Phosphorylation of ornithine decarboxylase by a polyamine ...

Proc. NatL Acad. Sci. USA'

Vol. 78, No. 9, pp. 5518-5522, September 1981. Biochemistry

Phosphorylation of ornithine decarboxylase by a polyamine-dependent protein kinase (polyamies/interferon/antizyme/nucleus) VALERIE J. ATMAR AND GLENN D. KUEHN* Department of Chemistry, Box 3G, New Mexico State University, Las Cruces, New Mexico 88003

Communicated by P. D. Boyer, June 11, 1981

ABSTRACT This paper presents evidence that a polyaminedependent protein idnase (EC 2.7A 1.37) purified from nuclei of the slime mold Physarum polycephalum catalyzes phosphorylation of ornithine decarboxylase (OrnDCase; L-ornithine carboxy-lyase, EC 4.1.1.17). The protein kinase had properties similar to OrnDCase antizyme. Phosphocellulose chromatography of nuclear preparations from P. polycephalum yielded the polyaminedependent protein kdnase of subunit M. 26,000 that, was resolved from a second fraction in whish the protein kinase copurified with a phosphate-acceptor protein of subunit Mr 70,000. At Nat concentrations less than -150 mM,. a complex formed between the protein kinase and the phosphate-acceptor protein. The complex did not demonstrate protein kinase or OrnDCase activity. The complex was dissociated by >150 mM Na' into its constituent proteins. The dissociated complex catalyzed phosphorylation of the Mr 70,000 component in the presence of spermidine and spermine, and it also demonstrated OrnDCase activity. The purified Mr 70,000 component from the complex and authentic OrnDCase, purified by procedures previously reported, were virtually identical with respect to OrnDCase activity, capacity to be phosphorylated by the polyamine-dependent protein kinase, amino acid composition, and immunological crossreactivity. Phosphorylation of OrnDCase by the polyamine-dependent protein kinase sharply inhibited OrnDCase activity. Thus, this is an example of posttranslational covalent modification of OrnDCase with concurrent alteration of its catalytic function. It is also an unusual example of control of the first enzyme in a biosynthetic pathway by a protein kinase that is, in turn, modulated by the immediate end products of the pathway.

zyme activity in vivo following treatment with pharmacological reagents. The precise loci of action for most of these reagents are unknown. Moreover, fluctuations in OrnDCase enzymatic activity as a consequence of treatment with these agents need not involve new enzyme synthesis or metabolic turnover by synthesis and degradation. Indeed, the original conclusion that ODC exhibited the extraordinarily short half-life of =z10 min hinted of an enzymatic process involving posttranslational modification. This possibility was further suggested when Heller'et aL (6, 7) discovered a protein factor of Mr 26,000 that complexed noncovalently with OrnDCase and noncompetitively inhibited its enzymatic activity. The polyamines-putrescine, spermidine, and spermineappear to induce the synthesis of the Mr 26,000 inhibitor protein. The inhibitor has been designated OrnDCase antizyme (6). The occurrence of OrnDCase antizyme appears to be ubiquitous. It has been detected in every eukaryotic cell line tested (8) and also in prokaryotes (9). In rat liver, 80% of the cellular antizyme is found localized in the nucleus partially bound to OrnDCase (7). Recently, Mitchell et aL (10) reported an antizyme-like factor in the slime mold Physarum polycephalum that requires spermidine or spermine to complex to OrnDCase. In earlier communications, we reported the discovery and characterization of a new type of protein kinase (EC 2.7.1.37) that is polyamine dependent (11-14). This polyamine-dependent protein kinase of subunit Mr 26,000 catalyzes phosphorylation of a unique acidic nucleolar protein of subunit Mr 70,000. The two proteins copurify from phosphocellulose chromatography (14). Phosphorylation of the Mr 70,000 nonhistone protein by the polyamine-dependent protein kinase yields a phosphoprotein that demonstrates numerous properties concordant with a specific regulatory role in rRNA gene transcription (12). During the characterization ofthe interaction of the polyaminedependent protein kinase with the Mr 70,000 nonhistone protein, we recognized many similarities between this system and that of OrnDCase antizyme and OrnDCase. We report here evidence showing the protein that we called Mr 70,000 nonhistone protein in our earlier work and the enzyme OrnDCase are indeed the same. Moreover, phosphorylation of OrnDCase by the polyamine-dependent protein kinase of Mr 26,000 sharply inhibits decarboxylation activity. This unique control mechanism represents an example of control ofthe first enzyme in a biosynthetic pathway by a protein kinase that is, in turn, modulated by the end products-of the pathway.

