Antizyme regulates the degradation of ornithine decarboxylase in fission yeast Schizosaccharomyces pombe. Study in the spe2 knockout strains.

The mechanism of the regulatory degradation of ornithine decarboxylase (ODC) by polyamines was studied in fission yeast, Schizosaccharomyces pombe. To regulate cellular spermidine experimentally, we cloned and disrupted S-adenosylmethionine decarboxylase gene (spe2) in S. pombe. The null mutant of spe2 was devoid of spermidine and spermine, accumulated putrescine, and contained a high level of ODC. Addition of spermidine to the culture medium resulted in rapid decrease in the ODC activity caused by the acceleration of ODC degradation, which was dependent on de novo protein synthesis. A fraction of ODC forming an inactive complex concomitantly increased. The accelerated ODC degradation was prevented either by knockout of antizyme gene or by selective inhibitors of proteasome. Thus, unlike budding yeast, mammalian type antizyme-mediated ODC degradation by proteasome is operating in S. pombe.

The mechanism of the regulatory degradation of ornithine decarboxylase (ODC) by polyamines was studied in fission yeast, Schizosaccharomyces pombe. To regulate cellular spermidine experimentally, we cloned and disrupted S-adenosylmethionine decarboxylase gene (spe2) in S. pombe. The null mutant of spe2 was devoid of spermidine and spermine, accumulated putrescine, and contained a high level of ODC. Addition of spermidine to the culture medium resulted in rapid decrease in the ODC activity caused by the acceleration of ODC degradation, which was dependent on de novo protein synthesis. A fraction of ODC forming an inactive complex concomitantly increased. The accelerated ODC degradation was prevented either by knockout of antizyme gene or by selective inhibitors of proteasome. Thus, unlike budding yeast, mammalian type antizyme-mediated ODC degradation by proteasome is operating in S. pombe.
Polyamines (putrescine, spermidine, and spermine) are biologically ubiquitous compounds that have been implicated in many aspects of growth and development in a wide range of organisms (1)(2)(3), but their precise function is largely unknown. The biosynthesis of polyamines in yeast and most of animals depends on the decarboxylation of ornithine to putrescine by ornithine decarboxylase (ODC, 1 EC 4.1.1.17). Subsequent attachment of an aminopropyl moiety forms spermidine, and the second aminopropyl transfer to spermidine yields spermine. Decarboxylated S-adenosylmethionine serves as the aminopropyl donor, which is produced by S-adenosylmethionine decarboxylase (AdoMetDC, EC 4.1.1.50). Both the decarboxylating enzymes are highly regulated and subject to control by cellular polyamines.
In mammalian cells, ODC is negatively regulated in response to the increase in cellular polyamines mainly through a unique mechanism mediated by a protein termed antizyme (4,5). Synthesis of antizyme requires translational frameshifting, which occurs at the end of the initiating frame or open reading frame 1 (ORF1) (6,7). The ribosome shifts its reading frame to the ϩ1 frame and continues to decode the second ORF to synthesize the entire antizyme protein. The second ORF of antizyme encodes all the known functions of the protein (8 -10). Polyamines stimulate the frameshifting and thus control the level of antizyme. Antizyme binds to ODC monomer preventing formation of the homodimeric active enzyme. The ODC associated with antizyme is rapidly degraded by the 26 S proteasome without ubiquitination (11)(12)(13). In addition, antizyme negatively regulates the polyamine transporter (14,15). Mammalian cells contain another regulatory protein, antizyme inhibitor, a homolog of ODC without the decarboxylating activity (16). It binds to antizyme with a higher affinity than ODC and releases active ODC from the inactive antizyme-ODC heterodimer.
At least in two lower eukaryotes, polyamine-induced destablization of ODC has been studied in detail. In Saccharomyces cerevisiae, spermidine-induced ODC destabilization seems to be dependent on protein synthesis (17,18), and the 26 S proteasome is responsible for the proteolysis of ODC (18,19). However, computer search of an antizyme homolog in the completed genome of S. cerevisiae has been negative (20). In a filamentous fungus, Neurospora crassa, spermidine and putrescine accelerate the turnover of ODC protein (21). Recently, a presumed antizyme homolog has been noted in the N. crassa genome (AF291578, MIPS Neurospora data base).
