| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of Clinical Medicine and
* Biochemistry, Trinity College, Dublin 2, Ireland
1Correspondence: Vitamin Research Laboratory, Sir Patrick Duns Trinity College Laboratory, Central Pathology, St. James’s Hospital, Dublin 8, Ireland. E-mail: jmcprtln{at}tcd.ie
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: folate catabolism • folic acid • antifolate • dihydrofolate reductase
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) of the diaminopyrimidines, including
trimethoprim (TMP). These are a group of potent antibiotics that
exhibit a high degree of specificity for bacterial dihydrofolate
reductase due to structural differences between the bacterial and
mammalian enzymes. It was for their work in this sphere that Hitchings
and Elion were awarded the Nobel Prize for Medicine in 1988
(2)
.
Folate coenzymes are required for the transfer of one-carbon units in
the biosynthesis of purines and pyrimidines, in amino acid
interconversions, and for the provision of methyl groups. The critical
function of dihydrofolate reductase (DHFR) in restoring
tetrahydrofolate (THF) concentrations (Fig. 1A
) after thymidylate biosynthesis established the basis for
antifolate drug therapy not only in infection, but also in cancer,
inflammatory, and autoimmune conditions (3)
. Though the
consequences of THF depletion (Fig. 1B
) include the
cessation in biosynthesis of purines and protein, the metabolic block
of greatest severity in animals and bacteria is thought to be the
interruption of thymidylate synthesis, so-called ‘thymineless death’
(4)
.
|
The 2,4-diaminopyrimidine derivative TMP is the preferred inhibitor of
bacterial dihydrofolate reductase in clinical and veterinary use. This
is because of its broad spectrum of antibacterial effects, its high
degree of selectivity for the bacterial enzyme, and its suitable
pharmacodynamic and pharmacokinetic properties (5)
. In
combination with a sulfonamide such as sulfamethoxazole, which inhibits
folate biosynthesis, the antibacterial spectrum of TMP was extended,
greater potency was achievable with lower doses resulting in fewer side
effects, and bactericidal power was increased compared to the more
bacteriostatic activity of the single components. With or
without a sulfonamide, however, a feature of antifolate action is the
cessation of nucleotide biosynthesis coincident with disruption of
folate metabolism.
The purpose of this study was to examine the hypothesis that antifolates such as TMP cause their effect through accumulation of dihydrofolate (DHF) and the consequent diminution of the tetrahydrofolate pool that occurs with inhibition of dihydrofolate reductase.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was
purchased as a lyophilized vial from the National Center for Industrial
and Marine Bacteria, Aberdeen, Scotland. Trimethoprim (Monotrim, BP 20
mg/ml) was from Solvay Healthcare, Southampton, U.K. Columbia agar (Lab
M, Bury, U.K.) supplemented with 7% horse blood was prepared by the
Microbiology Department, St. James’s Hospital, Ireland. Rat plasma was
prepared from blood collected in EDTA tubes and stored in 100 µl
aliquots at -70°C. Glucose media (8)
was composed of
(g/l): NaCl, 2.94; NH4Cl, 2.66;
KH2PO4, 3.4; CaCl, 7.4 x 10-3; MgSO4, 0.25;
glucose, 2.0; pH7.2. [3H]-para-aminoglutamic
acid ([3H]-pABA, 40.9 Ci/mmol) was purchased
from Moravek Biochemicals (Brea, Calif.). Folate monoglutamate
standards (DHF, THF, 5-CHO-THF, 10-CHO-THF,
5-CH3-THF,
5,10-CH2-THF, 5,10-CH=THF) were donated by
Eprova, Schaffhausen, Switzerland; para-aminobenzoylpolyglutamate
(pABGlun, n=1,3,5,7) standards
were purchased from Schircks Laboratories (Jona, Switzerland). All
other chemicals, including pABA and folic acid, were of analytical
grade or greater (Sigma, Poole, U.K.).
A schematic flow diagram of the experimental procedure is shown in
Fig. 2
.
|
Preparation of E. coli stocks on cryo-preserve
beads
The lyophilized E. coli was suspended in 400 µl
glucose media and 200 µl was used to inoculate a Columbia blood agar
incubated at 37°C. The resulting culture was stored at -70°C on
Protect beads (Technical Services Consultants Ltd., Lancashire, U.K.).
Bacterial cultures were reconstituted from -70°C, subcultured on
Columbia blood agar as required and checked for purity.