Ornithine decarboxylase (OrnDCase; L-ornithine carboxy-lyase, EC 4.1.1.17) catalyzes the first and rate-limiting reaction of polyamine biosynthesis, in which ornithine is decarboxylated to form the diamine, putrescine. Much experimental evidence has established that a rapid increase in OrnDCase activity is an invariant element in the response of all eukaryotic cells' stimulated to proliferate (1). Scrutiny of this phenomenon has identified two properties that make OrnDCase a unique representative of eukaryotic enzymes. First, a great number of diverse agents have the capacity to produce marked stimulation of the enzyme activity (2, 3). These agents include tumor-promoting agents, viral transformations, hormone administrations, and mechanical growth stimulation. Conversely, other reagents, such. as the polyamines and interferon, rapidly inactivate OrnDCase in a variety ofcell cultures. Second, OrnDCase has been ascribed the shortest apparent half-life of any known eukaryotic enzyme-i.e. "'10 min (4, 5). Nearly without exception, the conclusions derived from experiments regarding the inducibility and apparent rapid turnover rate of OrnDCase were obtained from studies on the rates of increase or decay of en-

MATERUILS AND METHODS Preparationof Nuclei, Nucleoli, and Crude Extracts. Nuclei were isolated from 48-hr shake cultures of microplasmodia (11,

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

Abbreviations: OrnDCase, ornithine decarboxylase. To whom reprint requests should be addressed.

*

5518

Biochemistry:

Atmar and Kuehn

12, 14, 15). Slime-free nucleoli were isolated by the Percoll (Pharmacia) gradient method (16). Purification of Enzymes. All enzymes were derived from P. polycephalum. Polyamine-dependent protein kinase (free or complexed to its phosphate-acceptor protein Of Mr 70,000) was purified from nuclei or nucleoli by phosphocellulose (Sigma; fine mesh) chromatography (14). The phosphate-acceptor protein of Mr 70,000 was purified in a form resolved from the polyamine-dependent protein kinase by column chromatography on AG3-X4A anion exchange resin (200-400 mesh, Cl- form; BioRad) (12). The dephosphoform of the Mr 70,000 protein was prepared by treatment with alkaline phosphatase agarose (12). Authentic OrnDCase was prepared from microplasmodia as described (17). Enzyme Assays. Protein kinase activity was measured by using the filter paper disk procedure with ['y-32P]ATP (14). OrnDCase activity was measured in a sealed vial by trapping 14Co2 onto Whatman 3-mm paper disks saturated with KOH after liberation of the radioactive gas from L-[1-'4C]ornithine (Amersham; specific radioactivity 57 mCi/mmol; 1 Ci = 3.7 10s becquerels) (18). The assay mixture was 125 mM Tris HCl, pH 8.3/0.25 mM Na2EDTA/75 AM pyridoxal phosphate/0.16 M NaCV0.25 mM L-[1-'4C]ornithine (specific activity, 1.50 mCi/mmol). The total reaction volume was 0.40 ml. Enzymatic assays were carried out at 30'C for 30 min. Reactions were terminated by addition of 1 ml of 2 M citric acid. A minimum of 4 hr was allowed for absorption of the `4C02 released by the KOH-saturated wick. Amino Acid Analyses. Analyses were performed with a Durram D400 amino acid analyzer (12). Preparation of Antibodies. Electrophoretically pure Mr 70,000 nonhistone protein from nucleoli of P. polycephalum (1 mg) was used to raise antibodies in New Zealand White rabbits by a multiple injection procedure (19). Immunodiffusion tests were conducted on 1% Bacto-Noble agar (Difco) plates containing 1% NaCl and 0.02% NaN3. Other Methods. Procedures for estimation of protein concentrations and analytical NaDodSO4polyacrylamide gel electrophoresis were as described (14). X

