Eur. J. Biochem. 161, 565-569 (1986) 0 FEBS 1986

Hexokinase PI1 from Saccharomyces cerevisiae is regulated by changes in the cytosolic Mg2 +-free ATP concentration Fernando MORENO, Teresa FERNANDEZ, Rosa FERNANDEZ and Pilar HERRERO Departamento Interfacultativo de Bioquimica, Facultad de Medicina, Universidad de Oviedo

(Received April 28/September 3, 1986) - EJB 86 0421

Hexokinase PI1 is not inhibited by high Mg-ATP concentrations if the MgZ+-free ATP is kept at low levels (0.01 mM) in the assay mixture. Hexokinase PI activity is not affected either by Mg2+-free ATP nor by free Mg2+ in the assay mixture. Thus, hexokinase PI and PI1 activities appear not to be regulated by substrate inhibition as proposed previously [Kopetzki, E. & Entian, K. D. (1985) Eur. J . Biochem. 146,657-6621. However, the level of Mg2+-free ATP in the hexokinase PI1 assay mixture strongly affects the enzyme activity by decreasing the V,,, and increasing the K , value for Mg-ATP from 0.15 mM to 5.0 mM.

The physiological role of this inhibition, which has not been described previously, was investigated by determining the cytosolic ATP and Mg2 + concentrations in yeast cells grown under derepressing and repressing conditions. Derepression is accompanied by an important loss of Mgz+ from the cells, maintaining the ATP concentration constant. This produces an increase of MgZ+-free ATP in the cytosol from 0.01 mM to 0.1 mM. This free ATP concentration would lead to a maximal inhibition of hexokinase PII.

Three glucose-phosphorylating enzymes, glucokinase and two hexokinase isoenzymes PI and PII, have been described in yeasts [l ,2]. The physiological role of these enzymes remains unclear. One of them is sufficient for growth on glucose, and either one of the two hexokinases for growth on fructose. In the last few years, special attention has been given to hexokinase PII; it has been described as an enzyme with both catalytic and regulatory properties [3 - 51.

Hexokinase PI1 can be inactivated by addition Of D-XylOSe to the culture medium, in which case the growth rate as well as rate of glucose consumption are similar to yeasts grown without D-xylose [5, 61. These results indicate that hexokinase PI1 is not essential for glucose phosphorylation. However, catabolic repression of invertase and maltase was relieved in yeast with xylose-inactivated hexokinase PI1 [5] and by a mutation in the hexokinase PI1 structural gene (hex 1) [A. This suggested a direct role of hexokinase PI1 in carbon catabolite repression of invertase and maltase. This hypothesis is supported by the isolation of nonrepressible mutants (hex lr) which were only defective in the regulatory domain with the hexokinase PI1 enzyme still active catalytically [8]. In such mutants the defect in catabolite repression was due to the defective hexokinase PI1 but independent of the capacity for glucose phosphorylation.

The interaction between hexokinase PI1 and an unknown signal mediating catabolite repression could produce a conformational change of the enzyme triggering derepression.

Correspondence fo F. Moreno, Departamento Interfacultativo de Bioquimica, Facultad de Medicina, Universidad de Oviedo, E-33071 Oviedo, Spain

Enzymes. Hexokinase, ATP: D-hexose 6-phosphotransferase (EC 2.7.1 . l ); invertase, P-D-fructofuranoside fructohydrolase (EC 3.2.1.26).

The detailed molecular mechanism is unknown at the mo- ment. Catabolite repression in cells being transferred from a non-repressing ethanol medium to a repressing glucose medi- um, does not depend on the accumulation of glycolytic or Krebs cycle intermediates [5 , 91.

Recently, both hexokinase isoenzymes were described to be strongly inhibited by high physiological concentrations of ATP (2mM) [lo]. Our results suggest that the substrate inhibition of hexokinase PI1 is produced by the Mg2+-free ATP accumulation in the assay mixture. This paper describes the strong decrease in the concentration of magnesium ions in the cytosol of cells being transferred from a repressing glucose medium to a non-repressing ethanol medium. The possible role of free ATP on the regulation of hexokinase PI1 is discussed.

