Phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase from Saccharomyces cerevisiae.

Phosphorylation of fructose-1,6-bisphosphatase with cyclic AMP-dependent protein kinase from yeast is accompanied by a 50% decrease in the catalytic activity (Pohlig, G. and Holzer, H. (1985) J. Biol. Chem. 260, 13818-13823). Using reactivation of phoshorylated fructose-1,6-bisphosphatase as assay, a protein phosphatase was about 2,000-fold purified to electrophoretic homogeneity from Saccharomyces cerevisiae. Upon incubation with phosphorylated fructose-1,6-bisphosphatase the purified protein phosphatase not only reverses the 50% inactivation caused by phosphorylation, but also the previously observed change in the pH optimum and in the ratio of activity with Mg2+ or Mn2+. The phosphatase is strongly inhibited by heparin and fluoride. L-Carnitine, orthophosphate, pyrophosphate, and succinate inhibit to 50% at concentrations from 1 to 10 mM. The molecular mass of the native phosphatase was found to be 180,000 Da. Sodium dodecyl sulfate-gel electrophoresis suggested four subunits with a molecular mass of 45,000 Da each. Half-maximal activity was observed with 5 mM Mg2+ or Mn2+, the pH optimum of activity was found at pH 7. Using polyclonal antibodies, disappearance of 32P-labeled fructose-1,6-bisphosphatase and concomitant liberation of the expected amount of inorganic [32P] phosphate was demonstrated.

* The costa of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "ucluertisemnt" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The designation of phosphorylated fructose-l,6-bisphosphatase follows the recommendations of the IUPAC-IUB Commission on Biochemical Nomenclature (J. Bwl. Chem. (1977) 252,5939-5941). that the phosphorylated fructose-1,6-bisphosphatase exhibits only about 50% of the catalytic activity of the nonphosphorylated enzyme (1,2,4,6) inspired us to develop an assay for phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase based on the measurement of the increase in catalytic activity which was expected to accompany the reaction: phosphorylated fructose-1,6-bisphosphatase + fructose-1,6-bisphosphatase + 4 orthophosphate. In the present paper such an assay and its application for the elaboration of a purification procedure and for characterization of phosphorylated fructose-l,6-bisphosphatase dephosphorylating protein phosphatase from Saccharomyces cerevisiue are described.

RESULTS
Using "assay 1" (cf. "Materials and Methods") phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase was purified from S. cereoisiae M1 in a six-step procedure as summarized in Table I. An about 2000-fold purification with a yield of 2% was obtained.
SDS3-polyacrylamide gel electrophoresis of the crude extract and of aliquots from the steps 4, 5, and 6 of the purification procedure is shown in Fig. 3. The electrophoretically homogenous enzyme resulting from step 6 was applied in its native form to gel filtration on Sephacryl S-300. With the assumption of the same relationship between Stokes radii and molecular weight for the standards and phosphorylated fructose-l,6-bisphosphatase dephosphorylating protein phosphatase, a molecular mass of 180,000 Da was calculated (Fig. 4). From SDS-polyacrylamide gel electrophoresis, a molecular mass of 45,000 Da was calculated. The native enzyme is therefore very probably a tetramer of four subunits with identical molecular mass.
When incubated at 50 "C in 10 min 50% of the activity of

TABLE I Purification of phosphorylated fructose-1,6-bisphosphatme dephosphorylating protein phosphatase from S. cerevisiae strain MI
The purification was performed as described under "Materials and Methods." Starting with 122 g, wet weight, of yeast cells about 45 pg of purified phosphorylated fructose-l,6-bisphosphatase dephosphorylating protein phosphatase were obtained in 4 days. Activity was assayed using assay 2.  Table I)  the purified enzyme disappear. The presence of 25 mM MgC1, somewhat stabilizes the enzyme: 50% inactivation in 10 min is now observed at 55 "C.
Mg2+ is a necessary cofactor for the phosphatase reaction. Half-maximal activity was observed at about 5 mM MgClz (Fig. 5). Maximal activation of the enzyme with Mn2+ leads to only about two-thirds of the activity of the maximal Mg2+ activated enzyme. Half-maximal activity was observed with 5 mM MnC12.
Dependence of phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase activity on the pH is shown in Fig. 6. Maximal activity was observed at pH 7.0.
To demonstrate dephosphorylation of phosphorylated fructose-1,6-bisphosphatase with the phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase contained in crude extract, 32P-labeled phosphorylated fructose-1,6-bisphosphatase was prepared as substrate from yeast cells preincubated with [32P]orthophosphate and then treated for 5 min with glucose as described previously (1). Immunoprecipitates with antibodies against fructose-1,6-bisphospha-  (Table I,   tase (reacting also with phosphorylated fructose-l,6-bisphosphatase) formed before and after 15,30, and 60 min incubation of F B p a~e ( 4~~P ) with the crude extract containing phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase were submitted to SDS-gel electrophoresis. As shown in Fig. 7, the 32P-labeled band of phosphorylated fructose-1,6-bisphosphatase disappears in the course of incubation. In the absence of MgCl,, the 32P band shows no significant change in its density during 60 min incubation ( l a n e 2 in Fig. 7). With phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase purified up to step 6 (cf . Table I) instead of crude extract the same time course of disappearance of the fructose-1,6-bisphosphatase 32P band as shown in Fig. 7 was obtained (data not shown). When phosphorylated fructose-1,6-bisphosphatase labeled in intact yeast cells with [3H]leucine instead of ["PJorthophosphate was incubated with phosphorylated fructose-l,6-bisphosphatase dephosphorylating protein phosphatase with or without MgC1, and then immunoprecipitated and submitted to SDS-gel electrophoresis no change in the density of the 3H band during 60 min incubation was observed (data not shown). This control experiment excludes proteolysis of F B p a~e ( 4~~P ) as an explanation for the disappearance of the "P-labeled fructose-l,6-bisphosphatase band shown in Fig. 7. To obtain additional evidence for dephosphorylation of F B P a~e ( 4~~p ) by the phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase, the liberated ["2P]orthophosphate was submitted to thin layer chromatography according to Takanami (8) and the density of the orthophosphate spot was determined on an autoradiogram. Parallelism of the time course of liberation of [32P]orthophosphate and of activation of phosphorylated fructose-1,6-bisphosphatase as determined in the standard assay is shown in Fig. 8. The phosphorylated and the nonphosphorylated forms of fructose-1,6-bisphosphatase exhibit characteristic differences in the pH optimum (5, 7) and in the dependence of activity on M e or Mn2+ (4,7). When phosphorylated fructose-1,6bisphosphatase is incubated with purified phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase in the presence of M$+, R(Mg/Mn) increases from 1.24 to 1.75 and Q(7/9) increases from 0.41 to 0.64 (Table 11). With preparations of unphosphorylated fructose-l,6-bisphosphatase, R(Mg/Mn) = 2.6 and Q(7/9) = 0.62 had been obtained (7). Concentrations of inhibitors necessary for 50% inhibition of phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase activity are depicted in Table 111. The inhibitors listed have also been found to inhibit protein phosphatases from mammalian tissues (11). In contrast to the inhibitory action on mammalian phosphatases no inhibitory effect on the yeast phosphatase was detected with spermine (10 mM), spermidine (10 mM), and poly-L-lysine (10 p~) (data not shown). Also a boiled extract from yeast did not inhibit. Incubation of the boiled yeast extract with ATP, M e , and cyclic AMP in the presence of protein kinase from beef heart did not produce inhibitory activity. Evidence for a heat stable, proteinaceous inhibitor of yeast phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase similar to the one found in mammalian tissues (11) was therefore not obtained.

