, and B. Rosenberg. 1970. Inhibition of Dunning ascitic
leukemia and Walker 256 carcinosarcoma with cis-diamminedichloroplatinum (NSC119875). Cancer Chemotherapy Reports. 54(5):325-328.
Konishi, H., T. Usui, H. Sawada, H. Uchino and Y. Kidani. 1981. Effects of anticancer
platinum compounds on Escherichia coli strains with normal and defective repair capacity.
Gann. 72:627-630.
Lin, X., H.-K. Kim, and S.B. Howell. 1999. The role of DNA mismatch repair in cisplatin
mutagenicity. J. Inorg. Biochem. 77:89-93.
Lippard, SJ. 1982. New chemistry of an old molecule: cis-[Pt(NH3) 2C12]. Science.
218:1075-1082.
Loehrer, PJ., and L.H. Einhorn. 1984. Drugs five years later. Cisplatin. Ann. Internal
Med. 100(5):704-713.
Lokich, J. 2001. What is the "best" platinum: cisplatin, carboplatin, or oxaliplatin? Cancer
Investigation, 19(7):756-760.
Mello, J.A., Trimmer E.E., Kartalou, M., and Essigmann, J.M. 1998. Nucleic Acids
Mol. Biol. 12:249-274. Conflicting roles of mismatch repair and nucleotide excision repair
in cellular susceptibility to anticancer drugs.
Nowosielska, A., M.A. Calmann, Z. Zdraveski, J.M. Essigmann, and M.G. Marinus.
2004. Spontaneous and cisplatin-induced recombination in Escherichia coli. DNA Repair.
3:719-728.
Pinto, A.L., and S.J. Lippard. 1985. Binding of the antitumor drug cisdiamminedichloroplatinum(II) (cisplatin) to DNA. Biochim. Biophys. Acta. 780:167-180.
Popoff S.C., D.J. Beck, and W.D. Rupp. 1987. Repair of plasmid DNA damaged in vitro
with cis- or trans-diamminedichloro-platinum(II) in Escherichia coli. Mutat. Res. 183:129137.
Roberts, JJ., and A.J. Thomson. 1979. The mechanism of action of antitumor platinum
compounds. Prog. Nucleic Acids Res. Mol. Biol. 22:71-133.
Rosenberg, B. 1973. Platinum coordination complexes in cancer chemotherapy.
Naturwissenschaften. 60(9):399-406.
Rosenberg, B. 1985. Fundamental studies with cisplatin. Cancer. 55:2303-2316.
Rosenberg, B., E. Renshaw, L. Van Camp, J. Hartwick, and J. Drobnik. 1967.
Platinum-induced filamentous growth in Escherichia coli.
18
Rosenberg, B., and L. Van Camp. 1970. The successful regression of large solid sarcoma
180 tumors by platinum compounds. Cancer Res. 30:1799-1802.
Rosenberg, B., L. Van Camp, and T. Krigas. 1965. Inhibition of cell division in
Escherichia coli by electrolysis products from a platinum electrode. Nature.
205(4972):698-699.
Rosenberg, B., L. Van Camp, J.E. Trosko, and V.H. Mansour. 1969. Platinum
compounds: a new class of potent antitumour agents. Nature. 222:385-386.
Shimizu, M., and B. Rosenberg. 1973. A similar action to UV-irradiation and a
preferential inhibition of DNA synthesis in E. coli by antitumor platinum compounds. J.
Antibiot. 26(4):243-5.
Talley R.W. 1970. Chemotherapy of a mouse reticulum cell sarcoma with platinum salts.
Proc. Amer. Assoc. Cancer Res. 11:78.
Walker, G.C. 1984. Mutagenesis and inducible responses in deoxyribonucleic acid
damage in Escherichia coli. Microbiol.Rev. 48(1):60-93.
Weiss, R.B., and M.C. Christian. 1993. New cisplatin analogues in development. Drugs.
46(3):360-377.
Welsh, C.W. 1971. Growth inhibition of rat mammary carcinoma induced by cis-platinum
diamminodichloride-II. J. Natl. Cancer Institute. 47(5): 1071-1078.
Williams, S.D., P.J. Loehrer, and L.H. Einhorn. 1984. Chemotherapy of advanced
testicular cancer. Sem. Urol. 2(4):230-237.
Zamble, D.B., and S.J. Lippard. 1995. Cisplatin and DNA repair in cancer
chemotherapy. Trends Biochem. Sci. 20:435-439.
Zdraveski, Z.Z., J.A. Mello, C.K. Farinelli, J.M. Essigmann, and M.G. Marinus.
2002; MutS preferentially recognizes cisplatin- over oxaliplatin-modified DNA. J. Biol.
Chem. 277(2):1255-1260.
Zdraveski, Z.Z., J.A. Mello, M.G. Marinus, and J.M. Essigmann. 2000. Multiple
pathways of recombination define cellular responses to cisplatin. Chem. Biol. 7(1):39-50.
Yarema, KJ., S.J. Lippard, and J.M. Essigmann. 1995. Mutagenic and genotoxic
effects of DNA adducts formed by the anticancer drug cis-diamminedichloroplatinum (II).
Nucl. Acids Res. 23(20):4066-4072.
Yarema, K.J., J.M. Wilson, SJ Lippard, and J.M. Essigmann. 1994. Effects of DNA
adduct structure and distribution on the mutagenicity and genotoxicity of two platinum
anticancer drugs. J.Mol.Biol. 236:1034-1048.
19
Chapter 2. Tetracycline Resistance in Bacteria.
2.1. Tetracyclines in Clinical Practice.
Tetracyclines, a broad-spectrum class of antibiotics, were discovered in the late 1940s
(Chopra and Roberts, 2001). The first members of this group to be discovered were
chlortetracycline (1948) and oxytetracycline (1948), and both were natural products of
antibiotic-producing bacteria: chlortetracycline was produced by Streptomyces aureofaciens
and oxytetracycline was a product of Streptomyces rimosus. Later, other tetracyclines were
identified: tetracycline (1953) as a natural product of Streptomyces aureofaciens, Streptomyces
rimosus and Streptomyces viridofaciens; and demethylchlortetracycline (1957) as a product of
Streptomyces aureofaciens. Other tetracyclines, such as methacycline (1965), doxycycline
(1967), and minocycline (1972), were synthesized; minocycline, at the time it was discovered,
was active against most strains that already had acquired resistance to other tetracyclines (Sum
et al., 1998). These drugs were referred to as 1st generation (1948-1957) and 2 nd generation
(1965-1972), respectively. Most recently, a 3 rdgeneration of tetracyclines, glycylcyclines, is
being developed, which will be discussed in a later section.
Tetracyclines quickly became widely used because of their multiple important
advantages (Col and O'Connor, 1987). The cost of production of tetracyclines is low (Liss and
Batchelor, 1987), they demonstrate good oral absorption, and they are active against a broad
range of traditional gram-negative and gram-positive bacterial pathogens, as well as against
bacteria lacking cell walls and those found intracellularly (such as mycoplasmas, chlamydiae,
and rickettsiae). They are also active against some eukaryotic protozoan parasites, such as
Toxoplasma gondii, Giardia lamblia, Plasmodium falciparum, Entamoeba histolytica,
Leishmania major, and Trichomonas vaginalis (Chang et al., 1990; Edlind, 1989; Katiyar and
Edlind, 1991). Tetracyclines are also relatively safe, and their side effects are few and minor
(Clendenning, 1965; Olson and Riley, 1966; Pflug, 1963). They cannot be used in pregnant
women and children because they cause temporary retardation of bone growth and can cause
permanent discoloration of teeth. Other side effects of tetracyclines include diarrhea due to the
high levels of antibiotic reaching the lower gastrointestinal tract. Tetracyclines may
accumulate in patients with renal insufficiency, leading to nephrotoxicity and/or diabetes
insipidus. In young adults treated for acne, tetracyclines may cause blurring of vision and
headache as a result of benign intracranial hypertension. Minocycline can cause vestibular
disturbance with vertigo and nausea, presumably due to its high lipid solubility and ability to
accumulate in high-lipid cells of the vestibular apparatus. However, these effects are reversible
and disappear when the therapy is discontinued (Chopra et al., 1992).
Typical tetracyclines inhibit bacterial growth by binding to the bacterial ribosome; this
binding prevents aminoacyl-tRNA from attaching to the A-site on the 30S ribosomal subunit,
thereby disrupting protein synthesis. However, the interaction of typical tetracyclines with a
ribosome is reversible; therefore, their action is bacteriostatic rather than bactericidal. In
contrast, atypical tetracyclines (such as chelocardin, thiatetracycline, anhydrotetracycline,
anhydrochlortetracycline) do not exert their effects through ribosomal interactions, but rather
they inflict cytoplasmic membrane damage. They are bactericidal rather than bacteriostatic;
20
however, they cannot be used clinically because of their high toxicity, which probably reflects
their ability to interfere with eukaryotic membrane functions as well (Roberts, 1996).
Tetracycline can also bind to 70S ribosomes found in the mitochondria, and inhibit
mitochondrial protein synthesis. It can only form a weak interaction with 80S ribosome of
eukaryotic cells, though, which may partially explain the selective antimicrobial activity of
tetracycline. Some of the eukaryotic protozoan parasites against which tetracyclines are active
have tetracycline-susceptible mitochondria; however, other susceptible parasites do not have
mitochondria. Therefore, the mechanism of action of tetracycline in eukaryotic parasites is still
not clear (Roberts, 2003).
Despite the widespread emergence of resistance in many instances, tetracyclines are
still extensively used in the treatment of a variety of bacterial and non-bacterial infections.
There are a number of conditions where tetracyclines are first-choice antibiotics. Tetracyclines
are used extensively against rickettsial infections: typhus, scrub typhus, and spotted fevers,
including Mediterranean spotted fever, Rocky Mountain spotted fever, and Q fever. Some rare
conditions for which tetracycline may be used are plague and tularemia (Russel et al., 1998a
and 1998b). The major area of clinical application of tetracyclines is the treatment of acne, and
periodontal disease is another area of their extensive use. Tetracyclines are also widely
employed in the treatment of genito-urinary infections (Portnoy, 1986). They offer effective
treatment against non-gonococcal urethritis, and may also be used for cervicitis, pelvic
inflammatory disease, syphilis and prostatitis as well (Chopra and Roberts, 2001). In addition,
tetracyclines are often used to treat other chlamydial infections, such as lymphogranuloma
venereum, trachoma, psittacosis and inclusion conjunctivitis (Chopra et al., 1992).
Tetracyclines are used against some bacterial gastrointestinal infections, for example, in
treatment of cholera and in prophylaxis of traveler's diarrhea (Rabbani et al., 1989; Islam,
1987). Gastritis and peptic ulcer disease associated with Helicobacter pylori have also been
treated with tetracyclines in multiple-drug regimens (Ribeiro et al., 2004).
Some applications of tetracyclines are quite recent. For example, tetracyclines are
among the first-choice antibiotics against Lyme disease, which is the most common tick-borne
infection in the United States (Nadelman et al., 2001; Luger et al., 1995). They are also used
for relapsing fever caused by Borrelia recurrentis, another tick-borne infection. Some studies
show that tetracyclines may be effective against leprosy (Ji et al., 1998; Ji et al., 1996).
Another important new application of tetracycline is prophylaxis and treatment of malaria
caused by Plasmidium falciparum, a eukaryotic parasite. Even mefloquine-resistant malaria at
this time is responsive to tetracycline therapy, and tetracyclines are currently the drugs of
choice (Pradines et al., 2000; Schwartz and Regev-Yochay, 1999). Also, activity of
tetracyclines against filarial nematodes has recently been demonstrated (Hoerauf et al., 1999;
Smith and Rajan, 2000); tetracycline is thought to affect the intracellular bacteria that co-exist
in a mutualistic relationship with the nematode and are essential for its survival. It is possible
that nematode-related diseases will be treatable by tetracycline in the future.
A significant aspect of the therapeutic applicability of tetracyclines is that they are the
agents primarily employed against bacteria that have high potential for use in bacterial
biological weapons (Navas, 2002). The three pathogenic organisms that are most likely to be
21
involved as biological weapons are Yersinia pestis (plague), Bacillus anthracis (anthrax), and
Francisella tularensis (tularemia); doxycycline is important in treatment and prophylaxis in
each case. (Inglesby et al., 2000; Dennis et al., 2001; Inglesby et al., 1999; Inglesby et al.,
2002).
2.2. Tetracycline Resistance in Bacteria.
Unfortunately, the emergence of microbial resistance to tetracyclines became a serious
problem limiting their use in clinical practice. Prior to the mid-1950s, the majority of bacteria
were susceptible to tetracyclines; only 2% of Enterobacteriaceae collected between 1917 and
1954 were resistant to tetracycline (Hughes and Datta, 1983). However, the situation has been
changing quickly. In 1953, shortly after the introduction of tetracycline therapy, the first
tetracycline-resistant bacterium, Shigella dysenteriae, the causative agent of bacterial
dysentery, was isolated in Japan. Multiple-drug resistant Shigella was first isolated in 1955,
and was resistant to tetracycline, streptomycin and chloramphenicol. The incidence of
multiple-drug resistant Shigella species resistant to tetracycline in 1955 in Japan was 0.02%;
by 1960 it represented almost 10% of the strains tested in Japan (Akiba et al., 1960). Over
60% of Shigella strains isolated between 1988 and 1993 in Brazil were multiple-drug and
tetracycline resistant (Lima et al., 1995). As reported in Boston in 1969, 38% of S. aureus,
61% of E. coli, 62% of Klebsiella sp., 58% of Enterobacter sp., 91% of Proteus sp., and 97%
of Serratia sp. were resistant to tetracyclines (Sabath, 1969). By the mid-1970s, increased rates
of resistance to tetracyclines were common in Enterobacteriaceae, Staphylococcus,
Streptococcus, and Bacteroides; high rates of tetracycline resistance were recorded by the mid1980s in Neisseria gonorrhoeae and Haemophilius influenzae (Roberts, 2003). Rates of
tetracycline resistance recorded in Streptococcus pneumoniae isolates tested in 1997 were as
high as 25.4 % in France and Belgium, 39.4% in Spain, 27.2% in Italy, 40.3% in Poland,
16.9% in USA, 22.2% in Mexico, 18.2% in South Africa, 22.6 % in Saudi Arabia, and 83.3%
in Hong Kong. (Felminhgam et al., 2000). Today, most genera examined have tetracyclineresistant isolates, but the percentage varies according to genus and species and geographic
location.
Due to the emergence of resistance, tetracyclines are no longer the antibiotics of choice
in the treatment of many conditions. For example, although tetracyclines were extensively
used in the treatment of bacterial respiratory infections, resistance as well as availability of
alternative drugs resulted in a decline in the use of tetracyclines for the treatment of
pneumonia and bronchitis where tetracyclines have long been drugs of choice (Roberts, 2003).
Since the emergence of tetracycline resistant strains of Neisseria gonorrhoeae (Heritage and
Hawkey, 1988; Waugh et al., 1988), tetracycline has also been discontinued as the first line of
therapy for gonorrhea (Speer et al., 1992).
Three major mechanisms of tetracycline resistance have been described so far. The two
predominant mechanisms involve either active efflux of the drug out of bacterial cells, or
protection of the bacterial ribosome by a resistance protein. The third mechanism, enzymatic
inactivation of tetracycline, involves chemical modification of tetracycline, in the presence of
oxygen and NADPH, by a 44 kDa cytoplasmic protein, which shares homology with NADPrequiring oxidoreductases. The tetX gene encoding this mechanism has so far only been found
22
in anaerobic Bacteroides species, and it did not confer resistance to B. fragilis species in which
it was originally found. The clinical relevance of this mechanism of resistance is therefore
doubtful, since it requires oxygen and should not be able to function in its natural anaerobic
Bacteroides host (Speer and Salyers, 1989; Speer at al., 1991).
The ribosomal protection genes confer resistance to tetracycline, doxycycline and
minocycline. They code for approximately 72.5 kDa cytoplasmic proteins, which bear
sequence similarity to ribosomal elongation factors and may be evolutionarily derived from
them. These proteins, in the presence of GTP, bind the tetracycline-blocked ribosome, release
tetracycline, supposedly through an allosteric mechanism, and dissociate from the ribosome,
which then returns to the elongation cycle. It is not clear yet whether ribosomal protection
proteins actively function to prevent tetracycline from rebinding the ribosome; it is speculated
that they may promote subtle rearrangements in ribosomal architecture that slow tetracycline
rebinding (Connell et al., 2003).
Active efflux of the drug is the most common tetracycline resistance mechanism in
Gram-negative bacteria. It is mediated by the expression of tetracycline-specific transmembrane efflux pumps, which transport tetracycline out of the cell before the drug can attack
its target, the ribosome. Each of the efflux genes codes for a 46 kDa inner membrane protein,
which has 12 (in Gram-negative) or 14 (in Gram-positive) hydrophobic membrane-spanning
helices. There currently are more than 20 recognized classes of tetracycline resistance
determinants encoding tetracycline efflux pumps, which share common genetic organization.
The most widespread tetracycline resistance determinant of Gram-negative bacteria is the class
B determinant associated with transposon TnlO. It encodes two genes in a divergent
orientation, tetR for the tetracycline-inducible repressor protein, and tetA for the efflux pump
protein. These two genes share a central regulatory region, with two overlapping tet promoters
and operators. In the absence of tetracycline, TetR protein binds to each of the two tet
operators, and blocks transcription of both the tetR and tetA genes. However, when
tetracycline is present, it binds the TetR repressor protein and induces a conformational change
in it, which results in the dissociation of the repressor protein from the operator DNA.
Subsequent expression of both TetR and TetA proteins ensures efficient export of tetracycline
(Hillen and Berens, 1994; Orth et al., 2000).
At this time, there are 36 known genes of resistance to tetracycline; 23 genes encode
membrane-associated, energy-dependent efflux pumps; of these, 21 are found exclusively in
Gram-negative bacteria, and only 2 (tetK and tetL) are found primarily in Gram-positive
isolates, although some Gram-negative isolates have been described with either tetK or tetL
(Roberts, 2003). The tetB gene has the widest range of distribution among the Gram-negative
bacteria, having so far been identified in more than 20 different genera (Chopra, 2002). It is
the only efflux pump that confers a high level of resistance to minocycline; other efflux pumps
provide poor protection against minocycline (Chopra et al., 1992; Speer et al., 1992).
Tetracycline efflux is the predominant mechanism of resistance in Enterobacteriaceae (Chopra
et al., 1992). Ten currently known genes encode a ribosomal protection mechanism; they
confer resistance to tetracycline, doxycycline, and minocycline. The ribosomal protection
mechanism is found in Gram-positive bacteria, Gram-negative anaerobic bacteria, and nonenteric Gram-negative bacteria, such as Neisseria gonorrhoeae and Haemophilus ducreyi.
23
(Roberts, 2003; Chopra and Roberts, 2001) The tetM ribosomal protection gene is the most
widespread tet gene in Gram-positive pathogens; so far, it has been found in clinical isolates
from 18 Gram-positive and 8 Gram-negative genera (Chopra and Roberts, 2001). Currently,
there are 39 genera of Gram-negative and 23 genera of Gram-positive bacteria described in
which the mechanism of tetracycline resistance has been established. Indeed, new genera
continue to be identified, and new tet genes continue to be described continuously (Chopra and
Roberts, 2001).
It is quite common for Gram-positive isolates to carry multiple genes of tetracycline
resistance from different classes. It is not common in Gram-negative isolates, though,
especially enteric species. The reason for this dissimilarity is not known (Roberts, 1996;
Chopra and Roberts, 2001). Gram-negative resistance genes encoding efflux mechanism are
generally found on transposons, large plasmids (most of which are conjugative and come from
different incompatibility groups), and integrons, while Gram-positive efflux genes are usually
carried on small plasmids. The ribosomal protection genes are usually found on transposons
integrated into the chromosome, or on plasmids. These mobile elements often carry other
genes of antibiotic resistance, and therefore selection for tetracycline resistance often leads to
selection for resistance to other agents, and vice versa (Chopra and Roberts, 2001).
Gram-negative efflux pumps are regulated through the expression of the repressor
proteins, which block transcription of the repressor and efflux genes in the absence of
tetracycline. In contrast, regulation of the Gram-positive efflux pumps (tetK and tetL) and
some of the ribosomal protection proteins does not involve a repressor, although expression of
resistance proteins is also inducible through transcriptional attenuation (Roberts, 1996;
Schnappinger and Hillen, 1996).