Determination of bacterial density
Turbidity was measured at 590 nm on 300 µl samples in a
96-well plate using a Multiskan Plus plate reader (Lab Systems,
Helsinki, Finland). Bacterial colonies from a Columbia blood agar were
suspended in glucose media to give a stock suspension with a turbidity
of 0.642 A.U. From this a dilution series was prepared from 0 to 100%
of the stock suspension in 10 equal intervals and the turbidity was
measured. The viable cell number in the stock suspension was determined
by making 1 in 104, 1 in
106, and 1 in 108 dilutions
of the initial bacterial suspension in glucose media. 100 µl of each
dilution was used to inoculate duplicate blood agar plates. Plates were
incubated overnight, colonies were counted and the bacterial density
(cells/ml) of the initial suspension was calculated. The bacteria
density of each standard dilution was calculated and a plot relating
bacterial density to turbidity was used to determine bacterial density
of experimental cultures.
Optimization of incubation periods for the first, second, and
third growth
All experimental procedures were performed on bacteria in log
phase growth. Approximately 50 E. coli colonies from a
Columbia blood agar were suspended in 500 µl of glucose media and
diluted so that 300 µl had a turbidity of 1.00 A.U. A 100 µl
aliquot of this suspension was inoculated into 5 ml of glucose media
and incubated at 37°C in an orbital incubator at 80 rpm. After
incubating for 11 h, 250 µl (460x106
(±20x106) cells) was inoculated into 5 ml of
glucose media containing 300 pmol [3H]-pABA.
After 9 h the bacteria was harvested by centrifugation at 4500
g for 8 min, and the pellet was resuspended in 2 ml glucose
media and repelleted. The washed pellet was resuspended in 1 ml glucose
media and 200 µl (1.8x109
(±0.1x109) cells) was inoculated into 5 ml of
glucose media containing the required concentration of TMP.
Determination of optimum concentration of [3H]-pABA
to be used for labeling bacteria during second growth
The first growth was incubated for 11 h before 250 µl was
inoculated into tubes containing 5 ml glucose media and 100, 200, 300,
or 400 pmol [3H]-pABA. After incubating for
9 h, the E. coli was harvested and the radioactive
content of each pellet was expressed as a percentage of total
[3H]-pABA added. 300 pmol
[3H]-pABA was found to be the highest
concentration giving maximum percentage radiolabel incorporation.
Determination of the minimum inhibitory TMP concentration
A 200 µl aliquot of the washed second growth pellet was added
to a duplicate series of tubes containing TMP serially diluted from 200
µg/ml to 200 ng/ml. The tubes were incubated for 4 h, when
bacterial density was determined and plotted against inhibitor
concentration. From this, the minimum concentration of TMP giving
maximum inhibition was determined.
Harvesting of bacteria, sample preparation, and high-performance
liquid chromatography (HPLC) analysis
TMP-treated bacteria were harvested at timed incubation periods
and the pellet was resuspended in 500 µl HEPES (100 mM, pH 7.3)
containing ß-mercaptoethanol (100 mM), and sodium ascorbate (2%
w/v). The bacteria were lysed by sonication on ice for 20 s x 3
(Sonifier 450, Branson Ultrasonics, Danbury, Conn.) fitted with a
microtip. Samples were deconjugated by incubating for 45 min. at 37°C
with 100 µl rat plasma and chromatographed on a 3 µ Kingsorb
(4.6x150 mm) ODS-2 column (Phenomenex). A SecurityGuard precolumn
module (Phenomenex) containing a disposable ODS insert (3x4 mm,
Phenomenex) was attached between the injection port and the column. The
column was eluted isocratically at 35°C with 5 mM Pic A (Waters,
Milford, Mass.) in 0.1M sodium acetate (pH 6.0), containing methanol at
16% (v/v) at 1 ml/min. Fractions were collected at 0.5 min intervals
using a Frac-100 fraction collector (Pharmacia Biotech, Uppsala,
Sweden), dissolved in EcoLite scintillation fluid (ICN, Costa Mesa,
Calif.) and counted on a 1500 Tricarb scintillation counter (Packard
Canberra, Reading, U.K.).
The magnitude of each peak was expressed as a percentage of total radioactivity injected onto the column and the retention times compared to those of folate standards detected at 280 nm on a SPD-6A UV spectrophotometer detector (Shimadzu, Kyoto, Japan).
Determination of DHF stability during folate extraction and HPLC
analysis
Third growth pellets were spiked with known amounts of standard
DHF, extracted, deconjugated, and prepared as usual for HPLC analysis.
The absorbance response measured at 280 nm was calculated as a
percentage of the original spike.
Identification of metabolically produced folic acid
Extracts from TMP-treated samples (24 h) were prepared and
chromatographed as described and the fractions corresponding to the
folic acid peak were retained and pooled. To 100 µl of this pool was
added 40 µl HEPES buffer (1M), 80 µl sodium ascorbate (10% w/v),
40 µl ß-mercaptoethanol (1M), and 40 µl NADPH (1M). The volume
was made up to 400 µl with H2O and divided in
two. To one half was added 2 µl bovine liver dihydrofolate reductase
(1 unit/20 µl, Sigma); the other was left untreated. Both samples
were incubated for 15 min at 25°C and chromatographed by HPLC.
