Polyamine stimulation of antizyme frameshifting

Translation normally occurs in a linear fashion reading nonoverlapping nucleotide triplets. Alternatively, some genes require translation in two overlapping reading frames. The ability of the ribosome to change to an alternative reading frame is called frameshifting. Frameshifting occurs at a specific site in the mRNA sequence, which usually allows the tRNA(s) bound to the ribosome to repair in the new reading frame. Frameshifting can be stimulated by both <italic> cis

POLYAMINE STIMULATION OF ANTIZYME FRAMESHIFTING by Lorin Marie Petros A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Human Genetics The University of Utah May 2006 Copyright © Lorin Marie Petros 2006 All Rig hts Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL SUPERVISORY COMMITTEE APPROVAL of a dissertation submitted by Lorin Marie Petros This dissertation has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory. /- 4- Ob !OhnAtkins ~~~~ &£~~ Robert Weiss THE UNIVERSITY OF UTAH GRADUATE SCHOOL FINAL READING APPROVAL To the Graduate Council of the University of Utah: I have read the dissertation Lorin Marie Petros in its final form and have found that (1) its fonnat, citations, and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the supervisory committee and is ready for submission to The Graduate School. Date I I Approved for the Major Department Mark F. Leppert, Ph.D. Co-Chair, Dept. of Human Genetics Approved for the Graduate Council David S. Chap Dean of The Graduate School ABSTRACT Translation normally occurs in a linear fashion reading nonoverlapping nucleotide triplets. Alternatively, some genes require translation in two overlapping reading frames. The ability of the ribosome to change to an alternative reading frame is called frameshifting. Frameshifting occurs at a specific site in the mRNA sequence, which usually allows the tRNA(s) bound to the ribosome to repair in the new reading frame. Frameshifting can be stimulated by both cis- and trans-acting elements. Frameshift stimulatory elements are thought to function by either pausing the ribosome at the frameshift site, decreasing the normal accuracy of the translating ribosome, or by aiding in the repositioning of the message into the new frame. A +1 frameshift is required for synthesis of functional antizyme protein. Antizyme frameshifting occurs at the site UCC UGA. The termination codon of the shift site is thought to pause the ribosome. Pausing at the shift site is likely aided by a 3' pseudoknot structure. The pseudoknot likely stimulates frameshifting by additional unknown mechanisms beyond the ribosomal pause. A 50 nucleotide sequence 5' of the shift site on antizyme mRNA is also stimulatory, though the mechanism is unclear. Finally, antizyme frameshifting is stimulated by polyamines. Antizyme is a key negative regulator of polyamine biosynthesis, and therefore the frameshift event acts as a sensor and regulator of cellular polyamine levels. Very little is known about how polyamines stimulate antizyme frameshifting. Analysis presented here demonstrates that polyamine stimulation of frameshifting requires a termination codon as part of the shift site. Polyamines appear to interfere with the termination process increasing the likelihood of the frameshift event. High levels of polyamine stimulation require the 5' element. In contrast, the pseudoknot stimulates frameshifting in a polyamine independent manner. The assay developed to studying polyarrline stimulation of antizyme frameshifting has many potential uses. It can be used to screen libraries of small molecules for their ability to stimulate frameshifting. A diverse array of polyamine analogs was shown to stimulate the antizyme frameshifting. The assay may also be useful in discovering a tmRNA-like activity, which targets aberrant proteins for degradation in eukaryotes. v TABLE OF CONTENTS ABSTRACT .......................................................................................... iv LIST OF TABLES ................................................................................. viii ACKNOWLEDGMENTS ......................................................................... ix Chapter 1. INTRODUCTION ............................................................................... 1 Translational recoding ................................................................................. 1 System for studying recoding in n1ammals .......................................... 9 Polyamine regulation ................................................................................. 12 References ................................................................................. 18 2. POLYAMINE SENSING DURING ANTIZYME mRNA PROGRAMMED FRAMESHIFTING ........................................................................... 26 Abstract .................................................................................... 27 Materials and methods ................................................................. 28 Results ..................................................................................... 29 Discussion .............. : .................................................................. 34 Acknowledgments ....................................................................... 36 References ................................................................................ 36 Supplementary n1aterial. ........................................................................... 39 3. SCREENING METHODS TO IDENTIFY POLYAMINE ANALOGS WHICH INDUCE ANTIZYME FRAMESHIFTING, DEPLETE NATURAL POLYAMINES, AND INHIBIT CELL GROWTH ...................................... 40 Abstract ..................................................................................... 40 Introduction ................................................................................ 41 Results ...................................................................................... 46 Discussion ................................................................................. 60 Material and methods ................................................................... 65 References ...................... a ••••••••••••••••••••••••••••••••••• , •••••••••••••••••••••• 68 Supplementary material ......... a •••••••••••••••••••••••••••••••••••••••••••••••••••••• 74 4. A CAUTIONARY TALE FOR USING THE DUAL LUCIFERASE SYSTEM TO MONITOR +1 FRAMESHIFTING ...................................... 81 Abstract ..................................................................................... 81 Introduction ................................................................................ 82 Results ...................................................................................... 87 Discussion ............................................................................... 1 00 Material and methods ................................................................. 107 References ............................................................................... 110 APPENDIX: SEQUENCE SPECIFICITY OF AMINOGL YCOSIDE-INDUCED STOP CODON READTHROUGH: POTENTIAL IMPLICATIONS FOR TREATMENT OF DUCHENNE MUSCULAR DYSTROPHY .................... 115 vii LIST OF TABLES Table 2.1 The 5' element effect on the competition between frameshifting, readthrough, and termination in the antizyme 2 context ....................... 35 2.2 The sequence of the top strand oligonucleotide for the test constructs ... 39 3.1 Structure, frameshifting, and growth rescue of the natural polyamines and select analogs ...................................................................... 48 3.2 Growth inhibition of HEK293 cells with MQTPA compounds and the natural polyamines ...................................................................... 55 3.3 Polyan1ine reduction after treatment with MQTPA compounds .............. 57 3.4 Polyamine metabolic enzyme activities in HEK293 cells ...................... 59 3.5 Characterization and comparison of the lead MQTPA compounds and spermidine (SPO) ........................................................................ 62 4.1 The sequence of the top strand oligonucleotides for the +1 constructs .. 1 08 ACKNOWLEDGMENTS I wish to thank Robert Schackmann for the synthesis of the oligonucleotides used in the creation of the constructs tested in the studies contained within this dissertation. CHAPTER 1 INTRODUCTION Translational recoding The synthesis of proteins by translation of mRNA is an essential process of all living organisms. Decoding of the mRNA into a chain of amino acids occurs by a series of steps requiring a complex structure of RNA and proteins called the ribosome (reviewed in Kapp and Lorsch, 2004; Liljas, 2004). The four main steps of translation are initiation, elongation, termination, and recycling (Figure 1.1). For simplicity, the processes described are for eukaryotic systems unless otherwise specified. The 43S complex, which is made up of the 40S subunit of the ribosome, an initiator tRNA, and other initiation factors, binds to the mRNA to begin initiation. The 43S complex scans the mRNA in a 5' to 3' direction until an initiation codon is reached (AUG), at which point the large (60S) ribosomal subunit joins the already bound small (40S) subunit. Elongation begins with a peptidyl tRNA (formyl methionyl tRNA for the initial codon) in the ribosomal P site and a vacant A site. A corresponding aminoacyl tRNA then binds to the three nucleotide codon in the A site. The ribosomal peptidyl transferase center catalyzes the formation of a peptide bond between amino acids on the tRNAs in the A and P sites. The P and A site tRNAs translocate to the E and P sites, and 2 B c ....... • ... ~ ---- Figure 1.1. A model of eukaryotic translation. (A) Initiation. (8) Elongation. (C) Termination and recycling. 3 the next three nucleotides of the mRNA translocate into the A site. Elongation is repeated until a stop codon (UAA, UAG, or UGA) occupies the A site. Eukaryotic release factor 1 (eRF1) recognizes all three stop codons and promotes the hydrolysis of peptidyl tRNA. Along with eRF3, eRF1 terminates translation and releases the completed polypeptide chain. Finally, the ribosomal subunits are recycled for use in another round of initiation. Translation is normally a very accurate process. In bacteria, processivity errors (failure to translate a full length protein) are estimated to occur about once every 30,000 codons with frameshifting errors occurring less than once every 300,000 codons (Kurland, 1992). While accurate translation is important, there are many exceptions to the "rules" of standard translation. The reprogramming of the genetic information by specific signals in the rrlRNA is called "recoding" (Gesteland et aI., 1992). The frequency of recoding at a specific site is much greater than the error frequency. The three main classes of recoding are frameshifting, redefinition, and bypassing (reviewed in Gesteland and Atkins, 1996; Ivanov et aI., 2003; Namy et aI., 2004). The most unusual class of recoding is translational bypassing (reviewed in Herr et aI., 2000). During bypassing the ribosome skips over non-coding nucleotides and then resumes standard decoding some distance downstream. Bypassing differs from frameshifting in that bypassing is frame independent. Bypassing is a three stage process. First, the ribosomes "takes-off' from the mRNA by the dissociation of the peptidyl tRNA from the mRNA, and the mRNA begins to move. The next stage is "scanning." Scanning involves the movement 4 of the ribosome relative to the mRNA. In the final stage, the tRNA "lands" and repairs with the mRNA at a matching codon to the take-off site. A 50 nucleotide translational bypass occurs during expression of the E. coli bacteriophage T 4 gene 60 (Huang et aI., 1988; Weiss et aI., 1990b). Four signals are important for this high level of bypassing: the nascent peptide, a stop codon following the take­off site, matching take-off and landing sites (Weiss et aI., 1990b), and a stem loop containing a pseudo Shine-Oelgarno (SO) sequence (FM Adamski, B Moore, RF Gesteland, JF Atkins, unpublished data). Another category of translational recoding is redefinition. Redefinition is changing the meaning of a codon in an mRNA specific manner. All known cases of redefinition specify an amino acid for a termination codon, though it is possible that an alternative amino acid could decoded a sense codon. Redefinition involves the lIreadthrough" of termination codons by a near cognate tRNA or the insertion of the 21 st amino acid, selenocysteine. Termination efficiency is determined by the nucleotide 3' of the stop codon, the last two amino acids incorporated, and stimulatory elements (reviewed in Bertram et aI., 2001). These factors effect the competition between readthrough and termination. Murine Leukemia Virus (MuLV) is an example of a retrovirus that uses readthrough of a termination codon to express the GagPol precursor protein. A glutamine is inserted for the stop codon at the end of the Gag ORF1 (Yoshinaka et aI., 1985). The efficiency of readthrough is about 50/0 and requires an RNA pseudoknot structure 3' of the stop codon (ten Dam et aI., 1990; Wills et aI., 1991; Feng et aI., 1992). The product of the pol gene interacts with eRF1, and this interaction increases readthrough at the Gag-Pol junction (Orlova et aI., 2003). 