RESULTS Preliminary Indications of Identity Between OrnDCase and the Mr 70,000 Nonhistone Protein. Our earlier work established that the Mr 70,000 nonhistone protein copurified from nuclei and nucleoli of P. polycephalum with a second protein of Mr 26,000. Separation of the two proteins followed by reconstitution experiments established that the smaller protein conferred the property of polyamine-dependent phosphorylation of the Mr 70,000 protein on a complex formed by the two proteins (14). The corresponding Mr values for the phosphate-acceptor protein and the protein kinase were the same as those previously reported for OrnDCase (20, 21) and OrnDCase antizyme (6, 20), respectively. OrnDCase antizyme had been found to be ubiquitously distributed among diverse eukaryotes (8). The Mr 70,000 nonhistone protein and the polyamine-dependent protein kinase from P. polycephalum were isolated from the nucleus (14), the same compartment previously reported in other eukaryotes to be the site of 35% to 80% of a bound form of OrnDCase antizyme (7). The isoelectric point of the Mr 70,000 nonhistone protein from P. polycephalum was determined to be 4.2, a value within the range (4.0-4.9) reported for OrnDCase from other eukaryotes (22). Finally, it was discovered that, at low Na+ concentrations, less than 150 mM, a complex formed between the Mr 70,000 and the Mr 26,000 proteins. The complex could be reversibly

Proc. NatL Acad. Sci. USA 78 (1981) 0.5 x

5519

NaCl, M

8

.

-

6.')

cd

. E . 4

14

$3

Fraction FIG. 1. Cofractionation of polyamine-dependent protein kinase (e and *) and OrnDCase (A) enzymatic activities from nuclear preparations by phosphocellulose chromatography. Crude nuclear extracts were treated batchwise with Bio-Rex 70 (Na' form, 400 mesh; Bio-Rad) (14) before column chromatography. The partially purified nonhistone proteins (120 mg) in 42 ml of 50 mM Tris HCl, pH 7.5, were applied to a phosphocellulose column (3.5 x 10 cm) equilibrated with 50 mM TrisHCI, pH 7.5. The column was eluted with this buffer until the absorbance of the effluent at 280 nm was 150 mM and up to at least 280 mM. The proteins of Mr 70,000 and Mr 26,000 could be resolved in the presence of 150-280 mM NaCl by Sephadex G-200 column chromatography (data not shown; see below). These experiments were essentially duplications of similar experiments previously performed between OrnDCase and OrnDCase antizyme (6). Copurification of OrnDCase Activity with the Mr 70,000 Nonhistone Protein on Phosphocellulose. NaCl gradient-elution chromatography of soluble proteins from nuclei or nucleoli on phosphocellulose yielded a fraction in which the polyaminedependent protein kinase and its unique substrate, the Mr 70,000 nonhistone protein, copurified (14). This fraction consistently eluted from phosphocellulose columns when the salt gradient reached 0.7 M NaCl. Fig. 1 illustrates such an isolation. Enzymatic activities for both polyamine-dependent protein kinase and OrnDCase were measured in the derived fractions. Clearly, those fractions that demonstrated a capacity to phosphorylate an endogenous substrate protein in the presence of the polyamines (Fig. 1, fraction b) also demonstrated OrnDCase activity in assay mixtures that contained 280 mM Na'. Removal of the Na' from assays for determining either OrnDCase or protein kinase activities abolished both activities (data not shown). Polyamine-dependent protein kinase activities measured in the absence of the polyamines, but in the presTable 1. Demonstration of OrnDCase activity in preparations of the Mr 70,000 nonhistone nucleolar protein Reaction mixture Activity* Complete 9.43 Complete lacking Mr 70,000 protein 0.055 Complete lacking pyridoxal phosphate 0.047 Results are represented as mol x 102 of 14CO2 released per mol of protein per min. The complete reaction mixture was 125 mM Tris HCl, pH 8.3/0.25 mM NaEDTA/0.075 mM pyridoxal phosphate/0.16 M NaCl/0.25 mM [1-1 Clornithine (specific radioactivity, 1.50 mCi/ mmol) containing 130 lsg of Mr 70,000 protein. The reaction time was 30 min. * Values are corrected for a zero-time control; those given here were 0.08-0.17 for different preparations of Mr 70,000 protein. Highly purifled authentic OrnDCase preparations, isolated by a different procedure (17), yielded values of 0.03-0.07.