Bacterial resistance to tetracyclines usually arises through the acquisition of resistance
genes. A low-level resistance to tetracyclines may be mediated by bacterial multi-drug efflux
pumps with broad substrate specificity (Chopra, 2002). Mutations in efflux pumps, 16S rRNA
sequences, and alterations in cellular membrane permeability can also lead to resistance,
although this is quite rare (Gerrits et al., 2002; Ross et al., 1998). Possibly, this is the reason
why obligate intracellular parasites such as Rickettsiae and Chlamydiae have not yet been
found to acquire resistance to tetracyclines. Since tetracycline resistance is generally a result of
acquisition of tetracycline resistance genes rather than mutations, intracellular organisms
would require concurrent infection of the cell with two genera in order for the gene transfer to
occur from one organism to another. While the possibility of this was demonstrated in vitro,
the likelihood of the same occuring in vivo is rather low (Chopra and Roberts, 2001; Roberts,
2003). Remarkably, no tetracycline-resistant protozoans have yet been described. It should be
noted that all drug-resistant protozoans described so far, including Plasmodiumfalciparum, are
resistant because of mutations in the genome. Therefore, if protozoan parasites develop
tetracycline resistance in the future, it can be expected that their resistance will be due to
mutations rather than acquisition of new genes (Roberts, 2003).
24
2.3. Current Research to Overcome Tetracycline Resistance.
Currently, there are two approaches that have been undertaken to circumvent bacterial
resistance to tetracycline. The first approach involves the development of molecules that
would act as efflux pump inhibitors and could be used in combination with earlier
tetracyclines to support their activity (Rothstein et al., 1993; Nelson et al., 1994; Nelson and
Levy, 1999). Particular emphasis has been given to tetracycline-like molecules that could
block the efflux by competitive inhibition of the transporter; a group of compounds with
inhibitory activity has been identified, and some showed synergistic activity when tested in
combination with doxycycline (Levy and Nelson, 1998; Sum et al., 1998). The second
approach involves finding tetracycline derivatives that have properties similar to tetracyclines,
but do not elicit bacterial resistance.
The approach involving development of efflux pump inhibitors is analogous to the use
of clavulanic acid in conjunction with earlier beta-lactam antibiotics, such as amoxycillin.
Clavulanic acid is a beta-lactamase inhibitor, and its use restores the potency of the betalactam antibiotic when the two are used in combination; by preventing the enzymatic
degradation of the antibiotic. In the development of efflux pump inhibitors, however, one
limitation is the wide variety of tetracycline efflux pumps that exist in tetracycline-resistant
bacteria. The homology between different pumps is not high, and therefore the likelihood of
finding a universal inhibitor active against all pumps is also low. Another limitation to this
methodology is that many tetracycline-resistant bacterial strains actually are resistant due to
the ribosomal protection resistance mechanism, rather than active efflux of the drug. Some
organisms possess both active efflux and ribosomal protection genes. In these cases, the use of
efflux pump inhibitors would not provide any therapeutic advantage, as the ribosomal
protection would still render cells resistant to tetracycline. Therefore, the development of
efflux pump inhibitors only targets a fraction of tetracycline-resistant bacteria (Chopra, 2002).
Development of tetracycline analogs with properties similar to tetracyclines holds more
promise at this time (Sum and Petersen, 1999; Testa et al., 1993; Petersen et al., 1999). The
analogs of tetracycline, called glycylcyclines, have the basic structural features similar to other
tetracyclines and have been shown to be at least as potent as earlier tetracyclines against a
broad spectrum of Gram-negative and Gram-positive bacteria. Importantly, they also display
activity against strains expressing different tetracycline resistance genes, including those that
encode efflux and ribosomal protection mechanisms (Chopra, 2002). Glycylcyclines have a
higher binding affinity for ribosomes than other tetracyclines, which is a likely explanation for
the inability of ribosomal protection proteins to provide cells with resistance to glycylcyclines.
Tetracycline efflux proteins, on the other hand, are likely to be unable to recognize
glycylcyclines, or to transport them out of the cell (Chopra, 2002; Sum et al., 1998).
Currently, the most developed glycylcycline is tigecycline (also known as GAR-936);
human phase I and phase II clinical trials have been completed for this compound (Zhanel et
al., 2004). Studies concluded that tigecycline was efficacious, well tolerated, and could be
safely used; its side effects were typical of other tetracyclines. So far, no naturally occurring
glycylcycline-resistant strains have been identified among human clinical isolates; however,
resistance to earlier glycylcyclines, associated with point mutations in the efflux protein, has
been reported in two veterinary isolates. Laboratory-derived mutations in the TetB tetracycline
25
efflux pump in E. coli have also been shown to confer resistance to earlier glycylcyclines
(Guay et al., 1994). Since these compounds are analogs of earlier antibiotics, bacterial
resistance is expected to arise sooner for them than would be expected for a novel class of
antibiotics.
A few novel tetracyclines have been isolated from natural sources. The
dactylocyclines, glycosides of tetracycline, are a group of tetracycline derivatives produced by
Dactylosporangium spp. The dactylocyclines are active against tetracycline-resistant Grampositive microorganisms, however, not against Gram-negative ones; moreover, some
tetracycline-susceptible bacteria are not inhibited by dactylocyclines (Sum et al., 1998; Speer
et al., 1992).
2.4. Novel Approach to Tetracycline Resistance.
In this work, we describe a finding that has potential for the development of novel antibacterial treatments for diseases caused by bacteria that have become resistant to the antibiotic
tetracycline and its analogs. We show here that tetracycline resistant bacteria expressing the
TnlO gene of tetracycline resistance, upon induction with tetracycline, became extremely
susceptible to killing by the anti-cancer drug cisplatin. Tetracycline-resistant bacteria grown in
tetracycline and subsequently treated with cisplatin in the presence of tetracycline were killed
about 105 -fold more effectively than wild-type bacteria, or those that were tetracycline resistant
but not exposed to tetracycline. We observed this phenomenon in different strains of
tetracycline-resistant E. coli, regardless of whether the tetracycline-resistance genes were
carried on the bacterial chromosome or on autonomous plasmids. The order of treatment was
important, as cells exposed first to cisplatin and then to tetracycline were not sensitized. Other
antibiotics (ampicillin, kanamycin, and chloramphenicol) tested with the respective antibioticresistant bacteria did not produce the same effect of sensitization to cisplatin. Therefore,
sensitization to cisplatin reflects a unique relationship among cisplatin, tetracycline and the
tetracycline resistance gene.
We determined that the total cellular platinum levels in sensitized tetracycline-resistant
cells were not increased; however the levels of platinum DNA damage were higher in
sensitized tetracycline-resistant cells than in wild-type cells. At this time, we do not know the
mechanism of increased DNA damage formation; nor do we know the significance of elevated
DNA cisplatin levels for enhanced toxicity. Our experiments, however, clearly indicate that
increased destruction of bacteria by cisplatin requires tetracycline resistance genes, in addition
to cisplatin and tetracycline.
Based on this work, which is described in detail in later chapters, we suggest that the
therapeutic power of tetracyclines could be restored and further advanced by using a
complementary drug which, in combination with tetracycline, induced selective destruction of
tetracycline-resistant cells. We propose an approach that, rather than avoid or suppress
tetracycline resistance, would harness it to combat infections. We expect that such an approach
could prove useful against a wide range of pathogenic bacteria, particularly Gram-negative
bacteria expressing the efflux type of tetracycline resistance. Although we have not yet
established the mechanism of toxicity observed in our experiments, it is likely that such
26
toxicity is actually dependent on a common quality of tetracycline resistance genes (especially
efflux pumps), rather than on a feature that is specific to the TnlO gene.
Bacterial strains that have acquired tetracycline resistance as a component of their
multi-drug resistant profile are common among bacterial pathogens, since genes of resistance
to tetracycline are frequently found on the same mobile elements as genes of resistance to
other antibiotics (O'Brien et al., 1987; Jahn et al., 1979; Ng et al., 1999; Inamine and Burdett,
1985). A successful approach would therefore target not only pathogenic bacteria resistant
exclusively to tetracycline, but possibly would also be powerful against bacteria that have
acquired resistance to multiple antibiotics, as long as they have acquired tetracycline resistance
as well. Since we describe the combination of two unrelated drugs, this approach would not be
as likely to elicit bacterial resistance as introduction of tetracycline analogs.
It is possible that molecules other than cisplatin, in combination with other
tetracyclines, would lead to more powerful destruction of bacteria expressing tetracycline
resistance. It would be desirable to achieve similar or better results with a compound that has
less overall toxicity than cisplatin. The side effects of cisplatin must be carefully weighed
when considering the benefits of its potential use against bacterial infections. Cisplatin
treatment may lead to gastrointestinal, renal, reproductive and neurological toxicity (Loehrer
and Einhorn, 1984); in addition, cisplatin is a suspected, although not proven, human
carcinogen (Greene, 1992). However, in a situation with a deadly multi-drug resistant infection
these side effects may be acceptable. One group of patients that is likely to benefit from
cisplatin-tetracycline treatment is cancer patients suffering from serious concurrent bacterial
infections. We would like to emphasize that we achieved selective destruction of tetracyclineresistant bacteria with cisplatin doses that are consistent with blood platinum levels found in
patients undergoing cisplatin chemotherapy (Sileni et al., 1992; Panteix et al., 2002; Nagai et
al., 1996).
In those instances where tetracyclines are still used, or even are currently the drugs of
choice, the possibility still exists for rapid development of bacterial resistance to this class of
antibiotics, and more conditions are likely to become refractory to tetracycline treatment. A
method that would afford destruction of tetracycline-resistant bacteria has an important future
potential, should resistance develop in conditions that are currently treated by tetracyclines
(Liss and Batchelor, 1987). In our study, cells expressing the genes of tetracycline resistance
actually became primary targets for destruction, since these genes in fact made cells more
susceptible to a second agent, cisplatin. We therefore suggest that tetracycline resistance can
be viewed not as a barrier to overcome, but rather as a property of bacterial cells to be
exploited in a fight against infection.
27
2.5. Structures of Tetracycline and its Analogs.
Tetracycline
Oxytetracycline
NH 2
NH2
OH
Chlortetracycline
O
OH
O
O
Doxycycline
H 3C.
OH
N(CH 3) 2
NH2
Demethylchlortetracycline
Methacycline
OH
N(CH3)2
OH
NH2
OH
~~OH
O
OH
OH
O
O
0
Minocycline
Tigecycline (GAR-936)
NH2
NH2
28
References:
Akiba, T., K. Koyama, Y. Ishiki, S. Kimura, T. Fukushima. On the mechanism of the
development of multiple-drug resistant clones of Shigella. 1960. Jpn. J. Microbiol. 4:219227.
Chang H.R., R. Comte, and J.C. Pechere. In vitro and in vivo effects of doxycycline on
Toxoplasma gondii. 1990. Antimicrob. Agents Chemother. 34(5):775-780.
Chopra, I. 2002. New developments in tetracycline antibiotics: glycylcyclines and
tetracycline efflux pump inhibitors. Drug resistance updates. 5:119-125.
Chopra, I., P.M. Hawkey, and M. Hinton. 1992. Tetracyclines, molecular and clinical
aspects. J. of Antimicrobial Chemotherapy. 29:245-277.
Chopra, I., and M. Roberts. 2001. Tetracycline antibiotics: mode of action, applications,
molecular biology, and epidemiology of bacterial resistance. Microbiology and Molecular
Biology Reviews. 65(2):232-260.
Clendenning, W.E. 1965. Complications of tetracycline therapy. Arch. Dermatol. 91:628632.
Col, N.F., and R.W. O'Connor. 1987. Estimating worldwide current antibiotic usage:
report of Task Force 1. Rev. Infect. Dis. 9(Suppl.3):S232-243.
Connell,
S.R., D.M. Tracz, K.H. Nierhaus,
and D.E. Taylor.
2003. Ribosomal
protection proteins and their mechanism of tetracycline resistance. Antimicrobial agents
and chemotherapy. 47(12):3675-3681.
Dennis, D.T., T.V. Inglesby, D.A. Henderson, J.G. Bartlett, M.S. Ascher, E. Eitzen,
A.D. Fine, A.M. Friedlander, J. Hauer, M. Layton, S.R. Lillibridge, J.E. McDade,
M.T. Osterholm, T. O'Toole, G. Parker, T. M. Perl, P.K. Russell, K. Tonat. 2001.
Tularemia as a biological weapon. Medical and public health management. JAMA.
285(21):2763-2773.
Edlind, T. D. 1989. Tetracyclines as antiparasitic agents: lipophilic derivatives are highly
active against Giardia lamblia in vitro. Antimicrobial Agents and Chemotherapy. 33(12):
2144-2145.
Felmingham, D., R.N. Grunenberg, and the Alexander Project Group. 2000. The
Alexander Project 1996-1997: latest susceptibility data from this international study of
bacterial pathogens from community-acquired lower respiratory tract infections. J.
Antimicrob. Chemother. 45:191-203.
29
Gerrits, M.M., M.R. de Zoete, N.L.A. Arents, EJ. Kuipers, and J.G. Kusters. 2002.
16S rRNA mutation-mediated tetracycline resistance in Helicobacter pylori. Antimicrob.
Agents Chemother. 46(9):2996-3000.
Greene, M.H. 1992. Is cisplatin a human carcinogen? J. Natl. Cancer Inst. 84(5):306-312.
Guay, G.G., M. Tuckman, D.M. Rothstein. 1994. Mutations in the TetA(B) gene that
cause a change in substrate specificity of the tetracycline efflux pump. Antimicrob. Agents
Chemother. 38(4):857-860.
Heritage J., and P.M. Hawkey. Tetracycline-resistant Neisseria gonorrhoeae. 1988. J.
Antimicrob. Chemother. 22(5):575-579.
Hillen, W., and C. Berens. 1994. Mechanisms underlying expression of TnlO encoded
tetracycline resistance. Annu. Rev. Microbiol. 48:345-369.
Hoerauf, A., K. Nissen-Pahle, C. Schmetz, K.Henkle-Duhrsen, M.L. Blaxter, D.W.
Buttner, M.Y. Gallin, K.M. Al-Qaoud, R. Lucius, and B. Fleischer. 1999. Tetracycline
therapy targets intracellular bacteria in the filarial nematode Litomosoides sigmodontis and
results in filarial infertility. The Journal of Clinical Investigation. 103(1): 11-18.
Hughes, V.M., and N. Datta. 1983. Conjugative plasmids in bacteria of the "preantibiotic" era. Nature. 302:725-726.
Inamine, J.M., and V. Burdett. 1985. Structural organization of a 67-kilobase
streptococcal conjugative element mediating multiple antibiotic resistance. J. Bacteriol.
161(2):620-626.
Inglesby, T.V., D.T. Dennis, D.A. Henderson, J.G. Bartlett, M.S. Ascher, E. Eitzen,
A.D. Fine, A.M. Friedlander, J. Hauer, J.F. Koerner, M. Layton, J. McDade, M.T.
Osterholm, T. O'Toole, G. Parker, T.M. Perl, P.K. Russell, M. Schoch-Spana,
K.
Tonat. 2000. Plague as a biological weapon. Medical and public health management.
JAMA. 283(17):2281-2290.
Inglesby, T.V., D.A. Henderson, J.G. Bartlett, M.S. Ascher, E. Eitzen, A.M.
Friedlander, J. Hauer, J. McDade, M.T. Osterholm, T. O'Toole, G. Parker, T.M.
Perl, P.K. Russell, K. Tonat. 1999. Anthrax as a biological weapon. Medical and public
health management. JAMA. 281(18):1735-1745.
Inglesby T.V., T. O'Toole, D.A. Henderson, J.G. Bartlett, M.S. Ascher, E.Eitzen,
A.M. Friedlander, J. Gerberding, J. Hauer, J. Hughes, J. McDade, M.T. Osterholm,
G. Parker, T. M. Perl, P.K. Russell, K. Tonat. 2002. Anthrax as a biological weapon,
2002. Updated recommendations for management. JAMA. 287(17): 2236-2252.
Islam, M.R. Single dose tetracycline in cholera. 1987. Gut. 28(8): 1029-1032.
30
Jahn, G., R. Laufs, P.-M. Kaulfers, and H. Kolenda. 1979. Molecular nature of two
Haemophilus influenzae R factors containing resistances and the multiple integration of
drug resistance transposons. J. Bacteriol. 138(2):584-597.
Ji, B., P. Jamet, E.G. Perani, S. Sow, C.Lienhardt, C. Petinon, J.H. Grosset. 1996.
Bactericidal activity of single dose of clarithromycin plus minocycline, with or without
ofloxacin, against Mycobacterium leprae in patients. Antimicrob. Agents Chemother.
40(9):2137-2141.
Ji, B., S. Sow, E. Perani, C. Lienhardt, V. Diderot, and J.Grosset. 1998. Bactericidal
activity of a single-dose combination of ofloxacin plus minocycline, with or without
rifampin, against Mycobacterium leprae in mice and in lepromatous patients.
Antimicrobial Agents and Chemotherapy. 42(5): 1115-1120.
Katiyar, S.K., and T.D. Edlind. 1991. Enhanced antiparasitic activity of lipophilic
tetracyclines: role of uptake. Antimicrob. Agents Chemother. 35(11):2198-2202.
Levy, S.B., L.M. McMurry, T.M. Barbosa, V. Burdett, P. Courvalin, W. Hillen, M.C.
Roberts, J.I. Rood, D.E. Taylor. Nomenclature for new tetracycline resistance
determinants. Antimicrobial Agents and Chemotherapy. 43(6): 1523-1524.
Levy, S.B., L.M. McMurry, V.Burdett, P. Courvalin, W. Hillen, M.C. Roberts, and
D.E. Taylor. 1989. Nomenclature for tetracycline resistance determinants. Antimicrobial
Agents and Chemotherapy. 33(8): 1373-1374.
Levy, S.B., and M. Nelson. 1998. Reversing tetracycline resistance. A Renaissance for the
tetracycline family of antibiotics. Advances in experimental medicine and biology. 456:1725.
Lima, A.A.M., N.L. Lima, M.C.N. Pinho, E.A.Barros, MJ.Teixeira, M.C.V. Martins,
and R.L. Guerrant. 1995. High frequency of strains multiply resistant to ampicillin,
trimethoprim-sulfamethoxazole, streptomycin, chloramphenicol, and tetracycline isolated
from patients with shigellosis in northeastern Brazil during the period 1988 to 1993.
Antimicrob. Agents Chemother. 39(1):256-259.
Liss, R.H., and F.R. Batchelor. 1987. Economic evaluations of antibiotic use and
resistance - a perspective report of Task Force 6. Rev. Infect. Dis. 9(Suppl. 3):S297-312.
Loehrer, P.J., and L.H. Einhorn. 1984. Drugs five years later. Cisplatin. Ann. Int. Med.
100(5):704-713.
Luger S.W., P.Paparone, G.P. Wormser, R.B. Nadelman, E.Grunwaldt, G. Gomez,
M.Wisniewski, J.J. Collins. Comparison of cefuroxime axetil and doxycycline in
treatment of patients with early Lyme disease associated with erythema migrans. 1995.
Antimicrob. Agents Chemother. 39(3):661-667.
31
Nadelman, R.B., J. Nowakowski, D. Fish, R.C. Falco, K. Freeman, D. McKenna, P.
Welch, R. Marcus, M.E. Aguero-Rosenfeld, D.T. Dennis, G.P.Wormser. Prophylaxis
with single-dose doxycycline for the prevention of Lyme disease after an Ixodes scapularis
tick bite. 2001. N. Engl. J. Med. 345(2): 79-84.
Nagai, N., M. Kinoshita, H. Ogata, D. Tsujino, Y. Wada, K. Someya, T. Ohno, K.
Masuhara,
Y. Tanaka, K. Kato, H. Nagai, A. Yokoyama,
and Y. Kurita. 1996.
Relationship between pharmacokinetics of unchanged cisplatin and nephrotoxicity after
intravenous infusions of cisplatin to cancer patients. Cancer Chemother. Pharmacol. 39(12):131-137.
Navas, E. 2002. Problems associated with potential massive use of antimicrobial agents as
prophylaxis or threrapy of a bioterrorist attack. Clin Microbiol Infect. 8(8):534-539.
Nelson, M.L., and S.B. Levy. 1999. Reversal of tetracycline resistance mediated by
different bacterial tetracycline resistance determinants by an inhibitor of the Tet(B)
antiport protein. Antimicrobial Agents and Chemotherapy. 43(7):1719-1724.
Nelson, M.L., B.H. Park, and S.B. Levy. 1994. Molecular requirements for the inhibition
of the tetracycline antiport protein and the effect of potent inhibitors on the growth of
tetracycline-resistant bacteria. J. Med. Chem. 37(9):1355-1361.
Ng, L.-K., M.R. Mulvey, I. Martin, G.A. Peters, and W. Johnson.
1999. Genetic
characterization of antimicrobial resistance in Canadian isolates of Salmonella serovar
Typhimurium DT104. Antimicrob. Agents Chemother. 43(12):3018-3021.
O'Brien, T.F. 1987. Resistance of bacteria to antibacterial agents: report of Task Force 2.
Rev. Infect. Dis. 9(Suppl. 3):S244-260.
Olson, C.A., and H.D. Riley. 1966. Complications of tetracycline therapy. J. Pediatr.
68(5):783-791.