Although S. cerevisiae and the fission yeast, Schizosaccharomyces pombe, are both ascomycetous fungi, they are almost as divergent each other as between these and mammals (22). S. pombe is often more similar to mammalian cells than S. cerevisiae in various aspects. Interestingly, the frameshift signal of mammalian antizyme directs the correct ϩ1 shift in S. pombe, but aberrant Ϫ2 shift in S. cerevisiae (23,24). Very recently, Ivanov et al. (25) identified an antizyme gene in S. pombe demonstrating its frameshift induction by polyamines, involvement in the cellular polyamine control, and inhibitory activity on ODC. Deletion of antizyme gene (SPA) from the fisson yeast genome did not bring about apparent phenotypes, but resulted in the increase in putrescine, spermidine, and cadaverine, which was more prominently observed in the cells at stationary phase than in exponentially growing cells, whereas overexpression of SPA led to large reduction in the cellular level of polyamines (25). A role for antizyme in ODC degradation in S. pombe cells, however, was not demonstrated.
In the present paper, we studied the mechanism of ODC repression by spermidine in S. pombe. In order to be able to regulate cellular spermidine experimentally, we cloned and disrupted S. pombe gene for AdoMetDC (spe2). We show that the null mutant of spe2 (⌬spe2) is devoid of spermidine and spermine and contains a high ODC activity, which is rapidly repressedinresponsetoadditionofspermidinethroughantizymedependent ODC degradation catalyzed by proteasome. To our knowledge, this is the first experimental evidence of antizyme being involved in the regulation of ODC degradation in singlecellular eukaryotes.

EXPERIMENTAL PROCEDURES
Materials-Proteasome inhibitors, clasto-lactacystin ␤-lactone and MG132 (Z-LLL-CHO), were purchased from Boston Biochem (Cambridge, MA) and Peptide Institute (Osaka, Japan), respectively. Restriction enzymes were obtained from New England Biolabs. Oligodeoxynucleotides were from Amersham Pharmacia Biotech. ␣-Difluoromethylornithine (DFMO) was kindly provided by Merrell Dow Research Institute (Cincinnati, OH). Media components were from Difco. Other chemicals and reagents were obtained from Sigma unless otherwise mentioned.
Yeast Strains and Culture-S. pombe strain JY745 (ura4-D18 leu 1-32 ade6-M210 h Ϫ ) was a kind gift from Dr. M. Yamamoto. An antizyme knockout strain, ⌬spa (ura4-D18 leu1-32 ⌬spa::Leu2 ade6-M216 h Ϫ ) (25) was generously provided by Drs. J. F. Atkins and I. P. Ivanov. Polyamine-free liquid MM (minimal medium) (26) was prepared in acid-washed glass containers and filter-sterilized with plastic disposable units. URA selection was performed on synthetic dextrose plates minus uracil. To measure cell growth rates in the polyamine-free condition, S. pombe cells were precultured in YES medium (0.5% yeast extract, 3% dextrose, and amino acid supplements), washed twice with, and diluted 1:100 in, polyamine-free MM. After growing for 24 h, the cells were again washed twice and diluted in fresh polyamine-free MM to the A 600 of 0.01. A part of the culture was taken for measuring A 600 . The cells for polyamine assay were taken from the same cultures 20 h after the second dilution. For the use in the other experiments, cells were pre-cultured overnight in YES, washed twice with, and diluted in polyamine-free MM to A 600 ϭ 0.2 unless otherwise mentioned. All the yeast culture was carried out at 30°C with shaking.
Cloning and Sequencing of Spe2-Degenerated oligonucleotide primers (sense (5Ј-TTYGARGGNYHNGARAARYTNYTNGA-3Ј) and antisense (5Ј-GTNGTNCCRCANGTNWTNARDAT-3Ј), where R ϭ A or G; Y ϭ T or C; W ϭ A or T; D ϭ A, G, or T; H ϭ A, C, or T; N ϭ A, G, C, or T), were designed based on multiple alignment of AdoMetDC from several eukaryotic organisms on the data base and used to amplify a fragment of spe2 gene from S. pombe genomic DNA with PCR. The product was labeled with [␣-32 P]dCTP (3,000 Ci/mmol, PerkinElmer Life Sciences) using Random Primer DNA Labeling kit (Takara). An S. pombe cDNA library in pCD2 vector (a gift from Dr. H. Nojima) was screened with the standard colony hybridization protocol (27). Positive clones were isolated, subcloned in the BamHI site of pBluescript SK (Ϫ) (Stratagene), and sequenced using Prism cycle sequencing kit (ABI) with T3 and T7 primers. Additional primers were designed to sequence the entire cDNA in both directions. The cDNA was then used in screening an S. pombe genomic DNA library constructed in pBluescript KS (Ϫ) vector (a gift from Dr. H. Nojima). One of the positive clones (R7A, 5.2 kb) was digested with EcoRI, subcloned into pBluescript SK (Ϫ), and sequenced as above. Sequence comparison was carried out with GENE-TYX-MAC program (Software Development).
