- Email: [email protected]
Studies on the acid-soluble nucleotide pool in thymine-requiring mutants of Escherichia coli during thymine starvation
lO4
BIOCHIMICA ET BIOPHYSICA ACTA
BBA 95480
STUDIES ON T H E ACID-SOLUBLE NUCLEOTIDE POOL IN THYMINEREQUIRING MUTANTS OF E S C H E R I C H I A COLI DURING THYMINE STARVATION III. ON THE REGULATION OF T H E DEOXYADENOSINE TRIPHOSPHATE AND DEOXYCYTIDINE TRIPHOSPHATE POOLS OF E S C H E R I C H I A COLI
JAN NEUHARD
Institute o/Biological Chemistry, University of Copenhagen, Copenhagen (Denmark) (Received March 3rd, 1966)
SUMMARY
Changes in the size of the deoxyribonucleoside triphosphate pool have been investigated in mutants of Escherichia coli subjected to different conditions of thymine restriction. The following results were obtained. I. Five different thymine-requiring mutants of E. coti, after removal of thymine from their growth media, all showed an immediate rise in dATP content reaching a maximum after about 60 min. 2. In two mutants (both E. coli 15 T - mutants) the dCTP pool increased linearly IO- to I5-fold after a lag of about 30 min. In these two mutants the maximal amount of dATP accumulated was only half of that reached by the other three. 3. Thymine prototrophs of E. coli also accumulated dATP when deprived of their intracellular thymidine nucleotides by addition of 5-fluorodeoxyuridine. 4. Inhibition of DNA synthesis in E. coli 15 T - A - U - by novobiocin in the presence of thymine did not result in any significant changes in the deoxyribonucleoside triphosphate pools. The results are discussed in relation to the regulation of deoxyribonucleotide synthesis.
INTRODUCTION
During the last few years work on the thymidine kinases (ATP: thymidine 5'-phosphotransferase, EC 2.7.1.21 ) from widely different sources have shown that dTTP is a specific feedback inhibitor of these enzymes1-5 in vitro. The possibility that thymidine nucleotides might have other regulatory functions in the deoxyribonucleotide metabolism of bacteria is supported by experiments with thymine-requiring Abbreviation: dFU,5-fluoro-2'-deoxyuridine.
Biochim. Biophys. Acta, 129 (I966) lO4-115
B!UCLEOTIDE POOLS IN E. coli
Io5
Escherichia coli auxotrophs. Thus, starvation of E. coli 7oV3-462 for thymine results in induction of high amounts of thymidine phosphorylase (thymidine :orthophosphate deoxyribosyltransferase, EC 2.4.2.4) paralleled by a substantial excretion of free deoxyribose into the growth mediums. Blsw~,s, HARDY AND BECK7 working with E. coli 15 T- have demonstrated that thymine deprivation causes a 6- to 8-fold increase in the cellular content of CDP reductase, the enzyme responsible for the conversion of CDP to dCDP. As a corresponding increase could be brought about by addition of 5-fhioro-2'-deoxyuridine (dFU) to the thymine prototroph E. coli B, they concluded that a thymidine nucleotide was acting as a corepressor for the ribonucleoside diphosphate reductase system in E. coll. In previous experiments on changes in the deoxyribonucleoside triphosphate pools of E. coli 15 T-A-U- (requiring thymine, arginine and uracil) we found~-1° that immediately after thymine was removed from the growth medium the dATP content of the cells rose, reaching a maximum value in about 6o min. The pool of dCTP, however, remained constant for the first 3° min of thymine starvation and then increased linearly, reaching a io- to IS-fold value after 9° min. While the increase in dATP was independent of concomitant protein or RNA synthesis, the dCTP content only rose when protein synthesis occurred simultaneouslyTM. The dGTP pool showed no changes under any conditions of thymine starvation investigated. In the present work, results will be presented that strengthen our earlier suggestionS, 1° that a thymidine nucleotide is involved directly in the regulation of the pool size of dATP in E. coll. Moreover, they seem to indicate that the ratio dATP/ATP plays an important role in the regulation of the pool size of dCTP. This makes us believe that the controlling effect of this ratio on the activity of the CDP reductase from E. coli B, recently shown by HOLMGREN, REICHARD AND THELANDER 11 to operate in vitro, may also be of importance in vivo.
MATERIALS
Pyrimidines, amino acids, ribo- and deoxyribonucleoside 5'-phosphates were .obtained from Sigma Chemical Company, St. Louis (U.S.A.). dFU was a gift from Hoffman-La Roche, Inc. (Switzerland) through the courtesy of Wmrum and Co. (Copenhagen). The sodium salt of novobiocin was obtained as Albamycin capsules from Upjohn (Belgium). Poly(ethyleneimine) (Polymin P) was a gift from Badische Anilin- und SodaFabrik, Ludwigshafen (Germany); cellulose powder for thin-layer chromatography .(MN 3o0) was from Macherey, Nagel and Co., Diiren (Germany), and methanol .(technical grade, containing less than 0.25 % water) from AKI (Copenhagen). Carrier-free [a*P~orthophosphate and ~2-14Clthymine were obtained from The Radiochemical Centre, Amersham (England).
Bacterial strains and growth conditions
The basal medium described previously8with glucose (o.2 %) as a carbon source was used throughout. The bacterial strains employed were all obtained on agar slants from the Institute of Microbiology, University of Copenhagen, through the Biochim. Biophys. Acta, 129 (1966) lO4-115
lO6
j. NEUHARD
courtesy of Dr. N. FILL. Table I shows the strains used (A) and their growth rates (C) in basal medium supphed with glucose and the necessary nutritional requirements (]3). Increase in cell mass was followed at 450 m/~ in a Zeiss M4 QIII spectrophotometer (I-cm light path). I ml bacterial culture with an absorbance at 450 m# of I.OOO corresponds to approx. 4" lO8 cells or 0.2 mg dry weight.
TABLE
I
BACTERIAL STRAINS USED IN THE P R E S E N T WORK
(A) Bacterial strain
(B) (C) Nutritional Growth rate* requirements(itg[ml ) (doublings per h)
(D) Re]eren6e
E. coli 15, T -
thymine, 2
1. 4
B A R N ~ R AND COHEN 12
E. coli 15, T - A - U -
thymine, 2 arginine, 20 uracil, i o
1.2-1. 3
I~ANAZIR et
E. coli KI2(~,), T -
t h y m i n e , 20
0.9
K O R N AND W E I S S B A C H 14
E. coli B/r, T - H -
t h y m i n e , 20 h i s t i d i n e , 5o
1. 3
E. 6oli B, M -
m e t h i o n i n e , 5o
1.2-1. 3
K U R L A N D AND 1V~AALOE15
E. coli B, M - T -
m e t h i o n i n e , 50 t h y m i n e , 2o
1.3-1. 4
F R I E S E N AND MAALOE le
al. a3
* I n b a s a l m e d i u m s u p p l i e d w i t h g l u c o s e (0.2 %) a n d n e c e s s a r y n u t r i t i o n a l r e q u i r e m e n t s .
METHODS
Determination o/ DNA synthesis Cells growing exponentially were harvested on 9-cm Millipore filters (Membran Filter, G6ttingen (Germany)), washed with basal medium and resuspended in a medium containing ~2-x4C]thymine (specific activity: 6.5/*C/#mole). At Io-min intervals I-ml samples were added to tubes containing 4 ml ice-cold 6 % trichloroacetic acid. The precipitates were collected on 27-ram Millipore filters, and washed three times with ice-cold 5 % trichloroacetic acid containing Io/~g [l~C]thymine per rnl. Subsequently the filters were glued to aluminum planchets, dried and counted in a Friesecke-Hoepfner thin-window gas-flow counter.
Quantitative deter~nination o[ acid-soluble nucleotides a. Labelling o/the nucleotides. In all experiments where the acid-soluble nucleotide pools were determined, cells labelled for at least 2 generations in complete medium containing [aZP]olthophosphate (specific activity: i-3/A2/k~ole ) were used. Subsequent incubations of the cells were performed in the desLved medinm containing [s~p]orthophosphate of the same specific activity as before. Biochim. Biophys. Acta, 129 (1966) lO4-115
NUCLEOTIDE POOLS IN E. coli
lO7
b. Preparation o/extracts. Samples of 5 ml were harvested by suction filtration on Millipore filters and treated as previously described1°. Before and after ether extraction and after lyophilization, aliquots were plated on planchets and counted in order to determine the recoveries of total ~*P radioactivity. While the ether extraction gave recoveries of 95-1oo %, the recoveries after lyophilization varied between 60 % and 9° %. All figures given in the following have been corrected for the corresponding losses. c. Fractionation o/ cell extracts. Separation of the different nucleotides was performed in two dimensions on poly(ethyleneimine)-impreguated cellulose layers on 20 cm ×2o cm glass plates. In most cases a stepwise elution was used in both dimensionsXT,is. Triphosphates: Separation of complex nucleoside triphosphate mixtures has previously been described1°. This procedure was followed except for minor modifications of the solvents. The solvents used throughout the present work were, in the first dimension: 2 M LiC1-2 M acetic acid (I :I, v/v) (3 cm) followed by 2.5 M LiC1-2 M acetic acid (I:I, v/v) (15 cm); and in the second dimension: 2.5 M ammonium acetate- 3.6 % boric acid (pH 7.o with ammonia) (5 cm) followed by 3-5 M ammonium acetate-5 % boric acid (pH 7.0 with ammonia) (I 5 cm). Diphosphates: A good separation of the different nucleoside diphosphates is achieved by modifying the solvents used for the triphosphate separation as follows: First dimension, 3 M LiCI-I M acetic acid (1:9, v/v) (2 cm) followed by 1.6 M LiC1-2 M acetic acid (I:I, v/v) (15 cm); second dimension, I M ammonium acetate2.5 % boric acid (pH 7.o with ammonia) (I cm) followed by 2 M ammonium acetate5 % boric acid (pH 7.0 with ammonia) (15 cm). Monophosphates: For the separation of nucleoside monophosphates and nucleotide-sugars, the method described by RANDERATHAND RANDERATH2° was used. To the start-spots of all chromatograms unlabelled nucleotides were added as markers. Compounds were located on the chromatograms with a Mineralight lamp and by autoradiography (exposure time 24 h on Kodirex X-ray film). There was always complete correspondance between the ultraviolet-detected spots (added markers) and the radioactive spots. Quantitative determinations of the radioactive compounds from the chromatograms were carried out as described previously1°,19.