[3H]-Folic acid was identified by its coelution
with an authentic standard and through enzymatic conversion by bovine
liver dihydrofolate reductase to tetrahydrofolate and coelution with
authentic standard on HPLC. The chromatographic profile of the
untreated sample remained unchanged.
Metabolic reutilization of folic acid upon removal of inhibitor
Two tubes each containing 5 ml of glucose media and TMP were
inoculated with 200 µl of the second growth pellet. After 2 h
incubation, both cultures were mixed and redivided before harvesting.
One pellet was prepared for HPLC analysis; the other was resuspended in
8 ml glucose media and centrifuged. The washed pellet was resuspended
in 5 ml glucose media and incubated for a further 4 h, then
harvested and prepared for HPLC analysis.
Identification of folic acid and pABGlu as polyglutamates
TMP-treated cultures were harvested at 4 h and extracted
for folic acid and pABGlu (see Fig. 3
). Without conjugase treatment, this extract was divided in two. A 200
µl aliquot of one half was incubated for 10 min with 20 µl HCl (6M)
and 10 mg zinc dust. The other half was left untreated. Both samples
were chromatographed on a µBondapak (3.9 mmx30 cm)
C18 column (Waters). A precolumn module
containing a disposable C18 insert (RCSS Guard
Pak, Waters) was attached between the injection port and the column.
The column was eluted isocratically at 2 ml/min with citrate-phosphate
(0.1 mM, pH 4.0): acetic acid (1%): methanol (80: 14: 6). Fractions
were collected at 0.5 min intervals and assayed for radioactivity by
scintillation counting. Authentic pABA-polyglutamates (n=1,
3, 5, 7) and folate monoglutamate standards were chromatographed to
establish their elution position on the chromatogram.
|
Determination of total folate in E. coli
Bacterial cultures were harvested, the cell pellets were
resuspended in 500 µl ascorbic acid (1%w/v), ß-mercaptoethanol (10
mM), and lysed by sonication. Extracts were deconjugated by incubation
with fresh human serum (100 µl) for 1 h at 37°C. Serial
dilutions in triplicate (from 1 in 4 to 1 in 46)
were made of the deconjugated sample in ascorbic acid (1%) and assayed
by microbiological assay (9)
. The serum folate
concentration was determined and subtracted to give the bacterial
folate value.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). However, treatment with 2 µg/ml TMP, the minimum concentration of
TMP giving maximum inhibition (98% growth inhibition), resulted in the
rapid but transient accumulation of dihydrofolate. Thus, within 10 min
of TMP treatment the radiolabeled THF cofactor content was reduced by
two-thirds, half the folate was in the form of dihydrofolate, and a
significant fraction had catabolized to pABGlu. Within 4 h of
treatment, however, the distribution of intracellular radiolabel
stabilized, with cells depleted not only of THF coenzymes but also of
dihydrofolate. Assay of total cellular folate by microbiological assay
(Table 2
) showed a substantial reduction in folate in TMP-treated cells at
4 h.
|
|
Figure 3
shows the HPLC assay and [3H]-folate
distribution (inset) before and after TMP removal from the bacterial
incubation medium. Removal of TMP inhibition resulted in a resumption
of cell division over a 4 h incubation period (data not shown).
Within that time the folic acid content of the cells was reduced to
one-third the TMP treatment level. The redistributed radiolabel was
found mainly in chromatography peaks corresponding to THF and
10-CHO-THF. Meanwhile, there was no change in the proportion of
radiolabel accruing to both pABGlu and DHF.
As demonstrated in Table 1
, incubation with TMP for 4 h
produced only two radiolabeled molecular species: pABGlu and folic
acid. Chromatography of cell extracts untreated with conjugase
(Fig. 4
) showed that the pABGlu portion, i.e., 57.1% (Fig. 4
, inset),
consisted mainly of pABA-polyglutamates(5–8).
