Arianna Donella

Preferential dephosphorylation of protein bound phosphorylthreonine and phosphorylserine residues by cytosol and mitochondrial “casein phosphatase”

The partial purification of a rat liver cytosol protein phosphatase is described, resulting in a preparation active on casein but not on phosvitin, cytosol phosphopeptides, ATP, ADP and p-nitrophenylphosphate, which, on the contrary, are still dephosphorylated by the protein phosphatase purified from rat liver mitochondria. Moreover the activity of the former enzyme on casein appears to involve only a limited amount of phosphoric sites which are also preferentially phosphorylated by soluble protein kinase. The isolation and evaluation of 32P-serine and 32P-threonine from protein-kinase-dependently labelled phosvitin and casein, before and after incubation with the two enzymes, led to the conclusion that mitochondrial protein phosphatase hydrolyzes more actively the phosphorylserine residues, while the cytosol “casein phosphatase” promotes a preferential breakdown of the 32P-threonine residues.

On the possible role of phosphoproteins as iron carriers.

The chemical analysis of purified cytosol phosphoproteins disclosed further similarities with phosvitin, i.e. a very high content of phosphate residues and of protein-bound Fe (see table 1). The amount of iron bound to phosvitin depends on the preparation procedure [3]: when prepared according to Mecham and Olcott [4] phosvitin exhibits an iron content fairly different from that of cytosol phosphoproteins. However both phosphoproteins when exposed to an excess of FeC13 reach the same saturation level corresponding to a Fe/P molar ratio of about 0.5. The effect of several conditions on the stability of [s9 Fe]-phosvitin complex is shown in table 2. It can be seen that only unlabelled Fe 3+ promotes a massive release of s 9 Fe, as expected through a mechanism of isotopic dilution. Among the remaining diand trivalent cations tested, a slight release of s9 Fe was observed only in the presence of high Co 2÷ concentrations, the remaining cations being quite ineffective, even those, like Mn 2+, Mg 2+ and protamine, which bind to phosvitin through electrostatic forces resulting in insoluble complexes. Very high ionic strength (2 M NaC1) was

Further characterization of mitochondrial protein phosphatase

Abstract Mitochondrial protein phosphatase from rat liver exhibits rather wide substrate specificity, since it readily dephosphorylates, besides phosvitin, casein, and cytosol phosphoproteins, also ATP, ADP, inorganic pyrophosphate, p -nitrophenylphosphate. Aliphatic phosphate esters (β-glycerophosphate, glucose-6- P , serine-phosphate) are not dephosphorylated to any detectable extent. Evidence for the participation of a single enzyme in the dephosphorylation of phosvitin and ATP is provided. However, the different affinity toward the two substrates and other evidence suggest that the enzyme has in vivo the biological role of dephosphorylating, at least preferentially, the phosphoproteins.

Structural requirements for the enzymatic phosphorylation of phosvitin

Some structural features required for the enzymatic phosphorylation of phosvitin by purified rat liver cytosol phosvitin kinase have been investigated by testing the activity of such an enzyme toward phosphopeptides differing in size and chemical composition, obtained by pronase or acid hydrolysis of phosvitin. The results obtained can be summarized as follows: (a) Phosvitin kinase phosphorylates even fairly simple phosphopeptides (mol.wt 1000-2000) at rates comparable with intact phosvitin. (b) Acetylation of both phosvitin and pronase phosphopeptides completely prevents their phosphorylation indicating that some lysine residues are strictly required for the phosvitin kinase reaction. (c) Accordingly polyphosphorylserine blocks Ser(P)n which are very actively phosphorylated in phosvitin and pronase phosphopeptides, do not undergo any more enzymatic phosphorylation once isolated as such in a form free of other amino acids. (d) The activity of phosvitin kinase toward substrates probably devoid of Ser(P)n blocks suggests that there are not required for the protein kinase reaction. However, they apparently enhance the phosphorylation rate of the peptide substrates, likely by making easier their binding to the enzyme. It is proposed therefore that the peptidic unit able to undergo phosphorylation by rat liver cytosol phosvitin kinase consists of one or more phosphorylserine residues having in their close proximity a lysine residue playing a critical role in the mechanism of transphosphorylation.

Involvement of polyphosphorylserine blocks in the Fe(III) binding by phosvitin.