5 A unique class of redefinition is specification of selenocysteine by "special" UGA codons (reviewed in Hatfield and Gladyshev, 2002; Driscoll and Copeland, 2003; reviewed in Berry, 2005). A tRNA with a UCA anticodon that is charged with a selenocysteine and its specific elongation factor are required. In addition, there is a distant signal in eukaryotic mRNAs that encode selenocysteine called the selenocysteine insertion sequence (SECIS) element. The SECIS-binding protein 2 (SBP2) is also required for selenocysteine insertion into proteins. The third class of recoding is frameshifting. Translational frameshifting is the ability of the ribosome to efficiently change to one of the two alternative reading frames. Translation elongation then continues normally in the new frame. A single nucleotide shift in the 3' (forward) direction is designated +1 frameshifting, and a single nucleotide shift in the 5' (backward) direction is called -1 frameshifting. Frameshifts of greater than a single nucleotide can occur (Weiss et aI., 1987; Weiss et aI., 1990a). In a natural context, -2 frameshifting appears to be a rare event, which does occur at least in the major tail protein genes of phage Mu (Xu et aI., 2004). Frameshifting in the -1 direction is more con1mon in viral genes, insertion sequences, and transposable elements than in stable cellular genes. -1 frameshifting often occurs by a tandem shift (Jacks et aI., 1988) at heptanucleotide sequences of the general form X XXV VYZ, where spaces represent zero frame codons, X is A, G, or U, Y is A or U, and Z is A, C, or U. 6 With tRNAs bound at the shift site in the A and P sites of the ribosome, the mRNA slips backwards by a single base. The nature of the shift site allows repairing of the tRNA to at least two of the three nucleotides at the new -1 frame codons. In some cases, -1 frameshifting occurs at hexanucleotide sequences with only the A site tRNA repairing. The P site tRNA either dissociates without repairing or detaches only from the third base of the codon (Licznar et aI., 2003). -1 frameshifting is thought to occur after the tRNA has entered the A site, but before peptidyl transfer (Harger et aI., 2002). An example of the tandem shift model of -1 frameshift can be found during translation of the genomic RNA of the infectious bronchitis virus (lBV, Brierley et aI., 1987). Frameshifting occurs at the sequence U UUA Me (Brierley et aI., 1992) and is stimulated by a 3' pseudoknot structure (Brierley et aI., 1989; Inglis et aI., 1990). The stimulation of frameshifting by 3' pseudoknots is a common feature of -1 programmed frameshifting (reviewed in Giedroc et aI., 2000), but the mechanism is not fully understood. Pseudoknots slow or pause ribosomes at the frameshift site, but pausing alone is not enough to stimulate frameshifting (Tu et a\., 1992; Somogyi et aI., 1993; Kontos et aI., 2001). +1 frameshifting is thought to occur with the frameshift site paused in the P site of the ribosome. Unlike -1 frameshifting, the shift takes place with a single tRNA bound in the P site. In addition, to pausing by pseudoknot structures, +1 frameshifting often occurs with a slow to decode codon in the A site. The A site codon is often a stop codon (Weiss et aI., 1987; Weiss et aI., 1990a). Alternatively, the A site codon can require a rare tRNA (Belcourt and Farabaugh, 7 1990; Sundararajan et aI., 1999). The mRNA can also be paused by 5' sequences (Larsen et aI., 1995). One model (Model 1) for +1 frameshifting proposes that a wobble pair between the P site tRNA and its codon disrupts ribosomal interactions, distorting the mRNA, and moving the +1 codon into the A site for decoding (Stahl et aI., 2002). Alternatively (Model 2), the P site tRNA is proposed to reposition into the +1 frame, similarly to -1 frameshifting. Once repositioned standard translation continues in the +1 frame (Figure 1.2, Baranov et a\., 2004) A well studied example of +1 translational frameshifting is that required for expression of E. coli release factor 2 (RF2, Craigen et aI., 1985). At the end of the first open reading frame (ORF1) of this gene, tRNALeu shifts from the zero frame CUU codon to a UUU codon in the +1 frame. The CUU codon is followed in the zero frame by the termination codon UGA. The frameshift event is favored by similar anticodon:codon interactions between the tRNA Leu and the mRNA in both the zero and +1 frames, by a poor termination context of the stop codon (UGA C), and by a SO sequence (Weiss et aI., 1987; Curran and Varus, 1988). The SO sequence, three nucleotides upstream from the frameshift site, pairs with the anti-SO sequence of the 16S rRNA (Weiss et aI., 1988). This pairing is thought to promote mRNA realignment in the +1 direction. The level of frameshifting is dependent on RF2 abundance (Adamski et aI., 1993). RF2 recognizes and terminates translation at UAA and UGA. When RF2 levels are high, termination at the end of ORF1 is favored, which lowers RF2 levels in the cell. With low RF2 levels, frameshifting is favored over termination, producing 8 5' ..... - ............ 3' / 3' ~ 3' Figure 1.2. Models of +1 frameshifting. A-G represents nucleotides in the mRNA. In Model 1 the P site tRNA has a wobble pair with nucleotide C, which distorts the mRNA to bring the EFG codon into the A site. In Model 2 the P site tRNA slips from the ABC codon to the BCD codon. This slippage moves the EFG codon into the A site. functional RF2 protein. The recoding event required for RF2 expression thus creates an autoregulatory mechanism for controlling the abundance of RF2. 9 In eukaryotes, another +1 translational frameshift event is used as an autoregulatory mechanism for controlling a protein's abundance. +1 frameshifting at the site UCC UGA is required for expression of the antizyme gene (Matsufuji et aI., 1995). Antizyme is a negative regulator of cellular polyamine levels, and polyamines stimulate the frameshift event. The mechanism behind antizyme frameshifting will be discussed below in more detail. System for studying recoding in mammals In order to study recoding events in higher eukaryotes the dual luciferase assay system was developed (Grentzmann et aI., 1998). The vector used in this system is p21uc (Figure 1.3). p21uc contains a Renilla luciferase gene (rluc) and a firefly luciferase gene (fluc). Fluc lacks an initiation codon, and its expression requires the translation of the upstream Rluc gene. Between the two genes is a polylinker region in which a sequence being studied for recoding efficiency can be inserted. Inserts are added to the p21uc vector so that synthesis of the firefly luciferase protein requires a recoding event. Translation is initiated at rluc. If the recoding event occurs, a Renilla luciferase-firefly luciferase fusion protein is synthesized. In the absence of recoding, only Renilla luciferase is translated. The Renilla and firefly luminescent activities are measurable, and the ratio of the firefly activity to the Renilla activity represents the percent of ribosomes which undergo the recoding event of interest. A corresponding insert is also cloned into SV40 early enhancer/promoter SV40 late poly (A) 10 Figure 1.3. The p21uc vector for the dualluciferase assay system. This figure was modified from Grentzmann et al. (1998). 11 the p21uc vector such that the Renilla luciferase-firefly luciferase fusion protein is synthesized by standard translation. This construct is called the in-frame control. The ratio of firefly to Renilla luciferase activity in the recoding construct is normalized to the ratio of firefly to Renilla luciferase activity in the in-frame control. The dualluciferase assay system has many advantages over other systems for the study of eukaryotic translational recoding. Luciferase assays are rapid and can be modified for high throughput approaches. The assays are also highly sensitive, allowing low levels of recoding to be measured. The two luciferase proteins are quantified in the same way (by the emission of light), but require different substrates. This allows the two assays to be performed in a single tube, reducing human errors. The p21uc vector contains features for transcription and translation to allow its use in both in vitro and in vivo man1malian systems. The original vector has also been modified for use in yeast (Harger and Dinman, 2003). The use of the same reporter genes in multiple translation systems allows for accurate comparison of recoding events in different organisms. Finally, translation of the recoded (fusion) protein and the control (Renilla luciferase alone) from a single transcript regulates mRNA levels, translation initiation, and transfection efficiencies. Controlling these factors allows for an accurate measurement of the effect of trans-acting factors on recoding, an important feature when studying polyamine stimulation. 12 Polyamine regulation Polyamines are ubiquitous organic cations with distributed charges (see Figure 1.4 for structures). Polyamines are involved in many cellular functions and are required for cell growth (reviewed in Childs et aI., 2003). Through their positive charges, polyamines can interact with DNA, RNA, and protein, and therefore affect a wide array of cellular processes including chromatin condensation, DNA replication, transcription, translation, RNA processing, ion transport, and apoptosis. Polyamines also playa role in embryonic development and have been shown to be essential at certain stages of larval growth (reviewed in Thomas and Thomas, 2001). Finally, polyamines are required for normal progression through the cell cycle (reviewed in Oredsson, 2003). While polyamines are essential for cell growth and cell cycle progression, increased levels have been correlated with the abnormal, rapid growth of cancer cells (reviewed in Pegg, 1988; reviewed in Cohen, 1998). It is not surprising then that polyamine metabolism through inhibitors of their biosynthesis and catabolism has often been a target for anti-cancer drugs (reviewed in Marton and Pegg, 1995; Thomas and Thomas, 2003; Wallace et aI., 2003). Furthermore, polyamine metabolism has been targeted as a strategy in cancer chemoprevention (Meyskens and Gerner, 1999; Gerner and Meyskens, 2004). Another strategy to fight cancer through polyamine reduction is by targeting polyamine regulation. Many polyamine analogs have been developed which deplete the intracellular polyamines and/or act as mimics of the natural polyamines by displacing them 13 A ~ ,N~N,H H I H B H I H'N~N~N .... I I H H H C H H H I I \N~N~N~N""H I I H H Figure 1.4. Structures of the natural polyamines. (A) Putrescine. (8) Spermidine. (C) Spermine. 14 from binding sites, without substituting for their functions in cell growth (reviewed in Huang et aI., 2005). Polyamine metabolism involves a complex series of enzymatic reactions (Figure 1.5). The three polyamines, putrescine, spermidine, and spermine, are synthesized from arginine and methionine. Arginase converts arginine into ornithine. Ornithine is then decarboxylated to form putrescine by ornithine decarboxylase (ODC). Methionine adenosyltransferase (MAT) coverts nlethionine to S-adenosylmethionine (SAM), which is decarboxylated by S­adenosylmethionine decarboxylase (SAM DC) to form decarboxylated S­adenosylmethionine (DC-SAM). DC-SAM combines with putrescine to form spermidine via spermidine synthase. A second aminopropyltransfer with DC­SAM and spermidine forms spermine through spermine synthase. Spermidine/spermine N1-acetyltransferase (SSAT) and polyamine oxidase (PAO) are involved in the back conversion of spermine to spermidine and spermidine to putrescine through acetylated intermediates (80lkenius and Seiler, 1981). Spermine can be directly converted back to spermidine by spermine oxidase (SMO, Vujcic et aI., 2002). Polyamines regulate many of the steps in their own biosynthesis and catabolism. ODC activity increases when polyamines are low and decreases when levels are high. Regulation of ODC is a mix of posttranscriptional regulation and protein degradation through the ODC inhibitor antizyme (discussed below). The activity of SAMDC is stimulated by putrescine (Pegg and Williams-Ashman, 1968; Williams-Ashman et aI., 1972) and inhibited by methionine tMAT (I) SAM s.(;MDC~ C02 S' DC-SAM SPERMIDINE SYNTHASE Methylthioadenosine Methylthioadenosine arginine t ARGINA~.E ornithine<. O~ loot I ANTIZVME !'-+C02 putrescine ~F ,AO Nt acetylspermidine SSAT~O Nt acetylspermine J tSAT~~ spermine C-/S ~ Figure 1.5. The pathway of polyamine metabolism. PA = polyamines, PUT = putrescine, SPD = spermidine, and SPM = spermine. Double arrowheads indicate stimulation, and a flat line as an arrowhead indicates inhibition. 