Polyamines Enhance Synthesis of the RNA Polymerase ς38 Subunit by Suppression of an Amber Termination Codon in the Open Reading Frame*

The mechanisms by which polyamines stimulate synthesis of the RNA polymerase ς38 subunit inEscherichia coli were studied. Polyamine stimulation was observed only in strains in which the 33rd codon of RpoS mRNA is a UAG termination codon instead of a CAG codon for glutamine in wild-typeE. coli. Readthrough of the termination codon by Gln-tRNAsupE was stimulated by polyamines. This stimulation was found to be caused by an increase in both the level of suppressor tRNAsupE and the binding affinity of Gln-tRNAsupE for ribosomes. The stimulatory effect was observed with a UAG termination codon but not with UGA and UAA codons. Readthrough of the UAG termination codon at the 270th amino acid position of RpoS mRNA was also stimulated by polyamines, indicating that polyamines stimulate readthrough of a UAG codon regardless of its location within the RpoS mRNA. When cell viability of anE. coli strain having a termination codon in the 33rd position of RpoS mRNA was compared using cells cultured with or without putrescine, it was higher in cells cultured with putrescine than in cells cultured without putrescine. The level of ς38 subunit in the cells cultured with putrescine was higher than that in cells cultured without putrescine on days 2, 4, and 8, but the level of ς70 subunit was almost the same in cells cultured with or without putrescine. These results confirm that elevated expression of the rpoS gene is important for cell viability at late stationary phase.

Polyamines, aliphatic cations present in almost all living organisms, are necessary for normal cell growth (1,2). Because polyamines interact with nucleic acids and mostly exist as polyamine-RNA complexes in cells (3,4), their proliferative effects are presumed to be caused by stimulation of nucleic acid and protein synthesis. In fact, it has been reported that polyamines stimulate the synthesis of some protein species in vitro (5)(6)(7) and in vivo (8,9), induce the in vivo assembly of 30 S ribosomal subunits (10 -12), and increase the fidelity of protein synthesis (13)(14)(15), altogether suggesting that polyamines reg-ulate protein synthesis at several different steps.
In Escherichia coli, the synthesis of OppA protein, a periplasmic substrate-binding protein of the oligopeptide uptake system, is strongly stimulated by the addition of putrescine to a polyamine-requiring mutant, MA261 (9). We found that (i) the stimulation of OppA synthesis takes place at the level of translation; (ii) the position and secondary structure of the Shine-Dalgarno (SD) 1 sequence (16) on OppA mRNA are important for this stimulation (17); and (iii) polyamines cause a structural change of the SD sequence and the initiation codon AUG of OppA mRNA, facilitating formation of the initiation complex (18). We also found that polyamines increase the translation of adenylate cyclase (Cya) mRNA by facilitating UUG codon-dependent initiation (19). Analysis of RNA secondary structure suggests that exposure of the SD sequence of mRNA is a prerequisite for polyamine stimulation of UUG codon-dependent initiation (19).
In the present study we found a novel mechanism of the polyamine stimulation, which is involved in the enhanced synthesis of RNA polymerase 38 subunit by polyamines (19). This type of the polyamine stimulation was observed only in E. coli strains in which the 33rd codon of RpoS mRNA is a UAG amber termination codon instead of a CAG codon for glutamine in wild-type E. coli strains. Readthrough of the termination codon by Gln-tRNA supE was stimulated by polyamines.