Disruption of spe2-One-step gene replacement was performed to disrupt spe2 as follows (Fig. 1B). S. pombe ura4 gene was taken from pREP2 (28) as a HindIII fragment and cloned into pBluescript SK (Ϫ) that had been modified by reversing the positions of the EcoRI and PstI sites with an oligonucleotide pair. The 3Ј-flanking region of spe2 (453 bp) was then inserted downstream of the ura4 gene at the PstI-EcoRI sites. The 5Ј-flanking region (840 bp) was first subcloned in pBluescript SK (Ϫ) at the EcoRI site, and then KpnI (on the vector)-BsaAI fragment (containing the 740-bp EcoRI-BsaAI fragment) was ligated upstream of ura4 at the KpnI-HincII sites of the earlier construct. In each step, subclones with the proper orientation were selected with restriction analysis and confirmed by sequencing. Thus in the final construct, almost the entire coding region of spe2 was replaced by ura4. The construct was digested with EcoRI followed by gel purification and used to transform JY745 or ⌬spa strains with modified lithium acetate method (29). The stable URA ϩ transformants were selected. Homologous recombination (spe2::ura4) was confirmed first by colony PCR, and then by Southern analysis of genomic DNA digested with HindIII using the standard techniques (27). The PstI-EcoRI fragment of S. pombe genome at the 5Ј vicinity of the disruption construct (1.0 kb) was used as a probe.
Enzyme and Polyamine Assays-For AdoMetDC assay, cells were grown as described above and harvested at an A 600 of 1.0. Cells from 3 ml of culture were collected with centrifugation at 1,000 ϫ g, 4°C for 10 min, and were suspended in 0.3 ml of the extraction buffer containing 50 mM Tris-HCl (pH 7.2), 1 mM DTT, 0.1 mM EDTA, and 2.5 mM putrescine. The suspensions were transferred to 2-ml screw-cap tubes (Sarstedt) together with 0.4 g of zirconia/silica beads (0.5 mm diameter, Biospec, Tokyo). The cells were broken by a Bead Beater Mini (Biospec) with three 1.5-min bursts. The cell lysate was centrifuged at 12,000 ϫ g, 4°C for 30 min with a microcentrifuge to collect the supernatant (extract). The reaction mixture included 40 mM Tris-HCl (pH 7.2), 1 mM DTT, 0.1 mM EDTA, 0.3 mg/ml bovine serum albumin, 0.1 mM putrescine, and 8 M S-adenosyl-4-[carboxyl-14 C]methionine (59 mCi/mmol, Amersham Pharmacia Biotech), and 50 l of cell extract in a total volume of 125 l. The 14 CO 2 released during the reaction at 37°C for 1 h was trapped onto a filter paper disc soaked with 10 l of 10% KOH and counted with scintillant.
ODC activity was measured using L-[ 14 C]ornithine as a substrate essentially as described (30). The cells were harvested at the indicated time and disrupted as above in extraction buffer containing 50 mM Tris-HCl (pH 7.5) and 1 mM DTT. One unit of ODC activity is defined as the amount releasing 1 nmol of 14 CO 2 from ornithine/h at 37°C.
To measure the polyamines, S. pombe cells were collected from 2 ml of the culture with centrifugation at 1,000 ϫ g, 4°C for 10 min and washed twice with phosphate-buffered saline (20 mM sodium phosphate, 140 mM NaCl, pH 7.4). Each pellet was resuspended in 0.2 ml of phosphate-buffered saline, and a part (50 l) was mixed with the same volume of 8% perchloric acid. The mixture was vortexed, kept on ice for 5 min, and centrifuged at 12,000 ϫ g, 4°C for 5 min. The supernatant (10 l) was subjected to polyamine analysis on high performance liquid chromatography and fluorometry as described (31).