Orth, P., D. Schnappinger, W. Hillen, W. Saenger, and W. Hinrichs. 2000. Structural
basis of gene regulation by the tetracycline inducible Tet repressor-operator system.
Nature. 7(3):215-219.
Panteix, G., A. Beaujard, F. Garbit, C. Chaduiron-Faye, M. Guillaumont, F. Gilly, P.
Baltassat, and F. Bressole. 2002. Population pharmacokinetics of cisplatin in patients
with advanced ovarian cancer during intraperitoneal hyperthermia chemotherapy.
Anticancer Res. 22(2B): 1329-1336.
Petersen, P.J., N.V. Jacobus, WJ. Weiss, P.E. Sum, and R.T. Testa. 1999. In vitro and
in vivo antibacterial activities of a novel glycylcycline, the 9-t-Butylglycylamido
derivative of minocycline (GAR-936). Antimicrobial Agents and Chemotherapy. 43(4):
738-744.
32
Pflug G.R. 1963. Toxicities associated with tetracycline therapy. Am. J. Pharm. Sci.
Support Public Health. 135:438-450.
Portnoy, D. A retrospective study of doxycycline in the treatment of genitourinary
infections. 1986. Clin. Ther. 9:3-5.
Pradines, B., A. Spiegel, C. Rogier, A. Tall, J. Mosnier, T. Fusai, J.F. Trape, and D.
Parzy. 2000. Antibiotics for prophylaxis of Plasmodium falciparum infections: in vitro
activity of doxycycline against senegalese isolates. Am. J. Trop. Med. Hyg. 62(1): 82-85.
Rabbani, G.H., M.R. Islam, T. Butler, M. Shahrier, and K. Alam. Single-dose
treatment of cholera with furazolidone or tetracycline in a double-blind randomized trial.
1989. Antimicrob. Agents Chemother. 33(9):1447-1450.
Ribeiro,
M.L.,
M.M.
Gerrits,
Y.H.B. Benvengo,
M. Berning,
A.P.O.
Godoy,
E.J.Kuipers, S. Mendonca, A.H.M. van Vliet, J. Pedrazzoli, and J.G. Kusters. 2004.
Detection of high-level tetracycline resistance in clinical isolates of Helicobacter pylori
using PCR-RFLP. FEMS Immunology and Medical Microbiology. 40:57-61.
Roberts, M.C. 1996. Tetracycline resistance determinants: mechanisms of action,
regulation of expression, genetic mobility, and distribution. FEMS Microbiology Reviews.
19:1-24.
Roberts, M.C. 2003. Tetracycline therapy: update. Clinical Infectious Diseases.
36(4):462-467.
Ross, J.I., E.A. Eady, J.H. Cove, and W.J. Cunliffe. 1998. 16S rRNA mutation
associated with tetracycline resistance in a gram-positive bacterium. Antimicrob. Agents
Chemother. 42(7): 1702-1705.
Rothstein, D.M., M. McGlynn, V. Bernan, J. McGahren, J. Zaccardi, N. Cekleniak,
and IKP. Bertrand. 1993. Detection of tetracyclines and efflux pump inhibitors.
Antimicrob. Agents Chemother. 37(8):1624-1629.
Russel, P., S.M. Eley, MJ. Fulop, D.L. Bell, and R.W. Titball. 1998b.The efficacy of
ciprofloxacin and doxycycline against experimental tularaemia. J. Antimicrob. Chemother.
41:461-465.
Russel P., S.M. Eley, M. Green, A.J. Stagg, R.R. Taylor, M. Nelson, RJ. Beedham,
D.L. Bell, D. Rogers, D. Whittington, and R.W. Titball. 1998a.Efficacy of doxycycline
and ciprofloxacin against experimental Yersinia pestis infection. J. Antimicrob.
Chemother. 41:301-305.
Sabath, LD. Drug resistance of bacteria. 1969. N. Engl. J. Med. 280(2):91-4.
33
Sanchez, A.R., R.S. Rogers III, and P.J. Sheridan. 2004. Tetracycline and other
tetracycline-derivative staining of the teeth and oral cavity. Int. J. of Dermatology. 43: 709715.
Schnappinger, D., and W. Hillen. 1996. Tetracyclines: antibiotic action, uptake, and
resistance mechanisms. Arch Microbiol. 165:359-369.
Schwartz, E., and G. Regev-Yochay. 1999. Primaquine as prophylaxis for malaria for
non-immune travelers: a comparison with mefloquine and doxycycline. Clin. Infect. Dis.
29:1502-1506.
Sileni, V.C., V. Fosser, P. Maggian, E. Padula, M. Beltrame, M. Nicolini, and P.
Arslan. 1992. Pharmacokinetics and tumor concentration of intraarterial and intravenous
cisplatin in patients with head and neck squamous cancer. Cancer Chemother. Pharmacol.
30(3):221-225.
Smith H.L, and T.V. Rajan. 2000. Tetracycline inhibits development of the infectivestage larvae of filarial nematodes in vitro. Experimental Parasitology. 95: 265-270.
Speer, B.S., L. Bedzyk, and A.A. Salyers. 1991. Evidence that a novel tetracycline
resistance gene found on two Bacteroides transposons encodes an NADP-requiring
oxidoreductase. J. Bacteriol. 173(1): 176-183.
Speer, B.S., and A.A. Salyers. 1989. Novel aerobic tetracycline resistance gene that
chemically modifies tetracycline. J. Bacteriol. 171(1):148-153.
Speer, B.S., N.B. Shoemaker, and A.A. Salyers. 1992. Bacterial Resistance to
tetracycline: mechanisms, transfer, and clinical significance. Clinical Microbiology
Reviews. 5(4):387-399.
Sum, P.E., and P. Petersen. 1999. Synthesis and structure-activity relationship of novel
glycylcycline derivatives leading to the discovery of GAR-936. Bioorganic and Medicinal
Chemistry Letters. (9): 1459-1462.
Sum, P.E., F.W. Sum, and S.J. Projan. 1998. Recent developments in tetracycline
antibiotics. Current Pharmaceutical Design. 4(2): 119-132.
Taylor, D.E., and A.Chau. 1996. Tetracycline resistance mediated by ribosomal
protection. Antimicrobial agents and chemotherapy. 40(1): 1-5.
Testa, R.T., P.J. Petersen, N.V. Jacobus, P.E. Sum, V.J. Lee, F.P. Tally. 1993. In vitro
and in vivo antibacterial activities of the glycylcyclines, a new class of semisynthetic
tetracyclines. Antimicrobial Agents and Chemotherapy. 37(11):2270-2277.
Tonat. 2000. Plague as a biological weapon. Medical and public health management.
JAMA. 283(17):2281-2290.
34
Zhanel, G.G., K. Homenuik, K. Nichol, A. Noreddin, L. Vercaigne, J. Embil, A. Gin,
J.A. Karlowsky, and DJ. Hoban. 2004. The glycylcyclines: a comparative review with
the tetracyclines. Drugs. 64(1):63-88.
Waugh, M.A., CJ. Lacey, P.M. Hawkey, J. Heritage, A. Turner, A.E. Jephcott. 1988.
Spread of Neisseria gonorrhoeae resistant to tetracycline outside the United States of
America. Br. Med. J. 296:898.
35
Chapter 3'. Discovery of Connection between Tetracycline Resistance
and Cisplatin Cytotoxicity in Escherichia coli.
In the course of our work with cisplatin and base excision repair-deficient E. coli
cells (see Appendix III), we noticed that in our experiments we did not observe the
previously reported sensitivity of mutY cells to cisplatin (Fig. 1). CC104 mutY strains
previously had been shown to be more sensitive to cisplatin than wild-type CC104 strains
(Maria Kartalou, Ph.D. Thesis, 2000); these cells now seemed to be as sensitive to cisplatin
as wild-type CC104 cells. The AB1157 mutY and AB1157 wild-type strains were also
equally sensitive to cisplatin. The discrepancy between the previously observed and
currently observed sensitivities led us to go back to the frozen stocks of E. coli in which
the sensitivity had been previously observed, and to grow cells from scratch in their
respective antibiotics.
When cells were grown from the frozen stocks in their respective antibiotics, we
noticed that now the cells from the two genetic backgrounds, CC104 and AB1157,
behaved differently (Fig. 2). Specifically, the AB 1157 wild-type and AB 1157 mutY strains
were still equally sensitive to cisplatin. However, we now observed that the CC104 mutY
was actually more sensitive to cisplatin than CC104 wild-type, in line with previous
observations from our lab.
We sought to find the origins of this apparently different behavior of the two mutY
strains. We checked the two frozen stocks of CC104 mutY, the one originally obtained by
the lab in 1992, and the one created from it in 2000, in order to see whether the difference
between the two stocks could explain the difference between the two strains. The original
1992 stock was the stock that did not show the differential sensitivity in our initial
experiments in 2003, and prompted this investigation. The 2000 stock was the one in
which differential sensitivity was reported in 2000, and in which differential sensitivity
was observed again when cells were grown from scratch in antibiotics. When we grew
cells from the two stocks side by side in antibiotic, however, we noticed that now both
CC104 mutY stocks displayed differential sensitivity to cisplatin (Fig. 3). Our assumption
that the two stocks were somehow different was therefore wrong: the two stocks did
behave the same when grown under identical conditions.
We started to suspect the influence of the antibiotic. After all, the gene of resistance
to tetracycline, inserted to disrupt the mutY gene in the CC104 strain, can only be induced
by the presence of tetracycline in the growth medium. Therefore, the presence or absence
of tetracycline in the growth medium creates a different biological environment, which
could affect sensitivity to cisplatin (Fig. 6).
On the other hand, we still had the AB1157 strains, which did not show differential
sensitivity to cisplatin. In the AB 1157 mutY strain, the mutY gene is disrupted by insertion
of a chloramphenicol resistance gene. While it was possible that the two genetic
backgrounds - the AB 1157 and the CC104 - affected sensitivity to cisplatin differently, it
'This Chapter presents introduction to Chapter 4 and discusses its historical context.
36
was also possible that the difference between the two genes used to disrupt the mutY- the
tet and the cam genes - could play a role as well.
James Delaney of this laboratory had the AB1157 mutY strains where mutY was
disrupted by a tet gene from the CC104 strain. Martin Marinus had the CC104 strains
where mutY was disrupted by a cam gene from AB 1157. We therefore had a complete set
of mutant strains at hand to check whether the sensitivity to cisplatin was specific to the
genetic background or to the genes used to create mutations.
The results of an experiment with AB1157 mutY::tet and CC104 mutY::cam strains
grown in their respective antibiotics answered this question unequivocally. The sensitivity
to cisplatin was not specific to the genetic background. Rather, mutY::tet conferred
sensitivity to cisplatin when introduced into the AB1157 strain; the CC104 strain, on the
other hand, lost sensitivity to cisplatin when the mutation was substituted for mutY::cam
(Fig. 4). This observation further suggested that the gene of resistance to tetracycline could
actually be responsible for the observed cisplatin toxicity.
We then went to check this hypothesis directly, and compared the survival of
mutY::tet strains grown with and without tetracycline in the growth medium. Consistent
with results described above, tetracycline-resistant strains grown with tetracycline acquired
high sensitivity to cisplatin, while tetracycline-resistant strains grown without tetracycline
were as sensitive to cisplatin as wild-type strains. This observation was valid for both
genetic backgrounds that we tested, AB1157 and CC 104 (Fig. 5).
To further confirm the role of the tetracycline resistance gene in conferring
sensitivity to cisplatin to E. coli, we tested a number of other mutant strains in which genes
were disrupted either with the tetracycline resistance gene, or with other genes. In each
case, we observed that growth in tetracycline, but not in other antibiotics, resulted in high
sensitivity to cisplatin. We also observed that tetracycline-resistant cells grown without
tetracycline were as sensitive to cisplatin as their counterpart wild-type cells. High
sensitivity to cisplatin was also observed in a wild-type MG1655 strain which had the same
gene of tetracycline resistance on the F' plasmid, rather than on a chromosome, and was
grown with tetracycline. This strain, when grown without tetracycline, was as sensitive to
cisplatin as a wild-type MG1655 without the F' plasmid. The same wild-type MG1655
strain with the F' plasmid containing gene of resistance to kanamycin, when grown in
kanamycin, was still as sensitive as a wild-type. These experiments proved that the
sensitivity to cisplatin was not dependent on the type of genetic rearrangement, but rather
on the presence of the tet gene and on its induction by tetracycline. These findings and
others are further discussed in later chapters.
37
Figure 1. Survival of mul Y parallels wI survival in both AB 1157 and CC 104 E. coli
strains grown without antibiotics.
~--'-I
100
I
-+-AB1157
Ci
>
___ AB1157 mutY
10
.~
-It-
:J
UJ
~
0
wt
CC104 wt
~CC104
1
0.1
0
25
50
75
J.1MCOOP
38
100
mutY
Figure 2. Same strains as in Fig. I, but grown with respective antibiotics. Survival
of ABl157 mutY::cam grown with chloramphenicol parallels that of AB1157 wt.
However, CC104 mutY::tet grown with tetracycline is more sensitive to cisplatin than
CC104 wt.
100
ii
>
.~
~
-+- AB1157 wt
10
___
AB1157
mutY::cam
-A-CC104wt
1
fI)
'#.
~CC104
0.1
0.01
o
50
100
150
200
J.lM COOP
39
mutY::tet
Figure 3. When grown with tetracycline, both old and new stocks of CC 104 mul Y
are more sensitive to cisplatin than CCI04 wI.
100
Ii
>
.~
~
en
~
-+-CC104
wt (1992)
___ CC104 wt (2000)
10
----.- CC104 mutY(1992)
0
--*- CC104 mutY(2000)
1
0
25
50
100
75
~M COOP
40
Figure 4. mutY::tet confers sensItIvIty to cisplatin. ABl157 with mutY::tet
acquired sensitivity to cisplatin, while CC 104 with mut Y::eam lost such sensitivity; mut Y
strains were grown with respective antibiotics (tetracycline or chloramphenicol).
100
-+-AB1157
wt
__e_AB1157 muty::tet
0.1
o
50
100
150
200
~M COOP
41
-..-
CC104 wt
~
CC104 mutY::cam
Figure 5. CCI04 and AB1157 mutY::tet strains acquired sensitivity to cisplatin
only when grown with tetracycline.
5a. CC 104 strains.
100
10
--+-
'ii
>
(I)
~
0
_
1
"E:s
CC1 04 wt
CC104 mutY::tetlNoTc
-.-CC104
0.1
mutY::tetITc
0.01
0
100
50
150
200
~M COOP
5b. ABl157 strains.
100
--+-
10
'ii
>
~
(I)
1
?fl.
0.1
o
50
100
150
200
~M COOP
42
AB1157 wt
_AB1157
mutY::tetlNoTc
-.-AB1157
mutY::tetITc
Figure 6. In the absence of tetracycline, tetracycline-inducible repressor TetR binds
to the tet operators and blocks expression of the two Tn] 0 tet genes, tetA and tetR. When
tetracycline is present, in the complex with magnesiwn it binds to TetR and induces
conformational change in the repressor protein. This leads to dissociation of the TetR from
tet operators and subsequent expression of TetA antiporter and TetR repressor proteins.
TetA is a trans-membrane efflux pump, which exports tetracycline out of the cell by
exchanging one proton for one [Tc-Mgt complex. When intracellular tetracycline levels
drop, TetR returns to its original conformation and prevents further expression of the tet
genes.
~
Tn10
tetA
~
tetR
TC~",n
~o~o0{7
lJ
H+
H+
[Tc-M9t
H+6.
6.
Tn10
TetA ~
~TetR
TetA
43
H+
H+
Chapter 41. Tetracycline Resistance Determinant from Transposon TnlO
Confers High Sensitivity to Cisplatin in Escherichia coll upon Induction
with Tetracycline.
4.1. Abstract
Tetracycline resistance associated with transposon TnlO is encoded by the most
widespread tetracycline resistance determinant among Gram-negative bacterial pathogens.
Transposon TnlO carries two class B genes responsible for inducible tetracycline
resistance: one for the tetracycline efflux protein TetA, and another for the tetracyclineinducible repressor protein TetR, which blocks transcription of both genes in the absence
of tetracycline. In the presence of tetracycline, however, both proteins are expressed as a
result of a conformational change in TetR that allows transcription to proceed. We found
that tetracycline sensitized tetracycline-resistant E. coli cells to the toxic effects of the anticancer drug cisplatin. E. coli cells carrying the TnlO tetracycline resistance gene acquired
high sensitivity to cisplatin upon growth in tetracycline-containing medium, and were
further sensitized to cisplatin when tetracycline was present during cisplatin treatment.
Enhanced sensitivity to cisplatin in tetracycline-resistant E. coli cells was accompanied by
increased levels of cisplatin accumulation in cellular DNA. Plasmid pBR322, which carries
the constitutively expressed class C tetracycline resistance gene, did not confer sensitivity
to cisplatin upon cellular growth in tetracycline-containing medium; however, it also
sensitized cells to cisplatin when tetracycline was present during cisplatin treatment. Since
tetracycline induced high susceptibility to therapeutically relevant doses of cisplatin in
tetracycline-resistant bacteria, these results suggest an important new potential for
overcoming tetracycline resistance in bacterial pathogens, and may have implications for
cisplatin-based therapies as well.
This Chapter is a manuscript in preparation.
44
4.2. Introduction
Tetracyclines, a broad-spectrum class of antibiotics, were discovered in the 1940s,
and since the 1950s, they have been used in the treatment of a wide variety of bacterial
infections caused by both Gram-negative and Gram-positive bacteria. However, the
emergence of bacterial resistance became a limiting factor shortly after the introduction of
tetracycline therapy. Today, resistance to tetracyclines is a major clinical issue, and
ongoing research efforts are currently directed towards finding ways to overcome
tetracycline resistance (Levy and Nelson, 1998; Sum et al., 1998; Chopra, 2002; Roberts,
2003).
Tetracycline is a bacteriostatic agent, which stops bacterial growth by reversibly
binding to the ribosome and inhibiting protein synthesis. Currently, three types of
tetracycline resistance are known: active efflux of the drug, ribosome protection and
chemical modification of tetracycline, but only the former two are of clinical significance
(Schnappinger and Hillen, 1996; Chopra and Roberts, 2001). Active efflux of drug is the
most common tetracycline resistance mechanism, mediated by expression of tetracyclinespecific trans-membrane efflux pumps. In Gram-negative bacteria, there are currently
several known classes of tetracycline resistance determinants encoding tetracycline efflux
pumps, which share a common genetic organization (Roberts, 1996; Butaye et al., 2003).
The most widespread and the best-studied tetracycline resistance determinant of Gramnegative bacteria is the class B determinant associated with transposon TnlO (Hillen and
Berens, 1994). It consists of two genes in a divergent orientation, one coding for TetR, the
tetracycline-responsive repressor protein, and another coding for TetA efflux pump protein
- an antiporter which counterbalances the export of [tetracycline-Mg]+ complex out of the
cell with the import of a single H+ . These two genes share a central regulatory region, with
two overlapping tet promoters and operators. In the absence of tetracycline, TetR protein
binds to each of the two tet operators, and blocks transcription of both tetR and tetA genes.
However, when tetracycline is present, it induces a conformational change in the TetR
repressor protein by binding it in a complex with Mg2+. The conformational change in
TetR induced by the [tetracycline-Mg]+ complex results in dissociation of the repressor
protein from the operator DNA. This leads to expression of both TetR and TetA proteins,
ensuring the efficient export of tetracycline before it can reach its target, the ribosome
(Kisker et al., 1995; Orth et al., 1998; Orth et al., 2000; Saenger et al., 2000). For
convenience, in this work we will refer to the entire tetracycline resistance determinant as a
tet gene.
Cisplatin, which is a well-established anti-neoplastic drug discovered in 1965 and
successfully used in treatment of testicular cancer, as well as cancers of the ovary, lung,
head and neck, is a concurrent focus of this report. The precise mechanism of cisplatin
toxicity is not known, although many pathways leading to toxicity have been elucidated in
both bacterial and mammalian cells (Fuertes et al., 2003; Chu, 1994; Zdraveski et al.,
2000). It is currently believed that toxicity of cisplatin stems from its ability to form toxic
DNA lesions, although the events leading from DNA lesions to cell death are still a subject
of continuing research.
45
We describe here an unexpected finding that cisplatin has the capability to destroy
selectively the tetracycline resistant bacterial cells expressing the TnlO tet gene. We
determined that the high susceptibility to cisplatin in tetracycline-induced cells expressing
the TnlO tet gene was neither due to the blocked efflux of tetracycline, nor to the increased
cellular levels of cisplatin. However, increased toxicity of cisplatin was accompanied by
increase in platinum accumulation in cellular DNA. The precise mechanism of increased
DNA damage formation and associated cellular susceptiblity to cisplatin in tetracyclineresistant E. col remains to be determined at this time, and may hold potential in the
development of new pharmaceutical regimens involving cisplatin and tetracycline.
