Hypusine modification for growth is the major function of in Saccharomyces cerevisiae polyamine auxotrophs grown in limiting spermidine

Manas K. Chattopadhyay*, Myung Hee Park†, and Herbert Tabor*‡

*Laboratory of Biochemistry and Genetics, National Institute of Diabetes, Digestive, and Kidney Diseases, †Oral and Pharyngeal Cancer Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, MD 20892

Contributed by Herbert Tabor, November 26, 2007 (sent for review October 15, 2007) Spermidine and its derivative, hypusinated eIF5A, are essential for functions that have previously been reported for spermidine the growth of Saccharomyces cerevisiae. Very low concentrations have also been described for hypusinated eIF5A in yeast and .(of spermidine (10؊8 M) are sufficient for the growth of S. cerevisiae mammalian cells (5, 11, 16–20 polyamine auxotrophs (spe1⌬, spe2⌬, and spe3⌬). Under these conditions, even though the growth rate is near normal, the Results internal concentration of spermidine is <0.2% of the spermidine Comparison of Intracellular Polyamine Content and Morphology of concentration present in wild-type cells. When spe2⌬ cells are Wild-Type Cells and Polyamine Auxotrophs Grown on Low Concen- grown with low concentrations of spermidine, there is a large trations of Spermidine. Wild-type Saccharomyces cerevisiae cells decrease in the amount of hypusinated eukaryotic initiation factor contain large amounts of polyamines (estimated internal con- 5A (eIF5A) (1/20 of normal), even though there is no change in the centration of Ϸ0.003 M) (Table 1, Y534; ref. 21). Surprisingly, amount of total (modified plus unmodified) eIF5A. It is striking we found that this high internal concentration of polyamines is that, as intracellular spermidine becomes limiting, an increasing not needed for yeast growth; indeed, near-normal growth is portion of it (up to 54%) is used for the hypusine modification of found when polyamine auxotrophs are grown in 10Ϫ8 M sper- eIF5A. These data indicate that hypusine modification of eIF5A is a midine and the internal concentration of spermidine is very low most important function for spermidine in supporting the growth (1/240 of wild-type cells) (Table 1, Y535, Y536, and Y537). Data of S. cerevisiae polyamine auxotrophs. presented in Table 1 also show that the absence of putrescine does not affect the growth rate of a spe1⌬ auxotroph (Y535), if eIF5A ͉ protein synthesis ͉ yeast the medium contains 10Ϫ8 M spermidine. As seen in Fig. 1A (for spe2⌬ mutant Y536), the growth rate olyamines (putrescine, spermidine, and spermine) (Scheme in 10Ϫ8 M spermidine was near normal (2-h doubling time). At P1) are present in micromolar-to-millimolar concentrations in this concentration (Fig. 1B, point A), there were no gross prokaryotic and eukaryotic cells and play important roles in cell morphological changes or changes in the nuclear DNA as growth and development (1, 2). Intracellular levels of polyamines observed by DAPI staining (Fig. 1B). When these cultures were are regulated at various steps, including synthesis, degradation, further diluted 1:10 into fresh medium without added spermi- uptake, and excretion, and cells have developed intricate mech- dine (calculated 10Ϫ9 M spermidine in the culture), growth anisms to ensure tight regulation of intracellular polyamine continued at the same rate for Ϸ2 h but then gradually de- pools (3, 4). We have found that, of the three amines, spermidine creased. No gross morphological changes were seen at 6 h (Fig. is most important for cell growth, and near-normal growth is 1B, point B), but after 24 h of growth (Fig. 1B, point C)in obtained in mutant cultures containing extremely low concen- amine-free medium, the growth rate was very slow and the cells trations of spermidine (5–7). showed the abnormal morphology, enlarged size, reduced bud- One of the important functions of spermidine is its role as a ding and diffuse nuclear staining that we described earlier (5, 19) substrate for the hypusine modification of the putative eukary- (Fig. 1B). After 60–90 h in the amine-deficient medium, growth otic translation initiation factor 5A (eIF5A) (Scheme 1). eIF5A stopped, and a large portion of cells was not viable (19). is a highly conserved and essential protein present in all organ- Because the spe2⌬ cells accumulated large amounts of pu- isms from archaebacteria to mammals (8–10), and the hypusine/ trescine that might be responsible for the growth and morpho- deoxyhypusine modification of eIF5A is essential in all eukary- logical defects, we carried out comparable studies on a spe1⌬ otic cells (11). Hypusine was first isolated from a brain extract by spe2⌬ mutant that contained no putrescine or other polyamines Shiba et al. (12), and later Folk and colleagues (13–15) showed and found the same results as described above for the spe2⌬ that the eIF5A precursor undergoes a unique posttranslational mutant (data not shown). modification to form the hypusine residue. This modification We wished to examine how depletion of cellular polyamine occurs exclusively in this protein and proceeds by two consec- levels and concomitant decrease in hypusinated eIF5A (see utive steps. First, precursor eIF5A is modified by deoxyhypusine below) affects protein synthesis in vivo by administering a pulse synthase, where the aminobutyl group of spermidine is attached of 3[H]-leucine to spe2⌬ cells (Y536) and to wild-type cells to a specific residue, followed by hydroxylation of the (Y534). The results (Fig. 1C) show there is only a small decrease intermediate by deoxyhypusine hydroxylase to form the active in protein synthesis in cells grown in 10Ϫ8 M spermidine com- hypusinated eIF5A (16). pared with wild type, but a 70% decrease in protein synthesis in Hypusine modification by spermidine is essential for growth and protein synthesis (8, 10, 11) and in our current studies, we show that when a polyamine auxotroph is grown in the presence Author contributions: M.K.C., M.H.P., and H.T. designed research, performed research, of very limiting concentrations of spermidine, most of the contributed new reagents/analytic tools, analyzed data, and wrote the paper. spermidine is used for the modification of eIF5A. Further The authors declare no conflict of interest. support for the concept that this modification of eIF5A is the ‡To whom correspondence should be addressed. E-mail: [email protected]. major function of spermidine, particularly when grown with This article contains supporting information online at www.pnas.org/cgi/content/full/ limiting concentrations of spermidine is that several of the 0710970105/DC1.