DISCUSSION
During studies on phosphorylation/dephosphorylation of proteins in yeast, evidence was obtained for enzymes dephos-

TABLE I11
Inhibition ofphosphorylated fructose-1,6-bisphosphataseprotein phosphatase activity Phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase was incubated with the inhibitors for 10 min in the mixture of assay 1 (cf. "Materials and Methods"). In contrast to the routine start with fructose-l,6-bisphosphate the assay was started with phosphorylated fructose-1,6-bisphosphatase. phorylating phosphorylated glycogen phosphorylase (12), glycogen synthase (13,14), trehalase (15), NAD-glutamate dehydrogenase (16), and fructose-1,6-bisphosphatase (1,2). Partially purified and characterized were the protein phosphatases which dephosphorylate NAD-glutamate dehydrogenase in Candida utilis (16) and glycogen phosphorylase in baker's yeast (17). The fructose-l,6-bisphosphatase dephosphorylating enzyme from S. cerevisiae described in the present paper is probably different from these two previously characterized protein phosphatases from the following reasons. The NADglutamate dehydrogenase dephosphorylating activity elutes from DEAE-ion exchange columns at 0.2-0.35 M KC1 (16), whereas the protein phosphatase described here elutes at 0.1 M NaC1. The glycogen phosphorylase dephosphorylating activity elutes at 0.35 M KC1 (17) and is therefore also very probably not identical with our enzyme eluting at 0.1 M NaCl. Furthermore, the severalfold activation of glycogen phosphorylase by incubation with ATP and Mg' observed with the glycogen phosphorylase phosphatase (17) is not seen with the phosphorylated fructose-l,6-bisphosphatase dephosphorylating protein phosphatase described here.4 Under conditions where 4 pg of phosphorylated fructose-1,6-bisphosphatase are dephosphorylated in 1 min, the purified phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase does not dephosphorylate, i.e. inactivate, 2 pg of phosphorylated trehalase purified from baker's yeast in 30 min.5 From the phosphatase activity of the crude extract shown in Table I it is calculated that 1.6 pg of phosphorylated fructose-1,6-bisphosphatase can be dephosphorylated per min and per g, wet weight, of yeast. With a content of 48 pg of fructose-l,6-bisphosphatase/g, wet weight (61, dephosphorylation of the total amount of fructose-1,6-bisphosphatase in yeast could take place in 48:1.6 = 30 min. In vivo the dephosphorylation of fructose-1,6-bisphosphatase takes about 120 min after transfer of glucose-treated yeast to a glucose-free acetate containing medium (1, 2, 7). An explanation for the discrepancy of the time calculated for the dephosphorylation in vitro and in vivo may be that in intact cells the conditions for dephosphorylation of phosphorylated fructose-1,6-bisphosphatase are not as optimal as they are at the in vitro assay. Furthermore, in vivo phosphorylated fructose-1,6-bisphosphatase dephosphorylating protein phosphatase may be inhibited under the conditions where dephosphorylation is observed, i.e. at incubation of glucose-treated cells in a glucose-free, acetate containing medium (7). The presence of inhibiting metabolites, such as orthophosphate, pyrophosphate, succinate (cf. Yeast Fructose-l,6-bisphosphatase Dephosphorylating Phosphatase cules, as well as the absence of activating ions, such as free M P , may be considered among other possibilities in this context.