46
4.3. Materials and Methods
Bacterial strains and reagents
The genotypes of the E. coli K-12 strains used for this work are listed in Table 1. Strain
GM4292 was constructed by mating exponentially growing XL1-Blue (Stratagene) with
MG1655 (both at 1-2 x 10 /ml at a 1:1 ratio) for 60 min and plating dilutions of the
mixture on MacConkey agar-tetracycline plates. A red tetracycline-resistant, nalidixic
acid-sensitive colony was selected, purified and designated GM4292. Strain GM4293 was
constructed by introducing F'42 (F'lac +) from AB1874 into KM81 by conjugation as
described above and selecting for Lac+ StrR colonies. A purified Lac+ StrR recombinant
overnight culture was diluted and plated on MacConkey agar-kanamycin and following
incubation overnight at 370 C, white colonies were selected. White colonies can arise either
by gene conversion, where the lac+ allele on the F'42 plasmid is converted to lacZ::Kan, or
by loss of the F'lac+ . The white colonies were tested for their ability to transfer
kanamycin-resistance at high frequency to MG1655, indicating retention of the F' plasmid.
One of these MG1655 recipients was designated GM4293.
Strains carrying the pBR322 plasmid were created using standard transformation by
electroporation protocol. Briefly, cells were grown to mid-log phase, spun down, and
washed 3 times in ice-cold sterile water. Cells were then resuspended in the final volume
of 100 l of ice-cold sterile water, and 10 lig pBR322 (New England BioLabs) was added.
The BTX electroporation system, ElectroCell Manipulator 600, was used to deliver the
pulse. After the pulse, cells were immediately resuspended in lml of LB broth.
Appropriate dilutions of cells were prepared in M9 minimal salts, plated on LB-ampicillin
plates, and allowed to form colonies overnight.
MacConkey agar and LB broth were purchased from Difco. LB broth contained 10
g tryptone, 5 g yeast extract, and 10 g NaCl per L. All antibiotics and cisplatin (cisdiamminedichloroplatinum(II), or CDDP) were purchased from Sigma-Aldrich.
Antibiotics were used at the following final concentrations: 15 jpg/mltetracycline (Tc), 100
glg/ml ampicillin, 50 plg/ml kanamycin, 100 ljg/ml streptomycin, 30 ljg/ml nalidixic acid
and 20 jlg/ml chloramphenicol. Cisplatin solutions were prepared by dissolution in
phosphate buffered saline (PBS) at 370 C for one hour with rotation, filtering the solution
through a 0.2jim Acrodisc filter, and determining the concentration by UV absorbance at
A301 using a Beckmann DU-65 spectrophotometer. Appropriate doses of cisplatin were
then prepared by dilution in PBS.
Survival curves
Cells were grown overnight from single colonies in LB broth. The next morning,
saturated overnight cultures were diluted 1:100 and grown for two hours to mid-log phase
(cell density approximately 3x108 cells/ml). At this time, the cells were spun down,
resuspended in M9 minimal salts, and treated with various concentrations of cisplatin at
370 C for 1 hour. Exposure to cisplatin was stopped by dilution of the cells 1:100 in M9
salts. Appropriate serial dilutions were prepared for each dose and plated on LB plates or
47
LB/Tet plates. Colonies were counted the next day after incubation at 370 C. Survival was
calculated as a percentage of cells forming colonies in treated populations relative to
untreated.
Atomic Absorption Spectroscopy
For atomic absorption experiments, cells were grown and treated according to the
procedure outlined above. Appropriate serial dilutions were prepared and the cells plated
on LB plates to determine cell counts. After cisplatin treatment, the cells were washed in
M9 minimal salts and then lysed using Bio-Rad cell lysis solution. The total cellular
platinum content was determined by atomic absorption spectroscopy using a Perkin Elmer
AAnalyst-300 spectrometer equipped with HGA-800 graphite furnace system and AS-72
autosampler.
For assessment of DNA platinum adduct levels, DNA isolation was carried out
according to manufacturer's instructions using Bio-Rad AquaPure Genomic DNA Isolation
Kit, and the DNA concentration was determined by UV absorbance at A260 using a
Beckmann DU-65 spectrophotometer. Platinum content was measured by atomic
absorption spectroscopy as specified above.
48
4.4. Results
We have observed that the presence of tetracycline in the growth medium of the
MG1655 carrying an F'TnlO plasmid affected strain survival upon subsequent treatment
with cisplatin (Fig. la). When MG1655/F'TnO10cells were grown without tetracycline,
their survival upon exposure to cisplatin was not different from that of MG1655 wild-type
cells not carrying the plasmid. However, when tetracycline was present in the growth
medium of overnight and log phase cultures of the MG1655/F'Tn0, subsequent treatment
with cisplatin caused a dramatic decrease in survival of these tetracycline-resistant
bacteria. In similar experiments with MG1655/F'TnlO::kan, the presence of kanamycin in
the growth medium of overnight and log phase cultures did not affect the survival upon
treatment with cisplatin, as compared to MG1655 (Fig. lb). These results indicated that
sensitization of bacteria to cisplatin was caused by the presence of tetracycline in the
growth medium, while neither the F' plasmid itself, nor the presence of kanamycin in the
growth medium had affected the sensitivity of E. coli to cisplatin.
Sensitization of bacteria to cisplatin by tetracycline was also observed when the
TnlO tet genes were present in the bacterial chromosome rather than on the F' plasmid
(Fig. 2). We tested three pairs of DNA repair mutants in three different genetic
backgrounds, and in each case we observed that the presence of tetracycline, but not other
antibiotics, in the growth medium sensitized them to subsequent cisplatin treatment.
Survival of CC104 mutY::cam, GM7330 fpg::amp and AB1157 mutS::kan mutants was
not affected by the presence or absence of chloramphenicol, ampicillin, and kanamycin,
respectively, in the growth medium. However, survival of tetracycline-resistant CC104
mutY::tet, GM7330 fpg::tet and AB1157 mutS::tet mutants upon treatment with cisplatin
dramatically decreased when these mutants were grown with tetracycline, while their
survival was not different from respective wild-types when tetracycline was absent from
the growth medium.
Sensitization of tetracycline-resistant bacteria to cisplatin by tetracycline pretreatment, however, appeared to be dependent on the type of tet gene. The pBR322 tet gene
did not produce similar effect of sensitization to cisplatin in AB1157, CC104 or MG1655
cells carrying pBR322 when they were grown with tetracycline prior to cisplatin treatment
(Fig. 3).
Treatment with tetracycline did not appear to have any effect on survival of
tetracycline-resistant cells when it was applied after cisplatin exposure (Fig. 4). In these
experiments, tetracycline-resistant cells were grown either with or without tetracycline,
treated with cisplatin, and then plated on either LB or LB/Tet plates. Survival of TnlO tet
gene-carrying cells was not different from the wild-type when they were grown without
tetracycline, and decreased when they were grown with tetracycline. However, in each
instance survival rates were not affected by post-treatment with tetracycline, i.e., survival
was identical on LB and on LB/Tet plates.
Residual quantities of tetracycline present at the time of cisplatin treatment had no
effect on survival, as demonstrated by experiments where tetracycline was either added to
wild-type cells, or washed out of tetracycline-resistant cells grown in tetracycline, at the
49
time of cisplatin treatment (Fig. 5c). Addition of tetracycline to wild-type cells during
cisplatin treatment did not affect their survival, and washing the residual tetracycline out of
tetracycline-resistant cells before cisplatin treatment did not improve their survival. These
observations suggest that residual tetracycline present at the time of cisplatin treatment
could not have been responsible for the increased toxicity in tetracycline-resistant cells
grown with tetracycline prior to cisplatin treatment.
Additional sensitization to cisplatin by tetracycline was observed in experiments
where tetracycline-resistant bacteria were grown with tetracycline and then co-treated with
cisplatin in the presence of tetracycline. We observed an additional 100-fold decrease in
survival in tetracycline-resistant cells that were not only pre-treated, but also co-treated
with tetracycline (Fig. 6). Interestingly, this effect was observed in MG1655 cells carrying
either F'TnlO or pBR322, unlike the pre-treatment sensitization, which was observed only
with the TnlO tet gene, but not the pBR322 tet gene.
Sensitization to cisplatin by co-treatment was not dependent on the presence of
tetracycline in the growth medium prior to cisplatin exposure, and was also observed with
both TnlO and pBR322 tet genes (Fig. 7). In these experiments, tetracycline-resistant cells
were grown without tetracycline, and then treated either with cisplatin alone, or with
cisplatin in the presence of tetracycline. We observed that when treated with cisplatin
alone, MG1655 cells carrying either F'TnlO or pBR322 tet genes were equally sensitive to
cisplatin as MG1655 cells not carrying any tet genes. However, when cells carrying either
plasmid were treated with cisplatin in the presence of tetracycline, their survival decreased.
One possible explanation of our results could be that the role of the tet geneencoded antiporter in observed toxicity was to increase the cellular concentration of
cisplatin. To test this possibility, we determined the total amount of cisplatin in MG1655
and tetracycline-induced MG1655/F'TnO10cells, at the highest cisplatin dose used in our
experiments (Fig. 8). The atomic absorption analysis was used to determine total levels of
platinum accumulated both inside the cells and on the cell surface. We found that platinum
levels were the same in MG1655 and in MG1665/F'TnO10cells that were grown in
tetracycline prior to cisplatin treatment (Fig. 8e). Co-treatment with tetracycline, however,
was accompanied by a small increase in total cellular levels of cisplatin in
MG1655/F'TnO10(p<0.0004), but not in MG1655 cells.
We next looked at the levels of cisplatin DNA damage by assessing the platinum
content in DNA isolated from MG1655 and MG1655/F'TnO10 cells induced with
tetracycline (Fig. 8f). We found that at the highest dose used in our experiments,
MG1655/F'TnO10cells grown with tetracycline accumulated more DNA damage than
MG1655 cells (p<0.01), and that co-treatment with tetracycline was associated with a
small further increase in levels of DNA damage in MG1655/F'TnO10cells (p<0.005), but
not in MG1655 cells.
Prolonged (4 hours) treatment with 10-fold lower doses of cisplatin produced results
that paralleled those obtained by -hour treatment with higher doses (Fig. 9). Tetracyclineresistant cells that were grown (pre-treated) with tetracycline and then co-treated with
cisplatin and tetracycline were the most sensitive to cisplatin treatment. Tetracycline50
resistant cells that were either pre-treated or co-treated with tetracycline were also more
sensitive to cisplatin than wild-type, with pre-treatment having a more pronounced effect
of sensitization than co-treatment. Finally, cisplatin treatment of tetracycline-resistant cells
that were neither pre-treated nor co-treated with tetracycline resulted in survival that was
not different from survival of wild-type cells.
51
4.5. Discussion
We describe here our finding that the combination of two unrelated drugs, the
antibiotic tetracycline and the anti-neoplastic drug cisplatin, results in high toxicity to
tetracycline-resistant bacteria. Tetracycline is a widely used antibiotic, but emergence of
tetracycline-resistant bacterial strains refractory to tetracycline treatment has become a
major clinical problem, and finding efficient ways to overcome tetracycline resistance is an
important goal in modem clinical research. Cisplatin is a successful chemotherapeutic
agent; however, the precise mechanism of its action remains elusive, and advances in
understanding cisplatin toxicity may hold promise for design of better chemotherapeutic
regimens (Trimmer and Essigmann, 1999).
In the course of our cisplatin research involving E. coli strains, we found that E.
coli cells carrying the Tn10 tetracycline resistance gene, when grown with tetracycline,
acquired high sensitivity to subsequent treatment with cisplatin. This phenomenon was
observed in several genetic backgrounds (Fig. 1 and 2), and it was displayed regardless of
whether the TnlO tet gene was present on the F' plasmid (Fig. 1), or within the host
chromosome (Fig. 2). The TnlO-carrying cells that were grown without tetracycline did
not differ from the respective wild-type E. coli cells in their sensitivity to cisplatin.
Sensitivity to cisplatin was not conferred to E. coli strains resistant to other antibiotics
(kanamycin, ampicillin and chloramphenicol) by growth of these strains in their respective
antibiotics (Fig. 1 and 2), suggesting that a unique relationship exists among cisplatin,
tetracycline, and TnlO tet gene expression, that results in high lethality for tetracyclineresistant cells.
Interestingly, sensitization to cisplatin by growth in tetracycline was not conferred
to wild-type E. coli cells carrying the pBR322 tet gene (Fig. 3), indicating that sensitization
to cisplatin was also dependent on the type of tet gene expressed. There are several major
differences between these two tetracycline resistance determinants and their expression
patterns. Although both tetA genes encode tetracycline efflux pumps with similar
membrane topology (Eckert and Beck, 1989; Allard and Bertrand, 1992; Kimura et al.,
1997), the two pumps are phenotypically distinct and only share 45% sequence homology
(Nguyen et al., 1983; Rubin and Levy, 1990), which puts respective tet genes in two
different classes of tetracycline resistance determinants: TnlO belongs to class B, and
pBR322 belongs to class C. These two classes of tetracycline resistance determinants share
similar genetic organization, and transcriptional regulation of both classes involves the
respective TelR repressor proteins (Klock et al., 1985; Hung and Pictet, 1989). However,
in the process of construction of pBR322 from its ancestral plasmid pSC101, which carried
the entire class C tetracycline resistance determinant, the tetR portion of the gene was lost
through genetic rearrangements (Bolivar et al., 1977; Brow et al., 1985). This resulted in
constitutive expression of the class C tet gene from the pBR322 plasmid, as opposed to
tetracycline-inducible expression of the class B TnlO tet gene from the F' plasmid. It
would be interesting to see whether the complete class C tetracycline resistance
determinant would confer sensitivity to cisplatin by growth in tetracycline, similarly to the
complete class B determinant. While it is possible that the variation in response to cisplatin
52
lies in the functional differences between the two distinct pumps, it is also possible that
expression of the TetR repressor protein may be intimately involved in cellular response to
cisplatin, which is known to interact with many DNA-binding proteins (Zlatanova et al.,
1998; Kartalou and Essigmann, 2001b). It is also possible that the TetA and/or TetR
proteins may become involved in other biochemical pathways in the cell, and the
difference between the two determinants reflects the difference in their alternative
biochemical functions (Eckert and Beck, 1989; Griffith et al., 1994; Valenzuela et al.,
1996b).
In order to understand the sequence of events leading to increased lethality of
tetracycline-resistant bacteria, we checked whether reversing the drug exposure pattern
would change the bacterial response to the combined drug treatment. We observed that
while growth of Tn10-carrying bacteria in tetracycline increased their susceptibility to
subsequent treatment with cisplatin, tetracycline exposure that followed cisplatin treatment
did not have any effect on survival. Regardless of whether the tetracycline-resistant cells
were plated on LB or on LB/Tet plates subsequent to cisplatin treatment, they displayed
the same survival rates (Fig. 4), and the only factor that actually changed their survival was
whether or not they were induced by tetracycline prior to cisplatin treatment. This implied
that the sequence of exposure to the two drugs was critical for the bacterial survival
outcome, and that sensitization of TnlO-carrying bacteria to cisplatin by tetracycline could
only take place before or during, but not after exposure to cisplatin. One interpretation of
these results could be that sensitization by growth in tetracycline involved enhanced uptake
of cisplatin by TnlO-expressing bacteria - however, this assumption was not supported by
the atomic absorption experiments. Another interpretation could be that once cisplatin has
entered cells and/or formed adducts with its cellular targets, it has lost its potential for
toxicity specific to tetracycline resistant cells, i.e., TnlO expression can only enhance the
toxicity of unbound cisplatin.
We explored several scenarios that could explain the sensitivity of TnlO tet geneexpressing cells to cisplatin. One explanation for the high sensitivity of tetracycline-grown
bacteria to cisplatin would be that free cisplatin could interfere with tetracycline efflux.
Cisplatin could block the antiporter directly (Fig. 5a). Alternatively, cisplatin could form a
complex with tetracycline and interfere with tetracycline-Mg2+ chelation in a manner that
would not prevent ribosome binding and induction of TetR by [tetracycline-Mg]+ , but
would nevertheless preclude [tetracycline-Mg]+ recognition and export by TetA (Fig. 5b).
The TnlO tet gene-expressing bacterial cells, normally tetracycline-resistant by virtue of
their ability to export tetracycline, would therefore become tetracycline-sensitive due to
unalleviated tetracycline accumulation. If this mechanism were operative, i.e., if cisplatin
treatment would lead to accumulation of tetracycline in otherwise tetracycline-resistant
cells, then wild-type cells, which are unable to export tetracycline, would be similarly
sensitive to tetracycline if it were present at the time of cisplatin treatment. Also, if this
assumption were true, then washing the tetracycline out prior to cisplatin treatment would
improve survival of tetracycline-resistant bacteria by eliminating the possibility of
tetracycline accumulation due to cisplatin-blocked export.
53
Since bacteria were resuspended in M9 minimal salts at the time of cisplatin
treatment in the above experiments, even those cells that were grown in tetracycline prior
to cisplatin treatment would only have had residual doses of tetracycline remaining in the
cisplatin treatment medium. To look into the possibility that residual doses of tetracycline
present at the time of cisplatin treatment could have had an effect on survival rates, we
checked whether survival of wild-type cells would be affected by the residual amounts of
tetracycline, and whether washing the residual tetracycline out prior to cisplatin treatment
would improve survival of TnlO-carrying cells grown in tetracycline. We found that
survival of wild-type cells was not affected by addition of the residual quantities of
tetracycline to the cisplatin treatment medium, and that survival of TnlO-carrying cells was
not improved by washing tetracycline out prior to cisplatin treatment (Fig. 5c). Therefore,
the presence of residual doses of tetracycline at the time of cisplatin treatment was ruled
out as a potential source of increased cisplatin toxicity, and consequently, cisplatin's role as
a blocker of residual tetracycline export during cisplatin treatment was also ruled out.
Unexpectedly, introducing standard concentrations of tetracycline (15 [tg/ml) to TnlO
tet carrying cells at the time of cisplatin treatment resulted in further decrease in their
survival rates. Cells that were both grown in tetracycline, and subsequently co-treated with
tetracycline and cisplatin, displayed an additional 100-fold decrease in survival (Fig. 6).
Interestingly, this phenomenon was common to both the TnlO and pBR322 tet genes. We
next looked into survival of cells that were not grown with tetracycline, but were subjected
to cisplatin and tetracycline co-treatment, and found that co-treatment alone affected
survival regardless of prior growth with tetracycline (Fig. 7). Again, this phenomenon was
observed for both TnlO and pBR322 tet genes, unlike the sensitization conferred by
growth in tetracycline, which was specific to the Tnl 0 tet gene only. It is important to note
that wild-type cells were unaffected by co-treatment with tetracycline (which is a
bacteriostatic rather than a bactericidal agent), indicating that involvement of tet genes was
essential in co-treatment sensitization, and confirming that tetracycline accumulation
during cisplatin treatment was not responsible for the observed toxicity.
Another scenario that could explain the high sensitivity of TnlO-carrying cells to
cisplatin upon growth in tetracycline would be that expression of the TnlO antiporter
pump, induced by tetracycline, could result in increased cellular accumulation of cisplatin
(Gately and Howell, 1993). Increased cisplatin levels have previously been reported to be
associated with increased cisplatin toxicity (Leibbrandt and Wolfgang, 1995; Welters et
al., 1997), and resistance to cisplatin is often known to be accompanied by reduced
cisplatin accumulation (Kartalou and Essigmann, 2001a; Chu, 1994; Johnson et al., 1996;
Rixe et al., 1996; Loh et al., 1992). Since the initially activated cisplatin species
[Pt(NH3)2Cl(H2 0)] + carry single positive charge due to exchange of a single chloride
ligand for a single H2 0 molecule, it is possible that the tetracycline antiporter could be
capable of exchanging one [tetracycline-Mg]+ complex for one mono-aquated cisplatin
molecule (Fig. 8a). Besides this possibility of direct antiport, a compelling possibility
exists that expression of the trans-membrane pump introduces enough disruption into the
membrane structure to increase its permeability to cisplatin molecules (Fig. 8b). It has been
demonstrated that the TnlO efflux pump is a water filled-channel with a thin permeability
barrier (Kimura-Someya et al., 1998; Iwaki et al., 2000; Tamura et al., 2001); therefore,
54
cisplatin molecules could gain easier access into bacterial cells that express such pumps.
Increased cisplatin concentration in the cells could then lead to decreased bacterial survival
through an increase in levels of lethal DNA damage, or enhancement of other toxic
cisplatin effects. In line with this hypothesis are reports of increased sensitivity of E. coli
cells to mM concentrations of nickel and cadmium salts (Griffith et al., 1982; Podolsky et
al., 1996; Valenzuela et al., 1996a; Stavropoulos and Strathdee, 2000), where sensitivity
was conferred to E. coli by expression of the TetA class C gene and paralleled, in the case
of nickel salts, by an increased accumulation of nickel ions by tetracycline-resistant cells
(Podolsky et al., 1996).