RESULTS Effects o/ thymine starvation in different thymine-requiring strains Very distinct changes in the nucleoside triphosphate pools have previously been reported to take place 1° when thymine is removed from the growth medium of the triple auxotroph E. coli 15 T-A-U-. In order to determine if these changes are a general phenomena in cells of E. coli deficient in their thymidine nucleotide pools, a number of other thymine auxotrophs were investigated for changes in their nucleoside triphosphate pools after withdrawal of thymine from their growth media. The strains used are listed in Table I. Table II gives the pool sizes of the eight nucleoside triphosphates in the different strains, when growing exponentially in complete media. Figs. I and 2 give the sizes of the nucleoside triphosphate pools of two different Biochim. Biophys. Acta, 129 (1966) lO4-115
108
J. NEUHARD
TABLE II POOL SIZES OF THE NUCLEOSIDE TRIPHOSPHATES IN CELLS OF D I F F E R E N T STRAINS OF E. coli G R O W I N G EXPONENTIALLY IN COMPLETE MEDIUM CONTAINING [32P]ORTHOPHOSPHATE
F o r all d e t e r m i n a t i o n s the procedure described in METHODS was used.
Bacterial strain
E. E. E. E. E. E.
coli coli coli coli coli coli
Number o~
15, T 15, T - A - U KI2(~), T B/r, T - H B, MB, M - T -
I~moles per g dry weight
experiments G T P
ATP
CTP
UTP
dGTP
dATP
dCTP
dTTP
I 4 1 I 2 8
3-7 3.4 5.4 6.0 4.8 5.3
o.9 o.9 1.4 1. 7 1.2 1.6
1.9 2.o 2.2 3.2 2.2 3.1
-o.17 o.16 o.2o o.17 0.22
o.18 o.24 0.35 0.28 o.25 0.28
o.22 o.28 0.29 o.37 o.29 o.39
o.32 o.41 0.4 ° o.35 o.23 o.41
1.4 1.3 1.9 1. 7 1. 9 1.7
2s[
A
7.0 .E 6.C ._o~
A
B
- 2A~ I
~'
~ 12
A ATP "~
•
/
e ~
;
1-0[~
I
dATP
UTP
o.~p ~O
T
/
ATP
•
/
~,-0
o..l /
o t'.o 2'.o 1i0 Retative time after thymine removal
I
2.0
I
dATP
~o
°~'~o
t'o
2~o
I
ti0
Relative time after thymine removal
z:0
Fig. I. A m o u n t s of t h e four ribonucleoside t r i p h o s p h a t e s in (A), E. coli B M - T - (doubling time: 42 min) a n d (B), E. coli 15 T - A - U - (doubling time: 47 min) after r e m o v a l of t h y m i n e f r o m t h e media. 0 - 0 , GTP; O - © , CTP; A - A , U T P ; A-LS, ATP. Abscissa: relative time, using t h e doubling t i m e of each m u t a n t as unity. Fig. 2. A m o u n t s of t h e different deoxyribonucleoside t r i p h o s p h a t e s in (A), E. coti B M - T (doubling time: 42 min) and (B), E. coli 15 T - A - U - (doubling time: 47 min) after r e m o v a l of t h y m i n e f r o m t h e media. 0 - 0 , dGTP; O - G , dCTP; LS-&, dATP. Abscissa: as Fig. i.
strains as a function of time after thymine removal. Figs. IA and 2A give the figures for the ribo- and deoxyribonucleoside triphosphates, respectively, from one experiment with E. coli B M - T - ; Figs. I B and 2B give the corresponding figures from an experiment with E. coli 15 T - A - U - . While no major changes in the ribonucleoside triphosphates occur (Figs. IA and IB), two things should be pointed out regarding dATP and dCTP (Figs. 2A and 2B). (I) The previously observed rise in the dATP pool of E. coli 15 T - A - U - (ref. IO) is much more pronounced in E. coti B M - T - and, in contrast, (2) no changes whatsoever are observed in the dCTP pool of this latter strain, while dCTP rises to a I2-fold value in E. coli 15 T - A - U - . As the absolute amounts of the different triphosphates v a r y significantly from one strain to another (see Table I I and Figs. I and 2), and the ribonucleoside triphosphates never showed major changes, we have chosen in most of the figures to Biochim. Biophys. Acta, 129 (I966) lO4-115
lO9
NUCLEOTIDE POOLS IN E . coli
give the ratios between corresponding deoxyribo- and ribonucleoside triphosphates in each sample, rather than the absolute amounts. By doing this, results with different mutants are directly comparable. The effects of thymine starvation on the deoxyribonucleoside triphosphate pools of five thymine-requiring E. coli strains are compared in Fig. 3. While all mutants behave similarly during the initial 75 % of the doubling times, i.e. marked increase in dATP/ATP and constant dCTP/CTP, their response to prolonged thymine starvation is of two types. In one, represented by the two E. eoli 15 T- strains, the dATP/ATP ratio levels off, while dCTP/CTP increases very steeply. In the other group, comprising the two E. coli B mutants, the dCTP/CTP ratio remains constant throughout, while the dATP/ATP ratio increases for another 75 % of a doubling time reaching a plateau about twice as high as that found in the E. coli 15 strains. The changes found in E. coli KI2 (~) T- indicate that it is an intermediary type, but we have not investigated this strain further in the present work. In none of the mutants, however, did the dGTP pool change significantly. A
1.5
o3IA
B
/
t.0
0.~
io
/ 0
~
T
~o.
+ /e,-
i
c:tD.
02
o:
1!
•
I('
FU
o!
Ot
0.1
B
/
o
!--.-.%-:'-I
1.0
2'.0
1.0
Retative time after thymine removal
2.0
I
1,0
~
2,0
I
i
1.0
J
2.0
Relative time after ,hift
Fig. 3. Changes in t h e pool sizes of (A), d A T P a n d (B), d C T P relative to A T P a n d CTP, respectively, in different t h y m i n e - r e q u i r i n g m u t a n t s of E. coli, s t a r v e d for t h y m i n e . A - ~ k , E. coli B M - T - ; A - A , E . coli B/r T - H - ; 0 - - © , E. coli i5 T - A - U - ; O - - Q , E . coli I 5 T - ; [ ] - [ ] , E . coli KI2(A) T - . Full line: m e a n values of t h e two E. coli B strains. Dotted line: m e a n values of t h e two E. coli 15 T - strains. Abscissa: as Fig. I. Growth rates of t h e different m u t a n t s are given in Table I. Fig. 4. Effects of different conditions of t h y m i n e restriction on t h e pool sizes of (A), d A T P a n d (B), d C T P relative to A T P a n d CTP, respectively, in different s t r a i n s of E. coll. lX-lX, E. coli B M - T - shifted to a m e d i u m lacking t h y m i n e ; G-(D, E. coli B M - shifted to a m e d i u m containing d F U (2o/,g]ml); 0 - 0 , E. coli B shifted to m e d i u m containing d F U (2o/~g]ml). Abscissa: as Fig. i.