6554–6559 ͉ PNAS ͉ May 6, 2008 ͉ vol. 105 ͉ no. 18 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0710970105 Downloaded by guest on September 26, 2021 Scheme 1.

cells grown in 10Ϫ9 M spermidine. Thus, the protein synthesis cells grown in 10Ϫ6 M spermidine, 0.027 mM in cells grown in data are consistent with the growth-rate data (Fig. 1A). 10Ϫ7 M spermidine, 0.002 mM in cells grown in 10Ϫ8 M spermidine, and 0.0002 mM in cells grown in 10Ϫ9 M spermidine, Spermidine and Hypusine Content of a spe2⌬ Auxotroph Grown in whereas that for the wild-type cells was estimated to be Ϸ2mM Different Concentrations of [3H]-Spermidine. We grew the spe2⌬ (see below). cells in medium containing [3H]-spermidine to obtain accurate For each concentration, a portion of intracellular radioactivity quantitation of the intracellular spermidine and hypusine con- was found in the TCA-insoluble fraction. After thorough wash- tent of cells grown on low spermidine concentrations. ing of the TCA precipitate, the proteins were hydrolyzed in HCl, spe2⌬ cells were grown in different concentrations of [3H]- and the protein-bound radioactive component (Ͼ95%) was spermidine and harvested when the OD600 nm was 1.0. At each identified as hypusine by ion exchange chromatography. The concentration, Ͼ85% of the added radioactivity was taken up by total amount of hypusinated eIF5A was calculated for each the cells. The cells were extracted with trichloroacetic acid culture by dividing the total radioactivity in the TCA-insoluble (TCA), and the extract was subjected to HPLC as described in fraction by the specific activity of the aminobutyl moiety (1.6 ϫ Materials and Methods. Most of the radioactivity (Ͼ90%) was 107 DPM/nmol) of [1,8-3H]-spermidine added to the culture. As found in the spermidine and diaminopropane fractions. The shown in Table 2, the amount of hypusinated eIF5A markedly spermidine area was collected, and the radioactivity [disintegra- decreases when cells are grown in lower concentrations of tions per minute (DPM)] in this fraction was measured. The spermidine. concentration of spermidine in nanomoles was calculated by A comparable experiment was also carried out with wild-type dividing the radioactivity in this fraction by the specific activity cells by growing the cells with Ϸ3.0 ϫ 10Ϫ8 M[3H]-spermidine. of the commercial spermidine (3.2 ϫ 107 DPM per nmol) added The spemidine content of wild-type cells was 20 nmol/mg protein to the culture, because in the spe2⌬ mutant, no endogenous cold (Table 2), which is equivalent to 2 mM intracellular concentra- spermidine is synthesized. Based on these assays, the internal tion. To obtain the specific activity of the spermidine in the cells, spermidine concentration was estimated as 0.36 mM in spe2⌬ we chromatographed the TCA extract on an HPLC column;

Table 1. Comparison of polyamine content of wild-type cells and of polyamine auxotrophs grown in 10؊8 M spermidine (nmol/mg protein) Strain Genotype* Doubling time, h Putrescine Spermidine Spermine BIOCHEMISTRY Y534 Wild type (BY4741) 1.75 1 24 2.3 Y535† KANMX::spe1⌬ (Y5034) 2.1 0 0‡ 0.03‡ Y536† KANMX::spe2⌬ (Y1743) 2.2 36 0.1‡ 0‡ Y537† KANMX::spe3⌬ (Y5488) 2.2 55 0.1‡ 0‡

One milliliter of cell culture at 1.0 OD600 contained Ϸ mg of wet weight cells (Ϸ200 ␮g of protein). *With reference to polyamine biosynthetic genes. Saccharomyces Genome Deletion Project, Stanford University, Stanford, CA; Identification number (Open Biosystem) is shown in parentheses. †The medium contained 10Ϫ8 M spermidine. ‡The limit of detection is 0.03 nmol per mg of protein, and the limit of accuracy for the assay method is 0.05 nmol spermidine or spermine per mg of protein.