It has been suggested that the binding of iron(III) by phosvitin involves the phosphoric radicals of phosphorylserine residues, many of which are arranged in rows of several consecutive phosphoamino acids. In this paper we present evident that, unlike free phosphorylserine which does not interact with Fe(III), polyphosphorylserine blocks--(Ser-P)n, with n greater than or equal to 4 -- bind Fe(III) like phosvitin, though less actively, to give complexes stable at very acidic pHs. The binding of iron does not cause any polymerization of the phosphopeptides, and the (Ser-P)n-Fe(III) complexes display an Fe/P ratio significantly lower than 0.5, found in Fe-saturated phosvitin. These findings indicate that polyphosphorylserine blocks play an important role in the binding of iron by phosvitin, and that in the intact protein their binding capacity is optimized by the conformation of the polypeptide chain.

Characterization of cytosol phosphopeptides

In a previous paper the isolation from rat liver cytosol of a phosphoprote[n fraction has been described [1 ]. Such a phosphoprotein fraction, containing both phosphorylserine and phosphorylthreonine, was found to have a phosphate content approaching 1%. However the feeling that such a preparation was still contaminated by non-phosphorylated proteins prompted us to improve the purification. In the present paper a new procedure involving polyacrylamide gel electrophoresis is described leading to a phosphorylated fraction which is apparently free of non-phosphorylated proteins. The average molecular weight of such a purified fraction, evaluated by gel filtration, is about 2000. Its P content is about 4%, the highest ever reported for any phosphoprotein except phosvitin. The amino acid composition indicates a very acidic molecule with about 30% of glutamic and aspartic acid residues, and virtually all its serine and threonine as phosphorylated residues. Moreover quite remarkable amounts of bound iron are constantly present. bation at room temperature the labelled protein was separated from [32p] ATP by filtration through a Sephadex G-25 column (1.9 × 140 cm) equilibrated with 0.05 M Tris-HC1 buffer pH 7.5 containing 0.1 M NaC1. The radioactive protein fraction eluted at the void volume was concentrated by ultrafiltration through Diaflo UM 2 membranes to a volume of about 5 ml.

Direct detection of protein kinases on electropherograms

Since their discovery protein kinases have been usually tested by taking advantage of the solubility of ATP in deproteinizing agents like 10% trichloroacetic acid, in which protein bound phosphorylserine and phosphorylthreonine are quite stable. By such procedures at the end of incubation the excess of [32P] ATP is easily removed from the 32 P-labelled protein either by centrifugation followed by several washings [l-3] or by applying the sample on chromatographic paper [4] or on filters [5] which are subsequently eluted by trichloroacetic acid solutions. The counting of the protein residue at the end of both these procedures gives a reasonably precise measure of the amount of 32P transferred by the protein kinase from [32P] ATP to the phosphoprotein. In the last few years more and more attention has been devoted to protein kinases which have been found to differentiate both for their substrate specificity and for their sensitivity to cyclic nucleotides (see [6] ). As expected, investigators make extensive use of analytical gel electrophoresis to characterize protein kinases, though the only known procedure to localize the enzyme activity is to cut the gel into slices and to test the protein kinase eluted from every single segment by one of the methods previously mentioned [.5,7,8]. In the present note we shall describe a much less complicated and more rapid procedure which proved quite satisfactory for a direct evaluation of protein kinase activities on both polyacrylamide and acetylcellulose electropherograms. In principle such a method consists in laying down the gel, after electrophoresis, on a strip of chromatographic paper imbibed with a sample containing a buffer system, [“‘PI ATP with

Identification of the endogenous substrates of rat liver cytosol “phosvitin kinase”

Abstract 1. 1. Highly purified “phosvitin kinase” from rat liver cytosol has been found to be active on a low mol. wt peptide fraction and not on larger phosphoproteins isolated from the same source and resolved by Sepharose 6B gel chromatography. 2. 2. Such endogenous substrates of phosvitin kinase are resolved by 45% polyacrylamide gel electrophoresis into at least two components, both in the mol. wt range between 2000 and 3000. 3. 3. Cytosol phosphopeptides are devoid of the (ser-P)n blocks characteristic of typical acidic phosphoproteins like phosvitin and casein; nevertheless their affinity for the kinase is even greater than that displayed by te foreign substrate phosvitin.