MATERIALS AND METHODS

Materials

Sigma (St Louis, MO) reagents and enzymes were used throughout. Yeast extract was obtained from Difco Lab. Hydroxyapatite HA-Ultrogel was purchased from LKB. DEAE-Sephadex and chromatofocusing gel PBE94 were obtained from Pharmacia.

Strains and growth conditions

Saccharomyces cerevisiae strain G-517 was used as a wild type; two mutants obtained from the Yeast Genetic Stock Center, P2T22D (aadel hxkl glkl) lacking hexokinase PI and glucokinase, and PlT8C (aadel his1 hxk2 glkl) lacking hexokinase PI1 and glucokinase, were investigated.

566

The cells were grown in flasks with 300 ml medium con- taining 1 % yeast extract and 2% glucose, supplemented with 30 mg/l adenine for growing mutants on a rotatory shaker at 28°C. Growth was followed by determination of the absorbance at 600 nm. Yeasts were harvested from the cultures at an A of 0.9 when the sugar concentration reached 50 mM (repressed cells), or at an A of 2.7 when the glucose was exhausted in the medium (derepressed cells). Under the former conditions, hexokinase PI1 represents between 64 - 70% of the total glucose-phosphorylating activity of the cell. Under the latter conditions, hexokinase PI1 represents only about 30% of the total glucose-phosphorylating capacity of the cell [6]. The cells were washed twice with distilled water and resuspended in 10 mM potassium phosphate buffer, pH 7.0 for further treatment.

Sources ojhexokinase PI and PII

Hexokinase PI was partially purified by hydroxyapatite chromatography of cell-free extracts from PlT8C mutant. Yeast cells were homogenized for 20 min in a cell homogenizer (Vibrogen Cell Mill) with glass beads (0.5 mm in diameter) in 25 ml of 10 mM potassium phosphate buffer, pH 7.0. The beads were removed and washed with 100ml buffer. The washing buffer and the extract were then pooled and centrifuged at 39 000 x g for 20 min. The pellet was discarded. The cell-free supernatant was applied to a hydroxyapatite HA-Ultrogel column (2.5 x 15 cm) equilibrated with 10 mM potassium phosphate buffer, pH 7.0. Hexokinase PI activity was eluted with 500 ml of a linear gradient of 10-350 mM potassium phosphate buffer, pH 7.0. The fraction size was 5 ml. Fractions containing hexokinase PI activity were pooled, stored at 4°C and used as enzyme source. Starting from a preparation with 90 U hexokinase PI (specific activity of 0.2 U/mg), partially purified hexokinase PI was recovered with a specific activity of 1.6 U/mg. Pure hexokinase PI specific activity was 60 U/mg.

Three different sources of hexokinase PI1 have been used with identical results : (a) partially purified hexokinase PI1 obtained from P2T22D mutant by the procedure described above for hexokinase PI; (b) Sigma hexokinase PI1 (H-5875), which was further purified by chromatography on hydroxy- apatite HA-Ultrogel as described above; (c) hexokinase PI1 from S. cerevisiae G-517 purified using a four-step procedure: cell-free extract preparation, hydroxyapatite chromatog- raphy, chromatofocusing and DEAE-Sephadex A-50 chroma- tography. Starting from a preparation with 228 U hexokinase PI1 (specific activity 0.4 U/mg), pure hexokinase PI1 was re- covered with a specific activity of 132 U/mg and a yield of 26%.

Differential extraction of soluble pools from cytosol and vacuole

Differential extraction of soluble pools was done essen- tially as described for Saccharomyces carlsbergensis in [ll]. This method is also similar to those described for Candida or S. cerevisiae [12, 131.

Wet cells (100 mg) were collected and washed five times with 1 ml double-distilled water. The washing of the cells and the extraction of soluble pools were carried out by centrifuga- tion at 12000 x g for 30 s in an Eppendorf microfuge at 4°C. Extracts were prepared as follows.

Extract I , Magnesium ions adsorbed on the cell wall. The cells were washed three times with 1 ml double-distilled water

followed by three washings with 1 ml of 0.25 M sodium ace- tate buffer, pH 4.8, and three of 1 ml of 0.7 M sucrose. No further release of magnesium ions was achieved by a second treatment of the cells with acetate buffer and sucrose.