E. coli MA261 and MA261 rpoS::tet were grown at 37°C in medium A supplemented with 5 amino acids (100 g/ml each of Gly, Leu, Met, Ser, and Thr) in the presence (100 g/ml) or absence of putrescine (9). HT283 and DR112 strains were cultured according to the method of Hafner et al. (21) and Linderoth and Morris (22), respectively, in the presence (100 g/ml) or absence of putrescine. Antibiotics used were 100 g/ml ampicillin, 50 g/ml kanamycin, 30 g/ml chloramphenicol, and 15 g/ml tetracycline. Cell growth was monitored by measuring the absorbance at 540 nm. * This work was supported by a grant-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science, and Technology, Japan, and by Research for the Future Program Grant JSPS-RFTF 97L00503 from the Japan Society for the Promotion of Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Plasmids-Plasmids pMRTCG (ϭpMW33TCG) and pMRTAG (ϭpMW33TAG) (26) were kind gifts from Dr. Y. Kamio (Tohoku University). Total chromosomal DNA from E. coli was prepared according to the method of Ausubel et al. (27). For construction of pMWrpoS(33CAG) (ϭpMW33CAG) or pMWrpoS(270TAG) (ϭpMW270TAG), PCR was performed using 5Ј-CAACGAATTCCGT-GACCTTGCTCAGCGCA-3Ј (P1) and 5Ј-TGCAAGCTTGTATGGGCG-GTAATTTGACC-3Ј (P2) as primers and total chromosomal DNA of W3110 types A and B, respectively, as templates. After cutting with EcoRI and HindIII, the 1.9-kb fragment was inserted into the same restriction sites of pMW119 (Nippon Gene, Japan). Site-directed mu-tagenesis by overlap extension using PCR (28) was performed to prepare pMWrpoS(33TGA) (ϭpMW33TGA) and pMWrpoS(33TAA) (ϭpMW33TAA). The template used for the first PCR was chromosomal DNA from E. coli W3110 type A. To make pMWrpoS(33TGA) primers used for the first PCR were P1 and 5Ј-TTCCTCTTCGGCCAAATCGT-TATCACTGGGTTCTCATTCTACTAA-3Ј (underlined base for CAG substitution with TGA) and 5Ј-GCCTTAGTAGAATGAGAACCCAGTG-ATAACGATTTGGCCGAAGAGGAACTGTTATCGCAG-3Ј and P2. The second PCR was performed using the first PCR products as templates and P1 and P2 as primers. To make pMWrpoS(33TAA) primers used for the first PCR were P1 and 5Ј-TTCCTCTTCGGCCAAATCGTTATCACT-GGGTTCTTATTCTACTAA-3Ј (underlined base for CAG substitution with TAA) and 5Ј-GCCTTAGTAGAATAAGAACCCAGTGATAACGAT-TTGGCCGAAGAGGAACTGTTATCGCAG-3Ј and P2. The second PCR was performed using the first PCR products as templates and P1 and P2 as primers. After cutting with EcoRI and HindIII, the 1.9-kb fragment was inserted into the same restriction sites of pMW119.
For preparation of pSTrpoS(33TAG)-His 6 , PCR was performed using MA261 chromosomal DNA as templates and 5Ј-AGGGAATTCGGGTA-GGAGCCACCTTATG-3Ј and 5Ј-AGCCAAGCTTGAGATTAGTGGTGG-TGGTGGTGGTGCTCGAGCTCGCGGAACAGCGCTTCG-3Ј as primers. The 1.1-kb PCR product was inserted into the EcoRV site of pSTBlue-1 using pSTBlue-1 Perfectly Blunt cloning kit (Novagen) according to the manufacturer's instructions. A plasmid in which the rpoS(33TAG)-His 6 gene is under the control of T7 promoter was selected.
Western Blot Analysis-Antisera against 38 and 70 subunits were prepared as described previously (29). Western blot analysis was performed by the method of Nielsen et al. (30) using Proto Blot Western blot AP system (Promega).
Dot Blot and Northern Blot Analysis of RNA-Total RNA was prepared from various E. coli strains according to the method of Emory and Belasco (31). Dot blot analysis of RpoS mRNA was performed according to the method of Sambrook et al. (32). The 1.1-kb PCR product prepared as described above was labeled with [␣-32 P]dCTP using BcaBEST labeling kit (Takara Bio Inc.) and used as a probe. Northern blot analysis of tRNAs encoded by the metT operon was performed using the 0.9-kb PCR product prepared similarly to the probe described above (19).

Measurement of Aminoacylation Level of tRNA Gln and Gln-tRNA
Binding to Ribosomes-Total RNA was prepared from E. coli cells according to the method of Chomczynski and Sacchi (33) using TRIzol reagent (Invitrogen). Polyacrylamide gel electrophoresis, blotting, and detection of Gln-tRNA Gln and tRNA Gln were performed according to the methods of Varshney et al. (34) and Kowal et al. (35) using the 5Ј-end 32 P-labeled primer 5Ј-CCTCGGAATGCCGGAATTAGAATCC-3Ј as a probe for hybridization. Assay of Gln-tRNA formation was performed as described previously (36) except that tRNA obtained from E. coli MA261/pUCsupE (2 A 260 units) and crude aminoacyl-tRNA synthetases (110 g of protein) (37) were used. [ 3 H]Gln-tRNA binding to ribosomes was measured as previously described (37)  Measurement of Cell Viability-Cell viability was determined by counting colony numbers in aliquots of the culture grown on LB-containing 1.5% agar plate at 37°C. Thus, the definition of viable cells is that the cells are able to grow on agar plate.