With these models in mind, we determined the total levels of platinum in cells in
order to establish whether the increase in cellular toxicity was paralleled by an increase in
cellular cisplatin content. We found that at the highest cisplatin dose used in our
experiments, TnlO-carrying cells grown in tetracycline did not have increased platinum
levels as compared to the wild type cells (Fig. 8e). Increased cellular levels of cisplatin
were therefore ruled out as a likely source of high cisplatin toxicity in tetracycline-induced
cells harboring TnlO tet genes. TnlO-carrying cells grown in tetracycline and subsequently
co-treated with cisplatin and tetracycline, however, had levels of cisplatin slightly higher
than those of wild-type (p<0.01) and those of TnlO-carrying cells treated with cisplatin
alone (p<0.0004). It is possible that proton flow during active export of tetracycline could
lead to reduced proton concentration outside the cellular membrane, and therefore to fewer
cisplatin molecules acquiring positive charge prior to their entry into the cells. A higher
proportion of the uncharged cisplatin molecules could lead to higher levels of passive
diffusion of cisplatin through the cellular membrane during active export of tetracycline
(Fig. 8c).
We further reasoned that while the total cellular levels of cisplatin were unaffected
by growth in tetracycline, cisplatin reactivity could nevertheless be affected by antiportermediated modification in the intracellular environment (Fig. 8d), such as a transient pH
change. Local decrease in intracellular pH would favor formation of highly reactive
aquated platinum molecules (Rosenberg, 1985; Lippard, 1982; Pinto and Lippard, 1985),
which could in turn lead to increased levels of damage to the DNA (Murakami et al., 2001;
Chau and Stewart, 1999). To investigate this possibility, we measured the platinum content
in bacterial DNA isolated after treatment at the highest cisplatin dose used in our
experiments (Fig. 8f). We determined that the platinum content in DNA of
MG1655/F'TnO cells grown in tetracycline was higher than that of MG1655 cells
(p<0.01), even though the total cellular platinum content was similar in both cell types.
Subsequent co-treatment with tetracycline led to a small further increase in the levels of
DNA damage in MG1655/F'Tn10 cells (p<0.005), and this increase correlated with a small
increase in total cellular platinum levels. Co-treatment with tetracycline, however, did not
increase platinum levels in the cellular DNA of MG1655 cells, indicating that involvement
of TnlO genes was required for increased accumulation of DNA damage.
As these results suggest, the high toxicity of cisplatin to the TnlO-expressing
bacteria is associated with increased levels of DNA damage, although the precise
mechanism of increased DNA damage formation and specific toxicity remains to be
55
determined. This finding is in line with the results of post-treatment experiments, which
suggested that expression of the tet gene can only affect cisplatin toxicity before or during
cisplatin entry into the cells, possibly, before the drug has an opportunity to form lesions
with cellular targets such as DNA. We are considering the possibility that the DNAbinding TetR protein facilitates delivery of cisplatin molecules to DNA and/or enhances
formation of DNA lesions. It is also possible that the interplay between cisplatin,
tetracycline and expression of the tet gene may lead not only to increased levels of DNA
damage, but also to formation of damage of a different nature (Khan et al., 2003; Buschfort
and Witte, 1994; Witte et al., 1994; Quinlan and Gutteridge, 1988).
While our results using 1-hour exposure to doses of cisplatin up to 200 CIM
provoked substantial toxicity, we wanted to confirm that longer exposures to lower,
therapeutically appropriate cisplatin doses would also result in high toxicity for
tetracycline-resistant bacteria. As expected, prolonged (4-hour) treatment of bacteria with
10-fold lower doses of cisplatin produced results that paralleled those obtained for shorter
treatments with higher doses (Fig. 9). This result is an important indication that destruction
of tetracycline-resistant bacteria is possible by therapeutically relevant levels of cisplatin.
The low doses employed for 4-hour treatment are consistent with platinum levels found in
the blood of patients undergoing cisplatin chemotherapy (Sileni et al., 1992; Panteix et al.,
2002; Nagai et al., 1996).
Since tetracycline resistance determinants, particularly versions of the transposon
TnlO, are widely used in bacterial genetics, we would like to emphasize the importance of
being aware of potential artifacts in many experimental systems associated with
tetracycline use. Expression of tetracycline efflux pumps may have various pleiotropic
effects in bacterial hosts (Eckert and Beck, 1989; Griffith et al., 1994; Valenzuela et al.,
1996b; Bochner et al., 1980). These effects must be carefully segregated in each instance
from the effects ascribed to the genetic rearrangements that involve tetracycline resistance
determinants.
The interaction between cisplatin, tetracycline, and TnlO-associated tetracycline
resistance in E. coli suggests potential therapeutic applications of this finding. We have
observed this relationship in various genetic backgrounds of E. coli, and regardless of
whether the TnlO tet gene was plasmid-borne or integrated into the bacterial chromosome.
The pBR322 class C tetracycline resistance determinant also conferred a certain degree of
sensitization to cisplatin by tetracycline, although under more narrowly defined conditions,
and differences between these two determinants will help define the mechanism behind
tetracycline-induced susceptibility to cisplatin. In order to apply this knowledge towards
the efforts to reverse tetracycline resistance in bacterial pathogens and to restore the
pharmacological potential of tetracycline, it will be important to examine other bacterial
species with various tetracycline resistance determinants, as well as other tetracyclines and
other platinum compounds. Furthermore, unraveling the mechanism of cisplatin toxicity
associated with tet gene expression may potentially help define a novel pathway for
cisplatin action. It is possible that the counterpart of such a pathway will later be
discovered in mammalian cells, which may lay the groundwork for the design of novel
anti-cancer agents.
56
4.6. Acknowledgements
This work was supported by grants CA86061 (J.M.E.) and GM63790 (M.G.M) from
the National Institutes of Health.
We would like to thank B. Engelward, K. Murphy, E.B. Konrad, M. Volkert, and J.
Miller for providing the bacterial strains used in this work. We would also like to thank
S.J. Lippard and K.R. Barnes for access to and assistance with the atomic absorption
facility, and J.C. Delaney and L.E. Frick for proofreading this manuscript.
57
4.7. Figures
Table 1. The genotypes of the E. coli strains.
Strain
AB1157
Source/Reference
Genotype
thr-1 araC14leuB6(Am) A(gpt-proA)62 lacYl
Lab stock
tsx-33 supE44(AS) galK2(0c) hisG4(0c) rfbD1
mgl-51 rpoS396(Am) rpsL31(StrR ) kdgK51 xylA5
mtl-I argE3(0c) thi-1
GM4799
As AB1 157 but mutS458::mTn 10
OKan
Ref. 63
GM7698
As AB1157 but AmutS465::Tet
K. Murphy
GM7330
A(lacY-lacZ)286 ((p80dllA lacZ9) ara thi (?)
E.B. Konrad
GM7619
As GM7330 but fpg::Tn10
M. Volkert
GM8085
As GM7330 but Afpg::Amp
B. Engelward
CC104
F'ac? lac
Ref. 11
CC104mutY
As CC104 but mutY::mTn10
J. Miller
MV4706
As CC104 but AmutY::Cam
M. Volkert
MG1655
rph-1
Ref. 19
XL1-Blue
recA endA 1 gyrA96 (NalR)thi-1 hsdR17 supE44 relAl lac
Stratagene
proAB+larathi A(gpt-lac)5
[F' proAB laclqZ AMI5 Tn10 (TetR)]
GM4292
As MG1655 but F'::Tn10 laclq lacZAM15 proAB+
This work
GM4293
As MG1655 but F'42 lacZ::Kan
This work
AB 1874
F'42 (F'lac+)/lac- 19
Lab stock
KM81
As AB1157 but lacZ::Kan
K. Murphy
.
58
Figure 1. Survival of MG 1655/F'TnJ 0, grown in the presence and in the absence of
antibiotics in the growth medium.
la. Survival of MG1655/ F'TnlO::fef
tetracycline in the growth medium.
, grown in the presence and in the absence of
I b. Survival of MG 1655/F'TnJ 0:: /can, grown in the presence of kanamycin
medium.
in the growth
la.
100
10 C;
1
>
.~
~
rn
~
0
0.1
-+-
MG1655
___
MG1655/FTn10/NoTc
-.-
MG1655/FTn10/Tc
0.01
0.001
0
50
100
150
200
J.1MCOOP
lb.
100
C;
>
.~
~
-+-MG1655
10
___
In
~
0
1
0
50
100
150
J.1MCOOP
59
200
MG1655/ F::kanlkan
Figure 2. Survival of various E. coli mutants carrying the chromosome-encoded TnlO tet
gene, grown in the presence and in the absence of tetracycline in the growth medium,
compared to survival of non-tetracycline resistant counterparts.
2a. Survival of CC104, CC104 mutY::cam, grown with and without chloramphenicol, and
CC 104 mutY::tet, grown with and without tetracycline.
2b. Survival of GM7330, GM7330 fpg::amp, grown with and without ampicillin, and
GM7330fg::tet, grown with and without tetracycline.
2c. Survival of AB1157, ABI157 mutS::kan, grown with and without kanamycin, and
AB 1157mutS::tet, grown with and without tetracycline.
60
2a.
100
-+-CC104
10
-a-mutY::eam/No
ftj
>
1
~
0.1
.~
'"
::.e
0
eam
---k- m utY::eam leam
0.01
~
m utY::tetiNo Tc
-?IE-
m utY::tetITc
0.001
0
50
100
200
150
J.lMCOOP
2b.
100
-+-GM7330
10
Ii
>
1
:;,
0.1
.~
(#)
__
---k- fpg ::am p/am p
0.01
~
fpg::am plNo am p
~
fpg::tetiNo Tc
-?IE-
fpg::tetITc
0
0.001
0.0001
0
50
100
150
200
IlMCDDP
2e.
100
-+-AB1157
ftj
10
__
>
mutS::kanlNo kan
~
~
---k- mutS::kanlkan
::.e 0.1
~
mutS::tetJNo Tc
-?IE-
mutS::tetITc
'"
0
0.01
0.001
0
50
100
150
200
J.lMcOOP
61
Figure 3. Survival of E. coli strains carrying pBR322, grown with tetracycline in the
growth medium.
3a. Survival of AB 157/pBR322,grown with tetracycline;
3b. Survival of CC104/pBR322, grown with tetracycline;
3c. Survival of MG1655/pBR322, grown with tetracycline.
62
63
Figure 4. Post-treatment with tetracycline does not affect survival of Tn] 0 carrying cells.
4a. Survival of MG 1655/ F'Tn] 0, grown with and without tetracycline and plated on LB
and LB/Tet plates.
4b. Survival of CCI04 mutY::tet, grown with and without tetracycline and plated on LB
and LB/Tet plates.
4a.
100
-+-MG1655
10
Ci
>
.~
::::J
__
----.- MG1655::Tn10, No TelTet
1
(I)
~
0
MG1655::Tn10, No TelLS
0.1
-0(-
MG1655::Tn10, Te/LB
~
MG1655::Tn10, TelTet
0.01
0
50
100
200
150
J.1MCOOP
4b.
100
Ci
>
.~
::::J
(I)
~
-+-CC104
10
__
1
CC104::Tn10, No TelLB
----.- CC104::Tn10, No TelTet
0.1
-0(-
CC104::Tn10, TelLB
~
CC104::Tn10, TelTet
0
0.01
0.001
0
50
100
150
200
J.1MCOOP
64
Figure 5. Could increased accumulation of tetracycline explain cisplatin toxicity?
5a. Cisplatin could block TetA, preventing efflux of [Tc-Mgt.
5b. Cisplatin interference with tetracycline-magnesium
and export of the complex by TetA.
chelation could prevent recognition
5e. Residual quantities of tetracycline present at the time of cisplatin treatment do not
affect survival of MG 1655 and MG 1655/ F'TnlO carrying cells, and therefore are not
responsible for toxicity.
5a.
5b.
[Tc-Mg-Pt]
[Tc-M9t
5e.
100
10
Cii
>
-+-MG1655
.~
1
:J
UJ
o
~ 0.1
0.01
o
50
100
150
200
J.1MCOOP
65
_
MG1655, residual
Tc
-.-
MG1655/F"Tn10
~
MG1655/F"Tn10, after
2 washes
Figure 6. Survival of tetracycline-resistant cells, grown with tetracycline and either treated
with cisplatin alone, or co-treated with cisplatin in the presence of tetracycline.
6a. Survival of MG 1655/F'TnJ 0, grown in tetracycline and treated either with cisplatin
alone, or with cisplatin in the presence of tetracycline.
6b. Survival of MG 1655/pBR322, grown in tetracycline and treated either with cisplatin
alone, or with cisplatin in the presence of tetracycline.
6a.
100
Ci
>
10
:J
1
.~
-+-MG1655
(D
___ MG16551F'Tn10, Tc/Pt
'#. 0.1
--.-
0.01
MG16551F'Tn10, Tc/Pt-Tc
0.001
0.0001
0
50
100
200
150
IlM COOP
6b.
100
-+-MG1655
10
___ MG1655/pBR322, TclPt
Ci
>
.~
:J
--.-
1
en
::::le
0
0.1
0
50
100
200
150
11M COOP
66
MG1655/pBR322, TclPt-Tc
Figure 7. Survival of tetracycline-resistant cells, grown without tetracycline and either
treated with cisplatin alone, or co-treated with cisplatin in the presence of tetracycline.
7a. Survival of MG 1655/F'Tnl 0, grown without tetracycline
cisplatin alone, or with cisplatin in the presence of tetracycline.
and treated either with
7b. Survival of MG 1655/pBR322, grown without tetracycline and treated either with
cisplatin alone, or with cisplatin in the presence of tetracycline.
7a.
100
-+- MG1655, pt
10
___
MG1655, PtlTc
-.-
MG1655/FTn10,
pt
~
MG1655/FTn10,
PtfTc
fti
>
.~
1
:J
In
~
0
0.1
0
50
100
200
150
J.1MCDOP
7b.
100
-+- MG1655,
_ 10
as
>
.~
:;:,
1
en
~0.1
0.01
0
50
100
150
200
J.1M COOP
67
pt
___
MG1655, PtlTc
-.-
MG1655/pBR322,
pt
~
MG16551pBR322,
PtfTc
Figure 8. Total cellular and DNA platinum levels in MG1655 and MG1655/F'TnlO.
8a. Mono-aquated cisplatin molecule [Pt(NH3)2Cl(H20)] + could be exchanged by the
antiporter for the [Tc-Mg]+ complex.
8b. Expression of trans-membrane aqueous channels could increase permeability of the
cytoplasmic membrane.
8c. More uncharged cisplatin species could accumulate outside the cell due to reduced
local proton concentration, leading to enhanced passive diffusion across the membrane.
8d. Cisplatin reactivity could be enhanced inside the cell due to increased local proton
concentration.
8e. Total cellular platinum levels, as determined by atomic absorption spectroscopy, in
MG1655 and in MG1655/ F'TnlO, grown with tetracycline and treated either with cisplatin
alone, or with cisplatin in the presence of tetracycline.
8f. DNA platinum levels, as determined by atomic absorption spectroscopy, in MG1655
and in MG1655/ F'TnlO, grown with tetracycline and treated either with cisplatin alone, or
with cisplatin in the presence of tetracycline.
68
8a.
8b.
Pt+
Pt+
[Tc-MgJ+
8e.
+H'il-
Pt
Pt
Pt
H
'
Pt(NH.JzC1(OH)
-H+
Pt(NH.JzClz"
~ IPt(NH.JzCI(HzOW4
~
Pt(NH.JzC1(OH)
+H+
69
8e.
2
1.8
1.6
.! 1.4
Gi
u
1.2
'"Q
~
...
Q.
"0
1
0.8
E 0.6
Q.
0.4
0.2
0
wtPt
wt Pt-Te
Tn10 Pt
Tn10 Pt-Te
8f.
1600
1400
«
z
c
Cl
...
.2Q.
Cl
Q.
1200
1000
800
600
400
200
0
wtPt
wt Pt-Te
Tn10 Pt
Tn10 Pt-Te
70
Figure 9. Survival ofMGI655 and MG1655/ F'TnlO upon prolonged (4 hours) exposure
to IO-fold lower doses of cisplatin.
100
-+- MG1655, Pt alone
10
'is
>
.~
~
en
~
0
1
___
MG1655, Pt- Te
--.tr-
FTn1 0, No TelPt alone
---*- FTn10, No TelPt-Te
-?IE- FTn1 0, TelPt alone
0.1
-.-
0.01
0
5
10
20
15
~M COOP
71
FTn10, Tel Pt-Tc
References:
Allard, J.D., and K.P. Bertrand. 1992. Membrane topology of the pBR322 tetracycline
resistance protein. J. Biol. Chem. 267(25):17809-17819.
Bertrand, K.P., K. Postle, L.V. Wray, and W.S. Reznikoff. 1983. Overlapping
divergent promoters control expression of TnlO tetracycline resistance. Gene. 23:149-156.
Bochner, B.R., H.-C. Huang, G.L. Schieven, and B.N. Ames. 1980. Positive selection
for loss of tetracycline resistance. J. Bacteriol. 143(2):926-933.
Bolivar, F., R.L. Rodriguez, P.J. Greene, M.C. Betlach, H.L. Heyneker, and H.W.
Boyer. 1977. Construction and characterization of new cloning vehicles. II. A
multipurpose cloning system. Gene. 2(2):95-113.
Brow, M.A.D., R. Pesin, and J.G. Sutcliffe. 1985. The tetracycline repressor of pSC101.
Mol. Biol. Evol. 2(1):1-12.
Buschfort, C., and I. Witte. 1994. Induction and mechanism of DNA single- and doublestrand breaks by tetracycline/Cu(II) in the absence of light. Carcinogenesis. 15(12):29272930.
Butaye, P., A. Cloeckaert, and S. Schwarz. 2003. Mobile genes coding for effluxmediated antimicrobial resistance in Gram-positive and Gram-negative bacteria. Int. J.
Antimicrob. Agents. 22:205-210 (2003).
Chau, Q. and D.J. Stewart. 1999. Cisplatin efflux, binding and intracellular pH in the
HTB56 human lung adenocarcinoma cell line and the E-8/0.7 cisplatin-resistant variant.
Cancer Chemother. Pharmacol. 44:193-202.
Chopra, I., and M. Roberts. 2001. Tetracycline antibiotics: mode of action, applications,
molecular biology, and epidemiology of bacterial resistance. Microbiol. Mol. Biol. Rev.
65(2):232-260.
Chopra, I. 2002. New developments in tetracycline antibiotics: glycylcyclines and
tetracycline efflux pump inhibitors. Drug Resist. Updat. 5:119-125.
Chu, G. 1994. Cellular responses to cisplatin. The roles of DNA-binding proteins and
DNA repair. J. Biol. Chem. 269(2):787-790.
Cupples, C.G., and J.H. Miller. 1989. A set of lacZ mutations in Escherichia coli that
allow rapid detection of each of the six base substitutions. Proc. Natl. Acad. Sci. U S A.
86:5345-9.
72
Eckert, B., and C.F. Beck. 1989. Topology of the transposon TnlO-encoded tetracycline
resistance protein within the inner membrane of Escherichia coli. J. Biol. Chem.
264(20):11663-11670.
Eckert, B., and C.F. Beck. 1989. Overproduction of transposon TnlO-encoded
tetracycline resistance protein results in cell death and loss of membrane potential. J.
Bacteriol. 171(6):3557-3559.
Eisenbraun, M.D., and J.K. Griffith. 1993. Effects of plasmid pBR322 on respiratory
and ATPase activities in Escherichia coli. Plasmid. 30:159-162.
Fuertes, M.A., J. Castilla, C. Alonso, and J.M. Perez. 2003. Cisplatin biochemical
mechanism of action: from cytotoxicity to induction of cell death through interconnections
between apoptotic and necrotic pathways. Curr. Med. Chem. 10(3):257-266.
Gately, D.P., and S.B. Howell. 1993. Cellular accumulation of the anticancer agent
cisplatin: a review. Br. J. Cancer. 67(6):1171-6.
Griffith, J.K., J.M. Buckingham, J.L. Hanners, C.E. Hildebrand, and R.A. Walters.
1982. Plasmid-conferred tetracycline resistance confers collateral cadmium sensitivity of
E.coli cells. Plasmid. 8(1):86-8.
Griffith, J.K., D.H. Cuellar, C.A. Fordyce, K.G. Hutchings, and A.A. Mondragon.
1994. Structure and function of the class C tetracycline/H+ antiporter: three independent
groups of phenotypes are conferred by TetA(C). Mol. Membr. Biol. 11:271-277.
Guyer, M.S., R.E. Reed, T. Steitz, and K.B. Low. 1981. Identification of a sex-factor-
affinity site in E. coli as gamma delta. Cold Spring Harb. Symp. Quant. Biol. 45:135-140.
Hillen, W., and C. Berens. 1994. Mechanisms underlying expression of TnlO encoded
tetracycline resistance. Annu. Rev. Microbiol. 48:345-69.