Extract 2. Magnesium ions in the cytosol. To achieve the specific permeabilization of plasma membrane to low- molecular-mass compounds and ions, the cells were treated for 9 min with 3 ml of 1% cytochrome c solution in 0.25 M sodium acetate buffer, pH 4.8, containing 0.7 M sucrose. The fluorescent dye berberine sulphate (final concentration 10 pg/ ml cell suspension) was added to cytochrome-c-treated cells and fluorescence was immediately controlled in a Zeiss fluo- rescence microscope. After 9 min of treatment, the cytoplasm of almost all cells fluoresced brightly, with vacuoles appearing as dark spots against this background. Afterwards the cells were washed three times with 1 ml of 0.25 M sodium acetate buffer, pH 4.8, containing 0.7 M sucrose. The total supernatant (6 ml) was collected as cytosolic extract.

Extract 3. Magnesium ions in the vacuoles. The vacuolar membranes were disrupted by osmotic shock. The cells were extracted three times with 1 ml double-distilled water followed by three washings with 1 ml 0.25 M sodium acetate buffer, pH 4.8. The vacuole ions were determined in the combined extract (6 ml).

Extract 4. Magnesium ions in large vacuoles and bound ions. The remaining magnesium ions were extracted with 1 ml 0.5 M perchloric acid for 5 min. This resulted primarily in the release of ions from large vacuoles which resisted the osmotic shock, as well as their bound and water-insoluble forms [l 11.

Analytical methods

Hexokinase was assayed as described by Bergmeyer in [14]. NADPH formation was determined at 340 nm in a l-ml cuvette containing: 10 mM glucose, 20 mM Tris/HCl buffer, pH 7.5, 7.5 mM MgC12, 0.5 mM ATP, 0.65 mM NADP', 2 units glucosed-phosphate dehydrogenase and 10 - 50 pl of a suitably diluted enzyme preparation.

The external invertase activity was assayed in whole cells by a two-step method as described previously [15]. Glucose in the culture supernatant was assayed as described in [16].

ATP was extracted as described in [17] and was determined in a coupled assay with 3 units hexokinase PI1 (Sigma H-5875), 2 units glucose-dphosphate dehydrogenase (Sigma G-7750) and 0.65 mM final concentration of NADP' (Sigma N-0505) in a 20mM Tris/HCl buffer, pH 7.5, containing 7.5 mM MgClz and 10 mM glucose in a total volume of 1 ml.

Arginine was determined as described in [18]. Magnesium was determined in an atomic absorption

spectrophotometer (Perkin Elmer model 2280).

RESULTS Effect of Mg2+-free ATP on hexokinase PI and PII activity

Physiological concentrations of 2 mM ATP have been re- ported to strongly inhibit the hexokinase isoenzymes [lo]. Fig. 1 shows the effect of different Mg-ATP, MgZ+-free ATP and free Mg2+ concentrations on the activity of hexokinase PI and PII. The concentrations of Mg-ATP, free ATP and free Mg2+ were calculated for each addition of ATP to the reaction mixture of hexokinase employing the association constant of the Mg-ATP complex (K, = 1 x lo4 M) [19]. For sources of the isoenzymes see Materials and Methods. The enzymes were saturated with glucose (10 mM) and the magnesium ion con-

567 I 1 I , I I I 1 I : I I I I I I

& - I o-//oL-o I --00-0-0-0-0.

O/'O / loo - /oOP-o-o-

i : o/o

:I i O\. -- O\.

i ip 1'3

fp \o

6o a o / / \ -

~ . - , , . - o 00-0'