E//ects o] d F U in thymine prototroph strains It has been shown by COHEN et al. zz, that dFU added to growing cultures of E. coli B caused a gradual cessation of growth (measured turbidometticaily), together with a rapid loss of viability, resembling the phenomenon called thymineless deathZZ,za. This bacteriocidM effect of dFU was readily explained, as it was shown that dFUMP, formed in the cells from dFU 2z, was a very potent inhibitor of thymidylate synthetaseZl,24,zS. Thus it is possible to induce a thymine requirement in a bacteria not otherwise requiting thymine, merely by adding dFU to the growth medium. Biockim. Biopkys. Acta, r29 (I966) xo4-zI5
IIO
j. NEUHARD
dFU (20 #g/ml) was added to exponentially growing cultures of the strains E. coli B and E. coti B M- (see Table I). At the same time a culture of E. coli B M - T - was transferred to a nmdium lacking thymine. The nucleoside triphosphate pools of the three strains were then followed with time. The results of such an experiment are shown in Fig. 4. It is immediately apparent that thymine deficiency induced by dFU also leads to an appreciable accumulation of dATP, though not as extensive as that seen in the corresponding thymine auxotrophic strain after removal of exogenous thymine. The ratio dCTP/CTP, however, did not change in any of the three strains. In addition to what is shown in Fig. 4, it was found that the two mutants receiving dFU accumulated high amounts of 5-fluorouridine triphosphate ~e (7.5 /,moles per g dry weight) (c/. Table II) concomitant with a reduction in their pools of UTP, CTP and dCTP. dTTP was always completely absent after addition of dFU. COHEN and co-workers ~z also showed that dFU was split by the bacteria (perhaps by thymidine phosphorylase ~7) to 5-fluorouracil, which could be transformed to 5-fluorouridine nucleotides. These may in turn be built into RNA and thereby cause the observed disturbances in growth. Our finding of high levels of 5-fluorouridine triphosphate after addition of dFU seems to support this explanation. However, simultaneous addition of uracil and dFU to growing cultures suppresses specifically the effects of 5-fluorouracil on the RNA metabolism, without initially affecting the loss in viable counts. Thus a culture of E. coli B M- was treated with dFU (20 #g/ml) in the presence of an equal amount of uracil. The changes with time in the absolute amounts of the deoxynucleoside triphosphates during the incubation is shown in Fig. 5 (full lines). Included in the figure are the curves showing the changes in dATP and dCTP of E. coli B M-T- (dashed lines) and E. coli 15 T - A - U - (dotted lines) taken from
2.4
/
2.2
ts
A
B
to)
!
g 1.4
~
0.3
/~xx
o,:
/
//
1.o
0.2
I.~
'? ,.jo
oo)
x
i.-o. o~.~ o
dATP 0.1
~
0.l
--
(i
q &
dTTP 1.0
~ ~e""--~-
2.0 1~0 Relative time after shift
o
dGTP 2.0
t .0
1:0
2.0 Retative time after shift
210
Fig. 5. - - , a m o u n t s of the four d e o x y r i b o n u c l e o s i d e triphosphates in /~. coli B M - a f t e r a shift from complete m e d i u m to a m e d i u m containing d F U (2o/*g/ml) and uracil ( i 8 $ , g / m l ) ; . . . . values from Fig. 2 B ; - - - , values from Fig. 2A. Abscissa: as Fig. i . Fig. 6. Changes in pool sizes of (A), d A T P and (B), d C T P relative to A T P and CTP, respectively, in E. coli B M - T - a f t e r a shift from c o m p l e t e m e d i u m to media containing: A - & , o / * g t h y m i n e per ml; G - C ) , o . 2 / ~ g t h y m i n e per ml; 0 - 0 , o . 5 / * g t h y m i n e per ml. N u m b e r s in parentheses give the pool sizes of d T T P in each sample in per cent of t h e zero values. Abscissa: as Fig. I.
Biochim. Biophys. Acta, 129 (1966) l O 4 - 1 1 5
NUCLEOTIDE POOLS IN
E. coli
III
Figs. 2A and 2B, respectively. Throughout the experiment of Fig. 5 the mass, measured turbidometrically, increased exponentially. Only minute amounts of 5-fluorouridine triphosphate could be detected (not shown). Three main features of Fig. 5 should be pointed out: (I). Exogenous uracil is able to reverse, to some extent, the depletion of the d T T P pool caused b y dFU (left part of Fig. 5)- This is in agreement with the finding of COHEN et al, (Fig. 4 of ref. 2I), as they showed that simultaneous addition of dFU and uracil to growing bacterial cultures resulted in loss of viable cells for about I h, possibly followed by an increase. (2). The dATP pool increases initially when d T T P is depleted, and as the d T T P pool starts to build up, there is a parallel fall in dATP (Fig. 5, left part). This behaviour of the dATP and dTTP pools suggests a close regulatory relationship between the sizes of these two pools. (3). Since this experiment was performed with an E. coli B strain one would have suggested the dCTP pool to be constant throughout the experiment (dashed line on right part of Fig. 5). This is not the case, however. Instead, dCTP increases steeply after a lag of about 75 % of a doubling time. B y comparison with the behaviour of the dCTP pool in E. coli 15 T - A - U - (dotted line in right part of Fig. 5), it seems as if dCTP accumulation only takes place when the dATP accumulation is kept down, possibly by a low d T T P pool. E[[ects o[ suboptimal exogenous thymine concentrations Experiments with dilute cultures of E. coli 15 T - A - U - have revealed zs,2g that when these cells are grown on media containing suboptimal amounts of thymine, the rate of DNA synthesis is reduced. It has now been shown that at the same time the d T T P pools of the cells are lowered 30. These observations prompted us to investigate whether the apparent relationship between the dATP and d T T P pools, indicated by Fig. 5, also exists in thymine-requiring bacteria growing on suboptimal concentrations of exogenous thymine. The pool sizes of the different nucleoside triphosphates were determined in three cultures of E. coli B M - T - transferred from a medium with a thymine concentration of 2o/~g/ml to media containing o, 0.2 and 0.5 #g of thymine per ml, respectively. The results given in Fig. 6A show that the rate of increase of the dATP pool is in fact dependent on the level of d T T P in the culture. The dCTP/CTP ratio, however, shows a more complex behaviour (Fig. 6B). The two curves with low, but significant d T T P pools, indicate a dependence on d T T P similar to that of dATP. Lowering the exogenous thymine concentration to zero, however, prevents any rise in the dCTP pool, although maximal dATP accumulation takes place. It is likely, therefore, that the size of the dCTP pool is under the regulatory influence of both the dATP and the d T T P pools, but in opposite directions. Thus a low d T T P pool is a prerequisite for an increase of the dCTP pool, but, on the other hand, a sufficiently high dATP content will prevent any further dCTP accumulation. The kinetics of the accumulation of dATP and dCTP in the experiment with 0.2 #g of thymine per ml (Fig. 6, open circles) further support this idea. Here the very steep rise in dCTP levels off as the dATP/ATP ratio increases beyond a certain level. EHects o[ novobiocin in E. coli 15 T - A - U A common feature in all the experiments described above is that the various conditions of thymine restriction imposed on the bacteria result in an inhibition of Biochim. Biophys. Acta, I29 (1966) I o 4 - I I 5
II2
J. NEUHARD
DNA synthesis. The possibility therefore remains that it is the inhibition of DNA synthesis per se, and not the lowered pool of thymidine nucleotides, which is responsible for the changes in the dATP and dCTP pools observed. To rule this out, effects on the nucleoside triphosphate pools have been investigated when DNA synthesis is inhibited by other means than thymine restriction. Recently SMITH AND DAVISal have shown that the antibiotic novobiocin in concentrations of IOO #g/ml is able to inhibit DNA synthesis in E. coli 15 T-Argto an extent of 81 °/o, apparently by interfering directly with the DNA polymerase (deoxynucleoside triphosphate :deoxyribonucleic acid deoxynucleotidyltransferase, EC 2.7-7.7) of the cells. RNA synthesis was inhibited rather less (54 %). Fig. 7 gives the results of an experiment where the effect of novobiocin (200 /~g/ml) on the incorporation of [2-z4C]thymine into DNA was determined in E. coli 15 T-A-U-. With this concentration DNA synthesis is more than 85 % inhibited. In Fig. 8 the effects of novobiocin on the dATP and dCTP pools are shown. Included for comparison are the corresponding curves from Fig. 2B. Apparently inhibition of DNA synthesis by other means than thymine restriction does not result in appreciable changes in the dATP and dCTP pools. This strengthens the suggestion that the primary cause of the changes during thymine restriction is the low intracelhilar amount of thymidine nucleotides. Novobiocin does not alter the pool size of dGTP significantly (not shown).
'o=<
,.o
¢ONTRO~
1.5
0.15[
/
§
o
~3
~o.
. , t <0.1(] ~
o
/. .....
II
........./..
/
"
/
: ' \aA'rP
]
t
/
1
1.0 •
~2 ~.5
/
dCTP
/
#
NOVOBIOCIN
Z 300 100 2()0 ' Mass increase (per cent)
~<2o~oL_.. i
i
1.0
2.0
Relativetimeaftershift
Fig. 7- I n h i b i t i o n of D N A s y n t h e s i s b y n o v o b i o c i n (200/~g/ml) in E. coli 15 T - A - U - , m e a s u r e d as described in METHODS. I n s e r t : T h e effect of n o v o b i o c i n on g r o w t h of t h e culture. Fig. 8. Effects of novobiocin on t h e pool-sizes of & - A , d A T P a n d 0 - 0 , d C T P relative to A T P and CTP, respectively, in E. coli 15 T - A - U - . A t t i m e zero t h e culture w a s shifted to a m e d i u m c o n t a i n i n g novobiocin (2oo/~g/ml). "" ", values of d A T P / A T P a n d - - - , values of d C T P / C T P f r o m Fig. 2]3.