(Chromatography of conjugase-treated or untreated control samples
showed that less than 5% of the radiolabel had eluted under the same
conditions; results not shown.) The first intact folate standard (THF)
did not elute until 29 min, by which time the pABGlu polyglutamates
(n=1–8) had eluted.
|
Reductive cleavage of the cellular folates showed not only pABGlu-derived polyglutamates, but also those arising from the cleavage of the only intact folate species present, namely, folic acid.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hitherto, the disruption of folate metabolism due to DHFR inhibition
has been postulated to be due to the accumulation of DHF (Fig. 1B
), a folate cofactor dependent on reduction to THF for its
further participation in the one-carbon cycle. In this paper we provide
evidence that the commonly observed attrition of the total folate pool
(10)
upon antifolate treatment is related to catabolism of
folate. Catabolism and depletion of folate have been demonstrated to
occur in rats (11
, 12)
and in humans (13)
and
are accelerated under conditions of rapid cell division. Ineluctable
losses of folate in this fashion may reflect the requirement for
constant replenishment of folate in those organisms that cannot
synthesize their own. In this study, DHF was shown to accumulate
acutely upon inhibition of DHFR but to diminish to undetectable levels
with time as the concentration of pABGlu, the product of C9-N10 bond
cleavage, correspondingly increased. Catabolism has been postulated to
occur enzymatically (P. Stover, personal communication), but it is also
possible that the parallel effects of antifolates on disruption of the
cell redox state and decrease in intracellular pH (14
, 15)
may contribute to the spontaneous cleavage of labile folates such as
DHF in the cell. Removal of DHF through scission to inactive
catabolites may enhance further the effect of TMP through decreased
competition for DHFR binding. These results also call into question the
proposed ‘thymineless death’ model (4)
of antifolate
action, which suggests that the folate cofactor most vulnerable would
be that involved in thymine synthesis, which is
5,10-methylenetetrahydrofolate. This is in contrast to our finding in
E. coli of a more generalized loss of folate, suggesting
disruption of all folate-dependent pathways. This was reflected not
only by the radiolabel loss in treated cells, but also in reduction of
total folate as shown in Table 2
.
From chromatography of cell extracts untreated with conjugase, we
demonstrated that folate scission caused polyglutamated pABA of up to
eight glutamate residues to accumulate in the cell. The greater
efficiency of polyglutamate forms of folate in metabolic reactions is
believed to relate to enhanced enzyme binding compared to folate
monoglutamates (16)
, suggesting that polyglutamated
catabolites present an additional disruption to folate-dependent
reactions. Figure 4
also demonstrates that the folic acid produced in
TMP-treated cells is present in polyglutamate forms.
The microorganism used in these studies, E. coli (NCIMB
8879), displayed another characteristic response to inhibitor
insult—the ability to store folate in its most stable chemical form,
namely, folic acid. Within 4 h of TMP treatment, it constituted
the only intact folate species in the cell. Furthermore, with removal
of the inhibitor, it was shown that the radiolabel accruing to folic
acid was distributed among other reduced folate coenzymes. This was
coincident with a resumption of cell proliferation. Folic acid has
previously been shown to occur to a minor extent in nature due to
nonspecific chemical oxidation of reduced folates during the extraction
process (17)
. Here, however, we show that not only does it
occur extensively, but it also exhibits metabolic activity through
reconversion to reduced folates. The accumulation of folic acid may
have physiological consequences for bacterial survival and
recolonization on premature termination of treatment of infection. This
duel effect of trimethoprim may reflect just one part of a spectrum of
microbial response, from total catabolism on the one hand to complete
sequestration of folates, as folic acid, on the other. The relative
degree of catabolism or oxidation to folic acid might vary depending on
such factors as species of bacteria, the redox state of the cell, or
the therapeutic regime of antifolate. These considerations may even
determine the response to be either bacteriostatic through oxidation of
DHF to folic acid or bactericidal through folate catabolism. Figure 1
illustrates schematically the contrast between the established
postulate of antifolate action whereby the disruption of folate
metabolism is due to DHF accumulation and the model of catabolism
and/or oxidation to folic acid we describe here.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 13, 1999. Accepted for publication June 8, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. L. Carter, L. Jager, L. Gardner, C. C. Hall, S. Willis, and J. M. Green Escherichia coli abg Genes Enable Uptake and Cleavage of the Folate Catabolite p-Aminobenzoyl-Glutamate J. Bacteriol., May 1, 2007; 189(9): 3329 - 3334. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. C. Basset, E. P. Quinlivan, J. F. Gregory III, and A. D. Hanson Folate Synthesis and Metabolism in Plants and Prospects For Biofortification Crop Sci., January 31, 2005; 45(2): 449 - 453. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and Zn/HCl-cleaved (•)
extracts of TMP-treated (2 µg/ml) E. coli incubated
for 4 h. Cell extracts were prepared and chromatographed as
described in the text. Inset: tabulation of the polyglutamate
distribution as a percentage of the total radiolabel chromatographed
for pABGlu (uncleaved extract) and pABGlu + FA (Zn-cleaved extract).
The percentage accruing to folic acid was found by subtraction. Peak
numbers in main diagram and numbers (n) in inset table refer to
polyglutamate chain length; Vo= column void volume.