E L 40

20 9 - A C

I I r I

0 1 2 3 s o 0.1 0 . i ' I 2 0 1 2 3 4 5 0 7

M g - A T P ImM1 Free A T P tmMl Free Mg2irnMJ

Fig. 1. Effect of dflerent concentrations of Mg-ATP, free ATP andfree Mg2+ on hexokinase PI and PII activity. Purified hexokinase PI (0) and hexokinase PI1 (0 ) were assayed in the presence of increasing ATP concentrations at constant concentrations of Mg2+ (7.5 mM) and glucose (10 mM). Identical results were obtained by using the three different sources of hexokinase PI1 indicated under Materials and Methods. The results actually presented in the figure were obtained using a partially purified hexokinase PI1 preparation. The concentrations of Mg- ATP, free ATP and free Mgz+ were calculated for each addition of ATP to the reaction mixture of hexokinase employing the association constant of the Mg-ATP complex (K, = 1 x lo4 M)

Table 1. Content of magnesium ions in various cell compartments

3 -

2 -

1 I v /

2

V

1

0 10 2 0

1 I M g - A T P t m M )

Fig. 2. Effect of different concentrations of Mg-ATP on the hexokinase PII activity at low and high concentrations of free ATP. Partially purified hexokinase PI1 was assayed in the presence of low free ATP concentrations (0) by using a constant MgZ+/ATP ratio of 10 or in the presence of high free ATP concentrations (0 ) by using a constant MgZ+/ATP ratio of 0.5. The data have been replotted in the inserted figure to show the typical Michaelis-Menten kinetics obtained. Identi- cal results were obtained by using the three different sources of hexokinase PI1 indicated under Materials and Methods

centration was kept constant (7.5 mM), as described by Bergmeyer [14].

Hexokinase PI shows a typical substrate saturation curve with perfect Michaelis-Menten kinetics (Fig. 1 A). Maximal activity was reached at 1.5 mM Mg-ATP. In our experience, hexokinase PI activity was not affected by high Mg-ATP concentrations, in contrast to the results described in [lo]. Neither high free ATP nor free Mg2+ concentrations affected the enzyme activity (Fig. 1 B,C).

Hexokinase PI1 activity, however, was strongly inhibited by high Mg-ATP concentration (Fig. 1 A), where maximal

Extract Mg2+ in Mg2+ in repressed cells derepressed cells

pmol/g wet wells (% total)

1. Cell wall 4.70 (16) 3.60 (30) 2. Cytosol 7.08 (24) 2.30 (19) 3. Vacuoles 9.25 (31) 1.66 (14) 4. Largest vacuoles 8.70 (29) 4.22 (37)

Total 29.73 (100) 11.78 (100) and bound ions

activity was obtained at 0.75 mM Mg-ATP. A further increase in the Mg-ATP concentration above 1 mM resulted in an inhibition of enzymatic activity with a decrease of 70% at 4 mM Mg-ATP. Under the conditions of this assay not only the Mg-ATP concentrations changed, but the concentrations of Mg2+-free ATP and free Mg2' also varied (Fig. 1B,C). Thus, the inhibition of hexokinase PI1 here could be due to an increase of free ATP as well as a decrease of free Mg2 + in the assay mixture. Submaximal doses of ATP showed gradual effects on hexokinase PI1 activity. Thus 0.06 mM free ATP suppressed activity by 50% while 0.15 mM ATP nearly led to a maximal inhibition of about 70%.

The K , values for ATP, using 10mM glucose as cosubstrate, were for hexokinase PI and PI1 0.2mM and 0.15 mM, respectively. These results are in the same range as the values described previously [lo, 201.

Using different Mg2+/ATP ratios it is possible to keep either the free ATP or the free Mg2+ concentrations constant. Thus, we used an Mg2+/ATP ratio of 10 to keep the free ATP concentration at 0.01 mM. The results obtained by varying the concentrations of the Mg-ATP complex from 0.024 mM to 10 mM and the free Mg2+ concentration from 0.023 mM to 90 mM are summarized in Fig. 2 . Evidently, the inhibition of hexokinase PI1 is counteracted, despite which the affinity of the enzyme for Mg-ATP remained unchanged (K , = 0.16 mM).