DISCUSSION
The results of the present work m a y be summarized as follows: (I). Cells of Biochim. Biophys. Acta, 129 (1966) l O 4 - I 1 5
NUCLEOTIDE POOLS IN E. coli
113
E. coli are able to regulate the sizes of their deoxyribonucleoside triphosphate pools in such a way that they remain nearly constant, even when DNA synthesis is inhibited by novobiocin 3x (see Fig. 7). (2). E. coli cells deprived of their intracellular thymidine nucleotides respond immediately by accumulating dATP, indicating that the size of the dATP pool is in some way regulated in vivo by the pool size of the thymidine nucleotides. It is impossible from the present experiments to decide which of the nucleotides is the actual regulator, but since other known regulatory mechanisms involving thymidine nucleotides seem to implicate dTTPX-5, 32-85, and since the triphosphate is the immediate precursor of DNA, it seems likely that dTTP is also the active compound in the regulation of the pool size of dATP. (3)- In most cases thymine deprivation also causes accumulation of dCTP, but apparently this is counteracted by high levels of dATP*. (4). Under no experimental conditions so far investigated does the size of the dGTP pool change. In E. coli the reduction of ribonucleotides to the corresponding deoxyribonucleotides is known to take place at the ribonucleoside diphosphate leveP 6. All four normally occurring ribonucleoside diphosphates may be reduced by extracts of E. coli B 3~-a9. Furthermore, it is known from studies on Ehrlich ascites-tumor cells4°, chick-embryo extracts 41 and mouse leukemia cells42, that these reactions are inhibited in vitro by one or more of the deoxyribonucleoside triphosphates. In view of these findings and the fact that the ribonucleoside diphosphate reductases are the enzymes that connect the ribonucleotide and deoxyribonucleotide metabolism, the altered pools of dATP and dCTP observed may be explained by changes in the activities of the corresponding reductases following thymine starvation. Recently BlSWAS, HARDY AND BECK~ have shown that thymine starvation of E. coli 15 T- and treatment of E. coli B with dFU result in an immediate rise in the CDP reductase activity of the cells. Working with E. coli 15 T - A - U - we have previously reported 1° that if no protein synthesis occurs throughout the period of thymine starvation, the increase in dCTP content is prevented, without affecting the dATP accumulation. It was therefore proposed 1° that derepression of CDP reductase was responsible for the increased dCTP pools observed in thymine-deprived cells. Since all the mutants used in the present work are able to accumulate dCTP under appropriate conditions of thymine deprivation, it seems likely that derepression of the CDP reductase is a general response to restrictions in the intraceUular pool of thymidine nucleotides in E. coll. Assuming that dATP acts as a specific inhibitor of the CDP reductase 4°, it is possible to explain why dCTP fails to accumulate under certain conditions of thymine starvation, since these are the conditions where by far the highest concentrations of dATP are achieved (Fig. 2A). Only in cases where the dATP accumulation levels off after 75 % of a doubling time (Fig. 2B), while CDP reductase activity continues to rise (Fig. 6 of ref. 7), is any expansion of the dCTP pool observed. The lag in dCTP accumulation observed in the two E. coli 15 T - strains (Fig. 3B) could likewise be explained by the inhibitory action of the initially increasing dATP pool on the derepressed CDP reductase (Fig. 3A). In agreement with this, HOLMGREN, REICHARD AND THELAND~R xx have shown " Unpublished experiments from our laboratory with E. coli B M - T - have shown t h a t neither the pool o1 dAMP nor t h a t of dADPchanges significantly during thymine starvation (see METHODS). Biockim. Biophys. Acta, I29 (I966) xo4-~x5
114
J. NEUHARD
that dATP is a strong inhibitor of the purified CDP reductase from E. coli B. They found that the inhibition was competitive with the activating effect of ATP so that the determinant of the reaction rate in vitro was the ratio dATP/ATP. The agreement is even more striking if one compares our dATP/ATP ratios determined in vivo (Fig. 3A) with those given by Fig. 7 in their publication n. A dATP/ATP ratio of 0.05, i.e., the value found in exponentially growing E. coli, would in their assay give an inhibition of about 25 %, while ratios of o.15 and 0.27, which are the final values obtained after 9o-min thymine starvation of E. coli 15 T - and E. coli B T- strains, would show inhibitions of about 80 % and 95 %, respectively, in their assay, In view of the findings discussed above it is tempting to assume that the effect of dTTP on the dATP pool could be explained analogously by an inhibitory action of dTTP on the reductase converting ADP to dADP. Since, however, all values given in this work represent pool sizes, and any pool is determined by different anabolic and catabolic reactions, it is evident that the regulatory influence of dTTP on the dATP pool might be explained in other ways.
ACKNOWLEDGEMENTS
The author wishes to thank Dr. A. MUNCH-PETERSEN for her encouragement and continued interest in this work, and Dr. J. L. INGRAHAM for reading the manuscript. The excellent technical assistance of Mrs. K. DJURHUUS is gratefully acknowledged. This work was supported by grants from Carlsberg-Fondet and NOVOFondet.
REFERENCES
I 2 3 4 5 6 7 8 9
IO ii 12 13 14 15 16 17 18 19 20 21 22 23
H. IVES, P. A. MORSE, Jr. AND VAN R. POTTER, J. Biol. Chem., 238 (i963) 1467. R. BREITMAN, Bioehim. Biophys. Acta, 67 (1963) 153. OKAZAKI AND A. KORNBERG, J. Biol. Chem., 239 (1964) 275. BRESNICK, U. B. THOMPSON, H. P. MORRIS AND A. G. LIEBELT, Biochem. Biophys. Res. Commun., 16 (1964) 278. M. GREEN, M. PINA AND V. CHAGOYA, J. Biol. Chem., 239 (1964) 1188. T. R. BREITMAN AND R. M. BRADFORD, Biochem. Biophys. Res. Commun., 17 (1964) 786. C. BISWAS, J. HARDY AND W. S. BECK, J. Biol. Chem., 24o (1965) 3631. A. MUNCH-PETERSEN AND J. NEUHARD, Biochim. Biophys. Acta, 80 (1964) 542. A. MUNCH-PETERSEN AND J. NEUHARD, Intern. Congr. Biochem., 6th, New York, 1964, Abstracts IX. J. NEUHARD AND A. MUNCH-PETERSEN, Bioehim. Biophys. Acta, 114 (1966) 61. A. HOLMGREN, P. REICHARD AND L. THELANDER, Proe. Natl. Acad. Sci. U.S., 54 (1965) 830. H. D. EARNER AND S. S. COHEN, J. Bacteriol., 68 (1954) 80. D. KANAZIR, H. D. EARNER, J. G. FLAKS AND S. S. COHEN, Biochim. Biophys. Acta, 34 (1959) 341 • D. KORN AND A. WEISSBACH, Biochim. Biophys. Acta, 61 (1962) 775. C. G. KURLAND AND O. MAALOE, J. Mol. Biol., 4 (1962) 193. J. D. FRIESEN AND O. MAALOE, Biochim. Biophys. Acta, 95 (1965) 436. K. RANDERATH AND E. RANDERATH, J. Chromatog., 16 (1964) i i i . E. RANDERATH AND K. RANDERATH, J. Chromatog., I6 (1964) 126. J. NEUHARD, E. RANDERATH AND K. RANDERATH, Anal. Biochem., 13 (1965) 211. K. RANDERATH AND E. RANDERATH, Anal. Bioehem., 13 (1965) 575. S. S. COHEN, J. G. FLAKS, H. D. EARNER, M. R. LOEB AND J. LICHTENSTEIN, Proc. Natl. Acad. Sci. U.S., 44 (1958 ) lOO4. S. S. COHEN AND H. D. EARNER, Proc. Natl. Acad. Sci. U.S., 4 ° (1954) 885. H. MENNIGMANN AND W. SZYBALSKI, Biochem. Biophys. Res. Commun., 9 (1962) 398.
D. T. R. E.
Biochim. Biophys. Acta, 129 (r966) lO4-115
NUCLEOTIDE POOLS IN E. coli 24 25 26 27 28 29 3° 31 32 33 34 35 36 37 38 39 4° 41 42
115
J. G. FLAKS AND S. S. COHEN, J. Biol. Chem., 234 (1959) 2981. K. O. HARTMANN AND C. HEIDELBERGER, J. Biol. Chem., 236 (1961) 3oo6. R. W . BROCKMANN, J. M. D A v i s AND P. STUTTS, Biochim. Biophys. Acta, 4 ° (196o) 22. W . E. RAZZEL AND P. CASSHYAP,J. Biol. Chem., 239 (1964) 1789. O. MAALOE, J. Cellular Comp. Physiol., Suppl. I, 62 (i963) 31. K . RASMUSSEN, p e r s o n a l c o m m u n i c a t i o n . S. HENNINGSEN AND J. NEUHARD, u n p u b l i s h e d results. D. H. SMITH AND B. D. DAVIS, Biochem. Biophys. Res. Commun., 18 (1965) 796. E. SCARANO,G. GERACI, A. POLZELLA AND E. CAMPANILE, J. Biol. Chem., 238 (1963) P C 1556. C. F. MALEY AND F. MALEY, J. Biol. Chem., 239 (1964) 1168. F. MALEY AND C. F. MALEY, J. Biol. Chem., 24o (1965) P C 3226. W . H. FLEMMING AND M. J. BESSMAN, J. Biol. Chem., 24 ° (1965) P C 41o8. P. REICHARD, A. BALDSTEN AND L. RUTBERG, J. Biol. Chem., 236 (1961) 115o. P. REICHARD, J. Biol. Chem., 237 (1962) 3513. L. E. BERTANI, A. H-~GGMARK AND P. REICHARD, J. Biol. Chem., 238 (1963) 3407 . A. LARSSON, J. Biol. Chem., 238 (1963) 3414 . H. KLENOW, Biochim. Biophys. Acta, 61 (1962) 885. P. REICHARD, Z. N. CANELLAKIS AND E. S. CANELLAKIS,J. Biol. Chem., 236 (1961) 2514. N. R. MORRIS, P. 1REICHARD AND G. A. FISHER, Biochim. Biophys. Acta, 68 (1963) 93.