15 16 spermidine and spermine (Mamont et aI., 1978; Alhonen-Hongisto, 1980; Mamont et aI., 1981). Polyamines positively regulate SSAT, which is superinduced by some polyamine analogs (Pegg et aI., 1990; Porter et aI., 1991). A polyamine response element (PRE) has been identified in the human SSAT gene (Wang et al., 1998). A key negative regulator of intracellular polyamine levels is the protein antizyme. Antizyme inhibits ODC by binding it and targeting it for degradation by the 26S proteasome (Murakami et aI., 1992; Li and Coffino, 1993; Zhang et aI., 2003). The protein was named antizyme because it is the "anti-enzyme" for ODC (Heller et aI., 1976). In addition, antizyme inhibits polyamine import (Mitchell et aI., 1994; Suzuki et aI., 1994; Sakata et aI., 1997) and stimulates polyamine export (Sakata et aI., 2000). Antizyme inhibitor (AZI) binds to and sequesters antizyme, preventing antizyme's function in polyamine homeostasis (Bercovich and Kahana, 2004). Antizyme also interacts with cyclin D (Newman et aI., 2004), and Drosophila antizyme is involved in the regulation of cyclin B (Vied et aI., 2003), suggesting a direct role for antizyme in cell cycle regulation. Antizyme expression requires translational +1 frameshifting, which is stimulated by polyamines, and thus antizyme serves as a sensor and regulator of polyamine levels (Miyazaki et aI., 1992; Matsufuji et aI., 1995; Ivanov et aI., 2000b). When polyamine levels are low, standard translation with termination at the end ofORF1 is efficient. When polyamine levels are high, a considerable proportion of the ribosomes shift into the +1 frame to translate ORF2, which is required to produce functional antizyme. Frameshifting occurs at the end of 17 ORF1 at the site UCC UGA. mRNA sequence 5' of the shift site (the 5' element) and an RNA pseudoknot 3' of the shift site, along with the ORF1 stop codon, are important stimulators of the frameshift event (Matsufuji et aI., 1995; Ivanov et aI., 2000a). A three nucleotide spacing between the shift site and the pseudoknots is much shorter than the six to nine nucleotide spacing found in -1 frameshifting. In humans three antizymes exist. Antizyme 1 (Miyazaki et aI., 1992) and antizyme 2 (Ivanov et aI., 1998; Zhu et aI., 1999) are expressed in most tissues. Antizyme 3 (Ivanov et aI., 2000b; Tosaka et aI., 2000) is expressed only in germ cells and lacks the 5' and 3' stimulatory elements. How polyamines stimulate frameshifting is a process that is poorly understood. The following is a study of polyamine stimulation of antizyme frame­shifting. Chapter 2 describes the requirements for polyamine stimulation of +1 frameshifting. A termination codon as part of the shift site and the antizyme 5' element were demonstrated to be important for polyamine stimulation. Polyamines also stimulated readthrough of stop codons. Chapter 3 uses the dual luciferase assay for studying the polyamine stimulation of the antizyme frameshift cassette of Chapter 2 to screen a polyamine analog library for potential anticancer drugs. This screen along with another assay identified polyamine analogs that mimic natural polyamines in their ability to enter the cell through the polyamine transporter and stimulate +1 frameshifting in the antizyn1e gene. These analogs do not substitute for the natural polyamines in growth functions and cannot rescue cells from DFMO (an inhibitor of ODC) treatment. Chapter 4 deals with the limitations of studying +1 frameshifting using the dualluciferase assay system. The Appendix presents a study of another class of small molecules, aminoglycosides, and their effect on translational readthrough. As was demonstrated in Chapter 2 with the polyamines, the ability of aminoglycosides to stimulate readthrough depends on the context of the termination codon. References Adamski, F.M., Donly, B.C. and Tate, W.P. (1993) Competition between frameshifting, termination and suppression at the frameshift site in the Escherichia coli release factor-2 mRNA. Nucleic Acids Res, 21, 5074- 5078. 18 Alhonen-Hongisto, L. (1980) Regulation of S-adenosylmethionine decarboxylase by polyamines in Ehrlich ascites-carcinoma cells grown in culture. Biochem J, 190,747-754. Baranov, P.V., Gesteland, R.F. and Atkins, J.F. (2004) P-site tRNA is a crucial initiator of ribosomal frameshifting. RNA, 10, 221-230. Belcourt, M.F. and Farabaugh, P.J. (1990) Ribosomal frameshifting in the yeast retrotransposon Ty: tRNAs induce slippage on a 7 nucleotide minimal site. Cell, 62, 339-352. Bercovich, Z. and Kahana, C. (2004) Degradation of antizyme inhibitor, an ornithine decarboxylase homologous protein, is ubiquitin-dependent and is inhibited by antizyme. J BioI Chem, 279, 54097-54102. Berry, M.J. (2005) Knowing when not to stop. Nat 8truct Mol BioI, 12, 389-390. Bertram, G., Innes, S., Minella, 0., Richardson, J. and Stansfield, I. (2001) Endless possibilities: translation termination and stop codon recognition. Microbiology, 147, 255-269. Bolkenius, F.N. and Seiler, N. (1981) Acetylderivatives as intermediates in polyamine catabolism. Int J Biochem, 13,287-292. Brierley, I., Boursnell, M.E., Binns, M.M., Bilimoria, B., Blok, V.C., Brown, T.D. and Inglis, S.C. (1987) An efficient ribosomal frame-shifting signal in the polymerase-encoding region of the coronavirus IBV. Embo J, 6, 3779- 3785. Brierley, I., Digard, P. and Inglis, S.C. (1989) Characterization of an efficient coronavirus ribosomal frameshifting signal: requirement for an RNA pseudoknot. Cell, 57, 537-547. 19 Brierley, I., Jenner, A.J. and Inglis, S.C. (1992) Mutational analysis of the "slippery-sequence" component of a coronavirus ribosomal frameshifting signal. J Mol Bioi, 227,463-479. Childs, A.C., Mehta, D.J. and Gerner, E.W. (2003) Polyamine-dependent gene expression. Cell Mol Life Sci, 60, 1394-1406. Cohen, S.S. (1998) A Guide to the Polyamines. Oxford University Press, New York. Craigen, W.J., Cook, R.G., Tate, W.P. and Caskey, C.T. (1985) Bacterial peptide chain release factors: conserved primary structure and possible frameshift regulation of release factor 2. Proc Natl Acad Sci USA, 82,3616-3620. Curran, J.F. and Yarus, M. (1988) Use of tRNA suppressors to probe regulation of Escherichia coli release factor 2. J Mol Bioi, 203, 75-83. Driscoll, D.M. and Copeland, P.R. (2003) Mechanism and regulation of selenoprotein synthesis. Annu Rev Nutr, 23, 17-40. Feng, Y.X., Yuan, H., Rein, A. and Levin, J.G. (1992) Bipartite signal for read­through suppression in murine leukemia virus mRNA: an eight-nucleotide purine-rich sequence immediately downstream of the gag termination codon followed by an RNA pseudoknot. J Virol, 66, 5127-5132. Gerner, E.W. and Meyskens, F.L., Jr. (2004) Polyamines and cancer: old molecules, new understanding. Nat Rev Cancer, 4,781-792. Gesteland, R.F. and Atkins, J.F. (1996) Recoding: dynamic reprogramming of translation. Annu Rev Biochem, 65, 741-768. Gesteland, R.F., Weiss, R.B. and Atkins, J.F. (1992) Recoding: reprogrammed genetic decod ing. Science, 257, 1640-1641. Giedroc, D.P., Theimer, C.A. and Nixon, P.L. (2000) Structure, stability and function of RNA pseudoknots involved in stimulating ribosomal frameshifting. J Mol Bioi, 298, 167-185. Grentzmann, G., Ingram, J.A., Kelly, P.J., Gesteland, R.F. and Atkins, J.F. (1998) A dual-Iuciferase reporter system for studying recoding signals. RNA, 4, 479-486. 20 Harger, J.W. and Dinman, J.D. (2003) An in vivo dual-Iuciferase assay system for studying translational recoding in the yeast Saccharomyces cerevisiae. RNA, 9,1019-1024. Harger, J.W., Meskauskas, A. and Dinman, J.D. (2002) An "integrated model" of programmed ribosomal frameshifting. Trends Biochem Sci, 27,448-454. Hatfield, D.L. and Gladyshev, V.N. (2002) How selenium has altered our understanding of the genetic code. Mol Cell BioI, 22, 3565-3576. Heller, J.S., Fong, W.F. and Canellakis, E.S. (1976) Induction of a protein inhibitor to ornithine decarboxylase by the end products of its reaction. Proc Natl Acad Sci USA, 73, 1858-1862. Herr, A.J., Atkins, J.F. and Gesteland, R.F. (2000) Coupling of open reading frames by translational bypassing. Annu Rev Biochem, 69, 343-372. Huang, W.M., Ao, S.Z., Casjens, S., Orlandi, R., Zeikus, R., Weiss, R., Winge, D. and Fang, M. (1988) A persistent untranslated sequence within bacteriophage T4 DNA topoisomerase gene 60. Science, 239, 1005-1012. Huang, Y., Pledgie, A., Casero, R.A., Jr. and Davidson, N.E. (2005) Molecular mechanisms of polyamine analogs in cancer cells. Anticancer Drugs, 16, 229-241. Inglis, S.C., Rolley, N. and Brierley, I. (1990) A ribosomal frameshift signal in the polymerase-encoding region of the IBV genome. Adv Exp Med BioI, 276, 269-273. Ivanov, I.P., Gesteland, R.F. and Atkins, J.F. (1998) A second mammalian antizyme: conservation of programmed ribosomal fran1eshifting. Genomics, 52, 119-129. Ivanov, I.P., Gesteland, R.F. and Atkins, J.F. (2000a) Antizyme expression: a subversion of triplet decoding, which is remarkably conserved by evolution, is a sensor for an autoregulatory circuit. Nucleic Acids Res, 28, 3185-3196. Ivanov, I.P., Gurvich, 0.1., Gesteland, R.F. and Atkins, J.F. (2003) Recoding: site- or mRNA-specific alteration of genetic readout utilized for gene expression. In Lapointe, J. and Brakier-Gingras, L. (eds.), Translation Mechanisms. Eurekah.com/Landes Bioscience, Georgetown, TX, pp. 354- 369. Ivanov, I.P., Rohrwasser, A., Terreros, D.A., Gesteland, R.F. and Atkins, J.F. (2000b) Discovery of a spermatogenesis stage-specific ornithine decarboxylase antizyme: antizyme 3. Proc Natl Acad Sci USA, 97, 4808- 4813. Jacks, T., Madhani, H.D., Masiarz, F.R. and Varmus, H.E. (1988) Signals for ribosomal frameshifting in the Rous sarcoma virus gag-pol region. Cell, 55,447-458. Kapp, L.D. and Lorsch, J.R. (2004) The molecular mechanics of eukaryotic translation. Annu Rev Biochem, 73, 657-704. 21 Kontos, H., Napthine, S. and Brierley, I. (2001) Ribosomal pausing at a frameshifter RNA pseudo knot is sensitive to reading phase but shows little correlation with frameshift efficiency. Mol Cell BioI, 21, 8657-8670. Kurland, C.G. (1992) Translational accuracy and the 'fitness of bacteria. Annu Rev Genet, 26, 29-50. Larsen, B., Peden, J., Matsufuji, S., Matsufuji, T., Brady, K., Maldonado, R., Wills, N.M., Fayet, 0., Atkins, J.F. and Gesteland, R.F. (1995) Upstream stimulators for recoding. Biochem Cell BioI, 73, 1123-1129. Li, X. and Coffino, P. (1993) Degradation of ornithine decarboxylase: exposure of the C-terminal target by a polyamine-inducible inhibitory protein. Mol Cell BioI, 13, 2377-2383. Licznar, P., Mejlhede, N., Prere, M.F., Wills, N., Gesteland, R.F., Atkins, J.F. and Fayet, O. (2003) Programmed translational-1 frameshifting on hexanucleotide motifs and the wobble properties of tRNAs. Embo J, 22, 4770-4778. Liljas, A. (2004) Structural Aspects of Protein Synthesis. World Scientific Pub. Co., Singapore. Mamont, P.S., Duchesne, M.C., Grove, J. and Tardif, C. (1978) Initial characterization of a HTC cell variant partially resistant to the anti­proliferative effect of ornithine decarboxylase inhibitors. Exp Cell Res, 115, 387-393. Mamont, P.S., Joder-Ohlenbusch, A.M., Nussli, M. and Grove, J. (1981) Indirect evidence for a strict negative control of S-adenosyl-L-methionine decarboxylase by spermidine in rat hepatoma cells. Biochem J, 196,411- 422. Marton, L.J. and Pegg, A.E. (1995) Polyamines as targets for therapeutic intervention. Annu Rev Pharmacol Toxicol, 35, 55-91. Matsufuji, S., Matsufuji, T., Miyazaki, Y., Murakami, Y" Atkins, J., Gesteland, R. and Hayashi, S. (1995) Autoregulatory frameshifting in decoding mammalian ornithine decarboxylase antizyme. Cell, 80, 51-60. 22 Meyskens, F.L., Jr. and Gerner, E.W. (1999) Development of difluoromethylornithine (DFMO) as a chemoprevention agent. Clin Cancer Res, 5, 945-951. Mitchell, J.L., Judd, G.G., Bareyal-Leyser, A. and Ling, S.Y. (1994) Feedback repression of polyamine transport is mediated by antizyme in mammalian tissue-culture cells. Biochem J, 299, 19-22. Miyazaki, Y., Matsufuji, S. and Hayashi, S. (1992) Cloning and characterization of a rat gene encoding ornithine decarboxylase antizyme. Gene, 113, 191- 197. Murakami, Y., Matsufuji, S., Kameji, T., Hayashi, S., Igarashi, K., Tamura, T., Tanaka, K. and Ichihara, A. (1992) Ornithine decarboxylase is degraded by the 26S proteasome without ubiquitination. Nature, 360, 597-599. Namy, 0., Rousset, J.P., Napthine, S. and Brierley, I. (2004) Reprogrammed genetic decoding in cellular gene expression. Mol Cell, 13, 157-168. Newman, R.M., Mobascher, A., Mangold, U., Koike, C., Diah, S., Schmidt, M., Finley, D. and Zetter, B.R. (2004) Antizyme targets cyclin D1 for degradation. A novel mechanism for cell growth repression. J Bioi Chem, 279, 41504-41511 . Oredsson, S.M. (2003) Polyamine dependence of normal cell-cycle progression. Biochem Soc Trans, 31, 366-370. Orlova, M., Yueh, A., Leung, J. and Goff, S.P. (2003) Reverse transcriptase of Moloney murine leukemia virus binds to eukaryotic release factor 1 to modulate suppression of translational termination. Cell, 115, 319-331. Pegg, A.E. (1988) Polyamine metabolism and its importance in neoplastic growth and a target for chemotherapy. Cancer Res, 48, 759-774. Pegg, A.E., Pakala, R. and Bergeron, R.J. (1990) Induction of spermidine/spermine N1-acetyltransferase activity in Chinese-hamster ovary cells by N1 N11-bis(ethyl)norspermine (corrected) and related compounds. Biochem J, 267, 331-338. Pegg, A.E. and Williams-Ashman, H.G. (1968) Stimulation of the decarboxylation of S-adenosylmethionine by putrescine in mammalian tissues. Biochem Biophys Res Commun, 30, 76-82. Porter, C.W., Ganis, B., Libby, P.R. and Bergeron, R.J. (1991) Correlations between polyamine analogue-induced increases in spermidine/spermine N 1-acetyltransferase activity, polyamine pool depletion, and growth inhibition in human melanoma cell lines. Cancer Res, 51, 3715-3720. Sakata, K., Fukuchi-Shimogori, T., Kashiwagi, K. and Igarashi, K. (1997) Identification of regulatory region of antizyme necessary for the negative regulation of polyamine transport. Biochem Biophys Res Commun, 238, 415-419. Sakata, K., Kashiwagi, K. and Igarashi, K. (2000) Properties of a polyamine transporter regulated by antizyme. Biochem J, 347, 297-303. 