Preparation of mRNA and in Vitro Translation-pSTrpoS(33TAG)-His 6 was linearized by HindIII, and RpoS(33UAG)-His 6 mRNA was synthesized by T7 RNA polymerase using AmpliScribe T7 High Yield transcription kit (Epicentre Technologies). The 30,000 ϫ g supernatant (S-30) of E. coli C600/pUCsupE was prepared as described previously (37). For in vitro translation, a reaction mixture (0.1 ml) containing 50 mM Tris-HCl, pH 7.5, 60 mM NH 4 Cl, 6 mM 2-mercaptoethanol, 2 mM ATP, 0.5 mM GTP, 4 mM phosphoenolpyruvate, 2.5 g of pyruvate kinase, 2 g of folinic acid, 50 kBq of [ 35 S]methionine (37 TBq/mmol), 0.2 mM each of 19 amino acids without methionine, 10 g of RpoS(33UAG)-His 6 mRNA, 17 A 260 units S-30, and magnesium acetete and spermidine at the specified concentration was incubated at 30°C for 1 h. A 45-l aliquot of each reaction mixture was placed on a 3MM paper disc (Whatman, 24 mm in diameter) and radioactivity insoluble in hot 5% trichloroacetic acid (TCA) was measured with a liquid scintillation spectrometer. A 10-l aliquot of nickel-nitrilotriacetic acid agarose suspension (50% V/V in water, Qiagen) was added to the residual 50-l aliquot, and the mixture was incubated for 30 min on ice. After removal of supernatant, the pellet was mixed with 30 l of sample buffer for SDS-polyacrylamide electrophoresis (38) and boiled for 2 min. A 20-l aliquot was used for 12% SDS-polyacrylamide electrophoresis and fluorography was performed according to the method of Laskey and Mills (39). Radioactivity of labeled 38 -His 6 was quantified using a Fuji Bas-2000II imaging analyzer. 38 Subunit Based on the Readthrough of an Amber Codon in the Open Reading Frame (ORF) of RpoS mRNA-We reported previously that synthesis of the RNA polymerase 38 subunit was stimulated by polyamines using one of the polyaminerequiring mutants, MA261 (19). To confirm this observation, the effect of polyamine addition on the synthesis of 38 subunit was examined using three independent isolates of polyaminerequiring E. coli mutants, MA261, HT283, and DR112 (20 -22). As shown in Fig. 1A, the increased synthesis of 38 subunit was observed for MA261 and HT283 but not for DR112. The level of 70 subunit was, however, nearly equal for all three mutants and was not affected by the addition of polyamines (Fig. 1A). The level of RpoS mRNA in E. coli MA261, as measured by dot blotting, was nearly equal in the presence and absence of polyamine addition (Fig. 1B), indicating that the stimulation of 38 synthesis by polyamines takes place at a post-transcriptional step(s).

Polyamine Stimulation of the Synthesis of RNA Polymerase
To understand why the polyamine stimulation takes place in only two strains (MA261 and HT283) among three polyaminerequiring mutants, the nucleotide sequence of the rpoS gene was determined. As shown in Fig. 2, a TAG termination codon (instead of CAG codon for glutamine) was present at the 33rd position of the ORF of the rpoS gene in MA261 and HT283. We also determined the nucleotide sequence of the rpoS gene in two kinds of E. coli W3110 from our laboratory stock. One (type D) has a CAG codon for glutamine, but the other (type F, new type) has the termination codon TAG (Fig. 2). In the E. coli strain WC196, the 33rd amber codon of the rpoS gene is suppressed by the supD gene, which translates the UAG codon into serine (26). Together these results suggest that a high incidence of mutations are accumulated at the 33rd codon of the rpoS gene.