Proc. Nad Acad. Sci. USA 78 (1981)

Biochemistry: Atmar and Kuehn

5520

Table 2. Polyamine-dependent phosphorylation of purified OrnDCase by polyamine-dependent protein kinase SSPi incorporated,* mol/mol of Spermine/spermidine polypeptide Protein substrate Mr 70,000 nucleolar 0.006 protein Mr 70,000 nucleolar 0.289 + protein Mr 70,000 nucleolar protein but not protein + 0.003 kinase 0.009 OmDCase + 0.418 OrnDCase OrnDCase but not protein + 0.008 kinase The assay mixture conteined the components given in the legend to Fig. 2A except that 0.5 mM spermine and 0.5 mM spermidine were added as indicated and proteins were added in the amounts 70 jtg of OrnDCase or 94 yg of Mr 70,000 nuclear protein. The reaction time was 20 mn. * The specific radioactivity of [v-32PIATP was 1.04 x 108 cpm/4&mol.

ence of 280 mM Na+, were -0.05 of those measured in the presence of the polyamines. These results indicated that the capacity to demonstrate OrnDC catalysis resided in the Mr 70,000 component of the kinase-phosphate acceptor protein complex (Fig. 1, fraction b). The Mr 70,000 nonhistone protein was purified and resolved from the protein kInase component of Mr 26,000 by anion exchange chromatography on AG3-X4A (12). Successful resolution was established by appearance of only one stained protein zone of Mr 70,000 in NaDodSO4polyacrylamide rod gels after electrophoretic separation (14). Table 1 shows that the resolved Mr Mr

15 20

10-3 30 50 70 90

X

2 -6 03

6 V-4

x

0

.-4

fi.'

I .0

&

CO

B B

0. 0

20

40 Gel slice

60

80

FIG. 2. Correspondence of protein and 32Pj in NaDodSO/polyacrylamide rod gels following labeling of authentic OrnDCase with [y-_2PIATP by purified polyamine-dependent protein kinase in an in vitro protein kinase assay. (A) The reaction mixture was 100 mM disodium (-glycerolphosphate, pH 6.8/20 mM NaF/l mM Na2EDTA/ 10 mM magnesium acetate/2 mM [3PIATP (specific radioactivity, 2.09 x 10 cpm/,umol)/60 mM NaCl, 0.5 mM spermidine/0.5 mM spermine containing 15 ,ug of polyamine-dependent protein kinase preparation (Fig. 1, fraction b) and 14 ,ug of authentic OrnDCase. The mixture was incubated at 30°C for 20 min, and reaction was stopped by addition of 1% NaDodSO/65 mM Tris HCl, pH 7.5/5% 2-mercaptoethanol. The mixture was heated in a boiling water bath for 2 min prior to application onto a 12.5% polyacrylamide gel. After electrophoresis, the gel was sliced into 1-mm transverse sections. Gel sections were dissolved in 30% H202 for scintillation spectrometry. (B) The protocol was the same as for A except that no polyamines were added.