23 Somogyi, P., Jenner, A.J., Brierley, I. and Inglis, S.C. (1993) Ribosomal pal.Jsing during translation of an RNA pseudoknot. Mol Cell BioI, 13,6931-6940. Stahl, G., McCarty, G.P. and Farabaugh, P.J. (2002) Ribosome structure: revisiting the connection between translational accuracy and unconventional decoding. Trends Biochem Sci, 27, 178-183. Sundararajan, A., Michaud, W.A., Qian, Q., Stahl, G. and Farabaugh, P.J. (1999) Near-cognate peptidyl-tRNAs promote +1 programmed translational frameshifting in yeast. Mol Cell, 4, 1005-1015. Suzuki, T., He, Y., Kashiwagi, K., Murakami, Y., Hayashi, S. and Igarashi, K. (1994) Antizyme protects against abnormal accumulation and toxicity of polyamines in ornithine decarboxylase-overproducing cells. Proc Natl Acad Sci USA, 91, 8930-8934. ten Dam, E.B., Pleij, C.W. and Bosch, L. (1990) RNA pseudoknots: translational frameshifting and readthrough on viral RNAs. Virus Genes, 4, 121-136. Thomas, T. and Thomas, T.J. (2001) Polyamines in cell growth and cell death: n10lecular mechanisms and therapeutic applications. Cell Mol Life Sci, 58, 244-258. Thomas, T. and Thomas, T.J. (2003) Polyamine metabolism and cancer. J Cell Mol Med, 7, 113-126. Tosaka, Y., Tanaka, H., Yano, Y., Masai, K., Nozaki, M., Yomogida, K., Otani, S., Nojima, H. and Nishimune, Y. (2000) Identification and characterization of testis specific ornithine decarboxylase antizyme (OAZ-t) gene: expression in haploid germ cells and polyamine-induced frameshifting. Genes to Cells, 5, 265-276. Tu, C., Tzeng, T.H. and Bruenn, J.A. (1992) Ribosomal movement impeded at a pseudoknot required for frameshifting. Proc Natl Acad Sci USA, 89, 8636- 8640. Vied, C., Halachmi, N., Salzberg, A. and Horabin, J.1. (2003) Antizyme is a target of sex-lethal in the Drosophila germline and appears to act downstream of hedgehog to regulate sex-lethal and cyclin B. Dev BioI, 253,214-229. Vujcic, S., Oiegelman, P., Bacchi, C.J., Kramer, O.L. and Porter, C.W. (2002) Identification and characterization of a novel flavin-containing spermine oxidase of mammalian cell origin. Biochem J, 367, 665-675. Wallace, H.M., Fraser, A.V. and Hughes, A. (2003) A perspective of polyamine metabolism. Biochem J, 376, 1-14. 24 Wang, Y., Xiao, L., Thiagalingam, A., Nelkin, B.D. and Casero, R.A., Jr. (1998) The identification of a cis-element and a trans-acting factor involved in the response to polyamines and polyamine analogues in the regulation of the human spermidine/sperrriine N 1-acetyltransferase gene transcription. J BioI Chem, 273, 34623-34630. Weiss, R.B., Dunn, O.M., Atkins, J.F. and Gesteland, R.F. (1987) Slippery runs, shifty stops, backward steps, and forward hops: -2, -1, +1, +2, +5, and +6 ribosomal frameshifting. Cold Spring Harb Symp Quant BioI, 52,687-693. Weiss, R.B., Dunn, D.M., Atkins, J.F. and Gesteland, R.F. (1990a) Ribosomal frameshifting from -2 to +50 nucleotides. Prog Nucleic Acid Res Mol BioI, 39, 159-183. Weiss, R.B., Dunn, D.M., Dahlberg, A.E., Atkins, J.F. and Gesteland, R.F. (1988) Reading frame switch caused by base-pair formation between the 3' end of 16S rRNA and the mRNA during elongation of protein synthesis in Escherichia coli. EMBO J, 7,1503-1507. Weiss, R.B., Huang, W.M. and Dunn, D.M. (1990b) A nascent peptide is required for ribosomal bypass of the coding gap in bacteriophage T 4 gene 60. Cell, 62,117-126. Williams-Ashman, H.G., Janne, J., Coppoc, G.L., Geroch, M.E. and Schenone, A. (1972) New aspects of polyamine biosynthesis in eukaryotic organisms. Adv Enzyme Regul, 10,225-245. Wills, N.M., Gesteland, R.F. and Atkins, J.F. (1991) Evidence that a downstream pseudoknot is required for translational read-through of the Moloney murine leukemia virus gag stop codon. Proc Nat! Acad Sci USA, 88, 6991- 6995. Xu, J., Hendrix, R.W. and Duda, R.L. (2004) Conserved translational frameshift in dsDNA bacteriophage tail assembly genes. Mol Cell, 16, 11-21. Yoshinaka, Y., Katoh, I., Copeland, T.D. and Oroszlan, S. (1985) Murine leukemia virus protease is encoded by the gag-pol gene and is synthesized through suppression of an amber termination codon. Proc Nat! Acad Sci USA, 82, 1618-1622. 25 Zhang, M., Pickart, C.M. and Coffino, P. (2003) Determinants of proteasome recognition of ornithine decarboxylase, a ubiquitin-independent substrate. EMBO J, 22, 1488-1496. Zhu, C., Lang, D.W. and Coffino, P. (1999) Antizyme2 is a negative regulator of ornithine decarboxylase and polyamine transport. J Bioi Chern, 274, 26425-26430. CHAPTER 2 POLYAMINE SENSING DURING ANTIZYME mRNA PROGRAMMED FRAMESHIFTING The following chapter is a reprint of an article coauthored by myself, Howard, M. T., Gesteland, R. F. and Atkins, J. F. It was originally published in Biochemical and Biophysical Research Communications, volume 338, pages 1478-1489, 2005 (copyright by Elsevier Inc.). 27 Available online at www.sdencedirect.com SCIENCE@DIRECTO BBRC ELSEVIER Biochemical and Biophysical Research Communications 338 (2005) 1478-1489 www.elsevier.com/locate/ybbrc Polyamine sensing during antizyme mRNA programmed frameshifting Lorin M. Petros a, Michael T. Howard a, Raymond F. Gesteland a, John F. Atkins a,b,;.; a Department of Human Genetics, University of Utah, 15 N, 2030 E, Rm 7410, Salt Lake City, UT 84112-5330, USA b Bioscience Institute, University College (ork, Cork, Ireland ReceIved 22 September 2005 Available online 27 October 2005 Abstract A key regulator of cellular polyamine levels from yeasts to mammals is the protein antizyme. The antizyme gene consists of two over­lapping reading frames with ORF2ln the + 1 frame relative to ORF1. A programmed + 1 ribosomal frameshift occurs at the last codon of ORFl and results in the production off ulHength antizyme protein. The efficiency off rameshifting is proportional to the concentration of polyamines, thus creating an autoregulatory circuit for controlling polyamine levels. The mRNA recodlng signals for frameshifting include an element 5' and a pseudoknot 3' of the shift site. The present work illustrates that the ORFl stop codon and the 5 D element are critical for polyamine sensing, whereas the 3' pseudoknot acts to stimulate frameshifting in a polyamine independent manner. We also demonstrate that polyamines are required to stimulate stop codon readthrough at the MuLV redefinition site required for normal expression of the GagPol precursor protein. © 2005 Elsevier Inc. All rights reserved. Keywords: Polyamine; Antizyme; Frameshifting; Readthrough; Termination; Recoding; Stop codon; Spermidine: Ornithine decarboxylase Polyamines are ubiquitous organic cations with distributed charges that interact with many macromole­cules, and affect normal cell growth and function at many levels including several aspects of protein synthesis (reviews (1-3]), There is a well-established link between polyamine content and human disease (reviews [4,5)). For example, significant success has been obtained in the treatment of some of parasitic infections by DFMO (a-ditluoromethyl­ornithine), which is known to alter intracellular polyamine levels by the irreversible inhibition of ornlthine decarboxyl­ase (DOC). In addition, persistent high polyamine levels are associated with an increase in cell proliferation, a decrease in apoptosis, and an increase in the expression of genes involved in tumor invasion and metastasis; and pre-clinical studies have sparked interest in certain poly­amine synthesis inhibition strategies for cancer chemopre­vention [61. As befits molecules involved in a broad range of cellular functions, their concentrations are highly regulated. • Corresponding author. Fax: + 1 801 585 3910. E-mail address: john.atkin~genetics.utah.edu OJ. Atkins). 0006-291X1S • see front matter © 2005 Els~vier Inc, All rights reserved. dOi:1O.l016/j,bbrc.200S.IO.115 The rate limiting enzyme in polyamine synthesis is ODC, which catalyzes the synthesis of putrescine (+2). Putrescine is then the substrate for the synthesis of spermi­dine (+3) from which spermine (+4) is derived. For conve­nience all will be referred to as polyamines, though putrescine is a diamine. One of the key regulators ofi ntra­cellular polyamine levels through DOC is the protein anti­zyme (Fig. 1 ). Antizyme binds to DOC and targets it for ubiquitin-independent degradation by the 265 proteasome [7-9). Although mammalian DOC degradation, at least under stress conditions, is also mediated by NAD(P)H qui­none oxidoreductase 1 [10], it is clear that antizyme plays a major regulatory role in polyamine synthesis. In contrast to its destabilization of DOC, antizyme serves to stabilize a protein, termed antizyme inhibitor [111 which sequesters antizyme. Antizyme also inhibits polyamine import [12- 14] and stimulates polyamine export [15}. Additional roles for antizyme have been proposed. For example, there is evidence that mammalian antizyme inter­acts with cyclln 0, at least in prostatic cells [161. Further, Drosophila antizyme [171 regulates nuclear entry and the overall levels of the sex determination master switch, 28 LM. Petros et at I Biochemical and Biophysical Research Communications 338 (2005) 1478-1489 1479 Antizyme mRNA +1 pR11 'It --.~~ \ l-......,o.afadation byProttuo I i CD ine+-Spemldine+-Putresclne+=-- Ornithine i Intracellular , Polyaminet eli~_~_~_~_e ___ -' Extra(l:elllJlar PoIyaminea Fig. 1. Sch~matlc diagram of antizyme regulation of cellular polyamine pools. 5ex.fethal. Drosophila antizyme, Hedgehog, and Sex-lethal function to regulate eyelln B [18~ In carrying out its differ­ent functions, antizyme has various cellular locations, per­haps including the mitochondria [19]. At least mammalian antizyme 1 has two nuclear export signals [201 and may be involved in nUcleocytoplasmic shuttling of OOC during different stages of mitosis [21]. Antizyme expression is induced by high Intracellular polyamine levels, and decreased with lowered levels thereby facilitating the standard maintenance of polyamine homeo­stasis. The polyamine level sensor is a non-standard decod­ing event that is required for antizyme synthesis. All antlzyme mRNAs contain two overlapping readlng frames, a short ORFl and a long ORF2 [22-241 ORF2 is in the +1 frame relative to ORF1, and it lacks a translational initia­tion codon. As the termination codon for ORFl enters the A~site of translating ribosomes, the outcome is dependent on polyamine levels. At low polyamine levels, termination at the end of ORFl is efficient; most ribosomes do not enter ORF2 and so do not synthesize antizyme. At high levels of polyamines, a substantial proportion of ribosomes shift to the +1 reading frame and then resume standard decoding to synthesize antizyme. In other words. the recod~ ing event involved is regulatory. The sequence of the shift site and its context are highly conserved betweenSchizosaccharomyces pombeitnd Homo sapiens [25]. In all mammals, the shift site is uee U(GA}. The stop codon, UGA, 3' adjacent to the shift codon, uee, is the ORFl terminator. Frameshifting at the anti­zyme mRNA shift site is stimulated by two cis-acting sig­nals. One of these, the S· element [23,251 encompasses approximately 50 bases upstream of shift site and is impor­tant for the polyamine effect [26}. The other cis-acting sequence is a pseudo knot located 3' of the shift site. The mammalian antizyme pseudoknot [23,27] and a different counterpart in a subset oft nvertebrate antizyme mRNAs [28] are the only pseudoknots known to act as stimulators for +1 frameshifting, though others have been proposed for a listeria phage [29]. Distinctive pseudoknots also act as recoding signals for programmed (in-phase) readthrough of stop codons, for example to synthesize the murine leukemia virus GagPol precursor, which is the source of reverse transcriptase [30-321 The Pol part of the product interacts with release factor eRF1. and this interaction probably regulates the amount of readthrough [33]. Readthrough at the mamma­lian antizyme ORFl terminator (Uee) UGA also occurs, at least in vitro with the addition of spermidine, to a level which is comparable to that off rameshifting (23J. Decod­ing at the end of antizyme mRNA ORF1 is therefore a competition between frameshifting, readthrough of the stop codon, and translational termination [34J. There is one gene for antizyme in yeasts {2S,35j, but three different antizyme genes in mammalian cells. Mam­malian antizyme 1 [24] and 2 [27,36] are expressed in all cells except in developing male germ cells where antizyme 3 [22,37} is expressed. However. unlike antizyme 1 and 2 mRNAs, antizyme 3 mRNA does not have the 5 and 3' recoding signals referred to above. Here we present a mutagenic analysis of the mammalian antizyme 2 frameshift region to investigate the relative importance and relationship of the 3' pseudo knot, the shift site (including the stop codon), and the S· element for poly­amine stimulation of the required frameshifting. While the 50 element has been implicated in polyamine sensing, the shift site including the stop codon, which is known to be important for high