Protein concentration of the cell extracts was measured by the Bradford method (32). Protein content of the cells in the suspension for polyamine analysis was assayed by BCA kit (Pierce). Bovine serum albumin was used as standard in both cases.
Expression of ODC in E. coli-We have isolated the cDNA and genomic DNA of S. pombe ODC. 2 Details of bacterial expression constructs of ODC are available from the authors upon request. Briefly, the 1296-bp fragment containing the ORF of S. pombe ODC cDNA was amplified with PCR and cloned into either pET3a (Novagen) at the NdeI-BamHI sites or pGEX-3T (Amersham Pharmacia Biotech) at the BamHI-EcoRI sites. The constructs were verified by sequencing. The resultant expression plasmids, pET3a-SPO and pGEX-SPO, were used for transformation of E. coli strains BL21 (DE3) and BL21, respectively.
Production of Antibody against ODC and Western Blotting-E. coli BL21 carrying pGEX-SPO was grown in the presence of 0.5 mM isopropyl-1-thio-␤-D-galactopyranoside at 18°C overnight to induce glutathione S-transferase (GST)-ODC fusion protein. The expression product was purified on a glutathione-Sepharose 4B-affinity column (Amersham Pharmacia Biotech) according to the manufacturer's instructions. Rabbits were injected with 0.5 mg of the protein mixed with Freund's complete adjuvant. Two booster doses were given with the same amount of protein using Freund's incomplete adjuvant 1 and 2 months later. The antisera were taken 7 days after the second booster and absorbed by an extract of ODC (spe1) gene-disrupted S. pombe cells. 2 The same S. pombe extracts (20 g) that were used for ODC assay were fractionated on 10% SDS-PAGE. Immunoblotting was performed according to a standard protocol using Immobilon-P membrane (Millipore). Immunodetection was carried out with the primary antibody (one of the antisera) against GST-ODC at a dilution of 1:5,000 and secondary antibody, alkaline-phosphatase conjugated anti-rabbit IgG (␥-chain-specific, Sigma), at a dilution of 1:10,000 essentially following the method of Mierendorf et al. (33).
Measurement of Inactive ODC Complex-The untagged ODC was induced by 0.3 mM isopropyl-1-thio-␤-D-galactopyranoside in BL21 (DE3) carrying pET3a-SPO and purified from the extract by DEAE-Cellulofine (Seikagaku Kogyo) chromatography as described previously (34). The partially purified ODC (7,500 units) was incubated with 25 M DFMO, in a mixture containing 10 M pyridoxal phosphate, 5 mM DTT, 40 mM Tris-HCl (pH 7.4), and 0.01% Tween 80 in a total volume of 100 l at 37°C for 2 h. The residual free DFMO was removed by gel filtration through a NAP 5 column (Amersham Pharmacia Biotech).
The resultant preparation did not contain detectable level of ODC activity. For each ODC assay, 15 units of DFMO-ODC (as initially determined) was mixed and kept on ice for 30 min before enzyme reaction. The amount of ODC complex was calculated from the gain of ODC activity as described previously (30). Alternatively, 0.1 g of recombinant rat antizyme inhibitor, expressed as GST fusion protein and purified as described, was mixed with each sample before enzyme reaction (16). The amount of ODC complex was assumed to be equal to the gain of ODC activity.

RESULTS
Cloning of S. pombe spe2 Gene-A region of spe2 containing the proenzyme cleavage site was amplified with degenerated PCR. The product (249 bp) was used in screening of an S. pombe cDNA library as the probe. Several clones with the identical cDNA insert of 1,421 bp were isolated. The cDNA contained the entire ORF that encodes a polypeptide of 378 amino acids corresponding to the AdoMetDC proenzyme with a calculated molecular mass of 42,711Da (Fig. 1A). The amino acid sequence deduced from the cDNA was 36% and 35% identical to the S. cerevisiae and human counterparts, respectively. The highly conserved proenzyme cleavage sequence YLLSE2SSMFV was noted (where the arrow indicates the cleavage site). S. pombe AdoMetDC also contains the residues known to be important for its activity, substrate binding, and putrescine-mediated activation (35). The 5Ј-untranslated region of S. pombe spe2 mRNA contained a small upstream ORF (uORF), which encodes a tetrapeptide, MTIF, located closely to the main ORF with only one nucleotide gap (Fig. 1A). There is no common feature in the peptide sequence specified by uORFs between S. pombe and other organisms (see Refs. 36 and 37 and references therein). S. cerevisiae spe2 transcript lacks uORF (38).