Biochim. Biophys. Acta, 129 (1966) lO4-115
BIOCHIMICA ET BIOPHYSICA ACTA
BBA 95480
STUDIES ON T H E ACID-SOLUBLE NUCLEOTIDE POOL IN THYMINEREQUIRING MUTANTS OF E S C H E R I C H I A COLI DURING THYMINE STARVATION III. ON THE REGULATION OF T H E DEOXYADENOSINE TRIPHOSPHATE AND DEOXYCYTIDINE TRIPHOSPHATE POOLS OF E S C H E R I C H I A COLI
JAN NEUHARD
Institute o/Biological Chemistry, University of Copenhagen, Copenhagen (Denmark) (Received March 3rd, 1966)
SUMMARY
Changes in the size of the deoxyribonucleoside triphosphate pool have been investigated in mutants of Escherichia coli subjected to different conditions of thymine restriction. The following results were obtained. I. Five different thymine-requiring mutants of E. coti, after removal of thymine from their growth media, all showed an immediate rise in dATP content reaching a maximum after about 60 min. 2. In two mutants (both E. coli 15 T - mutants) the dCTP pool increased linearly IO- to I5-fold after a lag of about 30 min. In these two mutants the maximal amount of dATP accumulated was only half of that reached by the other three. 3. Thymine prototrophs of E. coli also accumulated dATP when deprived of their intracellular thymidine nucleotides by addition of 5-fluorodeoxyuridine. 4. Inhibition of DNA synthesis in E. coli 15 T - A - U - by novobiocin in the presence of thymine did not result in any significant changes in the deoxyribonucleoside triphosphate pools. The results are discussed in relation to the regulation of deoxyribonucleotide synthesis.
INTRODUCTION
During the last few years work on the thymidine kinases (ATP: thymidine 5'-phosphotransferase, EC 2.7.1.21 ) from widely different sources have shown that dTTP is a specific feedback inhibitor of these enzymes1-5 in vitro. The possibility that thymidine nucleotides might have other regulatory functions in the deoxyribonucleotide metabolism of bacteria is supported by experiments with thymine-requiring Abbreviation: dFU,5-fluoro-2'-deoxyuridine.
Biochim. Biophys. Acta, 129 (I966) lO4-115
B!UCLEOTIDE POOLS IN E. coli
Io5
Escherichia coli auxotrophs. Thus, starvation of E. coli 7oV3-462 for thymine results in induction of high amounts of thymidine phosphorylase (thymidine :orthophosphate deoxyribosyltransferase, EC 2.4.2.4) paralleled by a substantial excretion of free deoxyribose into the growth mediums. Blsw~,s, HARDY AND BECK7 working with E. coli 15 T- have demonstrated that thymine deprivation causes a 6- to 8-fold increase in the cellular content of CDP reductase, the enzyme responsible for the conversion of CDP to dCDP. As a corresponding increase could be brought about by addition of 5-fhioro-2'-deoxyuridine (dFU) to the thymine prototroph E. coli B, they concluded that a thymidine nucleotide was acting as a corepressor for the ribonucleoside diphosphate reductase system in E. coll. In previous experiments on changes in the deoxyribonucleoside triphosphate pools of E. coli 15 T-A-U- (requiring thymine, arginine and uracil) we found~-1° that immediately after thymine was removed from the growth medium the dATP content of the cells rose, reaching a maximum value in about 6o min. The pool of dCTP, however, remained constant for the first 3° min of thymine starvation and then increased linearly, reaching a io- to IS-fold value after 9° min. While the increase in dATP was independent of concomitant protein or RNA synthesis, the dCTP content only rose when protein synthesis occurred simultaneouslyTM. The dGTP pool showed no changes under any conditions of thymine starvation investigated. In the present work, results will be presented that strengthen our earlier suggestionS, 1° that a thymidine nucleotide is involved directly in the regulation of the pool size of dATP in E. coll. Moreover, they seem to indicate that the ratio dATP/ATP plays an important role in the regulation of the pool size of dCTP. This makes us believe that the controlling effect of this ratio on the activity of the CDP reductase from E. coli B, recently shown by HOLMGREN, REICHARD AND THELANDER 11 to operate in vitro, may also be of importance in vivo.
MATERIALS
Pyrimidines, amino acids, ribo- and deoxyribonucleoside 5'-phosphates were .obtained from Sigma Chemical Company, St. Louis (U.S.A.). dFU was a gift from Hoffman-La Roche, Inc. (Switzerland) through the courtesy of Wmrum and Co. (Copenhagen). The sodium salt of novobiocin was obtained as Albamycin capsules from Upjohn (Belgium). Poly(ethyleneimine) (Polymin P) was a gift from Badische Anilin- und SodaFabrik, Ludwigshafen (Germany); cellulose powder for thin-layer chromatography .(MN 3o0) was from Macherey, Nagel and Co., Diiren (Germany), and methanol .(technical grade, containing less than 0.25 % water) from AKI (Copenhagen). Carrier-free [a*P~orthophosphate and ~2-14Clthymine were obtained from The Radiochemical Centre, Amersham (England).
Bacterial strains and growth conditions
The basal medium described previously8with glucose (o.2 %) as a carbon source was used throughout. The bacterial strains employed were all obtained on agar slants from the Institute of Microbiology, University of Copenhagen, through the Biochim. Biophys. Acta, 129 (1966) lO4-115
lO6
j. NEUHARD
courtesy of Dr. N. FILL. Table I shows the strains used (A) and their growth rates (C) in basal medium supphed with glucose and the necessary nutritional requirements (]3). Increase in cell mass was followed at 450 m/~ in a Zeiss M4 QIII spectrophotometer (I-cm light path). I ml bacterial culture with an absorbance at 450 m# of I.OOO corresponds to approx. 4" lO8 cells or 0.2 mg dry weight.
TABLE
I
BACTERIAL STRAINS USED IN THE P R E S E N T WORK
(A) Bacterial strain
(B) (C) Nutritional Growth rate* requirements(itg[ml ) (doublings per h)
(D) Re]eren6e
E. coli 15, T -
thymine, 2
1. 4
B A R N ~ R AND COHEN 12
E. coli 15, T - A - U -
thymine, 2 arginine, 20 uracil, i o
1.2-1. 3
I~ANAZIR et
E. coli KI2(~,), T -
t h y m i n e , 20
0.9
K O R N AND W E I S S B A C H 14
E. coli B/r, T - H -
t h y m i n e , 20 h i s t i d i n e , 5o
1. 3
E. 6oli B, M -
m e t h i o n i n e , 5o
1.2-1. 3
K U R L A N D AND 1V~AALOE15
E. coli B, M - T -
m e t h i o n i n e , 50 t h y m i n e , 2o
1.3-1. 4
F R I E S E N AND MAALOE le
al. a3
* I n b a s a l m e d i u m s u p p l i e d w i t h g l u c o s e (0.2 %) a n d n e c e s s a r y n u t r i t i o n a l r e q u i r e m e n t s .
METHODS
Determination o/ DNA synthesis Cells growing exponentially were harvested on 9-cm Millipore filters (Membran Filter, G6ttingen (Germany)), washed with basal medium and resuspended in a medium containing ~2-x4C]thymine (specific activity: 6.5/*C/#mole). At Io-min intervals I-ml samples were added to tubes containing 4 ml ice-cold 6 % trichloroacetic acid. The precipitates were collected on 27-ram Millipore filters, and washed three times with ice-cold 5 % trichloroacetic acid containing Io/~g [l~C]thymine per rnl. Subsequently the filters were glued to aluminum planchets, dried and counted in a Friesecke-Hoepfner thin-window gas-flow counter.
Quantitative deter~nination o[ acid-soluble nucleotides a. Labelling o/the nucleotides. In all experiments where the acid-soluble nucleotide pools were determined, cells labelled for at least 2 generations in complete medium containing [aZP]olthophosphate (specific activity: i-3/A2/k~ole ) were used. Subsequent incubations of the cells were performed in the desLved medinm containing [s~p]orthophosphate of the same specific activity as before. Biochim. Biophys. Acta, 129 (1966) lO4-115
NUCLEOTIDE POOLS IN E. coli
lO7
b. Preparation o/extracts. Samples of 5 ml were harvested by suction filtration on Millipore filters and treated as previously described1°. Before and after ether extraction and after lyophilization, aliquots were plated on planchets and counted in order to determine the recoveries of total ~*P radioactivity. While the ether extraction gave recoveries of 95-1oo %, the recoveries after lyophilization varied between 60 % and 9° %. All figures given in the following have been corrected for the corresponding losses. c. Fractionation o/ cell extracts. Separation of the different nucleotides was performed in two dimensions on poly(ethyleneimine)-impreguated cellulose layers on 20 cm ×2o cm glass plates. In most cases a stepwise elution was used in both dimensionsXT,is. Triphosphates: Separation of complex nucleoside triphosphate mixtures has previously been described1°. This procedure was followed except for minor modifications of the solvents. The solvents used throughout the present work were, in the first dimension: 2 M LiC1-2 M acetic acid (I :I, v/v) (3 cm) followed by 2.5 M LiC1-2 M acetic acid (I:I, v/v) (15 cm); and in the second dimension: 2.5 M ammonium acetate- 3.6 % boric acid (pH 7.o with ammonia) (5 cm) followed by 3-5 M ammonium acetate-5 % boric acid (pH 7.0 with ammonia) (I 5 cm). Diphosphates: A good separation of the different nucleoside diphosphates is achieved by modifying the solvents used for the triphosphate separation as follows: First dimension, 3 M LiCI-I M acetic acid (1:9, v/v) (2 cm) followed by 1.6 M LiC1-2 M acetic acid (I:I, v/v) (15 cm); second dimension, I M ammonium acetate2.5 % boric acid (pH 7.o with ammonia) (I cm) followed by 2 M ammonium acetate5 % boric acid (pH 7.0 with ammonia) (15 cm). Monophosphates: For the separation of nucleoside monophosphates and nucleotide-sugars, the method described by RANDERATHAND RANDERATH2° was used. To the start-spots of all chromatograms unlabelled nucleotides were added as markers. Compounds were located on the chromatograms with a Mineralight lamp and by autoradiography (exposure time 24 h on Kodirex X-ray film). There was always complete correspondance between the ultraviolet-detected spots (added markers) and the radioactive spots. Quantitative determinations of the radioactive compounds from the chromatograms were carried out as described previously1°,19.