level frnmeshifting, has not been inves­tigated for its role in sensing polyamine levels. In addition, we examine polyamines' ability to affect the competition between frameshifting, readthrough, and termination by testing polyamine stimulation of readthrough in the anti~ zyme context. Polyamines' effect on termination is further studied by measuring polyamine stimulation of read­through at a programmed readthrough site (MulV) and at non-recoding stop codons. The sequence of the antizyme 2 mRNA pseudoknot, unlike its antizyme 1 counterpart, has no stop codons in the two out-of-phase frames. Thus, without changing the nucleotide sequence, all three frames can be monitored in the antizyme 2 frameshift cassette. For this reason, and because polyamines are expected to stimu­late frameshifting by the same mechanism in the two anti­zyme mRNAs, antizyme 2 sequence was used in the studies reported here. Materials and methods Reporter plasmid cloning. Complementary oIkjonucleotides were syn­thesized at the University of Utah DNA/Peptide (ore Facility such that when annealed they would have Sail and 8amHI compatible ends. The sequences of the top strands are listed in supplementary material. AZ2-wt PKdel, AZl-wt, and r del were published previously as p2luCaz4 p2lucaz2pkdel, p2lucazl, and p2Lucauusdeli respectively [261 The controls for aU the antlzyme constructs were created by deleting the T of the first zero frame TGA coooo. MulV-PKdel sequence was published previously as MulVS [38L The (ontrol fOf MuLV-wt and MulV-PKdel is 29 1480 L.M. Petros et al. i Biochemical and Biophysical Research Communications 338 (2005) 1478 1489 the MuLV-wt construct with the T of the stop codon changed to a C. The UGA N. UAG N. and UAA N constructs were published previously and the control constructs have a C in place of the first nudeotide of tht' stop codon [391. TI1e annealed oligonucleotides were ligated into Sail and Bam HI digested p2lu;: vector [401 and transformed into E. coli strain DH5- a. Plasmjds were purified using the QIAfiltt'r Plasmid Midi Kit (Qiagen) according to the manufac.turer's speclfications. The 5equences were verified by autolMmwcyc:ler sequencing at the University of Utah Sequencing Core Facility. Cell culture and tratuf&tions. COS-7 cells were maintained as mono­layer cultures and were grown in Dulbe«o's modified Eagle's media (DMEM) with 100 mgll 0'9Iuc05e. L-glutamine, pyridoxine hydrochlo­ride. and 110 mgfl sodium pyruvate supplemented with 10% FSS. 50 Ulml penicillin, and 50 I g/ml streptomycin, Cells were gtown at 37 "C in an atmosphere of 5% CO l' All media and antibiotics were obtained from Gibeo Invitrogen (Carlsbad, CA), and all Sera were obtained from HyClone !logan. Un. For duallucifera~ aScsays, 1 x lO'l cells were plated in 90-wen, treated tissue culture plates and 2,5 mM DFMO was added at this time, Cells were grown for 24 h as described abOVE!. The cells were transfE!cted with OSII IipofectAMINE reagent (lnvltrogtmJ and 50 n9 plasmid DNA for 24 h In serum-free media In t~ presence of 2.5 rnM OFMO. Aftef the transfections. fresh media were added with 1 mM amlnoguanidine, 2.5 mM DFMO, and either no polyamines, 0.1 mM spermine, 1 mM spermidine, or 1 mM putrescine, The cells were incubated for an additional 48 h before being assayed. Dual hJdferase assay ilnd rooxling efficiency, Cells were lysed using 25,11 of puslve lysis buffer (Promegal, and hlt:iferase activity was deter­mlnEtd on 20 "I of the cell lysis buffer using the Dualluciferase reporter a~y (Prolmlga) on a Dynatech MLX Mkrotiter Plate luminometef as described in {40j. for all reactlons,light emission was measured between 2 and 12 s after 50 pJ of! uminescence substrate WaJ Injected. Statistics. For each construct assayed using the dual ludferase assay system, the data were checked to be normally distributed (a '" 0,05), outliers were eliminated, and the standard statistics were alculated as described in [411. The mean of the ratio of firefly to Renilla relative 1i9h1 units for the test constructs was then divided by the mean of the ratio of firefly to Renilla relative light units of the corresponding control construct to calculate the percent frame.hifting or readthrough, The fold stimulation was calculated by dividing the percent frameshifting or readthrough for a given polyamine treatment by the percent frame· shifting Of readthrou9h for the corresponding no polyamine treatment The standard deviation for the percent frameshifting, percent read· through, and t~ foid stimulation was ulculated using standard for­mulas for propagation of error. An unpaired two-sample t test was used to compare if differences in the calculated frameshifting or readthrough values were statistically significant. Those considered significantly dif, ferent had p values less than 0.05, most vafues being much ioweJ. Those with p values less than 0.05 or 0.01 are listed. If tlu: p value is not indicated, then it is less than 0.001. Results To assay the effect of polyamines on translation of an antizyme frameshift cassettel we used a reporter system which permits monitoring from a single vector (p2Iuc). both the product of standard triplet decoding and the transframe product derived from frameshifting. This sys­tem permits assessment of the proportion of ribosomes that shift frame. The reporter for the initial reading frame is Renilla luciferase, and the reporter for the 3' overlapping frame is firefly luciferase. Both luciferases are assayed in the same tube giving a sensitive internally controlled sys­tem [40}. Difluoromethylornithine (DFMO) is a competi­tive inhibitor of OOC and was present throughout the experiments to deplete the cellular polyamine pools, thus Increasing the sensitivity of the cells to the addition of exogenous polyamines. While DFMO reduces the levels of spermine, spermidine, and putrescine, the greatest reduc­tion is seen with spermidine, and the no externally added polyamine condition (NO PAl still has low levels of poly­amines [261. Spermine (0.1 mM), spermidine (1 mM), and putrescine {1 mM} were added to the cells' medium individ­ually to determine if any of these polyamines had different effects on the polyamine stimulation off rarheshifting. Since spermine, spermidine, and putrescine are products along the same pathway in polyamine biosynthesis, the addition of one of these polyamines affects the cellular polyamine pools for all three. Therefore, a lack of a differential effect by the polyamlnes does not eliminate the potential for the polyamines to stimulate frameshifting by distinct mecha­nisms. The concentrations used have been shown to cause near maximum levels off rameshift stimulation [26). High level frameshift stimulation by the polyamines was desired in order to see even very low levels of stimulation. It should be noted that the actual cellular concentrations with the given additions of polyamines were not determined, and the amount off fee and bound polyamines is unknown. The role of the 3' pseudoknot in polyamine stimulation The human antizyme 2 mRNA frameshift cassette, with and without the 3' pseudoknot. was cloned into the p2iuc vector and assayed in transiently transfected eOS-7 cells. The antizyme 2 wild type frameshift cassette (AZ2-wt) is defined as the sequence from the beginning of the 5 ele­ment (nt 197 from the 5' end of the mRNA) through the pseudoknot of the antizyme 2 coding sequence (nt 316 numbering as in [27]). In the wild type construct, poly· amines stimulated frameshifting 5.3-, 4.3-, and 3.5-fold for spermine, spermidine, and putrescine, respectively. When the pseudoknot was removed (PKdel), polyamines stimulated frameshifting (7.a-, 4.3-, and 25-fold for sperm­ine, spermidine, and putrescine, respectively) to a signifi­cant level over the NO PA condition (Fig. 2; p < 0.001 for all sign incant differences reported in this manuscript unless otherwise noted, see Materials and methods). The resulting conclusion, that the 3' pseudoknot is Significant for frameshifting but irrelevant for polyamine stimulation off rameshifting, is in agreement with previous studies [23,26]. The role of the shift site in polyamine stimulation The last ORFl zero frame sense codon is uee and is part of the shift site uee UGA. The shift codon, uec, was next examined for its role in polyamine stimulation off rameshifting, but in the absence of the pseudoknot based on the above results. The human antizyme shift site, vee UGA, was changed to UUU UGA and ece UGA, whlch are shift sites found in antizymes of some species of nematodes and fungi, respectively [25]. The shift site was also changed to CUU UGA and AM UAA. 30 LM. Petros et at (Biochemical and Biophysicbl Researc.h Communications 338 (2005) 1478-14S9 1481 .NOPA II &PM f '+ c~ Fig. 2. The effect of the antizyme 23' pseudokflots on poi)'amine induced frameshiftiog. COS-7 cells grown in the presence of DFMO were transiently transfected with the antizyme 2 wild t)'pe (AZ2·wt} and pseudoknot deletion {PKdel) constructs. Media were supplemented with 0.1 mM spermine (SPM. dark grey bars), 1 mM spermidine (SPO. light grey bars), 1 mM putresd~ (PUT. white bars), or no polyamines {NO PA. black bars}. Ba,s represent the average percent frameshiftlng as determined by the dual luciferase assay on at least two im:tependent experiments. Error bars represent the standard deviation. The indicated values are the fold stimulation over the NO PA condition. Frameshifting ranged from 1.0% at AAA UAA to 5.3% at CUU UGA when polyamine pools were depleted (NO PA, Fig. 3 A). Polyamines significantly increased frameshif­ting at all the shjft sites tested as compared to the NO PA condition (Fig. 3 A; p < 0.05 for 551, PUT and p < 0.01 for 552, PUT). With the additIon of spermine, frameshifting was stimulated 7.S-fold at uce UGA. the mammalian antizyme shift site. With the above changes to the shift site. spermine stimulation ranged from 4.0- to SA-fold. With the addition of spermidine. frameshifting was stimulated 4.3- fold at uce UGA and 2.0- to 3.S-fold at the other codons. Finally with the addition of putrescine, frameshifting was stimulated 2.S-fold at UCC UGA, and at the other shift codons frameshifting ranged from 1.3- to 2.4-fold. While the AAA UAA construct exhibHs a low base level off rame­shifting, polyamine stimulation is similar to the wild type shift site. These results show that there is versatility in the shift sites that can be used for polyamine stimulation off rameshifting. The ability of the tRNA anticodon to shift relative to the mRNA codon is most likely the reason for the differences in overall frameshlfting efficiency and polyamine stimulation. The dependence on the ORFl stop codon as part of the shift site for polyamine stimulation off rameshifting was investigated next. In cassettes with the wild type antizyme 2 context. the stop codon was changed to various sense codons by altering the second or third nucleotide. The U of the stop codon was preserved to maintain the potential for the P site tRNA to re-pair with the mRNA at the over­lapping +1 frame codon. The data for these experiments are represented as fold Increases instead of percent frame­shifting. because a decrease in Renilla ludferase protein A r i '+ ! D&PM .SPO DPUT Pl<det SS1 SS2 SS3 SS4 8MI BIte: uce UGA ceo UGA UUU UGA OUU UGA AltA UAA Al2-wl SC1 SC2 SC3 SC4 SC6 Shift aile: UCC UGA UCC UGG uee UGU uce UGC uce UCA UCC UUA Shift aile: AZ2-wt UCCUGA .top1 UCCUAG at0p2 UCCUAA fig. 3. Polyamine stimulation off rameshifting at djff~rent shjft S!t@S. (OS~7 cells grown in the presence of DFMO were transiently transfected with the indicated constructs. Media were supplemented with 0.1 mM spermine (SPM. dark grey bars). 1 mM spermidine (SPO. light grey bars). I mM putrescine (pUT, white bars), Or no polyamines (NO PA. black bm). (AJ The UCC UGA shift site was changed to other shift sites ISSI- 4) with the 5 el8lnent and no pseudoknQt. Bars represent the average percent frameshlfting as determined by the dualluclferase assay on at least two independent experiments. Error bars represent the standard deviation The indicated values are the fold stimulation ovef the NO PA condition, (Bl The UGA stop codon was (h,lnged 10 sense (odons (SC 1,51 in the wild type context. Bars represent the average fold stimulation over the NO PA condition done for at least two lndep"nd"nt experimenh. Error bars represent tht' nandard deviation. (C) The llGA stop codon was changi!d to the other two stop codons (stop 1-2) in the wild type c.ontext. Bars and value$ are the same is It) A. in the +1 constructs caused an artificial increase in overatl frameshifting. This artifact did not affect polyamine stimu­lation as confirmed by Western blotting (data not shown). In addition, this artifact did not occur with the correspond­ing constructs in the antizyme 1 context, but the same trend in polyamine stimulation was found (see below). 