The three E. coli strains, MA261, HT283, W3110 (type F), carry the TAG amber codon at the 33rd position of the rpoS gene, suggesting that these strains carry a suppressor for UAG and that the efficiency of suppression is enhanced by polyamines. To test the UAG suppression of the mutant rpoS gene in MA261 (one of the three mutants) we constructed an MA261 derivative with rpoS disrupted by insertion of the tet gene, and we tested the expression of a plasmid-encoded 38 after transfection of various plasmids containing the rpoS gene with various mutations at the 33rd codon. In the presence of polyamine addition, the level of 38 subunit increased only for the rpoS mutant with a UAG amber termination codon at the 33rd position of RpoS mRNA (Fig. 3B, 33TAG), whereas the 38 level remained unaltered for the rpoS mutant with a UGA codon at the same position (Fig. 3B, 33TGA). The 38 subunit was not detected for the mutant rpoS gene with the UAA termination codon (Fig. 3B, 33TAA), suggesting that the MA261 strain does not carry the UAA suppressor.
The expression level of 38 was essentially the same in the presence and absence of polyamines for the rpoS gene with CAG (Gln) or UCG (Ser) codons at the 33rd position (Fig. 3B,  33CAG and 33TCG). Taken together we conclude that neither CAG-dependent Gln-tRNA and UCG-dependent Ser-tRNA binding to ribosomes nor readthrough of the UGA termination codon were influenced by polyamines.
One of the W3110 strains (type B) carries a UAG termination codon at the 270th position of the RpoS mRNA (25). We checked whether polyamines enhance the level of suppression at this position. As shown in Fig. 4, polyamines enhanced the synthesis of full-length 38

. Effect of polyamines on the level of tRNA Gln encoded by supE and glnV (A) and of tRNAs encoded by metT operon (B) and effect of suppressor tRNA Gln on polyamine stimulation of the synthesis of 38 subunit in polyamine-requiring mutants of E. coli (C and D).
A, total RNA was prepared from E. coli MA261 cells cultured with or without 100 g/ml putrescine (PUT) and harvested at A 540 ϭ 0.2. Polyacrylamide gel electrophoresis was performed using 8 g of total RNA. OH Ϫ , RNA was incubated at 37°C for 2 h at pH 9.5. B, Northern blotting of metT operon tRNAs of E. coli MA261 was performed using 5 g of total RNA. Hybridized RNA is processed tRNAs. C, E. coli MA261 containing pUC119 or pUCsupE was cultured with or without 100 g/ml putrescine (PUT) and harvested at A 540 ϭ 0.2 and 0.6. Western blotting of 38 subunit was performed using 10 g of cell lysate protein. D, E. coli DR112, DR112rpoS::tet/pMW33TAG containing pACYCsupE or DR112rpoS::tet/pMW33TAG containing pUCsupE was cultured with or without 100 g/ml PUT and harvested at A 540 ϭ 0.2. Western blotting of 38 subunit was performed using 10 g of cell lysate protein. plasmid was expressed in the DR112 rpoS::tet strain, and the level of full-length 38 subunit in DR112 was very low (Fig. 4C,  DR112), suggesting that the level or activity of suppressor tRNA for UAG is low or that the UAG suppressor tRNA does not exist in the strain DR112.
Four different genes encoding tRNA Gln , glnU, glnW, glnV, and glnX, exist in the metT operon (40). A commonly occurring suppressor tRNA for the amber termination codon in E. coli is encoded by the supE gene, which can be generated after mutation in either glnV or glnX of the metT operon. To identify the nature of the amber suppressor in each of the three polyaminerequiring mutant strains, we determined the nucleotide sequences of the tRNA Gln genes for all three mutants. The nucleotide sequences of the tRNA GlnU , tRNA GlnW , and tRNA GlnV genes were the same among the three polyamine-requiring mutants (data not shown). However, the tRNA GlnX gene was changed to the tRNA supE gene in both MA261 and HT283 but not in DR112 (Fig. 5). These results indicate that tRNA supE is involved in the stimulation of 38 subunit synthesis by polyamines.