Screening of the S. pombe genomic DNA library identified seven overlapping clones encompassing an 8.2-kb region that covers the entire ORF. Comparison of the genomic and the cDNA sequences revealed that spe2 gene contains no intron (Fig. 1B). Searching the data base showed that spe2 is located at the close 3Ј vicinity of the cut14 gene on the S. pombe genome (39).
Disruption of spe2-The spe2 gene was disrupted by the single-step gene replacement (Fig. 1B). Southern hybridization confirmed the replacement of the single-copy spe2 gene with homologous recombination (data not shown). AdoMetDC activity in the extracts of exponentially growing cells was measured.
⌬spe2 mutant cells contained no detectable activity. Wild type cells exhibited 423 pmol of CO 2 /h/mg of AdoMetDC protein. ⌬spe2 mutant and wild type cells were grown in the polyamine-free medium for two passages. Twenty hours after the second passage, the mutant cells contained no detectable levels of spermidine and spermine and accumulated 11.7 times putrescine and 6.0 times cadaverine over those detected in wild type cells (Fig. 2). Thereafter, the mutant cells ceased growing without addition of spermidine or spermine (Fig. 3). When spermidine or spermine was added to the culture medium at 100 M concentration, they restored growing at a doubling time comparable to that of wild type cells after several hours of time lag. Addition of putrescine (Fig. 3), cadaverine, or diaminopropane (data not shown) had no supportive effect on the growth of ⌬spe2 cells.
Changes in ODC Activity in Wild Type and ⌬spe2 Cells-Wild type cells showed virtually no ODC activity after growing in polyamine-free medium overnight (Fig. 4A). At this time, the cells were collected and resuspended in the same volume of fresh polyamine-free medium. ODC activity was rapidly induced reaching a peak in 4 h, and then a sharp decay was observed. Addition of spermidine (100 M) 4 h after changing the medium caused faster ODC decay. On the other hand, ⌬spe2 cells contained a high ODC activity after overnight culture in the polyamine-free medium (Fig. 4B). Changing medium resulted in a much larger and more prolonged ODC induction in ⌬spe2 cells than in wild type cells. Addition of spermidine caused rapid decay of ODC activity also in ⌬spe2 cells. Thus, in ⌬spe2 cells, ODC was derepressed despite a large accumulation of putrescine. The mechanism of the accelerated ODC decay by spermidine was studied further in ⌬spe2 cells.
Protein Synthesis-dependent Degradation Accounts for Accelerating ODC Decay-Spermidine (100 M) was added to the ⌬spe2 culture 4 h after the medium change, and cells were harvested at the indicated time to measure ODC activity. In Fig. 5A the total ODC activities (the sum of free ODC and complex form ODC activities measured in the presence of DFMO-ODC, see below) are plotted. Addition of spermidine reduced 98% of ODC activity in 6 h with a half-life of 65 min. This decay was much faster than that after addition of cycloheximide (50 g/ml), the half-life of which was more than 6 h. When cycloheximide was added together with spermidine, the acceleration of ODC decay was completely prevented. Actinomycin D (2.5 g/ml), in contrast, did not change the effect of spermidine. Western blot analysis of the same cell extracts depicted that the decrease in ODC activity was associated with decrease in the amount of ODC protein (Fig. 5B). Addition of FIG. 4. ODC induction after changing medium and repression by spermidine. A, changes in ODC activity in wild type cells. B, changes in ODC activity in ⌬spe2 mutant cells. For each cell type, duplicate cultures were grown overnight in the polyamine-free medium and the medium was replaced by fresh medium at time 0. After 4 h of medium change, spermidine (100 M) was added to one of the cultures. Aliquots were removed from the culture at indicated times, and ODC activity was measured as described in the text. E, without spermidine; q, with spermidine.

FIG. 5. Accelerated degradation of ODC by spermidine.