31 1482 LM. Petro~ et al. / Biochemical and Biophysical Researc.h Communications 338 (:2005) 1478-1489 With the stop codon of the shift site changed to a sense codon, polyamine stimulation was eliminated (D.S- to 1.3- fold stimulation; Fig. 3 8). The only significant (p < 0,01) increase in frameshifting was seen with UCA in place of the SlOP and the addition of spermidine, but even this increase was three times less than that seen with UGA (1.3-fold for UCA as compared to 4.3·fold for UGA). This data provides evidence that a stop codon is required to achieve high level stimulation off rameshifting by polyamines. The ORFl stop codon is UGA in all known antizyme genes (25]. Previous in vitro data have demonstrated that the stop codon of antizyme 1 can be replaced with the other two stop codons [23L though frameshifting efficiency with UAG and UAA is somewhat lower. However, polyamine stimulation off rameshifting with either of the other two stop codons replaCing UGA at the shift site was not tested earlier. We replaced the UGA stop codon with the UAG and UAA stop codons in the antlzyme 2. context containing both the 5° element and the pseudo knot. Polyamines stim­ulated frameshifting to a significant extent at all 3 stop codons (Fig. 3 C). Spermine, spermidine, and putrescine stimulated frameshifting with a UAG stop codon 4,(}, 3.5-, and 2.1-fold, respectively. With a UAA stop codon, polyamines stimulated frameshifting 3.5- (spermine), 2.6- (spermidine) and 1.6- (putrescine) fold. Although poly­amine stimulation was preserved, it was reduced relative to stimulation with a UGA stop codon (7.8-, 4.3-, and 2.5-fold for spermine, spermidine, and putrescine, respec­tively). Frameshifting efficiency was also lower when the stop codon was UAA as compared to UGA (for example 29% for UGA and 16% for UAA with the addition of spermine). The ability of aU three stop codons to induce frameshifting and the absence of stimulation by any of A 50% 40% til c: I•e11 1 11,4 I! u.. ~ !! '" 0% AZ1-wt SC6 8C7 SC6 BC9 Shift site: lICC UGA UCClIGG UCCUGlI UCCUGC UCCUCA the five sense codons tested is consistent with one key fea­ture for polyamine sensing being related to termination. The Slop codon at the end of ORfl has previously been shown to be a key feature in the frameshifting of antizyme [23}. The question remains whether the lack of polyamine stimulation when the ORFl stop codon is changed to a sense codon is due to an inability of polyamlnes to stimu· lation frameshiftlng at the site UCC U when followed by a sense codon, or whether frameshifting itself! s unable to occur when the stop codon is replaced with a sense codon. Due to an artifact that o(curs when the stop is changed to a sense codon in the antizyme 2 context, we are unable to determine the actual levels off rameshifting, making this question difficult to address. We therefore made the same changes in the stop codon to sense codons in the antizyme 1 context (Fig. 4 A). In the antizyme 1 wild type construct (AZ1-wt) frameshifting is stimulated 8.4-. 6.8-. and 5.8-fold with spermine, spermidine, and putrescine respectively. When the stop codon is changed to UGG (5C6), UGU (SCn UGC (S(8), UCA (SC9), or UUA (SOO) poly-amine stimulation is virtually eliminated. With the stop codon changed to UGG. a low but significant increase in frameshlfting was seen with the addition of spermine (lA-fold as compared to 8A-fold in the wild type construct) and putrescine (p < 0.05; 1.2-fold as compared to 5.8-fold in the wild type construct). When the stop codon was chan­ged to UGU, a low (lA-fold for spermine and 1.2-fotd for spermidine and putrescine) but significant increase in frameshifting was seen with all three polyamines (p < 0.Q1 for spermidine and putrescine). No significant polyamine stimulation was seen when the sense codon was UG(, UCA. or UUA. Therefore, as seen in the antizyme 2 con­text, the stop codon as part of the shift site is required for high levels of polyamine stimulation. 8 10"'" .NOPA IISPM cSPD 8% o PUT .~ ~ ,ill;l 6% ~ oS ~ 4% !i: <!( 2% 0% SC10 AZ1-wt BS-wt AS-wt SC10 BS·UUA AS·UUA UCCUUA Fig.4. Frame-shifting and polyamine stimulation 10 the antlzyme 1 ront~t. C05-7 <ells grown in the presence of OfMO ~re transiently transfected with the antizyme 1 constructs. Bars represent the average percent frameshiftlng as determined by the dual luciferase assay on at least two independent el<periment>. Error bars represent the standard deviation. (A) The UGA stop codon (AZ1·wt) was changed to sense codons ($(6-10) in the wild type context. After transfections, media were $upplemented with 0.1 mM spermine (SPM, dark grey ban), 1 mM s~rmidine (SPD, light grey bars), 1 mM putrescine (PUT, white bars), or no polyamines (NO PA. black bars). The indicated values are the fold stimulation over the NO PA condition, IS} The average ~rcent frame,hiftlng was dete,mlned in the wild type and with the stop change to UUA {SC 10} contex:t with no polyamin~ added to the media. A zero ffame stop codon was added before the shift site (SS-INt, BS-UUA) and after the shift she (AS-wt, AS·UVA). 32 LM. Petlos et at / aiochemical and Biophysical Research Communications 338 (2005) 1478-1489 1483 In the antizyme 1 wild type context, when polyamlnes are depleted, 3.9% frameshifting occurs (Fig. 4 B). When the stop codon is changed to the sense codon UUA (SClO), frame­shifting decreases to 1.9%, demonstrating the importance of the stop codon in the A-site for frameshifting. By placing an in·frame stop codon just before the shift site, frameshif­ling should decrease because termination will occur before the shift site is reached. In both the wild type (BS-wt) and with the stop codon changed to UUA (BS-ULJA), changing the (odon immediately 5' of the shift site (odon (UCC) to a stop codon significantly decreases frameshifting to 1.7% and 1.4% (3 2.3- and l.4-fold decrease), respectively. These results demonstrate that a low level off rameshifting above the background for this assay system is occurring down­stream of the newly introduced stop codon. In addition, we placed an in-frame SlOP codon after the shift site in both the wild type (AS-wt) and With the stop change to UUA (AS·UUA). These changes should not decrease frameshif­ting; because once the frameshift event occurs, this stop codon will no longer be in-frame. In these constructs frame­shifting increased to 5.6% in the wild type context and 3.6% in the UUA context. Therefore, frameshifting must be occur­ring before this stop codon. The increase in frameshifting could be due to frameshifting occurring with the new stop codon in the A site or due to a more favorable context for frameshifting to occur at the uce U site. Regardless, when the stop codon is changed to UUA, frameshifting occurs at the uce or UUA codon in the P site, and this level off rame­shifting is above the background. These results demonstrate the ability off rameshifting to occur at the shift site when the stop codon is changed to a sense codon. but polyamines are unable to stimulate a high level off rameshifting. The role of the 5' element in polyamine stimulation Finally, the relative importance of the 5' element of anti­zyme 2 for polyamine stimulation was examined. The 5' element was removed in constructs with and without the $% CII 4% E !! i ~ ~ ': t c( Shlftsit&; AX UCCUOA 00110 UCCUGA SS1-5'de' cce UOA antizyme pseudoknot. With added spermine. spermidine or putrescine, the amount off rameshifting was stilt signif­icantly Increased from the NO PA condition in the absence of the 5' element and with the pseudoknot (S'del; Fig. 5; p < 0.01 for SPD). However. the greatest increase was only 2.1-fold as compared with S.3mfold in the presence of the 5' element (compare to Fig. 2). When both the 5 element and the pseudoknot were removed (shiftonly), only a 1.2-fold increase (p < 0.05) was seen with spermine and none with spermidine or putrescine. The great reduction of polyamine stimulation with deletion of the 5' element confirms the importance of the 5' element for polyamine stimulation shown earlier in HEK293 cells [26). While the S' element plays a major role in polyamine sensing, low polyamine stimulation is still evident in its absence. In addition to the human antizyme shift sfte. constructs with the ecc UGA, UUU UGA. CUU UGA, and AAA UAA shift sites were also tested for polyamine stimulation in the absence of the 5 element (Fig. 5). In these constructs. polyamine stimulation was almost completely eliminated (compare to Fig. 3 A). The only significant increase in frameshifting was seen at the shift site CUU UGA (5S3- 5' del) with the addition of spermidine, but this increase was almost four times lower than the stimulation seen with the 5" element (S.4-fold with the 5' element versus 1 A.-fold without the S' element), These results further support the important role of the 5' element in polyamine stimulation off rameshifting. Polyamine stimulation of readthrough When ribosomes encounter a stop codon in the ribosom­al A-site, there is competition between termination, read­through and frameshifting [34]. In an initial in vitro study. spermidine was shown to stimulate read through of the antizyme ORFl stop codon [23~ This investigation was extended here by monitoring polyamine stimulation of readthrough in the antizyme frameshift cassette in SS2-5'de' uuu UIlA 8S3.S·del CUUUGA -NOPA IIISPM CSPO IJPUT SS4-5'del AAAuM Fig, 5, Polyamine itimulation off rameshlftlng in the antlzyme 2 hameshift cassette with toe deletion of the 5 element, COS·7 cells grown in the presence of OFMO Wefe transiently transfected with antizyme :2 constructs with the S· element deleted (S del) and both the 5 element and 3' pseudoknot deleted (shift only). Cells were also uan&fected with constructs in which the 5' element and pseoooknot are absent And the wild t~ shift site (UCe UGAl was changed to CCC UGA (551·5 del). UUU UGA (552-5 . dell, CUU UGA (553·5' del), and AAA UAA (SS4·S· del). Media were supplemented with 0.1 mM spermine (SPM, dark grey ba(5), 1 mM spermidine (SPD, light grey bars), 1 mM putrescine (PUT, white bars), or no polyamiM$ (NO PA. black bars). Bars represent the average percent frameshifting as determined by the duallu.::lferase assay on at least two independent experiments. Error bars represent the standard deviation. The indicated valuecs are the fold stimulation over the NO PA condition. 33 1484 LM, Petros et at / Biochemical and Biophysical Research Communications 338 (200511478-1489 Fig. 6. Polyamine stimulation of readthrough at the ORFI stop codon of antizyme 2. CO$-7 cells grown In the presence of DFMO were transiently transfected with readthrough constructs for antizyme 2 wikl type (RT-wtl and with the pseudo knot (RT-PKdeU, the 5 element (RT·S 'del), or both (RT-shiftonty) deleted, Media were supplemented with 0,1 ruM spermine (SPM, dark grey bim). 1 mM spermidine (SPD, light grey bars), 1 mM putre$cine (PUT, white bars), or no polyamines (NO PA, black bars). Sars represent the average perCliH1t readthrough as determined by the dual luciferase assay on at least two independent experiments. Error bars represent the standard deviation. The indkated values are the fold stimulation over the NO PA condition. COS·7 cells. Constructs were created with the firefly lucif~ erase gene in the zero frame relative to theRenilia lucifer­ase gene. We tested the full-length antizyme 2 frameshift cassette (RT-wt) along with the pseudoknot (RT-PKdel), the 5' element (RT-S' deO, or both (RT-shift only) deleted to determine if these cis-acting sequences affected read­through. The antizyme stop codon was found to have a low level of readthrough close to background levels in the polyamine depleted condition (Fig. 6). Spermine, sper­midine, and putrescine significantly stimulated read­through of the UGA stop codon to approximately the same level regardless of the presence or absence of the stim­ulatory elements. These results rule out the possibility that the 5° element functions primarily during polyamine stimu­lation to shift the competition between readthrough and frameshifting in favor of the latter. The level of readthrough in the antizyme context was very low (less than 1 %). In order to determine if poly­amines could stimulate readthrough to higher levels, poly­amine stimulation of a programmed readthrough site was tested. We measured readthrough of the MuLV sequence with and without the pseudo knot (MuLV-wt and MulV­PKdel, respectively; Fig. 7). In the wild type construct, polyamines significantly stimulated readthrough from 1 % to 4-6% (4.7- to 7.0-fold). It is interesting to note that the amount of readthrough seen with the addition of poly­amines is similar to that of readthrough previously deter­mined in vitro (32}. In the absence of the pseudo knot. polyamines still significantly stimulated readthrough, but to a lower level, from 0.06% to 0.14-0.23% (2.4· to 3.9- fold). High levels of readthrough were stimulated by poly· amines at the programmed readthrough site, which depended on the cis-acting stimulator of this recooing event; however, the cis-ac.ting stimulators of antizyme -NO 6% IIISPM []SPD 5% o PUT ..I: QI t:::J 4% 3% I 2% Fig, ], Polyamine stimulation of readthrough at the GagPol junction of MuLV, COS-7 cells grown in the presence of DFMO were transiently tfilnsfected with read through con~tructs for the MulV wild type (MuLV· wtl and pseudoknot deletion (MuLV-PKdel), Media were supplemented with 0.1 mM spermine (SPM, dark grey bars), 1 mM spermidine (SPD, light grey bars), 1 mM putrescine (PUT, white bars), or no polyamines (NO PA, black bars). Bars represent the average percent readthrough as determined by the dual lucifera~ assay on at least two independent experlmliH1ts. Error bars represent the standard deviation, The indicated values are the fold stimulation over the NO PA condition. frameshifting do not facilitate polyamine stimulation of readthrough. Looking at the levels of polyarnines and their effect on readthrough and GagPol production in MuLV infected and uninfected cells warrants investigation. To test polyamines' ability to stimulate readthrough in the absence of recoding elements, we determined read­through levels with all three stop codon, and varied the 3D nucleotide in a randomized background. Consistent with prior studies in microorganisms. readthrough effi­ciency of stop codons in mammalian cells is determined by the identity of the stop codon and the 3' nucleotide 1 39,42-44}. Polyamines were able to stimulate read­through for all the constructs to varying degrees (Fig. 8). The greatest significant stimulation of read­through was seen with the UGA stop codon (2.3· to 13.5-fold; Fig. 8 A). The level of polyamine stimulation of readthrough varied by the 3' nucleotide and the poly­amine added. A lower, but still significant ( p < 0.01 for UAG A, SPD, and UAG U, SPM), level of readthrough was also stimulated by polyamines at the UAG stop codon (1.7- to 3.0-fold; Fig. 8 B). Readthrough of the UAA was also signincantly (p < 0.05) stimulated by poly­amines (1.4- to 2.3-fold), except at UAA G with the addi­tion of spermine (Fig. 8 C). The levels of readthrough with the addition of polyamines were similar to stop codon readthrough of the same constructs in untreated cells [39}. The stimulation of readthrough at all stop codons including programmed readthrough and frame­shift sites independent of the 5' element is strong evi­dence that elevated polyamine levels affect the termination process. 