RESULTS Effects o/ thymine starvation in different thymine-requiring strains Very distinct changes in the nucleoside triphosphate pools have previously been reported to take place 1° when thymine is removed from the growth medium of the triple auxotroph E. coli 15 T-A-U-. In order to determine if these changes are a general phenomena in cells of E. coli deficient in their thymidine nucleotide pools, a number of other thymine auxotrophs were investigated for changes in their nucleoside triphosphate pools after withdrawal of thymine from their growth media. The strains used are listed in Table I. Table II gives the pool sizes of the eight nucleoside triphosphates in the different strains, when growing exponentially in complete media. Figs. I and 2 give the sizes of the nucleoside triphosphate pools of two different Biochim. Biophys. Acta, 129 (1966) lO4-115
108
J. NEUHARD
TABLE II POOL SIZES OF THE NUCLEOSIDE TRIPHOSPHATES IN CELLS OF D I F F E R E N T STRAINS OF E. coli G R O W I N G EXPONENTIALLY IN COMPLETE MEDIUM CONTAINING [32P]ORTHOPHOSPHATE
F o r all d e t e r m i n a t i o n s the procedure described in METHODS was used.
Bacterial strain
E. E. E. E. E. E.
coli coli coli coli coli coli
Number o~
15, T 15, T - A - U KI2(~), T B/r, T - H B, MB, M - T -
I~moles per g dry weight
experiments G T P
ATP
CTP
UTP
dGTP
dATP
dCTP
dTTP
I 4 1 I 2 8
3-7 3.4 5.4 6.0 4.8 5.3
o.9 o.9 1.4 1. 7 1.2 1.6
1.9 2.o 2.2 3.2 2.2 3.1
-o.17 o.16 o.2o o.17 0.22
o.18 o.24 0.35 0.28 o.25 0.28
o.22 o.28 0.29 o.37 o.29 o.39
o.32 o.41 0.4 ° o.35 o.23 o.41
1.4 1.3 1.9 1. 7 1. 9 1.7
2s[
A
7.0 .E 6.C ._o~
A
B
- 2A~ I
~'
~ 12
A ATP "~
•
/
e ~
;
1-0[~
I
dATP
UTP
o.~p ~O
T
/
ATP
•
/
~,-0
o..l /
o t'.o 2'.o 1i0 Retative time after thymine removal
I
2.0
I
dATP
~o
°~'~o
t'o
2~o
I
ti0
Relative time after thymine removal
z:0
Fig. I. A m o u n t s of t h e four ribonucleoside t r i p h o s p h a t e s in (A), E. coli B M - T - (doubling time: 42 min) a n d (B), E. coli 15 T - A - U - (doubling time: 47 min) after r e m o v a l of t h y m i n e f r o m t h e media. 0 - 0 , GTP; O - © , CTP; A - A , U T P ; A-LS, ATP. Abscissa: relative time, using t h e doubling t i m e of each m u t a n t as unity. Fig. 2. A m o u n t s of t h e different deoxyribonucleoside t r i p h o s p h a t e s in (A), E. coti B M - T (doubling time: 42 min) and (B), E. coli 15 T - A - U - (doubling time: 47 min) after r e m o v a l of t h y m i n e f r o m t h e media. 0 - 0 , dGTP; O - G , dCTP; LS-&, dATP. Abscissa: as Fig. i.
strains as a function of time after thymine removal. Figs. IA and 2A give the figures for the ribo- and deoxyribonucleoside triphosphates, respectively, from one experiment with E. coli B M - T - ; Figs. I B and 2B give the corresponding figures from an experiment with E. coli 15 T - A - U - . While no major changes in the ribonucleoside triphosphates occur (Figs. IA and IB), two things should be pointed out regarding dATP and dCTP (Figs. 2A and 2B). (I) The previously observed rise in the dATP pool of E. coli 15 T - A - U - (ref. IO) is much more pronounced in E. coti B M - T - and, in contrast, (2) no changes whatsoever are observed in the dCTP pool of this latter strain, while dCTP rises to a I2-fold value in E. coli 15 T - A - U - . As the absolute amounts of the different triphosphates v a r y significantly from one strain to another (see Table I I and Figs. I and 2), and the ribonucleoside triphosphates never showed major changes, we have chosen in most of the figures to Biochim. Biophys. Acta, 129 (I966) lO4-115
lO9
NUCLEOTIDE POOLS IN E . coli
give the ratios between corresponding deoxyribo- and ribonucleoside triphosphates in each sample, rather than the absolute amounts. By doing this, results with different mutants are directly comparable. The effects of thymine starvation on the deoxyribonucleoside triphosphate pools of five thymine-requiring E. coli strains are compared in Fig. 3. While all mutants behave similarly during the initial 75 % of the doubling times, i.e. marked increase in dATP/ATP and constant dCTP/CTP, their response to prolonged thymine starvation is of two types. In one, represented by the two E. eoli 15 T- strains, the dATP/ATP ratio levels off, while dCTP/CTP increases very steeply. In the other group, comprising the two E. coli B mutants, the dCTP/CTP ratio remains constant throughout, while the dATP/ATP ratio increases for another 75 % of a doubling time reaching a plateau about twice as high as that found in the E. coli 15 strains. The changes found in E. coli KI2 (~) T- indicate that it is an intermediary type, but we have not investigated this strain further in the present work. In none of the mutants, however, did the dGTP pool change significantly. A
1.5
o3IA
B
/
t.0
0.~
io
/ 0
~
T
~o.
+ /e,-
i
c:tD.
02
o:
1!
•
I('
FU
o!
Ot
0.1
B
/
o
!--.-.%-:'-I
1.0
2'.0
1.0
Retative time after thymine removal
2.0
I
1,0
~
2,0
I
i
1.0
J
2.0
Relative time after ,hift
Fig. 3. Changes in t h e pool sizes of (A), d A T P a n d (B), d C T P relative to A T P a n d CTP, respectively, in different t h y m i n e - r e q u i r i n g m u t a n t s of E. coli, s t a r v e d for t h y m i n e . A - ~ k , E. coli B M - T - ; A - A , E . coli B/r T - H - ; 0 - - © , E. coli i5 T - A - U - ; O - - Q , E . coli I 5 T - ; [ ] - [ ] , E . coli KI2(A) T - . Full line: m e a n values of t h e two E. coli B strains. Dotted line: m e a n values of t h e two E. coli 15 T - strains. Abscissa: as Fig. I. Growth rates of t h e different m u t a n t s are given in Table I. Fig. 4. Effects of different conditions of t h y m i n e restriction on t h e pool sizes of (A), d A T P a n d (B), d C T P relative to A T P a n d CTP, respectively, in different s t r a i n s of E. coll. lX-lX, E. coli B M - T - shifted to a m e d i u m lacking t h y m i n e ; G-(D, E. coli B M - shifted to a m e d i u m containing d F U (2o/,g]ml); 0 - 0 , E. coli B shifted to m e d i u m containing d F U (2o/~g]ml). Abscissa: as Fig. i.