On the other hand, keeping the MgZ+/ATP ratio at 0.5 gives a free Mg2+ concentration of 0.08 mM, simultaneously increasing the Mg-ATP concentration from 0.01 mM to

568

Table 2. Intracellular distribution of magnesium ions as determined by cytochrome c fractionation The concentrations of ATP and Mg2+ were calculated on the basis that 1 g wet yeast contains 0.6 ml cell sap. The vacuolar volume is 25% of the protoplast volume

Cells (metabolic state) Glucose Invertase ATP Mg2+ in Gradients

cytosol vacuoles vacuoles) (supernatant) (whole cells) (cytosol/

mM mU/ml mM

Repressed 50 74 2.1 14 77 1:5 Derepressed 0 1330 2.1 4 14 1:3

Table 3. Changes in cytosolic free ATP concentration during derepression of S . cerevisiae The concentrations of Mg-ATP, free ATP and free Mg2+ were cal- culated employing the association constant of the Mg-ATP complex (K. = i x 104 M)

~

Cells Total Total Mg-ATP free ATP free Mgz+ ATP Mg2+

mM

Repressed 2.10 14.00 2.08 0.01 11.90 Derepressed 2.10 4.00 2.00 0.10 2.00

9.9 mM and the free ATP concentration from 0.04 mM to 10.1 mM. As shown in Fig. 2, high concentrations of free ATP led to an inhibition of hexokinase PI1 activity. Thus, the V,,, as well as the affinity for the Mg-ATP complex were decreased (with a K,,, of 5 mM for the latter).

These results suggest that the levels of free ATP can be of maximal importance in the regulation of hexokinase PII. The physiological role of this inhibition, which has not been de- scribed previously, was then studied by determining the levels of intracellular Mg2+ and ATP at different metabolic states of yeast cells.

Intracellular concentrations of Mg2+ and ATP in repressed and derepressed cells

As previously reported, the soluble pools of Mg2+ can be differentially extracted from the cytosol and the vacuoles of S. carlsbergensis by using cytochrome c 111,211. Accordingly, the treatment of S. cerevisiae cells with isotonic cytochrome c rendered the plasma membrane permeable to the fluorescent dye berberine sulphate, whereas the vacuoles remained impermeable to the dye.

The cytochrome c treatment breaks the plasma membrane, enabling the extraction of the cytosolic pool of small molecules. 7% of the total arginine was found in this extract, which is in agreement with a low arginine content in yeast cytoplasm [22]. The vacuolar membranes are only disrupted after osmotic shock. This extract containing the vacuolar pool is rich in arginine (80%), an amino acid known to be accumulated in these organelles [22]. From fluorescence microscope observations and from arginine determinations, it was inferred that the isotonic cytochrome c treatment only breaks the plasma membrane and that after the osmotic shock the vacuolar membranes were disrupted and the vacuolar pools liberated.

Cells of strain G-517 were harvested from cultures when the glucose concentration reached 50 mM (repressed cells)

or when glucose had been exhausted (derepressed cells). In repressed cells about 16% of the total Mg2+ was released, compared to 30% in derepressed cells successively treated with water, acetate buffer and 0.7 M sucrose (Table 1). Modification of the plasma membrane with the cytochrome c released 24% (repressed cells) and 19% (derepressed cells) of the total Mg"; 60% (repressed cells) and 50% (derepressed cells) of the cellular MgZf was found in the fraction stable to cytochrome c. This fraction was further extracted with double-distilled water and acetate buffer to determine the amounts of Mg2+ released by osmotic shock. All but the largest vacuoles burst and were no longer visible under the microscope. The Mg2+ ions from the remaining vacuoles were liberated by a 5-min incubation with 0.5 M perchloric acid (Table 1). The presence of glucose in the culture medium produces a threefold increase in the total Mg2' of the cell. This Mgz+ accumulation was also described by Lichko et al. [21] for S. carlsbergensis.

Table 2 shows the changes in the concentrations of Mg2' and ATP in each cellular compartment under repressing and derepressing conditions. To calculate the intracellular concen- trations of MgZf and ATP, we assumed that 1 g wet yeast contains 0.6 ml cell sap [23, 241 and that the vacuolar volume is 25% of the protoplast volume [ l l , 211. Thus, invertase was strongly repressed in yeasts grown on glucose and collected in the logarithmic phase of growth. Invertase synthesis was derepressed when the yeasts were grown on ethanol, either supplied externally or produced during glucose fermentation.