34 LM. Petros et at I Biochemical and Biophysical Research Communications 338 (2005) 1478-1489 1485 f l J UOAA UGAC UGAQ UGAU B &. f e I J UAQA UAGC UAGO UAGU C 1.0% r: 0.4% J UAAA UAAC UAAG UAAU Fig. 8. Pol)lilmine stimulation of readthrough of $top codons at nOn' re<:oding sites. (OS-7 c~H~ grown in the presence of OFMO were transIently transfected with UGA N {A), UAG N (B), and UAA N (el constructs where N "" A, C. G, or U. Media were supplemented with 0.1 mM spermine (SPM, dark grey bars), 1 mM spermidine (SPO, light grey bars), 1 mM putrescine (PUT, white bars), or no polyamines (NO PA. black bars), Bars represent the average percent readthrough liS determined by the dual IlJcifera~ assay on at least two independent experlmenu. Error b<lrs represent the standard deviation. The Indicated values are the fold stimulation over the NO PA conditi(>n. Discussion The programmed +1 ribosomal frameshift required for synthesis of antizyme occurs at the last sense codon of the short ORF1. This frameshifting is enhanced at elevated polyamine levels and serves as a sensor of polyamine levels. The present study shows the importance in mammalian cells of both the ORfl stop codon and the 5 . element for polyamine sensing. Polyamine sensing: stop codons versus slow-to-.decode sense codons Early work showed that a codon at which +1 frameshifting occurs becomes much rnore shift-prone when a stop codon is 3' adjacent ("shifty stops") [45,46J. A stop codon in the ribosomal A-site pauses ribosomes [471 which is presumed to facilitate the potential for codon: anticodon dissociation in the P-site, and following mRNA slippage, re-pairing of the anticodon to mRNA at the overlapping + 1 frame codon. Alterations of the levels of release factors, andlor addition of cognate (suppressor) tRNA, either in bacterial systems [46,48] or in rabbit reticulocyte Iysates [341 showed that tennination, +1 frameshiftlng, and readthrough are in direct competition. The most obvious relevant feature of stop codons is that they are slow-to-decode. In a- and b-globin mRNA, the time required to incorporate one amino acid to a growing peptide chain is about ten times faster than the time to ter­minate translation l49l. Certain rare sense codons in both Escherichia coli and Saccharomyces cerevisiae are also slow-to-decode as their corresponding aminoacyl tRNAs are either sparse or their anticodons have markedly less codon complementarity [501 Like stop codons, these sense codons are thought to stimulate + 1 frameshifting at 5' adjacent shift-prone codons by inducing ribosomal pausing [51,52], Previous work [23,53} dearly demonstrates that the anti­zyme mRNA ORFl stop codon, UGA, plays a key role in the frameshifting I'equired for ribosome entry to ORF2. We have shown here that not only is the ORFl stop codon critical for high level frameshifting, but it is also important for polyamine sensing. An interesting question is whether it is simply the slow-to-decode nature of stop codons or the termination process itself which is the important feature for polyamine stimulation off r~meshifting. Little to no polyamine stimulation was seen when the UGA codon was replaced by anyone of five sense codons, suggesting that the termination process is being affected. The choice of sense codons substituted in the present work was con­strained by the need to maintain U in the first position to preserve the potential for the P-site tRNA to re-pair to mRNA at the overlapping +1 frame codon. Although much less is known about the speed of decoding individual sense codons in mammalian systems than in E. coli (where marked differences have been found in even sense codons decoded by the same tRNA [54]), decoding rates are thought to be proportional to the frequency of codon usage. Based on the relative codon frequencies in highly expressed human genes l551 the leucine codon (UUA) and the serine codon (UeA) tested in the present work are the least frequently used codons for their respective amino acids. Both cysteine codons (UGC and UGU) were also tested and their relative abundance is 2-fold different, and like the sole codon for tryptophan (UGG), they are not amongst the most frequent codons. The lack of effect with these sense codoos suggests that a termination codon is required for significant polyamine stimulation of frameshifting. All known antizyme genes utilize a Similar shifty SlOp based mechanism off rameshifting for polyamine sensing indicating the importance of a stop codon in this process. 35 1486 LM. Petros et at / Biochemical and Biophysical Research Communications 338 (2005) 1478-1489 But elevated putrescine levels in an S. cerevisiae mutant increased Ty + 1 frameshifting (56) even though a shifty stop is not involved, suggesting that at least in this yeast polyamines may stimulate frameshifting at rare codons. It is possible that polyamines stimulate frameshifting by decreasing the rate of translation at stop codons and slow-to-decode codons, however, in mammalian systems only termination is slow enough to achieve the required frameshift levels. Of relevance to this issue is a recent observation in yeast that premature termination codons located distant from the normal 3' UTR cause aberrant termination (57J. It was shown that efficient termination of translation requires interaction between the terminating ribosome and a specific RNP structure which indudes poly(A}­binding protein and eRf3. Termination at stop cedons at the end of genes does not pause ribosomes or at least does not queue up ribosomes at the termination site(58J. Although it is unknown if a similar mechanism is at work in metazoans, the occurrence of the ORFl termina­tion codon far upstream of the 3' UTR may indicate that termination is altered relative to termination at stop codons found at the end of coding sequences. Inefficient or aberrant termination may contribute to the high levels of polyamine induced frameshifting observed at the anti· zyme shift site. In addition, antizyme frameshifting is responsive to interferon levels {59}. The anti-proliferative and anti-viral properties ofi nterferon are partly mediated by RNase l which associates with termination factor eRF3. Although the mechanism may be distinct from polyamlne regulation of antizyme, the involvement here of eRF3 further Illustrates that the levels off rameshifting at the antizyme shift site may be influenced by alterations in the termination process. The results presented show that elevated polyamlnes also increase readthrough of stop coOOns in non-antizyme contexts, as seen in earlier results in different systems [60,61 J. There is an optimal polyamine level for standard triplet decoding, though the precise level is not the same for different sequences (62-64i Above or below this opti­mal level it is likely that the ribosome is perturbed so that there Is reduced translational fidelity. Although there are no cognate tRNAs for stop codons, near~cognate tRNAs can read them inefficiently causing some read· through. A decrease in either tRNA discrimination, ter­mination efficiency, or both would tilt competition in the between .. 5' element NO PA (%) SPD F rameshlftlng 4.9 21.0 R eadthrougn 02 0,6 Termination 94.9 78.4 favor of near-cognate tRNA decoding (readthrough) in these cases, or readthrough and frameshifting in the anti­zyme context. The 5" element The 5 element is intriguing. It considerably increases frameshifting without elevated polyamine levels, and it is critical for the greatly enhanced frameshiftlng efficiency with polyamine addition (Table 1). If, as recently argued [65,66], antizyme programmed frameshifting involves anticodon detachment and re-pairing, then the conforma­tional change(s) presumed to be mediated by action of the 5' element promote one or more parts of the process in a manner that Is likely sensitive to occupancy of some secondary ribosome binding site for polyamines. Both the S. pombe [67J and mammalian (Matsufuji et at in prepa~ ration) 5' element sequences act at the mRN.A level rath· er than at their encoded product level, and at least the mammalian does so without involving intra-mRNA sec­ondary structure. By analogy with the effect on nternal Shine--Dalgarno sequences in E. coli pairing with the rRNA of translating ribosome to greatly boost ribosomal frameshifting {4S,68,691. the mammalian 5 element may also pair with the rRNA of translating ribosomes. This model cannot be tested the way its E. coli counterpart was, because of the multiplicity and distribution of mammalian rRNA genes. Nevertheless. there are a number of reports with suggestive evidence for mammalian mRNA pairing with components of the ribosome during initiation. One exam­ple of this pairing occurs during cap-independent transla­tion of the tobacco etch virus (TEV) mRNAs. where the internal ribosome entry site (IRES) is thought to bind to the 185 rRNA [701 In addition, direct binding was shown between the mouse Gtx homeodomain 5' untrans­lated region and the 18S rRNA [71 J. While it is likely that the SO element interacts with rRNA, either directly or indirectly through polyamines. the possibility oft nter­actions with the ribosomal proteins cannot be ruled out. Perhaps even more difficult than attempting to gain solid evidence for a putative mRNA:rRNA interaction is iden· tification of polyamine binding sites on mammalian ribo~ somes and trying to assess which may be the relevant one(s) for frameshift stimulation. It is not a task for the short-term future. and termination in the 2 context element NO PA (%) SPD (%) 1.3 1.7 0.1 0.4 98,6 97.9 The frameshifting values are from Fig. 2 and Fig. 5 . The readthrough values <IrE! from Fig. 6. +5 element are the antizyme 2 wHdtype constructs (AI2-wt. RT·wt), -S' element are the 5· deletion CQostfuct$ (S'del. RT-5 del), The percent terminati<m was cllkulated by subtracting the total per(ent hameshifting and readthrough from lOO. 36 LM. P~tros et aI-I Biochemk,al and Biophysical Re~ear(;h Communications 338 (2005) 147fH489 1487 Alignments of the known antizyme mRNA sequences 5' of the shift site (the (urrent total is over 100, LP. Ivanov. personal communication) indicate that the S· element is likely modular and that most ofi t only arose 150-200 mil­lion years ago [25]. Our results demonstrate that the shift site with a 3' stop codon alone can be a sensor of poly­amine levels (albeit very low levels) on whkh selection probably started operating at least 500 million years ago. These results, along with the above phylogenetic analysis, suggest that the 5' element is a refinement which increases the efficiency of sensing polyamines on the core shiftsrtel stop codon, and that the pseudoknot evolved to increase the overall levels off rameshifting. Acknowledgments We thank Drs. I. Ivanov and S. Matsufuji for ongoing discussions and their continuing support .• This work was supported by Grants from the National Institutes of Health, ROl GM71853 (RF.G'), J.F.A. was supported by the Science Foundation Ireland in the final stages of this work, and M.T.H. was supported by a development grant from the Muscular Dystrophy ASSociation. Appendix Au Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.10161 j.bbrc.200S.1 0.115 . References [11 A.C Childs, OJ. Mehta. E.W. Gerner. Polyamlne-dependent gene expreuion, Cell. Mol. Lire 50. 60 (2003) 1394-1406. [2] P. Coffino, Regulation of cellular poiyamines by antltyme, Nat. Rev. Mol. Cell BioI. 2 (2001) 188-194. [3] 5.5. Cohen, A Guide to the Polyamines. Oxford University Pmss, New York, 1998, [4J J. J<lnne. L Alhonen, M. Pietila, TA Keinamm, Genetic approaches to the o:l1ular functions of polyamlM$, In mammals. EUL J. Biochem. 