*

1 L

0

0.8 £

~~~0.6~ 0.4 ~

|10

0

0.2

0

0

20 Phosphorylation time, min 10

FIG. 3. Effect of phosphorylation of OrnDCase by polyamine-dependent protein kinase. Ten samples of OrnDCase, 29 jug each, were incubated individually at 300C in 0.30 ml of the protein kinase assay mixture described in the legend to Fig. 2A. The specific radioactivity of the [y-32P]ATP was 1.64 x 107 cpm/flmol. Each tube contained 65 jtg of polyamine-dependent kinase from fraction a (Fig. 1). Five reaction mixtures contained 0.5 mM spermidine/0.5 mM spermine (A), and five did not (W). At the indicated times, one tube that contained polyamines and one thatdid not were removed and chilled on ice. These reaction mixtures, which contained 282 mM Na', were immediately dialyzed for 90 min against 100 ml of buffer containing the components for the OrnDCase assay, except L[1-14C]ornithine; this process increased reaction volume by -20%. After dialysis, 0.3 ml of each dialyzed sample was assayed in a sealed reaction vial for remaining OrnDCase activity by addition of 0.25 mM L-[1-4C]ornithine (e and *). For those tubes that contained polyamines, 0.05 ml of each dialyzed solution was also spotted onto paper disks, and the disk were washed (14) and then assayed for 32Pi incorporated into protein (n).

70,000 nonhistone protein isolated from nuclei ofP. polycephalum has the capacity to catalyze decarboxylation of L-[114C]omithine. The reaction was dependent on the addition of pyridoxal phosphate. Phosphorylation of Authentic OrnDCase. Authentic OrnDCase was purified from plasmodia of P. polycephalum (17) Table 3. Amino acid analyses of the Mr 70,000 nonhistone protein from nuclei of P. polycephalum and OrnDCase Authentic M, 70,000 OmnDCase protein Amino acid 21 20 Lysine 8 8 Histidine 17 17 Arginine 0 0 Tryptophan* 37 34 Asparagine/aspartic acid 18 17 Threonine 24 23 Serine 46 45 Glutamine/glutamic acid 15 17 Proline 26 26 Glycine 16 16 Alanine 21 20 Valine 6 6 Methionine 16 16 Isoleucine 30 31 Leucine 10 9 Tyrosine 13 13 Phenylalanine Analyses were performed on samples hydrolyzed in 6 M HCl at 110°C for 22 hr. Results represent nmol of residue recovered per 35 yg of hydrolyzed sample; they could not be expressed as number of residues per Mr 70,000 molecule because hydrolysates contained two basic residues that were not identified. * Determined after hydrolysis in 2% (vol/vol) mercaptoacetic acid.

Biochemistry:

Atmar and Kuehn

Proc. NatL Acad. Sc-i. USA 78 (1981)

5521

common antigenic determinants. Moreover, the same antiserum abolished the catalytic capacity of either protein to decarboxylate L-[1-'4C]ornithine (data not shown).

FIG. 4. Double-diffusion precipitin reactions of the -globulin fraction of antiserum to the Mr 70,000 nonhistone nuclearprotein from P. polycephalum with purified samples of the Mr 70,000 protein and authentic OrnDCase. Sample wells: Ab, antibodies to electrophoretically pure Mr 70,000 nuclear protein; a, purified Mr 70,000 nuclear protein; b, purified authentic OrnDCase.

and tested for its capacity to serve as a substrate for the polyamine-dependent protein kinase. The Mr 26,000 polyaminedependent protein kinase component was purified and resolved from the phosphate-acceptor protein Of Mr 70,000 on the phosphocellulose column (peak a, Fig. 1). A full discussion of, and evidence for, this resolution has been presented earlier (14). Table 2 shows that authentic OrmDCase prepared by the method of Mitchell et aL (17) was an efficient phosphate-acceptor protein for the Mr 26,000 protein kinase in a reaction that was polyamine-dependent. Table 2 also presents the repetition of reconstitution experiments previously reported (14) between the resolved protein kinase and the Mr 70,000 nonhistone protein. These results established that the Mr 26,000 component that demonstrated polyamine-dependent protein kinase activity toward the Mr 70,000 nonhistone protein provided the same activity in the presence of authentic OmDCase. When protein kinase samples from fraction a, Fig. 1, were