The levels of ATP were identical in both conditions but the cytosolic Mg2+ levels were about fourfold higher under repressing conditions than during derepression. The vacuolar Mg2+ levels were about fivefold increased under the re- pressing conditions compaired to derepression.

Also of special interest is the unequal distribution of Mg2+ between cytosol and vacuoles. In vacuoles the magnesium content was 5.5 and 3.5 times increased in repressed and derepressed cells, respectively. These values underestimate the real concentrations since the largest vacuoles did not burst under the osmotic shock and their Mg2+ content was deter- mined only in 0.5 M perchloric acid. The calculated Mg2+ concentrations in vacuoles and cytoplasm points to the exis- tence of Mg2 + concentration gradients within the yeast cell.

From these results and using the association constant of the Mg-ATP complex [19], the cytosolic Mg-ATP, free ATP and free Mg2 + concentrations were calculated for repressed and derepressed conditions (Table 3). The concentrations of the Mg-ATP complex scarcely changed from repressed (2.08 mM) to derepressed cells (2.0 mM). However, the Mgz+- free ATP concentration was about tenfold higher in the derepressed conditions than during repression. Free Mg2 +

appeared to be sixfold higher in the repressed conditions.

569

DISCUSSION

Under our assay conditions, hexokinase PI1 is not in- hibited by high Mg-ATP concentrations if the Mg2 +-free ATP is keep at low concentrations in the assay mixture (Fig. 2). Thus, the hexokinase PIi activity appears not to be regulated by substrate inhibition as proposed previously [lo].

However, our results suggest that high Mg2+-free ATP or low free Mg2+ concentrations could be responsible for the inhibition of hexokinase PI1 (Fig. lB,C). Changes in free Mgz+ concentrations in the assay mixture from 4mM to 90 mM, at low constant Mg2+-free ATP levels (0.01 mM), had no effect on the enzymatic activity (Fig. 2). Changes in the free Mgz+ concentrations from 4 mM to 7 mM led to an about 100% activation of the enzyme, but under the conditions applied the free ATP also decreased from 0.1 mM to 0.01 mM (Fig. 1 C). Combining the present data, we suggest that free Mg” has no effect on the inhibition of hexokinase P 11.

The level of Mg2+-free ATP in the hexokinase PI1 assay mixture, strongly affects the enzyme activity. This hypothesis was supported by the results shown in Fig. 2. At constant free MgZ + concentrations and increasing free ATP concentrations, the hexokinase PI1 activity was inhibited both by decreasing the V,,, and increasing the K, for Mg-ATP (Fig. 2).

The physiological role of this inhibition observed in vitro was investigated by determining the cytosolic ATP and Mgz+ concentrations in yeast cells grown under repressing and derepressing conditions (Table 2) . Derepression is accom- panied by an important loss of MgZf from the cells maintaining ATP concentration constant. The change from repressing to the derepressing conditions resulted in variations of the intracellular Mg2+/ATP ratio from 7 to 2. This produces an increase of free ATP from 0.01 mM to 0.1 mM (Table 3). However, this cytosolic ATP concentration would lead to a maximal inhibition of hexokinase PI1 as concluded from the results shown in Fig. 1 B. It is necessary to take into account that the concentration of free and complexed ATP is (in the intact cell) not only the result of concentrations of ATP and Mg”, but also of the many other metabolites and coenzymes known to form complexes with MgZ+, but nucleotide and glycolytic or Krebs cycle intermediate concentrations do not change when cells are transferred from a non-repressing ethanol medium to a repressing glucose medium [5, 91.

Thus, our results indicate that MgZf-free ATP plays an important role in the in vivo regulation of hexokinase PI1 activity of yeast. A regulation by adenine nucleotides as de- scribed for hexokinase PIi has not been previously shown for other enzymes. Recently Mg2+-free ATP has been shown to be involved in the regulation of Kf and Ca2+ channels in the membranes of animal cells. Intracellular free ATP directly blocks K + channels in pancreatic B cells [25] or skeletal muscle cells [26] and activates calcium channels of sarcoplasmic retic- ulum membranes of skeletal muscle cells [27]. Blocking and activation of the channels is ATP-specific and appears to be independent of ATP metabolism.