Chattopadhyay et al. PNAS ͉ May 6, 2008 ͉ vol. 105 ͉ no. 18 ͉ 6555 Downloaded by guest on September 26, 2021 Fig. 1. Effect of spermidine deprivation on growth (A), morphological abnormalities (B), and rate of protein synthesis (C)ofspe2⌬ cells. (B) Cultures were grown Ϫ8 in 10 M spermidine and diluted at zero time to an OD600 of 0.001. The cultures were grown to an OD600 of 1 and at time A diluted into amine-free medium. The optical density values have been corrected for these dilutions. At times A, B, and C, cultures were taken for morphological studies. Cultures were harvested, fixed, and stained with DAPI and observed with a DAPI filter set (Lower). The same cells were also observed by DIC microscopy (Upper). (Scale bar ϭ 10 ␮m.) Parallel cultures were grown at higher spermidine concentrations as a control. In a separate experiment, to determine the protein synthesis rate, 3[H]-leucine was added into spe2⌬ cultures grown in different spermidine concentrations, and to wild-type yeast cells. The cultures were incubated for 20 min, harvested, washed, and 3[H]-leucine incorporation into protein was determined in TCA precipitates as described in Materials and Methods.

Ͼ90% of the counts migrated with the spermidine peak, whereas (Y534) cells. Strikingly, the cells grown in 10Ϫ8 M spermidine a small portion (Ͻ10%) was found in the diaminopropane and contained only Ϸ5% of the hypusinated eIF5A content of spermine areas. The amount of spermidine in the spermidine wild-type cells or 13% of that of spe2⌬ cells grown in 10Ϫ6 M peak was determined by the o-phthalaldehyde reaction; the spermidine. The spe2⌬ culture grown in 10Ϫ9 M spermidine radioactivity was measured and the specific activity calculated contained Ͻ2% of the hypusinated eIF5A found in spe2⌬ cells (this calculation is an approximation, because the specific activ- grown in 10Ϫ6 M spermidine. The finding that growth rates and ity of the spermidine isolated from the cells at the end of the protein synthesis are only slightly decreased in 10Ϫ8 M spermi- incubation was lower than that present earlier in the incubation dine, even though the cellular spermidine and hypusine are because of the synthesis of cold spermidine during growth). reduced by Ϸ1,000- and Ϸ20-fold, respectively, indicates that To obtain the number of nanomoles of hypusine present in the normal levels of both spermidine and hypusine are in large excess TCA-insoluble fraction of the wild-type cells, the number of over that needed for growth. counts in this fraction was divided by the specific activity of the Interestingly, the percentage of the radioactivity in the pro- aminobutyl moiety of the spermidine isolated from the cells (see tein-bound hypusine was remarkably higher in the cells grown in above). The amount of modified eIF5A present in wild-type cells the lower concentration of spermidine (Table 2). Thus, in spe2⌬ was estimated to be 0.28 nmol/mg protein in the wild-type cells grown in 10Ϫ8 M spermidine, 37% of the intracellular

Table 2. Labeling of yeast cells with [3H]-spermidine Spermidine* Hypusine* Hypusine/ DPM/ml nmol/mg DPM/ml nmol/mg (spermidineϩhypusine), Cultures per treatment cells protein cells protein %

Wild type (Y534) (with 5 ␮Ci/ml 4.86 ϫ 106 20 3.46 ϫ 104 0.28 1.4 [3H]-spermidine) Y536-10Ϫ6 M[3H]-spermidine 2.70 ϫ 107 4.2 3.15 ϫ 105 0.098 2.3 Y536-10Ϫ7 M[3H]-spermidine 1.84 ϫ 106 0.29 1.60 ϫ 105 0.05 15 Y536-10Ϫ8 M[3H]-spermidine 1.39 ϫ 105 0.022 4.13 ϫ 104 0.013 37 Y536-10Ϫ9 M[3H]-spermidine 1.10 ϫ 104 0.0017 6.35 ϫ 103 0.002 54

Spermidine and hypusine content of wild-type and spe2⌬ cultures grown in different concentrations of [3H]-spermidine. *The spermidine and hypusine content in the wild-type and spe2⌬ strains was estimated after chromatographic separation, as described in the text.