Mechanism of Polyamine Stimulation of Readthrough of the Amber Termination Codon in RpoS mRNA-One possible mechanism of the enhancement of amber suppression by polyamines is an increase in the level of suppressor tRNA. To test this possibility, we first measured the level of tRNA supE in E. coli MA261. Total cellular RNA was subjected to Northern blot hybridization using a probe that hybridizes to both tRNA supE and tRNA GlnV . As shown in Fig. 6A, the combined level of tRNA supE and tRNA GlnV in E. coli MA261 was higher for the culture with polyamines than that without polyamines. This was confirmed by Northern blot analysis of the combined levels of all seven species of tRNA (see Fig. 5A) encoded by the metT operon (Fig. 6B). The level was found to be 1.8ϫ higher in the presence of polyamines than that found in its absence. Most of the tRNA supE and tRNA GlnV was aminoacylated in E. coli MA261 cultured with or without the addition of polyamines (Fig. 6A). These results suggest that polyamines stimulate transcription of the metT operon or stabilize the tRNAs. The effects of high-level tRNA supE on the polyamine stimulation of the synthesis of 38 subunit were then examined. As shown in Fig. 6C, the degree of stimulation of the synthesis of 38 subunit by polyamines became smaller in both early and middle logarithmic phases by transforming a high-copy number plasmid (pUC119) containing the gene for tRNA supE . This was caused by the increase in the level of 38 subunit in cells cultured without polyamines. When E. coli DR112 rpoS::tet was transformed with the gene for tRNA supE in middle-copy number plasmid (pACYC184) but not in high-copy number plasmid (pUC119), it was found that polyamines stimulated the synthesis of 38 subunit (Fig. 6D). As for the cells transformed with pUCsupE, the decrease in polyamine stimulation of the synthesis of 38 subunit was also caused by the increase in the level of 38 subunit in cells cultured without polyamines. The results indicate the necessity of limiting amounts of tRNA supE for polyamine stimulation of the synthesis of 38 subunit.
Next we analyzed possible effects of polyamines on Gln-tRNA formation and RpoS mRNA-dependent protein synthesis in a cell-free system. Detailed analysis was carried out using spermidine because it is more effective than putrescine (at least for protein synthesis in vitro) (5,13). The tRNA preparation used for aminoacylation in vitro was rich in tRNA supE because tRNA was isolated from E. coli MA261 containing pUCsupE. As shown in Fig. 7A, Gln-tRNA formation was not influenced by the addition of 1 mM spermidine. Similar results were obtained in the presence of 2.5 mM spermidine (data not shown).
Possible effects of polyamines on mutant RpoS(33UAG) mRNA-dependent synthesis in vitro of the 38 subunit were then analyzed by measuring the incorporation of [ 35 S]methionine into the TCA-insoluble fraction. As shown in Fig. 7B, 1 mM spermidine significantly stimulated the overall activity of protein synthesis. Because the protein synthesis activity in the in vitro translation system employed depends on the addition of externally added mRNA (data not shown), the result indicates the stimulation of 38 subunit synthesis by spermidine. To confirm this interpretation the protein products were analyzed by SDS-PAGE, and the gels were subjected to fluorography. Fig. 7C shows one of the fluorograms, which indicates that the band intensity of 38 subunit increased, albeit at low levels, in the presence of spermidine. Stimulation of poly(U)-dependent polyphenylalanine synthesis by polyamines is attributable to the increase in aminoacyl-tRNA binding to ribosomes but not at the levels of peptide bond formation and translocation (5). Thus, we first examined the effect of polyamines on mRNA-independent binding of Gln-tRNA supE to ribosomes. Although the binding activity was low, it was clearly stimulated by 1 mM spermidine (Fig. 7D), indicating that the affinity of Gln-tRNA supE to ribosomes is increased by polyamines. Then, the effect of polyamines on AUGUAG-dependent binding of Gln-tRNA supE was examined. Polyamines also enhanced Gln-tRNA supE binding to ribosomes (data not shown). These results, taken together, suggest that polyamines stimulate the synthesis of 38 subunit through at least two steps: (i) the increased level of tRNA supE and (ii) the increased binding of Gln-tRNA supE to ribosomes.