A, decay in ODC. ODC had been induced by medium change, and spermidine (100 M) was added to the culture alone, in combination with cycloheximide (50 g/ml), or with actinomycin D (2.5 g/ml). Control cells were maintained with addition of cycloheximide alone or without any addition. Aliquots were removed from the culture at indicated times, and ODC activity was measured as described in the text. The total ODC activity (sum of free ODC plus complex ODC) is plotted. The growth was monitored with A 600 in the presence or absence of polyamines. E, ⌬spe2 without polyamine; OE, ⌬spe2 with putrescine; q, ⌬spe2 with spermidine; ϫ, ⌬spe2 with spermine; Ⅺ, wild type without polyamine.
cycloheximide, but not actinomycin D, prevented this accelerated decrease of the protein. These results are consistent with a model that a short-lived protein that is induced by spermidine at a posttranscriptional level mediates the accelerating decay of ODC. Since it has been shown earlier that S. pombe antizyme is expressed through polyamine-inducible translational frameshifting (25), the protein is very likely to be antizyme.
Detection of Complex Form of ODC and Increase in Its Ratio by Spermidine-In mammalian cells, antizyme binds to ODC, forming an inactive heterodimer from which active ODC can be released by adding an excess amount of DFMO-ODC (30) or antizyme inhibitor (16). To test if a similar ODC-antizyme complex exists in S. pombe, we prepared S. pombe ODC inactivated with DFMO and used it in the competitive assay. A significant increase in ODC activity was observed when DFMO-ODC was added to the cellular extracts (Fig. 6A). Addition of spermidine to the culture medium increased the ratio of the gain over the total ODC activity. Cycloheximide blocked the increase. A virtually identical result was obtained by the use of rat antizyme inhibitor instead of DFMO-ODC (Fig. 6B). Spermidine thus increased the fraction of S. pombe ODC that forms an inactive complex. The ratio of the complex form to the total ODC was up to 25% after spermidine treatment, although 98% of the initial ODC activity disappeared in 6 h. It is therefore likely that the ODC associated with antizyme is degraded rapidly.
Involvement of Antizyme in ODC Degradation-To confirm that antizyme is the protein factor that is induced by spermidine and promotes the degradation of ODC, we employed a knockout strain of S. pombe antizyme (⌬spa) (25). The spe2 gene in the mutant strain was further disrupted by the singlestep gene replacement. Addition of spermidine to the culture medium of ⌬spa-⌬spe2 double knockout cells caused only a very slow ODC decay (Fig. 7). Approximately 90% of the activity remained even 6 h after addition of spermidine in the double knockout mutant compared with 2% in the ⌬spe2 single mutant. The result clearly indicates that antizyme is necessary for a major part of ODC decay that is stimulated by spermidine. It was also noted that about 10% of the initial ODC activity reproducibly disappeared in 6 h after the addition of spermidine in the absence of antizyme. Almost no ODC reduction was observed in the double mutant to which cycloheximide was added together with spermidine. Simultaneous addition of actinomycin D did not block the ODC decay.
Effects of the Proteasome Inhibitors on ODC Degradation-The 26S proteasome has been shown to catalyze ODC degradation in both animals and S. cerevisiae (11,18,19). We tested if the proteasome is also involved in the accelerated ODC decay in S. pombe using inhibitors of proteasome. clasto-Lactacystin ␤-lactone or MG132, both at 50 M, was added 4 h before spermidine. Both did not affect the cell growth at this concentration. As shown in Fig. 8, the rapid decay of ODC caused by spermidine was effectively inhibited by both inhibitors. This result strongly suggests that the 26 S proteasome catalyzes spermidine -induced degradation of ODC in S. pombe.

DISCUSSION
Mammalian antizyme represents a unique regulatory protein with a number of novel features, namely (i) expression and induction through translational frameshifting, (ii) the function as a non-ubiquitin stimulator for proteolysis of ODC by the proteasome, and (iii) dual activity on ODC and polyamine Amounts of antizyme-ODC complex were calculated from gain of ODC activity by adding DFMO-ODC and expressed as percentage over the total ODC activity. B, changes in antizyme-ODC complex as determined by rat antizyme inhibitor. ODC activity was assayed with and without addition of recombinant rat antizyme inhibitor (0.1 g). Amounts of antizyme-ODC complex were assumed to be equal to the gain of ODC activity and expressed as percentage over the total ODC activity. q, spermidine; ϫ, spermidine and cycloheximide; OE, cycloheximide alone; E, none.

FIG. 7. Antizyme is required for accelerated ODC decay.