6556 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0710970105 Chattopadhyay et al. Downloaded by guest on September 26, 2021 (Fig. 2B), an indication that the unmodified eIF5A precursor accumulates in spermidine-deprived spe2⌬ cells. Thus spermi- dine depletion does not seem to inhibit de novo synthesis of eIF5A protein but limits only its hypusine modification.

Growth Effects of Spermidine Analogs in Complete Absence of Sper- midine. A number of polyamine analogs were also tested (at 10Ϫ7 M concentrations) for their ability to support growth of the spe2⌬ auxotroph in the absence of added spermidine. Only 1-methylspermidine supported growth of the S. cerevisiae poly- amine auxotroph (2.3-h as compared with 1.7-h doubling time for spermidine). Furthermore, the S-stereoisomer of 1-methyl- spermidine showed better growth than the R-isomer (2.1 h for the S-isomer as compared with 2.9 h for the R-isomer). Addition of 8-methylspermidine, caldine, and putrescine had no effect on growth. Discussion The data in this paper, plus our previous reports (5–7), dem- onstrate that very low concentrations of spermidine (10Ϫ8 M) are sufficient for near-normal growth and protein synthesis of a spe2⌬ polyamine auxotroph of S. cerevisiae. In view of the importance of spermidine as a precursor of hypusine (in eIF5A), we were interested in what effect, if any, this large decrease in the internal concentration of spermidine in polyamine auxotroph cells (grown in 10Ϫ8 M spermidine) might have on the concen- Fig. 2. Level of hypusinated eIF5A (A) and total eIF5A (B) in wild-type and tration of total eIF5A (i.e., modified and unmodified) and on the different spermidine-supplemented spe2⌬ cultures. The yeast spe2⌬ culture concentration of modified (hypusinated) eIF5A. As shown in Ϫ8 was grown in 10 M spermidine and then diluted to an OD600 of 0.001 in Ϫ6 Ϫ7 Ϫ8 Ϫ9 3 Fig. 2B, we found that the total amount of eIF5A was essentially media containing 10 ,10 ,10 , and 10 M[ H]-spermidine, as indicated. unchanged, but that there was a large decrease (1/20) in the The wild-type culture was also grown in the presence of 5 ␮Ci/ml [3H]- amount of hypusinated eIF5A (Table 2 and Fig. 2A). Because spermidine. Cultures were harvested at an OD600 of 1 and subjected to analysis by SDS/PAGE. The migration of [3H]-spermidine incorporated as hypusine in the polyamine auxotrophs grow at near-normal growth rates in Ϫ8 eIF5A, and standard yeast eIF5A were determined by fluorography and stain- 10 M spermidine, these data were surprising in indicating that ing. Note that the [3H]-spermidine level in wild-type cells is diluted by intra- normally the yeast cells contain a many-fold excess of hypusi- cellular biosynthesis of cold spermidine (A). Total eIF5A (hypusinated and nated eIF5A than is needed for optimum growth. These obser- precursor) was determined in the above cultures by Western blot analysis vations are important in emphasizing the need for marked against rabbit antiyeast eIF5A antibody, kindly provided by Sandro Valentini depletion of the hypusinated eIF5A in any experiments designed (Sa˜o Paulo State University, Universidade Estadual Paulista, Araraquara, SP, to test the biologic effect of this modification and may account Brazil) (B). Equal loading of proteins in the gel lanes in A and B was confirmed for the difficulties observed in the past experiments designed to by staining with Coomassie and ponceau-S, respectively [see supporting in- formation (SI) Fig. 4]. evaluate the importance of eIF5A on protein synthesis (11). Most interesting was the observation (Table 2) that in the spe2⌬ cells grown on the lower concentrations of spermidine, a spermidine was converted to hypusine and in the cells grown in much larger percentage of the available spermidine was used for 10Ϫ9 M spermidine, 54% of the radioactive spermidine was the synthesis of hypusine; i.e., 54% of the available spermidine converted to hypusine. This is in contrast to wild-type cells or was used for hypusine biosynthesis in cells grown in very limiting Ϫ9 Ϫ6 spe2⌬ cells grown in 10Ϫ6 M spermidine, where only 1.4% or spermidine (10 M) vs. 1–3% for cells grown in 10 M 2.3% of the spermidine was used for hypusine synthesis. That an spermidine or in a prototrophic strain. Considering the known increasing percentage of spermidine is mobilized for hypusine binding of spermidine to other acidic macromolecules in cells, it synthesis as spermidine becomes limiting suggests the most is remarkable that 54% of cellular spermidine can be used for important function of cellular spermidine in yeast is the hypusine hypusine synthesis under these limiting conditions. One expla- modification of eIF5A. nation for these results could be that the hypusine modification is an enzymatic irreversible reaction, and thus may act as a ‘‘sink’’ We examined the radio-labeled protein by fluorography after for any available spermidine. Our results are consistent with the SDS/PAGE (Fig. 2A). In all cases, the radioactivity in the observation of Gerner et al. (22) that when spermidine was TCA-insoluble fraction migrated as a single band corresponding added to ␣-difluoromethylornithine (DFMO)-treated rat hep- to yeast eIF5A. The intensities of the labeled eIF5A band in the atoma cells, it was immediately incorporated into eIF5A before fluorogram decreased in cells cultured in lower concentration of 3 the intracellular content of free spermidine was elevated. [ H]-spermidine and are consistent with the amounts of hypusine When the cells are incubated in an even lower concentration (Table 2) estimated from acid hydrolysates of TCA precipitated Ϫ9 of spermidine (10 M or lower), the intracellular spermidine BIOCHEMISTRY proteins. We also measured the amount of total eIF5A in these and modified eIF5A concentrations go below the critical levels, cells by Western blot analysis (Fig. 2B) using a rabbit polyclonal and growth gradually stops. The importance of hypusine mod- antibody against recombinant yeast eIF5Aa (Tif51a), which ification for cell growth is further substantiated by finding that, recognizes the unmodified eIF5A precursor as well as the of the various analogs tested, only 1-methylspermidine sup- hypusinated eIF5A. In contrast to the fluorogram (Fig. 2A), ported growth of the S. cerevisiae polyamine auxotroph, and this which showed a sharp decline in the hypusinated eIF5A content analog is the one that has been definitively shown in in vitro in spe2⌬ cells cultured in lower spermidine concentrations, the studies to act as a substrate for the modification of eIF5A (16, total eIF5A protein (unmodified plus modified) did not change 23). Also, of the R- and S-stereoisomers of 1-methylspermidine, in cultures grown with different concentration of spermidine the S-isomer was better in supporting growth, and this isomer