incubated with authentic OmDCase and [yt-P]ATP, only one protein band was labeled in a polyamine-dependent reaction as shown by NaDodSO4polyacrylamide gel electrophoresis (Fig. 2). The only phosphorylated moiety was the subunit polypeptide of OrnDCase that was Mr 70,000. Effect of Phosphorylation on the Catalytic Function of OrnDCase. In general, biosynthetic enzymes that have been found to be subject to phosphorylation-dephosphorylation control are inactivated by phosphorylation (23). As shown in Fig. 3, the capacity of OmDCase to catalyze decarboxylation of L[1-'4C]ornithine was inhibited by phosphorylation of the enzyme by the polyamine-dependent protein kinase (Fig. 1, fraction a) and [y-32P]ATP. Results similar to these were observed when isolates of the Mr 70,000 nonhistone protein were substituted for authentic OmDCase. The OrnDCase catalytic function of the Mr 70,000 protein was also inhibited. Amino Acid Analyses. Table 3 shows the amino acid compositions of authentic OrnDCase prepared by the method of Mitchell et aL (17) and the Mr 70,000 nonhistone protein isolated from nucleoli of P. polycephalum. Neither protein contained detectable tryptophan. Cysteine residues were not determined. Within experimental error (3-5%), the two proteins were virtually identical. Immunological Crossreactivity. Rabbit antiserum prepared against the Mr 70,000 nonhistone protein demonstrated a precipitin reaction with its respective homologous antigen and with purified authentic OrnDCase by the Ouchterlony double-diffusion technique (Fig. 4). Thus, the two proteins contained

DISCUSSION A mechanism for the regulation of OrnDCase activity has been shown by these results. This finding has broad implications in numerous areas. It immediately impacts on investigations of different biochemical functions for the phosphorylated and dephosphorylated forms of OrnDCase. It provides a unifying explanation for the regulation of this enzyme by polyamines and perhaps other pharmacological agents, and it suggests how future investigations might be conducted to search for similar control of other biosynthetic enzymes by putative end productactivated protein kinases. OrnDCase and the Mr 70,000 nonhistone protein, previously investigated by our laboratory (11-14, 24) are shown here to be the same protein. Identity was established in the following properties: catalytic capacity to decarboxylate L-ornithine, amino acid composition, subunit polypeptide Mr, capacity to serve as unique protein substrates for a polyamine-dependent protein kinase, and immunological crossreactivity. Thus, by reference to our previous work, OrnDCase appears to be a multifunctional protein depending on its existence in a dephosphorylated or phosphorylated form. In results to be published elsewhere, we have detected a phosphoprotein phosphatase activity in nuclei and nucleoli of P. polycephalum that efficiently dephosphorylated phosphorylated OrnDCase. In its dephosphorylated form, OrnDCase functions as the catalyst for which it is named. In its phosphorylated form, OrnDCase does not catalyze decarboxylation ofornithine. Our earlier investigations have shown that the phosphorylated form has numerous properties concordant with a specific regulatory role in rRNA gene transcription in the nucleolus (12). The phosphoprotein has been shown to be a component of a deoxyribonucleoprotein complex isolated from nucleoli that contain palindromic ribosomal DNA. The purified phosphoprotein binds with high specificity to a region ofpurified ribosomal DNA near the symmetry axis of the palindrome. It stimulates rRNA synthesis by RNA polymerase I within the deoxyribonucleoprotein complex. Both of these properties are dependent on the phosphorylation state of OrnDCase. Thus, it seems likely that phosphorylation of OrnDCase serves a broader role than merely preventing and controlling biosynthesis of the polyamines. Many details of this proposal, of course, require greater definition, but the broad outline is apparent. Recognition that OrnDCase is subject to