Changes in ODC activity after spermidine addition to ⌬spa-⌬spe2 double mutant cells and ⌬spe2 single mutant cells are shown. ODC had been induced by medium change. Spermidine was added to ⌬spa-⌬spe2 double mutant alone (f), in combination with cycloheximide (ϫ), or with actinomycin D (Ⅺ). Control culture was maintained without addition (OE). In parallel experiments, ⌬spe2 single mutant was treated with spermidine (q) or none (E).
transporter. The frameshift induction serves as a polyamine sensor. The feedback system allows both the maintenance of the cellular polyamines within a certain range and their appropriate fluctuation. The molecular mechanisms of antizyme functions, however, have not been fully understood. In addition to the availability of genetics, recent identification of antizyme in S. pombe (25) makes the organism particularly an attractive system to study the mechanisms of antizyme functions.
The wild type S. pombe cells contain a high endogenous level of spermidine, which is not readily changed by exogenous spermidine. To study the feedback control of ODC by exogenous polyamines, use of mutants lacking spermidine synthesis appears to be useful as shown in S. cerevisiae (17,18). Therefore, we intended to clone and disrupt S. pombe gene for AdoMetDC, spe2.
The ⌬spe2 mutant (spe2::ura4) cease dividing after several generations in polyamine-free medium. The mutant contained a large excess of putrescine and cadaverine, indicating indispensability of spermidine and spermine for growth in the organism. The essential role of spermidine and spermine was also demonstrated in S. cerevisiae and N. crassa (40,41). ⌬spe2 mutant of the budding yeast showed an increase in cell size, a decrease in budding, accumulation of vesicle-like bodies, and abnormal distribution of actin-like material (40). In the presence of oxygen, a rapid cessation of cell growth and associated cell death were observed (42). Similar studies in S. pombe are yet to be performed.
Wild type S. pombe cells contain very low basal level of ODC. It was induced by changing medium, but subject to rapid repression due probably to increase in the endogenous polyamines (Fig. 4A). In ⌬spe2 cells ODC is elevated (Fig. 4B) despite a large accumulation of putrescine (Fig. 2). Spermidine added to the culture medium could rapidly repress ODC (Fig.  4B). It is of interest to note that S. pombe ODC is repressed by spermidine, but not effectively by putrescine, in connection with the observation that spermidine or spermine, but not putrescine, supports cellular growth. In S. cerevisiae, ODC decay is accelerated by putrescine to much less extent than by spermidine (43), whereas N. crassa ODC is strongly destabilized by both spermidine and putrescine (21).
The half-life of ODC activity was greatly shortened by spermidine in ⌬spe2 cells (Fig. 5A). The ODC decay is mostly attributable to a change in ODC protein (Fig. 5B). The accelerated ODC decay was prevented by the inhibitor of protein synthesis, but not by the inhibitor of RNA synthesis. In addition, it was found that a part of ODC existed as an inactive complex, from which active ODC is released by the competitors such as DFMO-ODC or antizyme inhibitor (Fig. 6). Although the fraction of ODC forming the inactive complex was increased by spermidine, it only accounts for a part of ODC decay in the activity. All these results are compatible with a model that antizyme is induced through the stimulation of translational frameshifting by spermidine, bound to ODC forming inactive complex, and promotes ODC protein for rapid degradation. The half-life of ODC activity in the spermidine-treated cells, 65 min (Fig. 5A), is not as short as those observed in mammalian cells, but comparable to those in S. cerevisiae in the repressed conditions (17,43).
Using an S. pombe mutant lacking antizyme and AdoMetDC (spa::Leu2 spe2::Ura4), we show requirement of antizyme for the accelerated ODC decay. In the absence of functional antizyme gene, the initial ODC activity was slightly higher and the decay of ODC activity after spermidine addition was much retarded than in SPA ϩ cells (Fig. 7). It is noted, however, that spermidine still promoted ODC decay to a certain extent in the absence of the antizyme in a protein synthesis-dependent man-ner. This observation may suggest that spermidine not only induces de novo synthesis of the antizyme, but some other protein(s) or additional form(s) of antizyme to causes the slower decay of ODC. The presence of multiple forms of antizyme has been reported from mammalian systems (for review, see Ref. 44) and a lower vertebrate (45).
We show that S. pombe ODC is subject to the polyaminestimulated degradation catalyzed by the 26 S proteasome as previously observed in S. cerevisiae and the higher animals. The mechanism by which ODC is targeted for the degradation is very similar between S. pombe and mammals, but substantially different between the two yeast species. S. pombe will serve as a suitable model system to study both the mammaliantype polyamine regulation, and the molecular bases of the difference in ODC regulation among the yeast, which may reflect the difference in the critical features of ODC and the proteasome.