Hexokinase PI activity was affected neither by free ATP nor by free MgZf in the assay mixture (Fig. 1B,C). Hexokinase PI1 was suggested to play a unique role in the

regulation of carbon catabolite repression in yeast. In this background, the differences in the regulation of the hexokinase isoenzymes reported in this paper may be of major importance.

The authors thank Prof. S. Gascon and Dr F. Barros for their interest in this work and many interesting discussions and suggestions, and Prof. A. Panek and Dr J. Heinisch for the critical reading of the manuscript. This work was supported by a grant from Comisibn Asesora de Znvestigacibn Cientijica y TPcnica.

REFERENCES 1. Colowick, S. P. (1973) in The enzymes (Boyer, P. D., ed.) 3rd edn,

2. Barnard, E. A. (1975) Methodr Enzymol. 42C. 6-20. 3. Frohlich, K. U., Entian, K. D. & Mecke, D. (1984) Mol. Gen.

4. Entian, K. D., Kopetzki, E., Frohlich, K. U. & Mecke, D. (1984)

5. Fernkndez, R., Herrero, P. Gascbn, S. & Moreno, F. (1984) Arch.

6. Fernandez, R., Herrero, P. & Moreno, F. (1985) J. Gen.

7. Entian, K. D. & Mecke, D. (1982) J. Biol. Chem. 257,870-874. 8. Entian, K. D., Hilberg, F., Opitz, H. & Mecke, D. (1985) Mol.

9. Entian, K. D. & Zimmermann, F. K. (1980) Mol. Gen. Genet.

10. Kopetzki, E. & Entian, K. D. (1985) Eur. J. Biochem. 146, 657-

1 1 . Okorokov, L. A., Lichko, L. P. & Kulaev, I. S. (1 980) J . Bacteriol.

12. Eilarn, Y. , Lavi, H. & Grossowicz, H. (1985) J. Gen. Microbiol.

13. Huber-Walchli, V. & Wiemken, A. (1979) Arch. Microbiol. 120,

14. Rergmeyer, H. U . (1965) Methods of enzymatic analysis, 2nd ed,

15. Gascon, S. & Lampen, J. 0. (1968) J . Biol. Chem. 243, 1567-

16. Moreno, F., Rodicio, R. & Herrero, P. (1981) Cell Mol. Biol. 27,

17. Entian, K. D., Zimmermann, F. K. & Scheel, I. (1977) Mol. Gen.

18. Ramos, F., Thuriaux, P., Wiame, J. M. & Bechet, J. (1970) Eur.

19. Orentlicher, M., Brandt, P. W. & Reuben, J . P. (1977) Am. J .

20. Sols, A., de la Fuente, G., Villar-Palasi, C. & Asensio, C. (1958)

21. Lichko, L. P., Okorokov, L. A. & Kulaev, I. S. (1980)J. Bacteriol.

22. Wiemken, A. & Diirr, M. (1974) Arch. Microbiol. 101,45-57. 23. Conway, E. J . & Downey, M. (1950) Biochem. J. 47,347-355. 24. Eraso, P. & Gancedo, J. M. (1984) Eur. J . Biochem. 141, 195-

25. Cook, D. L. & Hales, C. H. (1984) Nature (Lond.) 311, 271-

26. Spruce, A. E., Standen, M. B. & Stanfield, P. R. (1985) Nature

27. Smith, J. S., Coronado, R. & Meissner, G. (1985) Nature (Lond.)

pp. 1-48, Academic Press, New York.

Genet. 194, 144-148.

Mol. Gen. Genet. 198, 50-54.

Microbiol. 139, 139- 142.

Microbiol. 131,2705 - 2709.

Cell Biol. 5, 3035 - 3040.

177.345 - 350.

662.

144.661 -665.

131,623-629.

141 - 149.

p. 117, Academic Press, New York.

1572.

589 - 592.

Genet. 156.99-105.

J. Biochem. 12,40-47.

Physiol. 233, C127-C134.

Biochim. Biophys. Acta 30,92-101.

144,666 - 671.

198.

273.

(Lond.) 316.736-738.

316,446-449.