phosphorylation control by a polyamine-dependent protein kinase provides a cogent explanation for reports indicating reversible posttranslational control of this enzyme by putrescine, spermidine, and spermine (1, 6, 10; refs. therein), and by pharmacological manipulation of polyamine biosynthesis (1, 25). Rapid inhibition of OrnDCase by the polyamines has been known for some time. It has been observed in a number of physiological systems, as well as in a variety of cell cultures. Heller et ad (6) provided a partial explanation for these effects with the discovery of OrnDCase antizyme. OrnDCase antizyme is a protein inhibitor of Mr 26,000. The polyamines induce the synthesis of the OrnDCase antizyme in virtually every eukaryotic and prokaryotic cell line examined (6, 8, 9). Mere complexation between OrnDCase and its antizyme in solutions containing low salt concentrations inhibits the catalytic activity of OrnDCase. Increased salt concentrations, such at 250 mM NaCl or 10% (NH4)2SO4, are sufficient to dissociate the complex and restore OrnDCase activity (6). The molecular weights and general

5522

Biochemiatry: Atmar and Kuehn

physical properties of the antizyme and the protein of Mr 26,000 that demonstrated polyamine-dependent phosphorylation of OrnDCase in the present communication are virtually the same. At low concentrations of Na' (< 150 mM), OrnDCase and the Mr 26,000 protein from P. polycephalum form a complex that has little or no OrnDCase activity. At 280 mM Na', the complex fractionates into two protein components, one of which contains active OrnDCase and the other ofwhich contains active polyamine-dependent protein kinase. However, the manner of interaction among OrnDCase, the Mr 26,000 protein kinase, spermidine, spermine, and Na' is complex. The interaction could be explained by the tendency of the two proteins to associate in low salt and to dissociate in high salt (e.g., >150 mM NaCl). For example, when the experiment in Fig. 3 was carried out in the presence of 0.5 mM spermidine/0.5 mM spermine but the absence of Na', no phosphorylation of OrnDCase was observed and no OrnDCase activity was detected. We interpret these results as described above. High Na' concentration (> 150 mM) dissociates the complex that forms spontaneously between OrnDCase and the Mr 26,000 protein. Under high Na' conditions, spermidine and spermine can interact with either the Mr 26,000 protein or OrnDCase (or both), thus activating the polyamine-dependent protein kinase reaction. Phosphorylation of OrnDCase follows. In the absence of high Na', the OrnDCase-Mr 26,000 protein complex cannot be activated by spermidine or spermine to effect phosphorylation of OrnDCase. The associative events among OrnDCase, the Mr 26,000 protein, and the polyamines appears to follow a strictly ordered sequence before the protein kinase reaction can occur; dissociation of the complex appears to be a prerequisite for phosphorylation of OrnDCase. These results, as presented, preclude definite conclusion that the Mr 26,000 protein per se is a polyamine-dependent protein kinase. It was only when this protein was combined with OrnDCase and the polyamines that the phosphorylation reaction was initiated. However, we have recently discovered that the Mr 26,000 protein can phosphorylate casein in the absence of polyamines at -0.1 the rate of OrnDCase phosphorylation. This conclusively demonstrates that the Mr 26,000 protein is a protein kinase. Several pharmacological agents can also inactivate OrnDCase in mammalian cells on a temporal scale similar to that of the polyamines. These include cytochalasin B (26), colchicine (27), retinoic acid (28), and interferon (26). It will be of interest to determine whether any of these compounds exert their inhibitory effects through the action of the polyamine-dependent protein kinase on OrnDCase. In the case of the interferon effect on mammalian cells, the facts are particularly suggestive. Interferon induces a protein kinase activity that phosphorylates a polypeptide of Mr &70,000 (29, 30). The Mr 70,000 polypeptide copurifies with the kinase. The appearance of the Mr 70,000 phosphoprotein is the most apparent difference between interferon-treated and control cell fractions, and it can serve as a marker of the action of interferon. It appears only under conditions in which the antiviral state is induced. In general, inhibition of the catalytic function of biosynthetic enzymes through phosphorylation by protein kinases has been well documented (23). However, the property of activation by spermidine and spermine of the highly specific protein kinase that phosphorylated and consequently inactivated OrnDCase in the present work was not specifically anticipated. This ex-

Proc. Nad Acad. Sci. USA 78 (1981)

ample represents a feedback control mechanism by end products of a biosynthetic pathway-the phenomenon may not be limited to OrnDCase alone but may extend to other regulatory enzymes in metabolism. We thank Tom Burnett for conducting the amino acid analyses. Tech.' nical assistance in the culturing of P. polycephalum, provided by Adrianna Uranga who was supported by U. S. Public Health Service Grant

RR08136, is gratefully acknowledged. This work was supported by U.S. Public Health Service Grant GM18538.

1. Maudsley, D. B. (1979) Biochem. Pharmacol, 28, 153-161. 2. Morris, D. R. & Fillingame, R. H. (1974) Annu. Rev. Biochem. 43, 303-325. 3. Raina, A. & Janne, J. (1975) Med. Biol 53, 121-147. 4. Russell, D. H. & Snyder, S. H. (1969) Mol Pharmacol 5, 253-262. 5. Manen, C. A. & Russell, D. H. (1975) Life Sci. 17, 1769-1776. 6. Heller, J. S., Fong, W. F. & Canellakis, E. S. (1976) Proc. Natl Acad. Sci. USA 73, 1858-1862. 7. Heller, J. S., Kyriakidis, D., Fong, W. F. & Canellakis, E. S. (1977) Eur. J. Biochem. 81, 545-550. 8. Heller, J. S., Chen, K. Y., Kyriakidis, D. A., Fong, W. F. & Canellakis, E. S. (1978)J. Cell Physiot 96, 225-234. 9. Kyriakidis, D. A., Heller, J. S. & Canellakis, E. S. (1978) Proc. NatL Acad. Sci. USA 75, 4699-4703. 10. Mitchell, J. L. A., Augustine, T. A. & Wilson, J. M. (1981) Biochim. Biophys. Acta 657, 257-267. 11. Atmar, V. J., Daniels, G. R. & Kuehn, G. D. (1978) Eur. J. Biochem. 90, 29-37. 12. Kuehn, G. D., Affolter, H. U., Atmar, V. J., Seebeck, T., Gubler, U. & Braun, R. (1979) Proc. Natl. Acad. Sci. USA 76, 2541-2545. 13. Daniels, G. R., Atmar, V. J. & Kuehn, G. D. (1980) Fed. Proc. Fed. Am. Soc. Exp. Biol 39, 2096 (abstr.). 14. Daniels, G. R., Atmar, V. J. & Kuehn, G. D. (1981) Biochemistry 20, 2525-2532. 15. Mohberg, J. & Rusch, H. P. (1971) Exp. Cell Res. 66, 305-316. 16. Affolter, H. U., Behrens, K., Seebeck, T. & Braun, R. (1979) FEBS Lett. 107, 340-342. 17. Mitchell, J. L. A., Carter, D. D. & Rybski, J. A. (1978) Eur. J. Biochem. 92, 325-331. 18. Kobayashi, Y. & Maudsley, D. V. (1974) Biological Applications of Liquid Scintillation Counting (Academic, New York), p. 163. 19. Vaitukaitis, J., Robbins, J. B., Nieschlag, E. & Ross, G. T. (1971) J. Clin. Endocrinol Metab. 33, 988-991. 20. Pegg, A. E., Conover, C. & Wrona, A. (1978) Biochem. J. 170, 651-660. 21. Raina, A. & Janne, J. (1975) Med. Biol 53, 121-147. 22. Campbell, R. A., Morris, D. R., Bartos, D., Daves, G. D. & Bartos, F., eds. (1978) Advances in Polyamine Research (Raven, New York), Vol. 1. 23. Cohen, P. (1981) Recently Discovered Systems of Enzyme Regulation by Reversible Phosphorylation. Molecular Aspects of Cellular Regulation (Elsevier-North Holland, Amsterdam), Vol. 1. 24. Atmar, V. J., Daniels, G. R., Kuehn, G. D. & Braun, R. (1980) FEBS Lett. 114, 205-208. 25. Jdnne, J., Pbso, H. & Raina, A. (1978) Biochim. Biophys. Acta 473, 241-293. 26. Lee, E. J., Larkin, P. C. & Screevalsan, T. (1980) Biochem. Biophys. Res. Commun. 97, 301-308. 27. Chen, K., Heller, J. & Canellakis, E. S. (1976) Biochem. Biophys. Res. Commun. 68, 401-408. 28. Verma, A. K. & Boutwell, R. K. (1977) Cancer Res. 37, 3196-3301. 29. Revel, M. & Groner, Y. (1978) Annu. Rev. Biochem. 47, 1079-1126. 30. Minks, M. A., West, D. K., Benvin, S. & Baglioni, C. (1979)J. Biol Chem. 254, 10180-10183.