Hillen, W., G. Klock, I. Kaffenberger,
L.V. Wray, and W.S. Reznikoff.
1982.
Purification of the TET repressor and TET operator from the transposon TnlO and
characterization of their interaction. J. Biol. Chem. 257(11):6605-6613.
Hung, A., and R. Pictet. 1989. Similarities between the regulatory sequences of the
unrelated tetracycline genes of pBR322 and TnlO. FEBS Lett. 245(1-2):57-60.
Iwaki, S., N. Tamura, T. Kimura-Someya,
S. Nada, and A. Yamaguchi.
2000.
Cysteine-scanning mutagenesis of transmembrane segments 4 and 5 of the Tn10-encoded
metal-tetracycline/H+ antiporter reveals a permeability barrier in the middle of a
transmembrane water-filled channel. J. Biol. Chem. 275(30):22704-22712.
Johnson, S.W., D.W. Shen, I. Pastan, M.M. Gottesman, and T.C. Hamilton. 1996.
Cross-resistance, cisplatin accumulation, and platinum-DNA adduct formation and
73
removal in cisplatin-sensitive and -resistant human hepatoma cell lines. Exp. Cell Res.
226:133-139.
Kartalou, M., and J.M. Essigmann 2001. Mechanisms of resistance to cisplatin. Mutat.
Res. 478 (1-2):23-43.
Kartalou, M., and J.M. Essigmann. 2001. Recognition of cisplatin adducts by cellular
proteins. Mutat. Res. 478 (1-2):1-21.
Katano, K., A. Kondo, R. Safaei, A. Holzer, G. Samimi, M. Mishima, Y.M. Kuo, M.
Rochdi, and S.B. Howell. 2002. Acquisition of resistance to cisplatin is accompanied by
changes in the cellular pharmacology of copper. Cancer Res. 62:6559-6565.
Khan, M.A., J. Mustafa, and J. Musarrat. 2003. Mechanism of DNA strand breakage
induced by photosensitized tetracycline-Cu(II) complex. Mutat. Res. 525:109-119.
Kimura-Someya, T., S. Iwaki, and A. Yamaguchi. 1998. Site-directed chemical
modification of cysteine-scanning mutants as to transmembrane segment II and its flanking
regions of the TnlO-encoded metal-tetracycline/H+ antiporter reveals a transmembrane
water-filled channel. J. Biol. Chem. 273(49):32806-32811.
Kisker, C., W. Hinrichs, K. Tovar, W. Hillen, and W. Saenger. 1995. The complex
formed between Tet repressor and tetracycline-Mg2+ reveals mechanism of antibiotic
resistance. J. Mol. Biol. 247:260-280.
Klock, G., B. Unger, C. Gatz, W. Hillen, J. Altenbuchner, K. Schmid, and R. Schmitt.
1985. Heterologous repressor-operator recognition among four classes of tetracycline
resistance determinants. J. Bacteriol. 161(1):326-332.
Leibbrandt, M.E.I., and G.H.I. Wolfgang. 1995. Differential toxicity of cisplatin,
carboplatin, and CI-973 correlates with cellular platinum levels in rat renal cortical slices.
Toxicol. Appl. Pharmacol. 132:245-252.
Levy, S.B., and M. Nelson. 1998. Reversing tetracycline resistance. A renaissance for the
tetracycline family of antibiotics. Adv. Exp. Med. Biol. 456:17-25.
Lippard, S.J. 1982. New chemistry of an old molecule: cis-[Pt(NH3 )2 C12 ]. Science.
218(4577):1075-1082.
Loh, S.Y., P. Mistry, L.R. Kelland, G. Abel, and KIR. Harrap. 1992. Reduced drug
accumulation as a major mechanism of acquired resistance to cisplatin in a human ovarian
carcinoma cell line: circumvention studies using novel platinum (II) and (IV)
ammine/amine complexes. Br. J. Cancer. 66(6): 1109-15.
Lomovskaya, O., and W.J. Watkins. 2001. Efflux pumps: their role in antibacterial drug
discovery. Curr. Med. Chem. 8:1699-1711.
74
Mistry, P., L.R. Kelland, S.Y. Loh, G. Abel, B.A. Murrer, and K.R. Harrap. 1992.
Comparison of cellular accumulation and cytotoxicity of cisplatin with that of tetraplatin
and amminedibutyratodichloro(cyclohexylamine)platinum(IV) (JM221) in human ovarian
carcinoma cell lines. Cancer Res. 52(22):6188-93.
Murakami, T., I. Shibuya, T. Ise, Z.S. Chen, S. Akiyama, M. Nakagawa, H. Izumi, T.
Nakamura, K. Matsuo, Y. Yamada, and K. Kohno. 2001. Elevated expression of
vacuolar proton pump genes and cellular pH in cisplatin resistance. Int. J. Cancer. 93:869874.
Nagai, N., M. Kinoshita, H. Ogata, D. Tsujino, Y. Wada, K. Someya, T. Ohno, K.
Masuhara, Y. Tanaka, K. Kato, H. Nagai, A. Yokoyama, and Y. Kurita. 1996.
Relationship between pharmacokinetics of unchanged cisplatin and nephrotoxicity after
intravenous infusions of cisplatin to cancer patients. Cancer Chemother. Pharmacol. 39(12):131-137.
Nguyen, T.T., K. Postle, and K.P. Bertrand. 1983. Sequence homology between the
tetracycline-resistance determinants of TnlO and pBR322. Gene. 25:83-92 (1983).
Orth, P., F. Cordes, D. Schnappinger, W. Hillen, W. Saenger, and W. Hinrichs. 1998.
Conformational changes of the Tet repressor induced by tetracycline trapping. J. Mol. Biol.
279:439-447.
Orth, P., D. Schnappinger, W. Hillen, W. Saenger, and W. Hinrichs. 2000. Structural
basis of gene regulation by the tetracycline inducible Tet repressor-operator system. Nature
Struct. Biol. 7(3):215-9.
Panteix, G., A. Beaujard, F. Garbit, C. Chaduiron-Faye, M. Guillaumont, F. Gilly, P.
Baltassat, and F. Bressole. 2002. Population pharmacokinetics of cisplatin in patients
with advanced ovarian cancer during intraperitoneal hyperthermia chemotherapy.
Anticancer Res. 22(2B):1329-1336.
Pinto, A.L., and S.J. Lippard. 1985. Binding of the antitumor drug cisdiamminedichloroplatinum(II) (cisplatin) to DNA. Biochimica et Biophysica Acta.
780(3):167-180.
Podoilsky, T., S.T. Fong, and B.T.O. Lee. 1996. Direct selection of tetracycline-sensitive
Escherichia coli cells using nickel salts. Plasmid. 36:112-115.
Reed,
E., R.F.
Ozols, R. Tarone,
S.H. Yuspa,
and M.C. Poirier.
1988. The
measurement of cisplatin-DNA adduct levels in testicular cancer patients. Carcinogenesis.
9(10):1909-11.
Rixe, O., W. Ortuzar, M. Alvarez, R. Parker, E. Reed, K. Paull, and T. Fojo. 1996.
Oxaliplatin, tetraplatin, cisplatin and carboplatin: spectrum of activity in drug-resistant cell
75
lines and in cell lines of the National Cancer Institute's Anticancer Drug Screen Panel.
Biochem. Pharmacol. 52:1855-1865.
Roberts, M. 1996. Tetracycline resistance determinants: mechanisms of action, regulation
of expression, genetic mobility, and distribution. FEMS Microbiol. Lett. 19(1):1-24.
Roberts, M. 2003. Tetracycline therapy: update. Clin. Infect. Dis. 36(4):462-467.
Rosenberg, B. 1985. Fundamental studies with cisplatin. Cancer. 55(10):2303-2316.
Rubin, R.A., and S.B. Levy. 1990. Interdomain hybrid Tet proteins confer tetracycline
resistance only when they are derived from closely related members of the tet gene family.
J. Bacteriol. 172(5):2303-2312.
Ryan, B.M., T.J. Dougherty, D. Beaulieu, J. Chuang, B.A. Dougherty, and J.F.
Barrett. 2001. Efflux in bacteria: what do we really know about it? Expert Opin. Investig.
Drugs. 10(8):1409-1422.
Quinlan, GJ., and J.M.C. Gutteridge. 1988. Hydroxyl radical generation by the
tetracycline antibiotics with free radical damage to DNA, lipids and carbohydrate in the
presence of iron and copper salts. Free Rad. Biol. Med. 5:341-348.
Saenger, W., P. Orth, C. Kisker, W. Hillen, and W. Hinrichs. 2000. The tetracycline
repressor - a paradigm for a biological switch. Angew. Chem. Int. Ed. 39:2042-2052.
Saier, M.H., I.T. Paulsen, M.K. Sliwinski, S.S. Pao, R.A. Skurray, and H. Nikaido.
1998. Evolutionary origins of multidrug and drug-specific efflux pumps in bacteria.
FASEB J. 12:265-274.
Schnappinger, D., and W. Hillen. 1996. Tetracyclines: antibiotic action, uptake, and
resistance mechanisms. Arch. Microbiol .165:359-369.
Sileni, V.C., V. Fosser, P. Maggian, E. Padula, M. Beltrame, M. Nicolini, and P.
Arslan. 1992. Pharmacokinetics and tumor concentration of intraarterial and intravenous
cisplatin in patients with head and neck squamous cancer. Cancer Chemother. Pharmacol.
30(3):221-225.
Stavropoulos, T.A., and C.A. Strathdee. 2000. Expression of the tetA(C) tetracycline
efflux pump in Escherichia coli confers osmotic sensitivity. FEMS Microbiol.
Lett. 190:147-150.
Sum, P.E., F.W. Sum, and SJ. Projan. 1998. Recent developments in tetracycline
antibiotics. Curr. Pharm. Des. 4:119-132.
76
Tamura, N., S. Konishi, S. Iwaki, T. Kimura-Someya, S. Nada, and A. Yamaguchi.
2001. Complete cysteine-scanning mutagenesis and site-directed chemical modification of
the TnlO-encoded metal-tetracycline/H antiporter. J. Biol. Chem. 276(23):20330-20339.
Thanassi, D.G., G.S.B. Suh, and H. Nikaido. 1995. Role of outer membrane barrier in
efflux-mediated tetracycline resistance of Escherichia coli. J. Bacteriol. 177(4):998-1007.
Trimmer, E. E., and J.M. Essigmann. 1999. Cisplatin. Essays Biochem. 34: 191-211.
Valenzuela, M.S., E.V. Ikpeazu, and K.A.I. Siddiqui. 1996a. E.coli growth inhibition by
a high copy number derivative of plasmid pBR322. Biochem. Biophys. Res. Commun.
219: 876-883.
Valenzuela, M.S., K.A.I. Siddiqui, and B.L. Sarkar. 1996b. High Expression of
plasmid-encoded tetracycline resistance gene in E.coli causes a decrease in membranebound ATPase activity. Plasmid. 36:19-25.
Welters, M.J.P., A.M.J. Fichitinger-Schepman, R.A. Baan, M.A.J.A. Hermsen,
WJ.F. van der Vijgh, J. Cloos, BJ.M. Braakhuis. 1997. Relationship between the
parameters cellular differentiation, doubling time and platinum accumulation and cisplatin
sensitivity in a panel of head and neck cancer cell lines. Int. J. Cancer. 71: 410-415.
Witte, I., G. Oetken, C. Buschfort, and A. Hartmann. 1994. A comparison of the DNAdamaging, the cytotoxic and genotoxic properties of tetracycline in human fibroblasts in
the presence and absence of light. Mutat. Res. 315:33-40.
Yamaguchi, A., T. Udagawa, and T. Sawai. 1990. Transport of divalent cations with
tetracycline as mediated by the transposon TnlO-encoded tetracycline resistance protein. J.
Biol. Chem. 265(9):4809-4813.
Zdraveski, Z.Z., J.A. Mello, M.G. Marinus, and J.M. Essigmann. 2000. Multiple
pathways of recombination define cellular responses to cisplatin. Chem. Biol. 7(1):39-50.
Zlatanova, J., J. Yaneva, S.H. Leuba. 1998. Proteins that specifically recognize
cisplatin-damaged DNA: a clue to anticancer activity of cisplatin. FASEB J. 12:791-799.
Wu, T.H., and M.G. Marinus. 1999. Deletion mutation analysis of the mutS gene in
Escherichia coli. J. Biol. Chem. 274(9):5948-52.
77
Chapter 5. Future Work.
5.1. Determine the contributions of individual genes (tetA vs tetR) to the observed
toxicity by using modified tetracycline resistance determinants. At this time, it is not clear
whether the tetR, tetA, or both genes encoded by the TnlO are responsible for the observed
results. Our current hypothesis is that, most likely, the pre-treatment sensitization is dependent
on the expression of TetR (and is therefore not observed with pBR322), while co-treatment
sensitization is dependent on the expression of TetA (and is therefore observed with both the
TnlO and pBR322 genes). To ascertain the roles of the two proteins in these processes, it
would be important to do experiments with two additional plasmids. One should carry only
the functional TnlO tetA gene, with the tetR gene disrupted or deleted (the plasmid with
functional TetR and tetA deletion would not confer tetracycline resistance and would
therefore not provide any information about TetR involvement). The other plasmid should
carry the pBR322 tetA gene as well as the corresponding tetR gene (class C), for example,
pSC101 parental plasmid from which pBR322 was derived. These plasmids should help to
answer definitively which proteins are involved in which processes and whether or not the
above stated hypothesis is correct.
5.2. Experiments with other platinum and other DNA-damaging compounds. It would be
necessary to try other platinum compounds in the similarly designed experiments. This study
would accomplish at least two goals. First, it would help determine which elements and which
qualities of the cisplatin molecule are important for the observed phenomenon, and therefore
are desirable in the design or choice of other molecules that could substitute for cisplatin in
the future. Second, it is possible that another platinum compound with milder side effects
(like carboplatin) will produce similar or higher toxicity in tetracycline-resistant E. coli. If we
establish through these experiments that the DNA-damaging domains of platinum compounds
are the most crucial to their activity, then we may suggest that other DNA-damaging
compounds should be tested for activity, preferably compounds that are established in clinical
practice, such as chlorambucil or bleomycin.
5.3. Experiments with other tetracyclines, other tetracycline resistance determinants,
and other bacterial species. It will also be important to assay other tetracyclines to determine
which ones will be able to produce similar toxicity in E. coli in combination with platinum
compounds. Since at this time the precise mechanism of the described toxicity is not clear, it
is not evident which parts of the tetracycline molecule are involved in interactions with
cisplatin. It is possible that existing modifications to the tetracycline molecule will be more
prone to interactions with cisplatin and will produce a stronger effect. We must underscore
here that the other antibiotics we tested (ampicillin, kanamycin and chloramphenicol) did not
produce the same effect of sensitization to cisplatin in combination with their respective
resistance genes, neither in pre-treatment, nor in co-treatment experiments.
In order to assess the scope of the described finding, it would be necessary to test other
tetracycline resistance determinants and bacterial species other than E. coli. We think that the
sensitivity to cisplatin conferred by TnlO may be dependent on a quality that is common
among some tetracycline-resistance determinants, rather than on a quality specific to TnlO. It
78
is likely that other tetracycline resistance genes that encode similarly regulated efflux pumps
will show a response similar to the TnlO gene. Since efflux pumps are predominantly found
among Gram-negative bacteria, especially enteric species, it is likely that this group of
tetracycline-resistant bacteria will be most susceptible to the combination of cisplatin and
tetracycline.
It will definitely be very interesting to test the bacterial strains that have acquired
multiple-antibiotic resistance, in addition to tetracycline resistance (Jahn et al., 1979; Ng et
al., 1999; Inamine and Burdett, 1985). We expect that such strains would be as susceptible to
cisplatin-tetracycline treatment as the strains that are tetracycline-resistant only. If this proved
to be the case, it would indicate a potential for such combination treatment against infections
that are refractory to many types of antibiotic treatments.
5.4. Animal experiments with tetracycline-resistant bacterial infections. It would be
important to determine whether a cisplatin-tetracycline co-treatment regimen could combat
tetracycline-resistant bacterial infections in laboratory animals. The likely target pathogens
would be Gram-negative, most likely enteric bacteria expressing tetracycline efflux genes.
The infections could also be multiple antibiotic-resistant, as long as they maintain effluxdependent tetracycline resistance. Platinum analogs with milder side effects could be tried, as
well as other tetracycline compounds, after tests are conducted with bacterial strains in culture
to see which compounds are the best candidates for combination therapy.
5.5. Consider effects of metal ions, such as magnesium, calcium, and copper salts. In
attempts to elucidate the pathway of observed toxicity, we must be mindful of the possible
connections between the properties of the two molecules in the cellular milieu. For example,
cisplatin is known to interfere with calcium and magnesium metabolism in cancer patients
(Goren, 2003; Stewart et al., 1985); tetracycline, on the other hand, is capable of chelating
divalent cations, most notably magnesium and calcium (Newman and Frank, 1976;
Yamaguchi, 1990). Another example of such a possible connection between the two
molecules is that cisplatin has been proposed to share intracellular trafficking pathways with
copper (Katano et al., 2002); tetracycline, on the other hand, is capable of undergoing redox
chemistry and inflicting DNA damage in the presence of copper salts (Quinlan and
Gutteridge, 1988; Khan et al., 2003). Although these observations do not yet suggest a clear
link, they nevertheless must be kept in mind for possible future revelations.
5.6. Determine the DNA platinum levels in pBR322-carrying cells. Our initial hypothesis
was that proton influx mediated by the antiporter pump could lead to transient mild
acidification of the local intracellular environment, which could result in increased cisplatin
reactivity and thus lead to increased DNA damage levels (Chau and Stewart, 1999; Murakami
et al., 2001). Although we do see increased DNA damage levels in tetracycline-grown cells,
we nevertheless doubt our initial hypothesis suggesting that this is a result of pump-mediated
intracellular acidification. The reason for this is that we do not observe the similar toxicity in
tetracycline-grown cells carrying the pBR322 plasmid. This plasmid encodes a similar
antiporter pump, and even though the level of homology between the two pumps is not high,
they both nevertheless exchange a tetracycline-metal complex for a proton, and activity of
either therefore may result in transient decrease in the intracellular pH. If increased DNA
79
damage levels were the result of intracellular acidification, and toxicity was directly
associated with increased DNA damage levels, then we should see a similar high toxicity in
pBR322-carrying strains. This question can be partially addressed by determining the DNA
platinum levels in strains carrying pBR322 grown with tetracycline. If there is no increase in
DNA damage levels in pBR322-carrying cells, then it is likely that toxicity in TnlO-carrying
cells may be directly (although not necessarily exclusively) associated with the increase in
DNA platinum levels in TnlO cells. That would also mean that an increase in DNA damage
levels in TnlO cells is not likely to result from intracellular acidification. If, however, pBR322
strains also have increased DNA damage levels, it is then possible that an increase in DNA
damage does occur by the same mechanism in strains carrying the two different pumps, i.e., it
may possibly result from intracellular acidification. In this case, toxicity in TnlO cells is
apparently not directly associated with increase in DNA damage (Janovska, et al, 2002).
5.7. Determine whether co-treatment alone leads to increased total cellular and DNA
platinum levels. At this time, we think that sensitization to cisplatin by co-treatment with
tetracycline is likely to be a phenomenon not only independent from sensitization by growth
in tetracycline, but also possibly occurring through a different mechanism. We already
demonstrated that the two phenomena are independent of each other; sensitization by cotreatment occurs in cells either grown or not grown with tetracycline. Then, unlike pretreatment sensitization, this phenomenon is observed in strains carrying either the pBR322 or
the TnlO tetracycline resistance gene. This suggests that it is dependent on some quality
shared between the two genes, in contrast to pre-treament sensitization, which likely is due to
a quality specific to the TnlO gene, but not the pBR322 gene. And finally, unlike pretreatment, co-treatment results in a small increase in total cellular platinum levels, which
correlates with a small increase in DNA damage levels. The last two observations may reflect
a mechanism of toxicity that depends on increased cellular accumulation of cisplatin through
the action of the antiporter, leading to a subsequent mild increase in DNA damage levels. To
establish whether this mechanism may be operative, it would help to determine whether cotreatment alone would lead to an increase in total cellular and DNA damage levels in strains
carrying the TnlO and pBR322 genes.
5.8. Search for TetA and TetR homology to mammalian proteins. Examine HeLa cells
transfected with TetR or TetA. At this time, we believe that the toxicity described in E. coli
cells will be restricted to bacterial pathogens since it is actually dependent on the presence of
bacterial genes of resistance to tetracycline. It is possible, though, that a counterpart of the
TnlO gene - or one of its constituents, such as the tetA or tetR gene - will be identified, for
example, in mammalian tumor cells. If this were to happen, then it would be possible that a
combination of drugs similar to tetracycline and cisplatin would prove lethal to such cells
through a pathway similar to one in bacterial cells. It is worth trying to determine whether the
tetracycline-cisplatin combination would be lethal to HeLa cells, for example, by expressing
TetA or TetR from a eukaryotic vector. TetR variants are widely used in eukaryotic gene
regulation (Berens and Hillen, 2003; Blau and Rossi, 1999; Shockett and Schatz, 1996).