E//ects o] d F U in thymine prototroph strains It has been shown by COHEN et al. zz, that dFU added to growing cultures of E. coli B caused a gradual cessation of growth (measured turbidometticaily), together with a rapid loss of viability, resembling the phenomenon called thymineless deathZZ,za. This bacteriocidM effect of dFU was readily explained, as it was shown that dFUMP, formed in the cells from dFU 2z, was a very potent inhibitor of thymidylate synthetaseZl,24,zS. Thus it is possible to induce a thymine requirement in a bacteria not otherwise requiting thymine, merely by adding dFU to the growth medium. Biockim. Biopkys. Acta, r29 (I966) xo4-zI5
IIO
j. NEUHARD
dFU (20 #g/ml) was added to exponentially growing cultures of the strains E. coli B and E. coti B M- (see Table I). At the same time a culture of E. coli B M - T - was transferred to a nmdium lacking thymine. The nucleoside triphosphate pools of the three strains were then followed with time. The results of such an experiment are shown in Fig. 4. It is immediately apparent that thymine deficiency induced by dFU also leads to an appreciable accumulation of dATP, though not as extensive as that seen in the corresponding thymine auxotrophic strain after removal of exogenous thymine. The ratio dCTP/CTP, however, did not change in any of the three strains. In addition to what is shown in Fig. 4, it was found that the two mutants receiving dFU accumulated high amounts of 5-fluorouridine triphosphate ~e (7.5 /,moles per g dry weight) (c/. Table II) concomitant with a reduction in their pools of UTP, CTP and dCTP. dTTP was always completely absent after addition of dFU. COHEN and co-workers ~z also showed that dFU was split by the bacteria (perhaps by thymidine phosphorylase ~7) to 5-fluorouracil, which could be transformed to 5-fluorouridine nucleotides. These may in turn be built into RNA and thereby cause the observed disturbances in growth. Our finding of high levels of 5-fluorouridine triphosphate after addition of dFU seems to support this explanation. However, simultaneous addition of uracil and dFU to growing cultures suppresses specifically the effects of 5-fluorouracil on the RNA metabolism, without initially affecting the loss in viable counts. Thus a culture of E. coli B M- was treated with dFU (20 #g/ml) in the presence of an equal amount of uracil. The changes with time in the absolute amounts of the deoxynucleoside triphosphates during the incubation is shown in Fig. 5 (full lines). Included in the figure are the curves showing the changes in dATP and dCTP of E. coli B M-T- (dashed lines) and E. coli 15 T - A - U - (dotted lines) taken from
2.4
/
2.2
ts
A
B
to)
!
g 1.4
~
0.3
/~xx
o,:
/
//
1.o
0.2
I.~
'? ,.jo
oo)
x
i.-o. o~.~ o
dATP 0.1
~
0.l
--
(i
q &
dTTP 1.0
~ ~e""--~-
2.0 1~0 Relative time after shift
o
dGTP 2.0
t .0
1:0
2.0 Retative time after shift
210
Fig. 5. - - , a m o u n t s of the four d e o x y r i b o n u c l e o s i d e triphosphates in /~. coli B M - a f t e r a shift from complete m e d i u m to a m e d i u m containing d F U (2o/*g/ml) and uracil ( i 8 $ , g / m l ) ; . . . . values from Fig. 2 B ; - - - , values from Fig. 2A. Abscissa: as Fig. i . Fig. 6. Changes in pool sizes of (A), d A T P and (B), d C T P relative to A T P and CTP, respectively, in E. coli B M - T - a f t e r a shift from c o m p l e t e m e d i u m to media containing: A - & , o / * g t h y m i n e per ml; G - C ) , o . 2 / ~ g t h y m i n e per ml; 0 - 0 , o . 5 / * g t h y m i n e per ml. N u m b e r s in parentheses give the pool sizes of d T T P in each sample in per cent of t h e zero values. Abscissa: as Fig. I.
Biochim. Biophys. Acta, 129 (1966) l O 4 - 1 1 5
NUCLEOTIDE POOLS IN
E. coli
III
Figs. 2A and 2B, respectively. Throughout the experiment of Fig. 5 the mass, measured turbidometrically, increased exponentially. Only minute amounts of 5-fluorouridine triphosphate could be detected (not shown). Three main features of Fig. 5 should be pointed out: (I). Exogenous uracil is able to reverse, to some extent, the depletion of the d T T P pool caused b y dFU (left part of Fig. 5)- This is in agreement with the finding of COHEN et al, (Fig. 4 of ref. 2I), as they showed that simultaneous addition of dFU and uracil to growing bacterial cultures resulted in loss of viable cells for about I h, possibly followed by an increase. (2). The dATP pool increases initially when d T T P is depleted, and as the d T T P pool starts to build up, there is a parallel fall in dATP (Fig. 5, left part). This behaviour of the dATP and dTTP pools suggests a close regulatory relationship between the sizes of these two pools. (3). Since this experiment was performed with an E. coli B strain one would have suggested the dCTP pool to be constant throughout the experiment (dashed line on right part of Fig. 5). This is not the case, however. Instead, dCTP increases steeply after a lag of about 75 % of a doubling time. B y comparison with the behaviour of the dCTP pool in E. coli 15 T - A - U - (dotted line in right part of Fig. 5), it seems as if dCTP accumulation only takes place when the dATP accumulation is kept down, possibly by a low d T T P pool. E[[ects o[ suboptimal exogenous thymine concentrations Experiments with dilute cultures of E. coli 15 T - A - U - have revealed zs,2g that when these cells are grown on media containing suboptimal amounts of thymine, the rate of DNA synthesis is reduced. It has now been shown that at the same time the d T T P pools of the cells are lowered 30. These observations prompted us to investigate whether the apparent relationship between the dATP and d T T P pools, indicated by Fig. 5, also exists in thymine-requiring bacteria growing on suboptimal concentrations of exogenous thymine. The pool sizes of the different nucleoside triphosphates were determined in three cultures of E. coli B M - T - transferred from a medium with a thymine concentration of 2o/~g/ml to media containing o, 0.2 and 0.5 #g of thymine per ml, respectively. The results given in Fig. 6A show that the rate of increase of the dATP pool is in fact dependent on the level of d T T P in the culture. The dCTP/CTP ratio, however, shows a more complex behaviour (Fig. 6B). The two curves with low, but significant d T T P pools, indicate a dependence on d T T P similar to that of dATP. Lowering the exogenous thymine concentration to zero, however, prevents any rise in the dCTP pool, although maximal dATP accumulation takes place. It is likely, therefore, that the size of the dCTP pool is under the regulatory influence of both the dATP and the d T T P pools, but in opposite directions. Thus a low d T T P pool is a prerequisite for an increase of the dCTP pool, but, on the other hand, a sufficiently high dATP content will prevent any further dCTP accumulation. The kinetics of the accumulation of dATP and dCTP in the experiment with 0.2 #g of thymine per ml (Fig. 6, open circles) further support this idea. Here the very steep rise in dCTP levels off as the dATP/ATP ratio increases beyond a certain level. EHects o[ novobiocin in E. coli 15 T - A - U A common feature in all the experiments described above is that the various conditions of thymine restriction imposed on the bacteria result in an inhibition of Biochim. Biophys. Acta, I29 (1966) I o 4 - I I 5
II2
J. NEUHARD
DNA synthesis. The possibility therefore remains that it is the inhibition of DNA synthesis per se, and not the lowered pool of thymidine nucleotides, which is responsible for the changes in the dATP and dCTP pools observed. To rule this out, effects on the nucleoside triphosphate pools have been investigated when DNA synthesis is inhibited by other means than thymine restriction. Recently SMITH AND DAVISal have shown that the antibiotic novobiocin in concentrations of IOO #g/ml is able to inhibit DNA synthesis in E. coli 15 T-Argto an extent of 81 °/o, apparently by interfering directly with the DNA polymerase (deoxynucleoside triphosphate :deoxyribonucleic acid deoxynucleotidyltransferase, EC 2.7-7.7) of the cells. RNA synthesis was inhibited rather less (54 %). Fig. 7 gives the results of an experiment where the effect of novobiocin (200 /~g/ml) on the incorporation of [2-z4C]thymine into DNA was determined in E. coli 15 T-A-U-. With this concentration DNA synthesis is more than 85 % inhibited. In Fig. 8 the effects of novobiocin on the dATP and dCTP pools are shown. Included for comparison are the corresponding curves from Fig. 2B. Apparently inhibition of DNA synthesis by other means than thymine restriction does not result in appreciable changes in the dATP and dCTP pools. This strengthens the suggestion that the primary cause of the changes during thymine restriction is the low intracelhilar amount of thymidine nucleotides. Novobiocin does not alter the pool size of dGTP significantly (not shown).
'o=<
,.o
¢ONTRO~
1.5
0.15[
/
§
o
~3
~o.
. , t <0.1(] ~
o
/. .....
II
........./..
/
"
/
: ' \aA'rP
]
t
/
1
1.0 •
~2 ~.5
/
dCTP
/
#
NOVOBIOCIN
Z 300 100 2()0 ' Mass increase (per cent)
~<2o~oL_.. i
i
1.0
2.0
Relativetimeaftershift
Fig. 7- I n h i b i t i o n of D N A s y n t h e s i s b y n o v o b i o c i n (200/~g/ml) in E. coli 15 T - A - U - , m e a s u r e d as described in METHODS. I n s e r t : T h e effect of n o v o b i o c i n on g r o w t h of t h e culture. Fig. 8. Effects of novobiocin on t h e pool-sizes of & - A , d A T P a n d 0 - 0 , d C T P relative to A T P and CTP, respectively, in E. coli 15 T - A - U - . A t t i m e zero t h e culture w a s shifted to a m e d i u m c o n t a i n i n g novobiocin (2oo/~g/ml). "" ", values of d A T P / A T P a n d - - - , values of d C T P / C T P f r o m Fig. 2]3.