Chattopadhyay et al. PNAS ͉ May 6, 2008 ͉ vol. 105 ͉ no. 18 ͉ 6557 Downloaded by guest on September 26, 2021 had been reported to be a better substrate for modification of Materials and Methods eIF5A in mammalian cells (24). Previously, we have shown that Strains and Growth Conditions. The strains used in this study are listed in Table in a polyamine auxotroph that cannot convert spermine to 1 (BY4741 background, Mat a his3 leu2 met15 ura3) and were maintained on spermidine because of a deletion in the FMS1 gene (spermine yeast extract peptone adenine dextrose (YPAD) agar plates. Before use, the oxidase), spermine does not support growth (7), and it has been strains were partially deprived of endogenous amines by incubation in min- Ϫ shown that spermine cannot be used as a substrate for deoxy- imal medium containing 10 8 M spermidine. The minimal medium consisted hypusine synthase (16). Further support for the major impor- of 0.67% yeast nitrogen base (Q-BIOgene), 2% glucose, and any required tance of hypusine formation in spermidine-limited cultures is supplements; to each liter of medium, 10 ml of1MK2HPO4 was added, resulting in an initial pH of 6.5. All incubations were carried out at 30°C with that the morphological defects we observed previously and in our shaking in air. The cells were diluted 1:1,000 into the same medium (contain- current studies (Fig. 1B) during polyamine deprivation are ing 10Ϫ8 M spermidine) to give a calculated optical density at zero time of similar in eIF5A-depleted cells (containing a normal amount of Ϸ0.001. The cultures were then incubated and harvested by centrifugation