Although expression of bacterial membrane proteins in mammalian cells may not be
straightforward, functional expression of bacterial antibiotic-resistance membrane protein in
mammalian cells has been demonstrated (van Veen et al., 1998).
80
5.9. Compare the transcriptional profiles of wild-type and tet cells with TnlO and
pBR322. Check whether survivors maintain their resistant phenotype, and if so, how
their transcriptional response is different. Likely, the pathway through which toxicity is
exerted by the combination of tetracycline, cisplatin and tetracycline-resistance genes will be
elucidated through a variety of approaches. It may help to conduct gene chip experiments to
compare the transcriptional profiles of wild-type and tetracycline-resistant cells subjected to
cisplatin-tetracycline treatment. It is certain that the transcriptional response of these cells
must be different, and it is possible that gene chips will lend a hand in identifying the cellular
pathways leading to toxicity. It would be interesting to compare the responses of TnlO- and
pBR322-carrying cells; however, this would be best done after it is determined which genes
are responsible for which effects (see above). It also would be interesting to determine
whether survivors of the cisplatin-tetracycline treatment maintain their resistant phenotype
and how their transcriptonal response to subsequent treatments is different. An important
result of such studies would be that they could help elucidate alternative functions that TetA
and TetR may have in bacterial cells. For example, DNA-binding TetR may be a
transcriptional regulator for a variety of other genes. The trans-membrane efflux pump may
also have other roles, such as signal transduction or regulation of other functions.
5.10. Repair shielding hypothesis is not supported by our data. Although shielding of
cisplatin adducts by TetR protein could seem like an attractive explanation for our results, we
believe that at this time our data do not support such a mechanism. On the contrary, our data
seem to suggest that the higher DNA damage levels in tetracycline-resistant cells result from
increased adduct formation rather than decreased adduct repair. First, the increase in DNA
damage levels is not likely to be a result of repair shielding since in our experiments DNA
was isolated right after treatment, i.e., additional time was not allowed for extended adduct
formation and/or repair. Second, if the repair shielding mechanism were operative, then
expressing TetR after cisplatin treatment would produce the toxic effect. However, in our
experiments where the tetracycline resistance gene was induced by tetracycline after cisplatin
treatment, toxicity was not enhanced in tetracycline-resistant cells compared to the wild-type.
We therefore believe that the increase in DNA damage levels stems from a different
mechanism, although this mechanism may possibly involve interactions with TetR, such as
forming DNA-TetR cross-links.
81
References:
Berens, C., and W. Hillen. 2003. Gene regulation by tetracyclines. Constraints of resistance
regulation in bacteria shape TetR for application in eukaryotes. Eur. J. Biochem. 270:31093121.
Blau, H.M., and F.M.V. Rossi. 1999. TetB or not TetB: advances in tetracycline-inducible
gene expression. Proc. Natl. Acad. Sci. USA. 96:797-799.
Chau, Q., and DJ.Stewart. 1999. Cisplatin efflux, binding and intracellular pH in the
HTB56 human lung adenocarcinoma cell line and the E-8/0.7 cisplatin-resistant variant.
Cancer Chemother. Pharmacol. 44:193-202.
Goren, M.P. 2003. Cisplatin nephrotoxicity affects magnesium and calcium metabolism.
Med. Pediatr. Oncol. 41:186-189.
Inamine, J.M., and V. Burdett. 1985. Structural organization of a 67-kilobase streptococcal
conjugative element mediating multiple antibiotic resistance. J. Bacteriol. 161(2):620-626.
Jahn, G., R. Laufs, P.-M. Kaulfers, and H. Kolenda. 1979. Molecular nature of two
Haemophilus influenzae R factors containing resistances and the multiple integration of drug
resistance transposons. J. Bacteriol. 138(2):584-597.
Janovska, E., O. Novakova, G. Natile, and V. Brabec. 2002. Differential genotoxic effects
of antitumor trans-[PtCl2(E-iminoether)2 ] and cisplatin in Escherichia coli. J. Inorg. Biochem.
90:155-158.
Katano, K., A. Kondo, R. Safaei, A. Holzer, G. Samimi, M. Mishima, Y.-M. Kuo, M.
Rochdi, and S.B. Howell. 2002. Acquisition of resistance to cisplatin is accompanied by
changes in the cellular pharmacology of copper. Cancer Research. 62:6559-6565.
Khan, M.A., J. Mustafa, and J. Musarrat. 2003. Mechanism of DNA strand breakage
induced by photosensitized tetracycline-Cu(II) complex. Mutation Research. 525:109-119.
Newman, E.C., and C.W. Frank. 1976. Circular dichroism spectra of tetracycline complexes
with Mg+ 2 and Ca+2 . J. Pharm. Sci. 65(12):1728-1732.
Ng, L.-K., M.R. Mulvey, I. Martin, G.A. Peters, and W. Johnson. 1999. Genetic
characterization of antimicrobial resistance in Canadian isolates of Salmonella serovar
Typhimurium DT 104. Antimicrob. Agents Chemother. 43(12):3018-3021.
Murakami, T., I. Shibuya, T. Ise, Z.-S. Chen, S. Akiyama, M. Nakagawa, H. Izumi, T.
Nakamura, K. Matsuo, Y. Yamada, and K. Kohno. 2001. Elevated expression of vacuolar
proton pump genes and cellular pH in cisplatin resistance. Int. J. Cancer. 93:869-874.
82
Shockett, P.E., and D.G. Schatz. 1996. Diverse strategies for tetracycline-regulated
inducible gene expression. Proc. Natl. Acad. Sci. USA. 93:5173-5176.
Stewart, A.F., T. Keating, and P.E. Schwartz. 1985. Magnesium homeostasis following
chemotherapy with cisplatin: prospective study. Am. J. Obstet. Gynecol. 153(6):660-5.
Quinlan, G.J., and J.M.C. Gutteridge. 1988. Hydroxyl radical generation by the
tetracycline antibiotics with free radical damage to DNA, lipids and carbohydrate in the
presence of iron and copper salts. Free Radical Biology and Medicine. 5:341-348.
van Veen, H.W., R. Callaghan, L. Soceneantu, A. Sardini, W.N. Konings. C.F. Higgins.
1998. A bacterial antibiotic-resistance gene that complements the human multidrug-resistance
P-glycoprotein gene. Nature. 391:291-295.
Yamaguchi, A., T. Udagawa, and T. Sawai. 1990. Transport of divalent cations with
tetracycline as mediated by the transposon TnlO-encoded tetracycline resistance protein. J.
Biol. Chem. 265(9):4809-4813.
83
Appendices.
Appendix I. Analysis of the "Transcription Factor Hijacking" Hypothesis in vivo.
The aim of this project was to investigate whether the "transcription factor
hijacking" phenomenon previously observed in our lab in vitro was taking place in vivo as
well. Human upstream binding factor (hUBF), an HMG-box protein which is a
transcription factor involved in the transcription of ribosomal RNA by RNA polymerase I,
was shown in our lab to interact specifically with cisplatin adducts in vitro (Treiber et al.,
1994). The binding affinity of hUBF to cisplatin adducts was remarkably strong (Kd-60
pM), and merely 3-fold lower than the affinity of hUBF to the rDNA promoter (Kd-18
pM), its natural binding site. This observation suggested that cisplatin adducts could act as
molecular decoys for hUBF, titrating it away from its natural sites of action. In further
experiments, it was shown that cisplatin adducts were indeed able to effectively compete
with rDNA promoters for hUBF binding, suggesting that functional inhibition of rRNA
transcription could result from such competition.
Subsequent work provided more evidence for the possibility that cisplatin adducts
may form molecular decoys for hUBF (Zhai et al., 1998). Ribosomal RNA transcription in
a reconstituted system in vitro was specifically inhibited by cisplatin adducts in a dosedependent manner. Moreover, increasing the levels of hUBF in the system, while keeping
the levels of cisplatin adducts constant, restored the level of transcription nearly to the
level seen in the absence of cisplatin adducts. These data together strongly suggested that
cisplatin adducts were highly effective in disrupting hUBF function in vitro.
Since hUBF could be titrated away by cisplatin adducts from the multi-factor
transcription complex assembled in vitro, the possibility of a similar event occurring in a
multi-component system in vivo also exists. Additional support for this notion is the fact
that the natural site of action of hUBF is the GC-rich nucleolar DNA, which could be
preferentially targeted by cisplatin and therefore provide ample alternative binding sites for
hUBF. If such an event were to take place in cells, rapidly growing cancer cells would
likely be more severely affected, since they rely heavily on increased levels of ribosomal
RNA synthesis to maintain rapid growth. However, normal cells would also be affected,
leading to non-specific cisplatin toxicity.
Hijacking could in principle also occur with other DNA-binding proteins,
disrupting not only rRNA synthesis, but also other essential processes as well. However,
this phenomenon would be limited to those proteins which have affinities for cisplatin
adducts comparable to the affinities for their natural binding sites; hUBF is the best
candidate described for such a mechanism so far, although other candidates may be
suggested (Trimmer et al., 1998; Vichi et al., 1997).
After pursuing a number of different approaches (which will not be described here)
to detecting hijacking in vivo, we followed the experimental design described below that
we believe would be the best and most direct way to detect hijacking in vivo, if such a
84
mechanism were to take place. We set out to use the chromatin immuno-precipitation
(ChIP) methodology to determine the changes in hUBF localization in cellular DNA that
may occur upon treatment with cisplatin. In this method, upon exposure to cisplatin, live
cells are treated with formaldehyde, which cross-links hUBF to the DNA sites it is bound
to at the time of treatment. Cells are then lysed, and chromatin is sheared and extracted;
nucleases may also be used to help digest unbound DNA, while protein-protected DNA
will stay intact. Using hUBF-specific antibody, hUBF is immuno-precipitated together
with the DNA fragments bound to it; the DNA fragments are then purified, amplified by
PCR if necessary, and analyzed by slot-blot (quantitative PCR may be used instead). If
substantial re-distribution of hUBF in cellular DNA is taking place in cells as a
consequence of cisplatin exposure, then we would expect to see a decrease in the amount
of the rDNA promoter immuno-precipitated with hUBF, and an increase in non-specific
DNA binding in response to cisplatin treatment. In this particular system, additional
advantage can be taken of the fact that hUBF is a nucleolar protein. Isolation of nucleoli
instead of nuclei should leave the bulk of the DNA out and help the enrichment of DNA
for the hUBF-bound fragments (O'Sullivan et al., 2002).
In parallel, we must monitor ribosomal DNA transcription by measuring transient
levels of freshly synthesized ribosomal RNA by a run-off assay. In this method, upon
treatment with cisplatin, nuclei are isolated and incubated with four ribonucleotides,
including radiolabeled UTP. Freshly synthesized radiolabeled RNA is then purified and
hybridized to DNA probes (slot-blot); RNA-DNA hybrids are then analyzed.
It is important to monitor by run-off assay the levels of transcription by all three
RNA polymerases. If the level of transcription by RNA polymerase I drops, but
transcription by other polymerases remains relatively unaffected, that would suggest that
pol I is selectively inhibited by cisplatin, providing support for the hijacking theory.
However, if other RNA polymerases were affected as well, that would reflect a nonspecific cellular response to cisplatin, in which transcription is deregulated. Good controls
to monitor would could be 5S ribosomal RNA synthesized by polymerase III, and mRNAs
for ribosomal proteins synthesized by polymerase II.
One of the factors that could impede detection of the a hijacking mechanism in
vivo, is that rRNA synthesis is subject to multiple regulatory controls in cells, which
involve regulation of the levels of hUBF (Glibetic et al., 1995). Cells that require stable
production of ribosomal RNA may be able to upregulate expression of hUBF to prevent
the drop in ribosomal RNA production; such upregulation may mask hUBF hijacking. On
the other hand, cells subjected to cisplatin assault may halt ribosomal RNA synthesis
because cellular growth is inhibited due to events other than hijacking; this would lead to a
drop in hUBF expression. This complexity means that controls for the levels of hUBF
expression must also be carefully designed.
We would like to emphasize the interesting results reported by Jordan and CarmoFonseca in 1998; these authors observed that upon treatment with cisplatin, hUBF was
redistributed within the nucleoli of the human cells. That redistribution was similar to the
one observed after selective inhibition of rRNA synthesis by actinomycin D, where
85
components of the polymerase I transcription machinery co-localized together upon
inhibition of rRNA synthesis, and remained associated with inactive rRNA genes (Jordan
et al., 1996). Similarly, upon treatment with cisplatin, hUBF redistributed in the nucleoli
along with other components of the polymerase I transcription machinery, suggesting that
such redistribution could follow inhibition of rRNA synthesis by cisplatin, rather than be
due to the cisplatin effects specific to hUBF. The authors also showed selective inhibition
of rRNA synthesis by cisplatin. These results together demonstrate that although disruption
of rRNA synthesis by cisplatin may be an important contributor to cisplatin cellular
toxicity, it may not necessarily be explained by hUBF hijacking in cells. It is certainly
possible that the redistribution of hUBF is a direct cause of redistribution of other
components of the polymerase I transcription machinery; however, it is also likely that,
conversely, such redistribution is a consequence of inhibition of rRNA synthesis by
cisplatin through a different mechanism.
86
Appendix II. Analysis of DNA Damage Following Treatment of Estrogen ReceptorPositive and Estrogen Receptor-Negative Breast Cancer Cells by Experimental Drug
E27a.
This work was part of a bigger "Fatal Engineering" project - a group effort to
design and synthesize drugs selectively targeting certain tumor types. The basis for the
"Fatal Engineering" project is the "repair shielding" hypothesis first proposed to account
for the mechanism of action of the successful anticancer drug cisplatin. According to this
hypothesis, cisplatin adducts may be shielded from cellular DNA repair machinery by
other cellular proteins with high affinity for cisplatin adducts. The result of such shielding
would be that cisplatin adducts would persist in the genome, leading to increased cellular
toxicity. Cells expressing higher levels of such cisplatin-binding proteins would therefore
accumulate more unrepaired adducts, leading to selective toxicity of the drug to these cells
(Huang et al., 1994; McA'Nulty et al., 1996; Brown et al., 1993).
Since some cancer types are known to overexpress certain proteins, it is indeed an
attractive idea to try to design compounds that would be selectively retained in the tumor
cells with an abundance of specific proteins that would shield them from repair. For
example, the estrogen receptor is overexpressed in approximately half of the breast cancer
tumors; DNA damaging compounds selectively recognized by the estrogen receptor could
be shielded from repair in estrogen receptor-positive tumors, leading to accumulation of
DNA adducts in tumor cells and selective toxicity (Essigmann et al., 2001).
The molecule designed to attract a cellular protein must possess several features critical
for its success in the cellular milieu. It should have a DNA-reactive functionality which
would form DNA adducts; a protein-binding domain which would be able to successfully
trap the protein of choice; and a covalent linker between these two moieties, such that it
would not be destroyed by the cellular enzymes. The choice of a target protein is also
essential. Estrogen receptor (ER) is a logical target protein for this project, since it is
overexpressed in a number of tumors, and also because multiple molecules are known to
interact well with the estrogen receptor and therefore could be used as a ligand-binding
domain in the designed compound.
A number of experimental compounds were synthesized in our lab according to these
principles, and were tested for their activity in estrogen receptor-positive and estrogen
receptor-negative breast cancer cell lines (Rink et al., 1996; Mitra et al., 2002; Sharma et
al., 2004). The goal of the project described here was to determine whether treatment with
one of the experimental compounds, called E27a in our lab (estradiol, the natural ligand
for the estrogen receptor, tethered to bis-chloroethyl aniline mustard), actually resulted in
greater accumulation of DNA damage in ER-positive cells, as compared to ER-negative
cells.
The method of choice employed to address this question was single cell gel
electrophoresis, also known as the comet assay (Fairbairn et al., 1995; Tice et al., 2000;
87
Moller et al., 2000). This assay, originally developed to detect double-strand DNA breaks
in individual cells, is currently advanced enough to also detect, through a variety of
modifications, single-strand breaks, alkali-labile sites, certain types of base modifications,
and inter-strand DNA cross-links. DNA strand breaks detected by the comet assay can
originate from direct impact of chemical or physical agents, or be products of incomplete
DNA excision repair, replication, recombination, or apoptosis. DNA damage measured by
the comet assay represents a mixture of frank strand breaks and DNA damage that is
converted to strand breaks by alkaline treatment combined with other modifications (DNA
repair enzymes, y-irradiation). The comet assay gets its name from the characteristic
"comet" tails that form when DNA migrates out of the cells during electrophoresis. The
more breaks are present in the cellular DNA, the higher is the number of DNA fragments
that are able to migrate, and the further is the distance they travel during electrophoresis.
The relative amounts of DNA in the "tail" and in the "head" of the comet (tail intensity), as
well as the distance the DNA migrated (tail length), then serve as an indication of the
extent of DNA strand breakage.
The comet assay used in this project to assess the levels of DNA damage and repair
consisted of the following basic steps. ER-positive and ER-negative cells were treated with
the experimental compound and a control compound. The control compound had the same
basic structure as the experimental compound, except that it lacked the estrogen receptor
recognition domain, and therefore was not expected to produce differential responses in
ER+ and ER- cells. Appropriate time intervals were allowed for repair, and cells were then
harvested for analysis. Single cell suspensions in low melting point agarose were prepared
for each sample and fixed on glass slides. Cells embedded in agarose on glass slides were
lysed in detergents in order to liberate DNA and disperse cellular components. Next, slides
with cells were subjected to alkaline treatment, which denatured DNA and converted the
alkali-labile sites into strand breaks. Slides were then subjected to electrophoresis under
the same alkaline conditions. After electrophoresis, slides were washed, fixed using
ethanol, stained with DNA-specific fluorescent dye, and analyzed using fluorescent
microscopy; the extent of DNA migration in the agarose gel reflected the extent of DNA
damage expressed as strand breaks.
The comet assay was initially used to assess the formation of inter-strand DNA
crosslinks by E27a; the compound, however, was not a very efficient inter-strand crosslinker, based on comet assay results. The comet analysis, at various pH values, of single
strand breaks resulting from E27a exposure suggested that the majority of such breaks
likely originated from the conversion of alkali-labile sites into strand breaks. Efficient and
specific conversion of E27a base damage into alkali-labile sites would therefore greatly
facilitate comet assay detection of base damage produced by E27a. Evidently, a
modification of the method was desired that would allow such efficient and specific
conversion - physical, chemical, or enzymatic - of the base damage into the alkali-labile
sites detectable by the comet assay.
As mentioned above, the comet assay can be modified to detect certain types of DNA
base damage through the use of DNA repair enzymes. Examples of such modifications to
the basic procedure are the use of lesion-specific endonucleases such as Fpg, Endo m, and
88
UVDE to estimate levels of 8-oxoguanine/formamidopyrimidine lesions, oxidized
pyrimidines, and pyrimidine dimers, respectively (Collins et al., 1997; Horvathova et al.,
1999). In principle, any lesion for which a specific repair enzyme exists may be detected
enzymatically; however, with an experimental compound alternative approaches must also
be explored. We had shown that heat was a very efficient agent for physical conversion of
E27a lesions into abasic sites.
In vitro experiments with an E27a-damaged plasmid indicated that heat promoted
specific degradation of E27a-damaged DNA. Temperatures of 50-600 C applied for 1-3
hours led to extensive formation of abasic sites in E27a-damaged plasmid, but not in
untreated plasmid. These results translated well into whole cells, where controlled thermal
treatment also produced preferential formation of alkali-labile sites in E27a treated cells.
Conditions were easily adjusted to produce maximal difference between the treated and
untreated cells, in order to eliminate the effects of non-specific depurination. Interestingly,
in vitro results also suggested that, while plasmid degradation at 370 C was also occurring
preferentially in the E27a-damaged plasmid, such a plasmid could be partially protected
from heat-induced degradation by the estrogen receptor ligand-binding domain.
Although we observed no consistent, reproducible differences between estrogen
receptor-positive and -negative human breast cancer cells lines with respect to E27a DNA
damage formation and repair, we nevertheless want to emphasize the importance of the
novel modification to the detection method. The use of controlled thermal treatment to
induce preferential formation of abasic sites from damaged bases makes the comet assay
suitable for work with many compounds that may form DNA base damage, but not strand
breaks or alkali-labile sites.
We would like to mention that preliminary evidence supporting the hypothesis that
repair of the E27a DNA adducts may be slower in ER-positive than in ER-negative cells
has now been obtained in our lab using accelerator mass spectrometry (AMS). According
to an AMS experiment, the E27a adducts were retained longer in ER-positive than in ERnegative cells, while the rate of removal of DNA adducts formed by control compound
melphalan was the same in both cell types (Hillier, 2005).