DISCUSSION
The results of the present work m a y be summarized as follows: (I). Cells of Biochim. Biophys. Acta, 129 (1966) l O 4 - I 1 5
NUCLEOTIDE POOLS IN E. coli
113
E. coli are able to regulate the sizes of their deoxyribonucleoside triphosphate pools in such a way that they remain nearly constant, even when DNA synthesis is inhibited by novobiocin 3x (see Fig. 7). (2). E. coli cells deprived of their intracellular thymidine nucleotides respond immediately by accumulating dATP, indicating that the size of the dATP pool is in some way regulated in vivo by the pool size of the thymidine nucleotides. It is impossible from the present experiments to decide which of the nucleotides is the actual regulator, but since other known regulatory mechanisms involving thymidine nucleotides seem to implicate dTTPX-5, 32-85, and since the triphosphate is the immediate precursor of DNA, it seems likely that dTTP is also the active compound in the regulation of the pool size of dATP. (3)- In most cases thymine deprivation also causes accumulation of dCTP, but apparently this is counteracted by high levels of dATP*. (4). Under no experimental conditions so far investigated does the size of the dGTP pool change. In E. coli the reduction of ribonucleotides to the corresponding deoxyribonucleotides is known to take place at the ribonucleoside diphosphate leveP 6. All four normally occurring ribonucleoside diphosphates may be reduced by extracts of E. coli B 3~-a9. Furthermore, it is known from studies on Ehrlich ascites-tumor cells4°, chick-embryo extracts 41 and mouse leukemia cells42, that these reactions are inhibited in vitro by one or more of the deoxyribonucleoside triphosphates. In view of these findings and the fact that the ribonucleoside diphosphate reductases are the enzymes that connect the ribonucleotide and deoxyribonucleotide metabolism, the altered pools of dATP and dCTP observed may be explained by changes in the activities of the corresponding reductases following thymine starvation. Recently BlSWAS, HARDY AND BECK~ have shown that thymine starvation of E. coli 15 T- and treatment of E. coli B with dFU result in an immediate rise in the CDP reductase activity of the cells. Working with E. coli 15 T - A - U - we have previously reported 1° that if no protein synthesis occurs throughout the period of thymine starvation, the increase in dCTP content is prevented, without affecting the dATP accumulation. It was therefore proposed 1° that derepression of CDP reductase was responsible for the increased dCTP pools observed in thymine-deprived cells. Since all the mutants used in the present work are able to accumulate dCTP under appropriate conditions of thymine deprivation, it seems likely that derepression of the CDP reductase is a general response to restrictions in the intraceUular pool of thymidine nucleotides in E. coll. Assuming that dATP acts as a specific inhibitor of the CDP reductase 4°, it is possible to explain why dCTP fails to accumulate under certain conditions of thymine starvation, since these are the conditions where by far the highest concentrations of dATP are achieved (Fig. 2A). Only in cases where the dATP accumulation levels off after 75 % of a doubling time (Fig. 2B), while CDP reductase activity continues to rise (Fig. 6 of ref. 7), is any expansion of the dCTP pool observed. The lag in dCTP accumulation observed in the two E. coli 15 T - strains (Fig. 3B) could likewise be explained by the inhibitory action of the initially increasing dATP pool on the derepressed CDP reductase (Fig. 3A). In agreement with this, HOLMGREN, REICHARD AND THELAND~R xx have shown " Unpublished experiments from our laboratory with E. coli B M - T - have shown t h a t neither the pool o1 dAMP nor t h a t of dADPchanges significantly during thymine starvation (see METHODS). Biockim. Biophys. Acta, I29 (I966) xo4-~x5
114
J. NEUHARD
that dATP is a strong inhibitor of the purified CDP reductase from E. coli B. They found that the inhibition was competitive with the activating effect of ATP so that the determinant of the reaction rate in vitro was the ratio dATP/ATP. The agreement is even more striking if one compares our dATP/ATP ratios determined in vivo (Fig. 3A) with those given by Fig. 7 in their publication n. A dATP/ATP ratio of 0.05, i.e., the value found in exponentially growing E. coli, would in their assay give an inhibition of about 25 %, while ratios of o.15 and 0.27, which are the final values obtained after 9o-min thymine starvation of E. coli 15 T - and E. coli B T- strains, would show inhibitions of about 80 % and 95 %, respectively, in their assay, In view of the findings discussed above it is tempting to assume that the effect of dTTP on the dATP pool could be explained analogously by an inhibitory action of dTTP on the reductase converting ADP to dADP. Since, however, all values given in this work represent pool sizes, and any pool is determined by different anabolic and catabolic reactions, it is evident that the regulatory influence of dTTP on the dATP pool might be explained in other ways.
ACKNOWLEDGEMENTS
The author wishes to thank Dr. A. MUNCH-PETERSEN for her encouragement and continued interest in this work, and Dr. J. L. INGRAHAM for reading the manuscript. The excellent technical assistance of Mrs. K. DJURHUUS is gratefully acknowledged. This work was supported by grants from Carlsberg-Fondet and NOVOFondet.
REFERENCES
I 2 3 4 5 6 7 8 9
IO ii 12 13 14 15 16 17 18 19 20 21 22 23
H. IVES, P. A. MORSE, Jr. AND VAN R. POTTER, J. Biol. Chem., 238 (i963) 1467. R. BREITMAN, Bioehim. Biophys. Acta, 67 (1963) 153. OKAZAKI AND A. KORNBERG, J. Biol. Chem., 239 (1964) 275. BRESNICK, U. B. THOMPSON, H. P. MORRIS AND A. G. LIEBELT, Biochem. Biophys. Res. Commun., 16 (1964) 278. M. GREEN, M. PINA AND V. CHAGOYA, J. Biol. Chem., 239 (1964) 1188. T. R. BREITMAN AND R. M. BRADFORD, Biochem. Biophys. Res. Commun., 17 (1964) 786. C. BISWAS, J. HARDY AND W. S. BECK, J. Biol. Chem., 24o (1965) 3631. A. MUNCH-PETERSEN AND J. NEUHARD, Biochim. Biophys. Acta, 80 (1964) 542. A. MUNCH-PETERSEN AND J. NEUHARD, Intern. Congr. Biochem., 6th, New York, 1964, Abstracts IX. J. NEUHARD AND A. MUNCH-PETERSEN, Bioehim. Biophys. Acta, 114 (1966) 61. A. HOLMGREN, P. REICHARD AND L. THELANDER, Proe. Natl. Acad. Sci. U.S., 54 (1965) 830. H. D. EARNER AND S. S. COHEN, J. Bacteriol., 68 (1954) 80. D. KANAZIR, H. D. EARNER, J. G. FLAKS AND S. S. COHEN, Biochim. Biophys. Acta, 34 (1959) 341 • D. KORN AND A. WEISSBACH, Biochim. Biophys. Acta, 61 (1962) 775. C. G. KURLAND AND O. MAALOE, J. Mol. Biol., 4 (1962) 193. J. D. FRIESEN AND O. MAALOE, Biochim. Biophys. Acta, 95 (1965) 436. K. RANDERATH AND E. RANDERATH, J. Chromatog., 16 (1964) i i i . E. RANDERATH AND K. RANDERATH, J. Chromatog., I6 (1964) 126. J. NEUHARD, E. RANDERATH AND K. RANDERATH, Anal. Biochem., 13 (1965) 211. K. RANDERATH AND E. RANDERATH, Anal. Bioehem., 13 (1965) 575. S. S. COHEN, J. G. FLAKS, H. D. EARNER, M. R. LOEB AND J. LICHTENSTEIN, Proc. Natl. Acad. Sci. U.S., 44 (1958 ) lOO4. S. S. COHEN AND H. D. EARNER, Proc. Natl. Acad. Sci. U.S., 4 ° (1954) 885. H. MENNIGMANN AND W. SZYBALSKI, Biochem. Biophys. Res. Commun., 9 (1962) 398.
D. T. R. E.
Biochim. Biophys. Acta, 129 (r966) lO4-115
NUCLEOTIDE POOLS IN E. coli 24 25 26 27 28 29 3° 31 32 33 34 35 36 37 38 39 4° 41 42
115
J. G. FLAKS AND S. S. COHEN, J. Biol. Chem., 234 (1959) 2981. K. O. HARTMANN AND C. HEIDELBERGER, J. Biol. Chem., 236 (1961) 3oo6. R. W . BROCKMANN, J. M. D A v i s AND P. STUTTS, Biochim. Biophys. Acta, 4 ° (196o) 22. W . E. RAZZEL AND P. CASSHYAP,J. Biol. Chem., 239 (1964) 1789. O. MAALOE, J. Cellular Comp. Physiol., Suppl. I, 62 (i963) 31. K . RASMUSSEN, p e r s o n a l c o m m u n i c a t i o n . S. HENNINGSEN AND J. NEUHARD, u n p u b l i s h e d results. D. H. SMITH AND B. D. DAVIS, Biochem. Biophys. Res. Commun., 18 (1965) 796. E. SCARANO,G. GERACI, A. POLZELLA AND E. CAMPANILE, J. Biol. Chem., 238 (1963) P C 1556. C. F. MALEY AND F. MALEY, J. Biol. Chem., 239 (1964) 1168. F. MALEY AND C. F. MALEY, J. Biol. Chem., 24o (1965) P C 3226. W . H. FLEMMING AND M. J. BESSMAN, J. Biol. Chem., 24 ° (1965) P C 41o8. P. REICHARD, A. BALDSTEN AND L. RUTBERG, J. Biol. Chem., 236 (1961) 115o. P. REICHARD, J. Biol. Chem., 237 (1962) 3513. L. E. BERTANI, A. H-~GGMARK AND P. REICHARD, J. Biol. Chem., 238 (1963) 3407 . A. LARSSON, J. Biol. Chem., 238 (1963) 3414 . H. KLENOW, Biochim. Biophys. Acta, 61 (1962) 885. P. REICHARD, Z. N. CANELLAKIS AND E. S. CANELLAKIS,J. Biol. Chem., 236 (1961) 2514. N. R. MORRIS, P. 1REICHARD AND G. A. FISHER, Biochim. Biophys. Acta, 68 (1963) 93.
Biochim. Biophys. Acta, 129 (1966) lO4-115