Physiological Significance of Polyamine Stimulation of Suppression of Amber Mutation in RpoS mRNA-The rpoS gene is essential for cell viability in the stationary phase (41). Thus, we compared cell viability between MA261 cells cultured with or without putrescine. The rate of cell growth was enhanced by A, E. coli MA261 (E and q) and MA261rpoS::tet (‚ and OE) cells were cultured in medium A supplemented with five amino acids in the presence (closed symbols) and absence (open symbols) of 100 g/ml putrescine, and cell growth was followed by measuring A 540 . B, viable cells were counted at designated times as described under "Experimental Procedures." Each value is the average of duplicate determinations. C, Western blotting of subunits was performed using 10 g of cell lysate protein for 38 or 5 g of cell lysate protein for 70 . putrescine as reported (17). When rpoS was disrupted, cell growth slowed slightly, but polyamine enhanced cell growth greatly (Fig. 8A). As for cell viability, determined by colony formation on a rich plate, it increased greatly when the cells were cultured in the presence of putrescine. However, cell viability of MA261 cultured in the absence of putrescine was higher than that of MA261 in which rpoS was disrupted (Fig.  8B). Cell viability was parallel with the level of 38 subunit (Fig. 8C). The results suggest that elevated expression of rpoS gene is important for cell viability. DISCUSSION Polyamines stimulate the synthesis of a set of proteins such as oligopeptide-binding protein (OppA) (17), adenylate cyclase (Cya) (19), and 38 subunit (RpoS) (this paper). Up to now two different mechanisms have been identified: a structural change of OppA mRNA, leading to enhanced template activity for translation (17), and stimulation of initiation codon UUG-dependent fMet-tRNA binding to Cya mRNA-ribosomes (19). The results herewith described show a novel mode of the polyamine stimulation of protein synthesis. Polyamines stimulate readthrough of the amber codon by enhancing the binding of amber codon UAG-dependent Gln-tRNA supE on ribosome-associated RpoS mRNA (see Fig. 9). Thus polyamines modulate protein synthesis not only at the level of initiation but also at the level of elongation of translation. We propose that genes whose expression is modulated by polyamines at the level of translation are referred to as "polyamine modulon." Experi-ments are in progress to find other members of the polyamine modulon.
The increase in readthrough of the amber codon by polyamines has been found for translation of the mutant gene 1 protein mRNA from T7 phage (42). Stimulation of readthrough of both UAG amber and UGA opal codons by polyamines also has been reported in E. coli and eukaryotic cell-free systems (43)(44)(45). In these cases, however, the mechanism was not studied in detail. As for polyamine stimulation of the synthesis of 38 subunit, the translation suppression was found to be caused by stimulation of Gln-tRNA supE binding to ribosomes via two processes, i.e. an increase in the level of tRNA supE and an increase in the binding affinity of Gln-tRNA supE to ribosomes. As to the increase in tRNA supE level, it is of interest to know whether polyamines stimulate transcription of the metT operon or stabilize tRNA supE . If the latter is the case, the increase in both intracellular level and intrinsic function of tRNA supE can be explained by a structural change of tRNA supE by polyamines.
Readthrough in vivo of the UGA opal codon at the 33rd codon of RpoS mRNA was not stimulated by polyamines in cells under our experimental conditions (see Fig. 3). The frequency of the use of UAG, UGA, and UAA as the termination codon in E. coli is 7.6, 29.3, and 63.1%, respectively (46). Furthermore, 316 among 4288 genes in E. coli K12 use tandem termination codons (47). Because not so many genes use UAG as the real termination codon at the end of full-length reading frames, the FIG. 9. Polyamine effects on three kinds of protein syntheses directed by OppA mRNA, Cya mRNA, and RpoS mRNA. A, polyamines cause a structural change of the SD sequence and the initiation codon AUG of OppA mRNA, facilitating formation of the initiation complex. B, polyamines stimulate interaction between the initiation codon UUG of Cya mRNA and the anticodon CAU of fMet-tRNA i . In this case, exposure of the SD sequence of the mRNA is probably a prerequisite for the stimulation. C, polyamines stimulate the readthrough of amber codon of RpoS mRNA probably through its stimulation of Gln-tRNA supE binding to ribosomes. RF1, release factor 1.
gene expression as a whole may not be influenced strongly by polyamines even if polyamines stimulate the readthrough of UAG at the natural termination sites.
We found that a mutation at the 33rd position of the ORF of RpoS mRNA occurs frequently. In such strains, polyamines enhance cell viability. The results confirmed that the 38 subunit is important for cell viability (41) and indicate that polyamines play an important role in cell viability of E. coli cells having an amber codon at the 33rd position of the ORF of RpoS mRNA.