polyamines). For example, either eIF5A or spermidine depletion when the OD600 reached 1. results in G1-S arrest in the cell cycle of yeast and mammalian cells (11, 16, 18); abnormalities in actin polarization have been Polyamine Analysis by HPLC. The cell pellet was extracted with 5 volumes of reported because of deprivation of either spermidine or eIF5A 10% perchloric acid or 10% trichloroacetic acid. Aliquots (usually 50 ␮l) were (5, 18, 20, 25). Recent studies from our laboratory (19) and by assayed for polyamines by HPLC chromatography essentially as described by Schrader et al. (26) have reported apoptotic cell death in yeast Murakami and Hayashi (35) using a Shim-pack column (Shimadzu, catalogue during spermidine depletion (19) or in eIF5A temperature no. ISC-05/S0504). Postcolumn fluorometric determination of polyamines was performed by reaction with o-phthalaldehyde (6), and data were collected by sensitive mutants (26). Like eIF5A mutants, spermidine- using Powerchrome software (eDAQ). Intracellular polyamine concentrations depleted yeast cells are highly sensitive to protein synthesis were calculated by assuming that 90% of the yeast cells wet weight is water. inhibitors, such as paromomycin (27, 28). Taken together, these findings strongly suggest that hypusine modification of eIF5A is Determination of Intracellular Spermidine and Hypusine Content of a Wild-Type a most critical function for spermidine in supporting the growth and a spe2⌬ Mutant Cultured in the Presence of [3H]-Spermidine. The yeast cells of S. cerevisiae. These findings are also important in indicating were cultured in 10 ml of minimal medium containing the indicated amounts of the need for reinvestigating the concentrations of spermidine [1,8-3H]-spermidine (16.6 Ci/mmol, Perkin–Elmer/NEN, catalogue no. NET522; required for the many reported functions for polyamines and for HPLC analysis showed Ͼ90% of the counts migrated in the spermidine area) evaluating whether they are primary effects of the polyamines or overnight. Cells (at 1.0–1.6 OD600) were harvested and washed with fresh culture secondary to the growth defects resulting from inadequate medium. To each cell pellet, 0.2 ml of 10% TCA was added and kept on ice for 20 ϫ eIF5A modification. min. After centrifugation (14,000 g) for 5 min, the TCA supernatant was separated for polyamine and radioactivity analyses. The TCA precipitates were The function of polyamines in mammalian cells may be much resuspended in 1 ml of 10% TCA containing 1 mM each unlabeled polyamines more complex, in view of their implicated roles in the regulation (putrescine, spermidine, spermine) to remove any radioactive spermidine non- of diverse cellular activities such as transcriptional, translational, covalently bound to the TCA precipitates. This wash was repeated (Ͼ3ϫ) until and posttranslational levels affecting proliferation, transforma- there was no radioactivity in the TCA wash. A portion of the washed TCA tion, differentiation, and apoptosis. Indeed, Nishimura et al. precipitate (1/10 of the total) was resuspended in 0.1 ml of 0.2 M NaOH, and an (29), using inhibitors, have demonstrated an independent role aliquot was used for protein determination by the bicinchoninic acid protein for polyamines and eIF5A in supporting the proliferation of assay (Promega). Another portion of the precipitate (1/5 of the total) was dis- mouse mammary carcinoma cells. In addition, Nagarajan et al. solved in the SDS sample buffer, subjected to SDS/PAGE, and used for fluoro- 3 (30) have shown that either spermidine or spermine derivatives graphic detection of [ H]-labeled hypusine-containing protein, eIF5A and for detection of total eIF5A protein (unmodified precursor and modified eIF5A) by can support growth of polyamine depleted SV-3T3 cells without Western blotting (11). A third portion of the washed TCA precipitate (1/5) was the need for metabolic conversion. Furthermore, cytostasis of hydrolyzed in 6 M HCl at 108°C overnight, and the amount of radioactive mammalian cells induced by the decarboxylase (ODC) hypusine present was determined after ion exchange chromatography as inhibitor DFMO (24) or a S-adenosylmethionine decarboxylase described, with slight modification (13). inhibitor (25) has been reported to have two phases, an acute initial phase due to reduction of total cellular polyamines and a Assay of Protein Synthesis in Wild-Type and spe2⌬ Mutant Grown in Different delayed second phase due to eventual deprivation of eIF5A. Spermidine Concentrations. The wild-type (Y534) and mutant strains (Y536) Because the polyamine auxotrophs are able to grow at near- were cultured in minimal medium containing the indicated concentrations of Ϫ normal rates in media containing 10 8 M spermidine even with spermidine. Duplicates of 3.75 ml of the exponentially growing cells (OD600 nm a low internal concentration of spermidine and of eIF5A, these of 0.3–0.4) were harvested by centrifugation at 2,000 ϫ g for 5 min at room findings lead to the unanswered question of why wild-type cells temperature, and 2.75 ml of medium was removed leaving 1 ml of medium in which cells were resuspended. 30 ␮Ci of [3H]-leucine [L-[4,5-3H(N)]leucine, 60 have a large excess of internal spermidine (1,000-fold more than Ci/mmol, Perkin–Elmer/NEN catalogue no. NET135H] was added, and cells needed for near-optimum growth) and of hypusinated eIF5A Ϸ were incubated for 20 min at 30°C with shaking. The reaction was stopped by ( 20-fold more than needed for growth). It is also noteworthy addition of 0.5 ml of ice-cold stop solution (1 mg/ml of cycloheximide and 3 that prokaryotes such as Escherichia coli do not have hypusine mg/ml of leucine). Cells were kept on ice for 1 min and centrifuged in the or eIF5A and yet contain large amounts of putrescine and microfuge. The cell pellets were resuspended in 1 ml of 15% TCA, and the spermidine that appear to have a number of physiological effects mixture was heated at 95°C for 10 min. The samples were cooled on ice and such as transcription, translation, and protection from stress (4, after centrifugation, the TCA precipitate was washed three times with 10% 31–34). It is also surprising that in vivo protein synthesis (Fig. 1B) TCA to remove free [3H]-leucine. A portion of the washed TCA precipitate (1/5) was nearly as efficient in cultures grown in 10Ϫ8 M spermidine was hydrolyzed in 6 M HCl at 108°C overnight, and the radioactivity in the acid 3 as in wild-type cells, in view of the reported effects of spermidine hydrolysate was counted to determine the amount of [ H]-leucine incorpo- rated into proteins. Another portion (1/5) of the TCA precipitate was resus- and polyamines on the stabilization of nucleic acids and on the pended in 0.1 ml of 0.2 M NaOH, and an aliquot was used for protein efficiency and fidelity of in vitro translation. Further studies are determination by the BCA method. needed to test whether the excess amounts of cellular polyamines in yeast and may be important for the fidelity of Microscopic Studies. For nuclear staining, cells were grown as above in differ- replication, transcription, and translation and for the long-term ent concentrations of spermidine, harvested, and fixed with 70% ethanol for genetic preservation and survival of the organism. 1 h on ice, and the cells were resuspended in PBS containing 1 ␮g/ml DAPI