89
Appendix III. Involvement of Base Excision Repair Proteins in the Processing of
Cisplatin Lesions in Escherichia coli.
The aim of this project was to determine whether base excision repair of DNA adducts
was involved in E. coli responses to cisplatin, with the long-term goal of finding out
whether such involvement could have a bearing on the responses of mammalian cells to
cisplatin as well. The background for this project is best detailed in the Ph.D. thesis of
Maria Kartalou. Briefly, the findings upon which this project was based were the
following.
Electrophoretic mobility shift experiments indicated that bacterial MutY DNA
glycosylase, which is known to excise A bases from A:G and A:8-oxoG mismatches, was
able to recognize all three cisplatin intrastrand cross-links as well as an A:G mismatch in
vitro; moreover, the presence of a platinum adduct did not abolish the ability of MutY to
remove adenine from the mismatches tested. Specifically, MutY was able to incise DNA at
the platinated adenine of the 1,2-d(ApG) adduct when it was mispaired with a G or C base.
Furthermore, when adenine was positioned opposite the platinated G of the 1,2-d(ApG)
adduct, excision of adenine was very efficient. However, excision of adenine was reduced,
compared to an A:G mismatch, when A was misincorporated opposite either the 5' or 3'
platinated G of the 1,2-d(GpG), or 5' platinated G of the 1,3-d(GpTpG) adduct, indicating
that the presence of the platinum adduct did hinder the ability of MutY to excise adenine.
In vivo experiments indicated that mutY cells were resistant to cisplatin-induced
mutagenesis, indicating that functional MutY was probably required for induction of
mutagenesis by cisplatin. Results also suggested that adduct recognition by MutY was
sufficient for mutagenesis, and excision ability of MutY glycosylase was not required
(Kartalou, 2000).
These observations suggested the hypothesis that MutY DNA glycosylase could be
competing with the mismatch repair protein MutS (which is responsible for binding to
DNA sequences containing mispaired bases and initiating repair process) for recognition of
the mismatched base associated with the cisplatin adduct (the so-called compound lesion)
in vivo, and could thereby be preventing MutS from initiating repair of the mismatched
cisplatin adduct. Such a competition would lead to persistence of the mismatch in the
genome, subsequently leading to the observed mutagenesis. To address this question, it
would be interesting to perform a glycosylase assay with MutY glycosylase in the presence
of MutS protein, in order to determine whether MutS could interfere with the ability of
MutY to bind and excise the cisplatin adduct from a mismatched base in vitro.
To determine whether competition between MutY and MutS proteins could be taking
place in vivo, we explored the following approach. According to the "futile repair"
hypothesis put forward to explain the sensitization of E. coli to cisplatin by mismatch
repair, in the absence of the methylation signal in a dam deficient background, mismatch
repair machinery engages in cycles of incomplete repair attempts, which eventually lead to
accumulation of DNA strand breaks and cell death. This process requires the involvement
of functional MutS, and mutation of the mutS gene, which eliminates mismatch repair,
90
leads to complete recovery of the wild-type phenotype with respect to cisplatin toxicity. In
the event that the MutY protein could be competing with MutS for binding to the
mismatched lesions, such a competition would be equivalent to the absence of MutS at the
mismatched cisplatin lesion sites. Therefore, overexpression of the MutY protein in the
dam background should have the same effect as a deletion of the mutS gene, and should
afford at least partial recovery of the wild-type phenotype.
We did experiments with dam methylation deficient strains overexpressing the MutY
protein; recovery of the wild-type phenotype was not observed, i.e., the high sensitivity of
dam mutants to cisplatin was paralleled by the high sensitivity of the same mutants
overexpressing the MutY protein. Therefore, such competition is either not taking place in
vivo, or is not sufficient to rescue the cells with functional MutS in dam background from
cisplatin toxicity.
It is interesting to note that, coincidentally, another group was also considering a
hypothesis that competition could be taking place between bacterial MutS and MutY
proteins for repair of the compound lesions involving an A:G mismatch and a cisplatin
adduct. In a recently published paper (Fourrier et al., 2003), the authors found that, in
contrast to results from our lab, MutY was not able to excise adenine mispaired with either
5' or 3' guanine of the 1,2-d(GpG) cisplatin crosslink. Further analysis indicated that this
lack of glycosylase activity resulted from the inability of MutY to specifically recognize
the A:G mismatch in the context of 1,2-d(GpG) cisplatin cross-link, rather than from
inhibition of glycosylase activity by the cisplatin lesion. Based on their observations, the
authors concluded that competition between MutS and MutY for binding to compound
lesions involving an A:G mismatch and a major cisplatin crosslink could not be taking
place, since MutY, unlike MutS, was not able to recognize such lesions in their
experiments.
On another note, building upon observations from our lab (Kartalou, 2000) that MutY
DNA glycosylase could catalyze cleavage at the sites of cisplatin modification in vitro, we
set out to determine whether cisplatin could also induce glycosylase activity in vivo, i.e.,
whether MutY could initiate repair events upon cisplatin treatment in E. coli. Using mutY
deficient E. coli strains, as well as strains over-expressing MutY, we hoped to determine
changes in the levels of AP sites resulting from cisplatin treatment in E. coli. Upon
treatment with cisplatin, AP sites in DNA would be covalently labeled with biotin-tagged
aldehyde reactive probe; the latter would be detected and quantified using avidin-tagged
horseradish peroxidase. We would expect to see a difference in the number of AP sites
upon cisplatin treatment in E. coli strains with different MutY status, if MutY were indeed
involved to some extent in the initiation of cisplatin adduct repair. We would expect that
the difference in the numbers of abasic sites due to presumed MutY processing of cisplatin
lesions would not be high; therefore, conditions for such an assay would be best worked
out with a known oxidative damage inducer that would produce natural substrates for
MutY. Experiments with cisplatin would probably benefit from the use of an AP
endonuclease deficient E. coli background for increased sensitivity, since abasic sites
would not be further processed, and would be more readily available for detection in such a
background.
91
It must be mentioned that the rationale and strong support for initiating this work came
from previous unpublished observations that mutY cells were more sensitive to cisplatin
than wild-type E. coli (i.e., that MutY protected cells from cisplatin toxicity), that
overexpression of MutY in a mutY deficient background rescued the wild-type phenotype
with respect to cisplatin toxicity, and that the excision ability of MutY was required for
such protection. These observations were later proved to be an unfortunate experimental
artifact, and therefore should not be considered as a basis for the described work.
Nevertheless, the involvement of MutY in cellular responses to cisplatin has not been ruled
out by any of the available data. Moreover, the evidence discussed below may provide
some support for the notion that BER may be involved in cellular responses to cisplatin in
E. coli.
Another indication of potential BER involvement in the processing of cisplatin
lesions in E. coli came from observations that MutM DNA glycosylase (also known as
FAPY glycosylase), which excises ring-opened purines and 8-oxoG lesions from DNA,
was capable of binding to all three cisplatin intrastrand crosslinks in electrophoretic
mobility shift assays, although it did not initiate excision of the adducts in vitro. In
addition, mutM-deficient E. coli strains were reported to be less sensitive to cisplatin than
wild-type, suggesting that MutM sensitized cells to cisplatin. Although we later found that
mutM strains were not necessarily less sensitive to cisplatin than wild-type, the results with
AP endonuclease-deficient strains discussed below seem nevertheless to support the notion
that MutM might be involved in the processing of cisplatin adducts in vivo and might be
sensitizing cells to cisplatin under certain conditions. The E. coli strains deficient in AP
endonucleases Exo III and Endo IV were shown to be more sensitive to cisplatin than the
corresponding wild-type strains, and an additional mutation in the mutM gene abolished
this sensitivity, returning survival to wild-type levels. This suggested that interactions of
MutM with cisplatin may occur in vivo (although the mechanism of such interactions is not
clear at this time), and lead to the creation of toxic abasic sites, which would persist in an
AP endonuclease-deficient background. Lack of functional MutM would therefore protect
cells from the formation of abasic sites upon cisplatin treatment, rescuing cells from
toxicity in an AP endonuclease-deficient background. As suggested above, these results
seem to support the idea that BER may be involved in the cellular response to cisplatin in
E. coli.
In addition to work in E. coli, we were taking these studies further with the intent to
find out whether hOOG1, the human homologue of E. coli MutM DNA glycosylase, was
also capable of binding cisplatin adducts in vitro, and if so, whether cisplatin adducts could
compete hOGG1 away from 8-oxoguanine lesions in vitro. As a continuation of these
studies in vivo, it would be important to examine the responses of hOGG1-deficient
mammalian cell lines to cisplatin.
92
References:
Brown, S.J., P.J. Kellett, and S.J. Lippard. 1993. Ixrl, a yeast protein that binds to
platinated DNA and confers sensitivity to cisplatin. Science. 261:603-605.
Collins, A.R., V.L. Dobson, M. Dusinska, G. Kennedy, and R. Stetina. 1997. The
comet assay: what can it really tell us? Mutat. Res. 375:183-193.
Essigmann, J.M., S.M. Rink, H.-J. Park, and R.G. Croy. 2001. Design of DNA
damaging agents that hijack transcription factors and block DNA repair. Adv. Exp. Med.
Biol. 500:301-313.
Fairbairn, D.W., P.L. Olive, and K.L. O'Neill. 1995. The comet assay: a comprehensive
review. Mutat. Res. 339:37-59.
Fourrier, L., P. Brooks, and J.M. Malinge. 2003. Binding discrimination of MutS to a
set of lesions and compound lesions (base damage and mismatch) reveals its potential role
as a cisplatin-damaged DNA sensing protein. J. Biol. Chem. 278(23):21267-21275.
Glibetic, M., L. Taylor, D. Larson, R. Hannan, B. Sells, and L. Rothblum. 1995. The
RNA polymerase I transcription factor UBF is the product of a primary response gene. J.
Biol. Chem. 270(9):4209-4212.
Hillier, S. 2005. Ph.D. Thesis. Novel genotoxins that target estrogen receptor and
androgen receptor expressing cancers: identification of DNA adducts, pharmacokinetics,
and mechanism. Massachusetts Institute of Technology.
Horvathova,
E., D. Slamenova,
and A. Gabelova.
1999. Use of single cell gel
electrophoresis (comet assay) modifications for analysis of DNA damage. Gen. Physiol.
Biophys. 18:70-74.
Huang, J.-C., D.B. Zamble, J.T. Reardon, S.J. Lippard, and A. Sancar. 1994. HMGdomain proteins specifically inhibit the repair of the major DNA adduct of the anticancer
drug cisplatin by human excision nuclease. Proc. Natl. Acad. Sci. USA. 91:10394-10398.
Jordan, P., and M. Carmo-Fonseca. 1998. Cisplatin inhibits synthesis of ribosomal RNA
in vivo. Nucleic Acids Research. 26(12):2831-2836.
Jordan, P., M. Mannervik, L. Tora, and M. Carmo-Fonseca. 1996. In vivo evidence
that TATA-binding protein/SL1 colocalizes with UBF and RNA polymerase I when rRNA
synthesis is either active or inactive. J. Cell Biol. 133(2):225-234.
Kartalou, M. 2000. Ph.D. Thesis. The role of base excision repair proteins in the cellular
responses to the anticancer drug cisplatin. Massachusetts Institute of Technology.
93
McA'Nulty, M.M., and S.J. Lippard. 1996. The HMG-domain protein Ixrl blocks
excision repair of cisplatin-DNA adducts in yeast. Mutat. Res. 362(1):75-86.
Mitra, K., J.C. Marquis, S.M Hillier, P.T. Rye, B. Zayaz, A.S. Lee, J.M. Essigmann,
and R.G. Croy. 2002. A rationally designed genotoxin that selectively destroys estrogen
receptor-positive breast cancer cells. J. Am. Chem. Soc. 124(9): 1862-1863.
Moller, P., L.E. Knudsen, S. Loft, and H. Wallin. 2000. The comet assay as a rapid test
in biomonitoring occupational exposure to DNA-damaging agents and effect of
confounding factors. Cancer Epidemiology, Biomarkers, and Prevention. 9:1005-1015.
O'Sullivan, A.C., G.J. Sullivan, and B. McStay. 2002. UBF binding in vivo is not
restricted to regulatory sequences within the vertebrate ribosomal DNA repeat. Mol. Cell.
Biol. 22(2):657-668.
Rink, S.M., K.J. Yarema, M.S. Solomon, L.A. Paige, B.M. Tadayoni-Rebek,
J.M.
Essigmann, and R.G. Croy. 1996. Synthesis and biological activity of DNA damaging
agents that form decoy binding sites for the estrogen receptor. Proc. Natl. Acad. Sci. USA.
93:15063-15068.
Sharma, U., J.C. Marquis, A.N. Dinaut, S.M. Hillier, B. Fedeles, P.T. Rye, J.M.
Essigmann, and R.G. Croy. 2004. Design, synthesis, and evaluation of estradiol-linked
genotoxicants as anti-cancer agents. Bioorg. Med. Chem. Lett. 14:3829-3833.
Tice, R.R., E. Agurell, D. Anderson, B. Burlinson, A. Hartmann, H. Kobayashi, Y.
Miyamae, E. Rojas, J.-C. Ryu, and Y.F. Sasaki. 2000. Single cell gel/comet assay:
guidelines for in vitro and in vivo genetic toxicology testing. Environmental and Molecular
Mutagenesis. 35:206-221.
Treiber, D.K., X. Zhai, H.-M. Jantzen, and J.M. Essigmann. 1994. Cisplatin-DNA
adducts are molecular decoys for the ribosomal RNA transcription factor hUBF (human
upstream binding factor). Proc. Natl. Acad. Sci. USA. 91:5672-5676.
Trimmer, E.E., D.B. Zamble, S.J. Lippard, and J.M. Essigmann. 1998. Human testisdetermining factor SRY binds to the major DNA adduct of cisplatin and a putative target
sequence with comparable affinities. Biochemistry. 37(1):352-362.
Vichi, P., F. Coin, J.-P. Renaud, W. Vermeulen, J.H.J. Hoeijmakers, D. Moras, and
J.-M. Egly. 1997. Cisplatin- and UV-damaged DNA lure the basal transcription factor
TFIID/TBP. The EMBO Journal. 16(24):7444-7456.
Zhai, X., H. Beckmann. H.-M. Jantzen, and J.M. Essigmann. 1998. Cisplatin-DNA
adducts inhibit ribosomal RNA synthesis by hijacking the transcription factor human
upstream binding factor. Biochemistry. 37(46): 16307-16315.
94
Biographical Note and Acknowledgements.
I was born on February 21st, 1973 into a Jewish family in the city of Kishinev, the
capital of the Republic of Moldova, in what then was the Soviet Union. Today, you would
not find this city on a map: its name has been changed back to the original Romanian
spelling Chisinau. My parents, Nelly and Vladimir, were a biologist and a mathematician,
respectively. After their divorce, for a number of years I lived in Kishinev with my
grandmother, Luyba, an incredible chemistry teacher and a remarkable, amazing woman,
with whom we are very close to this day. At the age of 14, I moved to Moscow to live with
my mother. In Moscow, I graduated from high school and attended the first two years of
college at Moscow Institute of Fine Chemical Technology, in the Department of
Biotechnology and Bioorganic Synthesis.
At the age of 18, I immigrated to America alone, temporarily leaving my family
behind. Many deep thanks to my much-loved mother, Nelly, who had the courage to let me
go alone while she herself could not join me at the time. That being the year of 1991, the
year of the most turmoil in changing Russia, it was unclear at the time whether she would
ever be able to follow me, or would become trapped behind the new Iron Curtain.
Fortunately, things changed for the better, and a few years later, the rest of my family also
came to America for good.
I came to live to Boston in December 1991, and in September 1992 I was already
enrolled as a transfer student at Brandeis University, in the Department of Biochemistry. I
graduated
from Brandeis
in May,
1994. Soon, I was employed
at Hybridon,
a
pharmaceutical company that directed its efforts towards drug development based on antisense DNA technology. I was very fortunate to work with wonderful people in a unique
environment created by Dr. Aaron Cohen, in an atmosphere that both emphasized creative
research and nurtured good relationships between lab members. My coworkers became
very close friends. Eva Budman, a warm, sincere and wise woman, was a wonderful
companion both at work and outside the lab. Alexei Belenky, on top of being my scientific
adviser, was also the best friend. I cannot emphasize enough how much I owe to Alexei,
how much I have learned and continue to learn from him, both in my professional and
personal life. These friendships continued well beyond our work at Hybridon. I sought
advice from Alexei for a long time after leaving Hybridon, and I still seek out and treasure
his wise word. The Hybridon gang still gathers every year at Alexei's house. Many thanks
to Dr. Aaron Cohen, the chief of our analytical lab, who brought all of us together.
Sadly, in a few years Hybridon was forced to undergo substantial downsizing; our
analytical lab was practically eliminated, and most of us ended up at various research labs
scattered around Massachusetts. I was fortunate to land a position at the Genome
Center/Whitehead Institute, MIT, where I worked with then post-doc and now professor at
MGH/Harvard and the Broad Institute, Dr. David Altshuler. David, an extremely bright
and energetic young man, was also an interesting and enthusiastic work companion. Work
with him was always highly charged and exciting, his enthusiasm was contagious - it was a
pleasure and honor to work together.
95
I hadn't been at the Whitehead long; soon I was accepted as a graduate student at
MIT, in what then was the Division of Toxicology, and in the summer of 1998 I started my
graduate research, working on the Transcription Factor Hijacking hypothesis. I am very
much thankful to the people who encouraged me to go to graduate school - most of all,
Alexei Belenky, who never allowed me to think that my professional life could do well
otherwise. I am indeed very thankful to my adviser John Essigmann, who accepted me as a
graduate student in his lab and gave me this unique chance to become an MIT graduate.
My scientific life at MIT has not been straightforward. I went through both exciting
and extremely difficult times. But, as much as I wish more of my work would have turned
out more publishable results, I have to admit that I probably learned more from having to
work on multiple projects that were frequently unrewarding. I very much want to thank all
of my family, especially my mother and my grandmother, for their understanding and
wisdom. They encouraged me when I needed it, but they also were always ready to accept
things as they were, without disappointment. I also would like to thank my father for being
proud of me -- I am not sure what it is that he is proud of, but it flatters me anyway.
In the meantime, my personal life was booming. In my second year at MIT, I got
married to Leonid Perlov, a long-time acquaintance, who became the best and the closest
friend in the world. Every day of our marriage, my husband kept begging for children, and
I kept saying that first I needed two papers. Well, I still don't have the two papers, but we
are expecting our second child soon. I must say, I am glad my husband got it his way.
Aside from many things that I could thank my husband for, I am very much
indebted to him for all the support and encouragement he gave me during my graduate
studies. His compassion and sense of humor helped release stress when it seemed like it
was just too much. And, how could I get through writing my thesis had he not put aside
much of his interests and had he not made my thesis a priority. Here, I also want to thank
my incredible in-laws, Margarita and Michael Perlov. I am blessed with a truly wonderful
mother-in-law; I really don't know how I would have managed without her help, without
her taking care of our baby while I was working on my thesis. Needless to say, my long
hours writing translated into her long hours babysitting, and it's hard to imagine where
things would be had she not been there for me. Many thanks also go to our nanny, Galina
Alekseyevna. She has been like a grandmother to our son Yakov, and she gave me the best
thing a babysitter can give a mother: peace of mind while the baby is in someone else's
hands. We have been very fortunate to have such an awesome nanny, and indeed her
contribution to my thesis writing is also invaluable.
I cannot thank enough my wonderful mother who came a long way to help me
while I was finishing my thesis. Her thoughtful presence was magical and peaceful, and
everything she did in her kind and selfless manner has always been just so right, so wise,
so timely, and so helpful. Deep thanks to her husband Lev and to my half-sister Esther,
now 12, whose understanding and patience made Nelly's commitment possible.
96
Finally, I would like to thank everyone who contributed, directly or indirectly, to
the work presented in this thesis. I admire the courage and open-mindedness of my adviser,
John Essigmann; I believe these are the qualities that allowed this work to bear fruit. At the
time, accepting the "tetracycline hypothesis" meant going out of one's way of thinking;
John went this long way towards acknowledging science that was so new to our lab.
I am deeply indebted to Martin Marinus, whose contribution to this work is
invaluable. Martin's broad expertise and assistance with technical issues greatly helped
propel this project; however, I am even more thankful to him for giving me the confidence
to proceed. Many thanks to Jim Delaney, whose technical expertise and willingness to help
always amazed me; it was always Jim to whom I turned first for help and advice. I would
like to thank Jennifer Robbins for her companionship, for the moral support she gave me
throughout this work, and for the times that we simply enjoyed chatting with each other.
Also thanks go to Kaushik Mitra, who is no longer at in the lab, but who had been a good
friend and who helped me get through some of the most difficult times. I wish to thank all
of my current and former lab members for the friendly and non-competitive atmosphere in
the lab - and indeed, many thanks to John Essigmann for knowing how to nurture this rare
spirit.
97