6558 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0710970105 Chattopadhyay et al. Downloaded by guest on September 26, 2021 (Sigma) and 1 mg/ml p-phenylenediamine (Sigma) for 30 min. The cells were ACKNOWLEDGMENTS. We thank Dr. Senya Matsufuji of Jikei University then washed twice in PBS and observed in a fluorescence microscope (Zeiss School of Medicine, Tokyo, for the Shim-pack column and the detailed HPLC Axiophot) equipped with a digital camera (Cool Snap HQ, Photometrics), and protocol for polyamines. We thank Drs. Alex R. Khomutov and Nikolay Gri- gorenko (Russian Academy of Sciences, Moscow) for the R- and S- IP Lab software (for capture) through a DAPI filter set. The cells of the same stereoisomers of 1-methylspermidine. This research was supported by the fields were also visualized by differential interference contrast (Nomarski) Intramural Research Program of the National Institutes of Health (National microscopy. Institute of Diabetes, Digestive and Kidney Diseases).

1. Tabor CW, Tabor H (1984) Polyamines. Annu Rev Biochem 53:749–790. 19. Chattopadhyay MK, Tabor CW, Tabor H (2006) Polyamine deficiency leads to accumu- 2. Pegg AE (1986) Recent advances in the biochemistry of polyamines in . lation of reactive oxygen species in a spe2⌬ mutant of Saccharomyces cerevisiae. Yeast Biochem J 234:249–262. 23:751–761. 3. Wallace HM, Fraser AV, Hughes A (2003) A perspective of polyamine metabolism. 20. Chatterjee I, Gross SR, Kinzy TG, Chen KY (2006) Rapid depletion of mutant eukaryotic Biochem J 376:1–14. initiation factor 5A at restrictive temperature reveals connections to actin cytoskeleton 4. Cohen SS (1998) A Guide to the Polyamines (Oxford Univ Press, New York), pp 1–595. and cell cycle progression. Mol Genet Genomics 275:264–276. 5. Balasundaram D, Tabor CW, Tabor H (1991) Spermidine or spermine is essential for the 21. Chattopadhyay MK, Tabor CW, Tabor H (2005) Studies on the regulation of ornithine aerobic growth of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 88:5872–5876. decarboxylase in yeast: effect of deletion in the MEU1 gene. Proc Natl Acad Sci USA 6. Hamasaki-Katagiri N, Tabor CW, Tabor H (1997) Spermidine biosynthesis in Saccharo- 102:16158–16163. myces cerevisae: polyamine requirement of a null mutant of the SPE3 gene (spermidine 22. Gerner EW, Mamont PS, Bernhardt A, Siat M (1986) Post-translational modification of synthase). Gene 187:35–43. the protein-synthesis initiation factor eIF-4D by spermidine in rat hepatoma cells. 7. Chattopadhyay MK, Tabor CW, Tabor H (2003) Spermidine but not spermine is essential Biochem J 239:379–386. for hypusine biosynthesis and growth in Saccharomyces cerevisiae: spermine is con- 23. Byers TL, Ganem B, Pegg AE (1992) Cytostasis induced in L1210 murine leukaemia cells verted to spermidine in vivo by the FMS1-amine oxidase. Proc Natl Acad Sci USA by the S-adenosyl-L-methionine decarboxylase inhibitor 5Ј-([(Z)-4-amino-2-butenyl]m- 100:13869–13874. ethylamino)-5Ј-deoxyadenosine may be because of hypusine depletion. Biochem J 8. Schnier J, Schwelberger HG, Smit-McBride Z, Kang HA, Hershey JW (1991) Translation 287:717–724. initiation factor 5A and its hypusine modification are essential for cell viability in the 24. Hyvonen MT, et al. (2007) Role of hypusinated eukaryotic translation initiation factor yeast Saccharomyces cerevisiae. Mol Cell Biol 11:3105–3114. 5A in polyamine depletion-induced cytostasis. J Biol Chem 282:34700–34706. 9. Chen KY, Liu AY (1997) Biochemistry and function of hypusine formation on eukaryotic 25. Zanelli CF, Valentini SR (2005) Pkc1 acts through Zds1 and Gic1 to suppress growth and initiation factor 5A. Biol Signals 6:105–109. cell polarity defects of a yeast eIF5A mutant. Genetics 171:1571–1581. 10. Wolff EC, Kang KR, Kim YS, Park MH (2007) Posttranslational synthesis of hypusine: 26. Schrader R, Young C, Kozian D, Hoffmann R, Lottspeich F (2006) Temperature-sensitive evolutionary progression and specificity of the hypusine modification. Amino Acids eIF5A mutant accumulates transcripts targeted to the nonsense-mediated decay path- 33:341–350. way. J Biol Chem 281:35336–35346. 11. Kang HA, Hershey JW (1994) Effect of initiation factor eIF-5A depletion on protein 27. Balasundaram D, Tabor CW, Tabor H (1999) Sensitivity of spermidine-deficient Sac- synthesis and proliferation of Saccharomyces cerevisiae. J Biol Chem 269:3934–3940. charomyces cerevisiae to paromomycin. Antimicrob Agents Chemother 43:1314–1316. 12. Shiba T, Mizote H, Kaneko T, Nakajima T, Kakimoto Y (1971) Hypusine, a new amino 28. Zanelli CF, et al. (2006) eIF5A binds to translational machinery components and affects acid occurring in bovine brain. Isolation and structural determination. Biochim Biophys translation in yeast. Biochem Biophys Res Commun 348:1358–1366. Acta 244:523–531. 29. Nishimura K, et al. (2005) Independent roles of eIF5A and polyamines in cell prolifer- 13. Park MH, Cooper HL, Folk JE (1981) Identification of hypusine, an unusual , in a protein from human lymphocytes and of spermidine as its biosynthetic precursor. ation. Biochem J 385:779–785. Proc Natl Acad Sci USA 78:2869–2873. 30. Nagarajan S, Ganem B, Pegg AE (1988) Studies of nonmetabolizable polyamines that 14. Cooper HL, Park MH, Folk JE (1982) Posttranslational formation of hypusine in a single support growth of SV-3T3 cells depleted of natural polyamines by exposure to alpha- major protein occurs generally in growing cells and is associated with activation of difluoromethylornithine. Biochem J 254:373–378. lymphocyte growth. Cell 29:791–797. 31. Igarashi K, Kashiwagi K, Kishida K, Kakegawa T, Hirose S (1981) Decrease in the S1 15. Cooper HL, Park MH, Folk JE, Safer B, Braverman R (1983) Identification of the protein of 30-S ribosomal subunits in polyamine-requiring mutants of Escherichia coli hypusine-containing protein hyϩ as translation initiation factor eIF-4D. Proc Natl Acad grown in the absence of polyamines. Eur J Biochem 114:127–131. Sci USA 80:1854–1857. 32. Minton KW, Tabor H, Tabor CW (1990) Paraquat toxicity is increased in Escherichia coli 16. Park MH (2006) The post-translational synthesis of a polyamine-derived amino acid, defective in the synthesis of polyamines. Proc Natl Acad Sci USA 87:2851–2855. hypusine, in the eukaryotic translation initiation factor 5A (eIF5A). J Biochem (Tokyo) 33. Jung IL, Oh TJ, Kim IG (2003) Abnormal growth of polyamine-deficient Escherichia coli 139:161–169. mutant is partially caused by oxidative stress-induced damage. Arch Biochem Biophys 17. Park MH, Joe YA, Kang KR (1998) Deoxyhypusine synthase activity is essential for cell 418:125–132. viability in the yeast Saccharomyces cerevisiae. J Biol Chem 273:1677–1683. 34. Chattopadhyay MK, Tabor CW, Tabor H (2003) Polyamines protect Escherichia coli cells 18. Chattopadhyay MK, Tabor CW, Tabor H (2002) Absolute requirement of spermidine for from the toxic effect of oxygen. Proc Natl Acad Sci USA 100:2261–2265. growth and cell cycle progression of fission yeast (Schizosaccharomyces pombe). Proc 35. Murakami Y, Hayashi S (1985) Role of antizyme in degradation of ornithine decarbox- Natl Acad Sci USA 99:10330–10334. ylase in HTC cells. Biochem J 226:893–896. BIOCHEMISTRY

Chattopadhyay et al. PNAS ͉ May 6, 2008 ͉ vol. 105 ͉ no. 18 ͉ 6559 Downloaded by guest on September 26, 2021