271 (2004) sn-894. [5) H.M, Wallace. A.V. Fraser. A, Hughes, A pers~ve of polyamine metaboUsm. BJIXMm. J. 376 (2003) 1-14. [61 E.W. Gerner, F.l. Meyskens Jr. Potyamines and cancer: old molecules, new understanding, Nat Re\', Cancer 4 (2004) 781- 792. [7] X. li, P. Coflino. Degradation of ornithine decarboxylase: exposure of the C-terminal target by a polyamine-inducible inhibitory protein. Mol. Celt Blot 13 (1993) 2377-2383 .. [aJ y, Murakami. S. Matsufuji, T. Ka~jiK S, Ha.ya.shi. K, Igarashi. T, Tamura, K. Tanaka, A. khihara, Ornithine decarboxylase is degraded by the 265 proteasome without ubiquitination, Nature 360 (1992) 597~599. [9} M. Zhang. eM. Pickart, P. (offino, De(~rminant$ of proteasome: nKognition of ornithine de<:atboxy1a~. a ubiquitin-lndependent $ubstrate. EMBO 1 22 (2003) 1488~ 1496. nO] G. Asher. Z. Bercovlch, P. TsvetKolI, Y. Shaul, C. Kahana. 20S proteasomal degradation €If omithine decarboxylase is regulated by NQ01, Mol. Cell 17 (2005) 645-655. [111 Z. Bercovich. C Kahana. Degradation of antizyme inhibitor, an ornithine decarboxylase homologous protein, i~ ubiquitin-oependent and is Inhibited by antizyme, J. Blot. Chem. 27912oo4} 54097-54102 (12) lL Mitchell, G.G. Judd, A. Bareyaf-Leyser, S.Y. Ling, Feedback repression of polyamine transport is medhlted by antizyme In mamm<llian tisslR<u!ture cells, Bioc::hem.). 299 0994l19-22. [J 31 T. Suzukt Y. He, K. Kashiwilgi. Y. Murakami, S. Hayashi. K. Igarashi. Antlzyme protects a9all1$( abrnnmal accumulation and of polyamines in ornithine decarboxylase-overproducing cells, Nat!. Acad. Sci. USA 91 (1994) 8930··8934. [14J K. Sakata. T, Fukuchi·Shimogori, K. Kashiwilgi. K Igarashi. Identification of regulatory region of antizyme necessary for the negative regulation of polyamine transport, Biochem, Biophys. Res. Commun, 238 (1997) 415-419, [IS} K. Sakata. K. KashlwagJ, K. Igarashi, Properties of a polyamine transporter regulated by antizyme, Biachem. J. 347 (2000) 297-303. [161 R.M. Newman, A. MobaKher, U. Mangold, C. Koike, $, Diah. M. SChmidt, D. Finley, B.R. Zetter. Anti.~yme targets cydin 01 for d~~raljatiion. iii novel rn~hanism for cell gfOwth repression. J. 8iol. 279 C2004} 41504-41511. [171 t Ivanov. K. Simin, A. Letsou, J. Atkins. R. Gesteland. The Drosophila gene for antizyme requires ribosomal fral1le'$hifting for expression and contains an intronic gene tor snRNP Sm 03 on the oppoSite strand. Mol. Cell. BioI. 18 (1998) 1553-1561. [lS} c. Vied, N. Halachmi, A. Salzberg. 11. Horabln, Antizyme is a target of sex-lethal in the Drosophila germline and appears to act down­stream of hedgehog to regulate sex·h~thal and cyclin Il, De\'. BioI. 253 (2003) 214-229. (191 J.I... Mitchell, G.G. Judd, Antizyme: modification, affecting palyami~ homoeostasls, Biochem. Soc Trans, 26 (1998) 591-595. (20) N, Murai, y, Murakaml, 5. Matsufujl. Identification of nuclear export signals in antizyme-l. J. Bioi. Chem. 278 (2003) 44791-44798. [21) R.G. SChipper. V.M. CuijJ)tlrs, LH. De Groot. M. Thlo, AA Verhofstad. Intracellular Iocalizatlon of ornithine decalboxylase and its (~ulatQf)' pro«'in, antlzyme·t J. Histochem. (}'tochem. 52 (2004) 1259-1266. [22} LP. Ivanov. A. Rohl'Hasser, D.A. Te:rreros., R.F. Gesteland. LF. Atkins, Discovery of it spermatogenesis stage-specific ornithine de<arbo)(yllls~ antlzyme: antizyme ), Prot. N,ll!. Acad. Sci. USA 91 (2000j4806"4813. (23) S. Matsufuji. T. Matsufuji. Y. Miyazaki. Y. MUrakami, J. Atkins, R. Gesteland, S, Hayashi, Autoregulatory fmmeshifting in decod· ing mammalian ornithine decarboxylase antizyme, Cell 80 (1995) 51-60. [.24] Y. Miyazaki. S. M..m ufuji. S. Hayashi, Cloning and characterization of a rat gent' encoding ornithine decarboxylase antizyme. Gem: 113 119921191-197, [2S} LP. Ivanov, R,F. Gesteland, J.F. Atkins. Antlzyme expression: a subversion of tripl~ decoding, wlll<:h is remarkably w!lserved by evolution, is a Sensor for an autoregUlatory circuit, Nuclei<: Acids Re$. 28 (2000) 3185-31%, [26J M. Howard, B. Shirts, ), Zhou. C. Carlson. S. Mat5ufuji, R, Gesteland, R. Weeks, J. Atkins, Cell culture analysis of the regulatory frame~hift ev~nt required for the expre.~sion of mamma Uan lIntizyl1le's. Genes Cell 6 (2001) 931,941. [27} I.P. Ivanov, A..F, Gesteland, IF, Atkins, A se<:ond mammalian antizy~; conservation of programmed ribosomlll frameshifting, ~hOmks S2 (1998) n9~129. (28) I.P.lvanov, e.8. Anderson. RJ. Gesteland, J.F. Atkins. Identifica-tion of a new antizyme mRNA -11 frameshifting stimulatory pseudoknot in a sub~t of diverse invertebrates and Its apparent absence In Intermediate spedfi. J. Mol. BioI. 339 (2004) 495-504. [29] M. Zimmer, E. Sattel~rger, R.B. Inman. R. Calendar, M.J. Loes~ner, Genorne and ptoteome of listeria monocytogeM$, phage PSA: an unusual case for programl1le'd -t 1 translational frameshiftiog in structural protein synthesis, Mol. Mlcrobicl. 50 (2oo)} 303··317. flO} V.X, Feng. H. Yuan, A. Rein. J.G. levin. Bipartite signal for read­through suppression in mUrine leukemia virus mRNA: an eight­nucleotide purine-rich $equence im~iately downstream of the gag termination codon followed by an RNA pseudoknot.. ]. Virol. 66 (1992) 5127-5132. 37 1488 LM. Petr<)s et at! Biochemical and Biophysical Research Communkations .338 i2OO5/147S-1489 [31} I:.B. ten Dam, C.W. Pleij, t, Bosch, RNA pseudoknots: translational frameshlfting and reachhfough on viral RNAs, Virus Gene~ 4 (1990) 121-136. [321 N.M. Wills. R,F. Ger.teland. J.F. Atkins, Evidence that a downstream pseudoknot Is required for translational read-through of the Moloney murlne leukemia virus gag stop (odOl\ Proc, Nat!. Aca<i. Sci. USA 88 (1991) 6991-6995. [33} M. Orlova, A. Yueh, J. leung, S.P. Goff. Reverse transcriptase of Moloney murine leukemia virus binds to eukaryotic release factm 1 to modulate suppression of translational termination, Cell 115 (2003) 3HH3J. r341 Z.N. Karamys~va, AL Karamyshev. K. Ito. T. Yokogawa, K. Nishikawa, Y. Nakamura, S. MatsuflJji. Antizyme framesh1ftin9 as a functional probe of eukaryotic translational termination, Nucleic Adds Res. 31 (2003) Sg49-5956. [35j R. Palanimurugan, rt Scheel, K. Hofmann. R. Jurgen Dohmen. Polyaminfl regu~a~ their synthes!$ by indUCing expreSSion and blocking degradation or DOC antizyme, EMBO J. 23 (2004) 4857- 4867. [361 C. Zhu, D.W. Laflg, P. Coffina, Antizyme2 Is a negative regulator of ornithine decarboxylase and polyamine transport, J. BioI. Chem. 274 (1999)26425-26430. [37j 'f. Tosaka. H. Tanaka. Y. Vanc, K. Masai, M. Nozakl. It Yomogida. S. Otani, H. NoJima. Y. Nishimullf1. Identification and characterization of testis specific ornithine decarb0i<Ylase anmyme (OAZ-t) gene: expression in haploid germ (ells and polyamine· induced frameshifiing, Genes Ceils 5 (2000) 265-276. £381 M.T. Howard, G. Ag9iuwal, ca. Anderson. S. KhatrI. KM. Flanigan. J.f, Atkins, Receding elements located adjacent to a sub!K!t of eukaryal sef.enocysteine-speclfylng UGA todons. EMBO J. 24 (2005) 1596-1607. [39} M.T. Howard, a.H. Shirts. LM. Petro $, K.M. Flanigan, RF. Gesteland, U. Aikins, Sequence specificity of amln09lycoskle­induced stop coodon readthrough: potential implications for treat­ment of Duchenne muscular dystrophy. Ann. Neurol. 48 (2000) 164· 169. [401 G. Grentzmann, J.A. Ingram. PJ. Kelly, R.t. Gesteland. J.F. Atkins, A dual-Iudferase reporter system for studying recoolng signals, RNA 4 (1998) 419-486. [41J J.L Jacobs. ).0. Dlnmart, Systematic analysis of bicistronic reporter Quay data, Nucleic: Acids Res. 32 (2004) 160. [42} C.M .. Brown, PA. Stockwell, c.f\I. Trotman, W.P. Tate, Sequence analySiS suggests that tetra-nudeotldes signal the termination of protein synthesis 11'1 eukaryotes, Nudeic Acids Res. 18 (990) 6339- 6345. [43} K.1t M<:Caughan, eM. Brown. M.E. Oalphln, MJ. Berry, W.P. Tate. Translational termination efficiency in mammals is infl\renced by the base following the stop codon, Proe. Natl, Acad. Sci. USA 92 (1995) 5431-5435. [44j W.P. Tate, SA Manllefing. Three, four or more: the translational stop signal at length. Mol. MtClooiol. 21 (1996) 213-219. [451 R.S. Weiss. D.M. Dunn, J.F. Atkins. R.F. Gesteland, Slippery runs, shifty stops, backward steps, and forward hops: -2, -t,. +1, +2. +5. and +6 ribosomal frame~hiftiIl9, Cold Spring Harb. Symp. Quant. BioI. 52 (l987} 687-693. [461 R.S. Weiss, P.M. Dunn. J.F. Atkins. RF. Ges1E!land. Ribosomal franurshifting from -2 to +50 nudeotides, Prog. Nudeic Acid Res. Mol. Siol. 39 09901159-11:13. [47l S.l. WoUn, P. Walter, Ribosome pausing and stacking during trao$lilltion of a eukalyotic mRNI\ EMBO J. 7 (1988) 3559-3569. [481 F.M. Adam$ki, B.C. Donly, W.P, Tate, Competition betweef1 frallUtShiftiog, termination and suppression at the frame shift site in the Escherkhia coli release factor-2 mRNA, Nucleic Acids Res. 21 (1993) 5074-5078. [49) H.F. Lodish. M. Jacobsen, Regulation of hemoglobin synthesis. Equal rates of tfan~lation and termination of a- and b-globln chains, 1. BioI. Chem. 241 (1972) 3622-3629. [SOl C. Gurthne, J. Abel$Qf\. Organization and expression of tliNA genes In Sacchatomyc~s cerevislaf: In: J.N. Strathern, E.W. Jones, JJt Broach (Ed:>.), The Molecular Biology of the yeast 5accharomYC1ts: Metabolism and Gene Expression, ColdSpring Halbor laboratory Preu, Cold Spring Harbor, NV, 1982, pp, 487-528. (51) M.F. Bekourt, P.J. Farabaugh. Ribosomal frameshifting in the yeast retrotransposon 1y: tRNAs induce slippage on a 7 nudeotide minimal site, Cell 62 (1990) 339-352. [52} A. Sundararajan, W.A. Michaud, Q. Qian. G. Stahl. PJ. Farabaugh, Near-cognate peptldyl-tRNAs promote +1 programmed translational framesMting in yeast, Mol. Cell 4 {1999) 1005-1015. [53} E. Rom. C. Kahana. Po1yamines regulate the expressIon of omithlne decarboxylase antizynur in vitro by Inducing ribosomal frame­shifting, Proe. Natl. Aead. Set USA 91 (1994) 3959-3963. [54} M.A. Sorensen. S. Pedersen, Absolute in vivo translation fates of individual eodem In Escl~ichia coli. The two glutamic acid codons GAA and GAG are translatEM:! with a threefold difference III raw, J. Mol. BioI. 222 (1991) 265-280. [55) A. Cid·Arrt!9ui, V. Juarez, H. lur Hausen. A synthetk E7 gene of human paplUomavirus type 16 that yields enhanced expression of the protein in mammalian cells and Is useful fOf DNA immunization studies, J. Vito!' 71 (2003)4928-4937. [56) D. Salasundaram, 10. Dinman. C.W. Tabor. H. Tabor, SPEl and SPE2: two e»ential genes in the biosynthesis of polyamines that modulate + 1 ribosomal frameshifting In Saccharomyces eerevisi~ J. BacterioL 176 (1994) 7126-7128. [5'] N. Amranl, It Ganesan. S. Kervestin, OA Mangus. S. Gh¢sh. A. Jawb!'.ort, A faux 3 o-UTR promotes aberrant termination and triggefs noos~mse·mediatEM:! mRNA decay, Nature 432 (2004) 112-118. [581 Y. Arava, F.E. Boas. PD. Brown. O. Herschtag. Dissecting eukary· otic translation and its control by ribosorM den1ilty mapping, Nude!( Adds Res. 33 (2005) 2421~2432. [591 F. Le Roy, T. Salehzada, C. Bisbal, J.P. Dougherty, S.w. Peitz, A newly discover4d function for RNase L In regulating translation termination, Nilt. Struct. Mol. Bioi. 1:2 (2005) 505·512. tOOl M.M. l'iryniewicz, R.A. Vonder Haar, Polyamines enhance read" through of the UGA termination codon in a mammalian llUtS$enger RNA. Mol, Gen. Genet 190 (\983) 336-343. (61) M. Yoshida, K. Kashiwagi, G. Kawaii A. Ishihama, K. Igarashi, Polyamines enhan<:e synthesis of the RNA polymerase sigma 38 subunit by suppression of an amber termination codon in the open r~jng frame, J. BioI. Chem. 277 (2002) 37139-31146. [62} J. Atkins. J. lew!!'., C. Anderson. R. Gesteland, Enhanced differential synthesis of proteins in a mammalian cell-free system by addition of PQlyamloes, J. BioI. Chem. 250 (1975) 5688-5695. 163} K. Igarashi. K. Kas.hiwagl, R. Aoki. M. Kojima, S. HirOiE', Comparative studies on the increase by polyamioes of fidelity of protein ~ynthesls in Escherichia coli and wheat germ cell-free systems. Biocht'm. Biophys. Res. Commun. 91 (1919) 440-448. [641 CWo Tabor, H, Tabor. Polyamines, Annu. Rev. Biochem. 53 {19M} 749-790, (65) P.V. Baranov, R.f. Gt'.steland, IF. Atkins, P·site tRNA is a crucial initiator of ribosomal frameshlftlng. RNA 10 {2004l 221-230. [66} I.P. Ivanov, 0.1. Gurl/Ieh, R.F. Gesteland. J.F. Alkins, Recoding: Site- or mRNA-Specific Alteration of GeMtic Readout Utilized fOf Gene Expfession. in; J. lapointe, L. Srakier-Gingras (Eds.). Tran.~­lation Mechanisms, Eurekah.com/Landes Bioscience. Georgetown, TX. 2003, pp. 354-369. [6711.P.lvan()v, R.F. Gesteland, S. Matsutuji, J.F. Atkins, Programmed frameshtfting in the synthesis of mammalian antlzyme is + 1 in mammals, predominantly + 1 in fission yeast, but ·2 in budding yeast, RNA 4 (1998) 1230-123S. [68) 8. Larsen, N,M. Wills, R.F. Gesteland, J.F. Atkins, rRNA-mRNA ba>e pairin9 stimulates a pr09fammed ·1 ribosomal frameshift J. Bacteriol. 176 (1994) 6842-685L [691 R.B, Weiss. D,M. Dunn, A,E. Dahlberg. J.F. Atkins. R.F. Gesteland. Reading frame switch caused by base-pair formation between the i 38 LM. Petros et al. I Biochemical and Biophysical Research Communi<ations 338 (2(05) 1478-1489 1489 end of 16S rRNA and the mRNA during elongation of protein $ynt~5i$ in Escherichia (:011. EMBO 1. 7 (1988) 1503-1507. l70.1 V, Zeeoko. D,R, Gallie, Cap-independent translation of tobacco Etch virus is conferred by an RNA pseudo knot in the 5 '-leader, ). BioI. Chem, 280 (2005) 26816-26824, [711 M.e. Hu, p, Tranque. G.M. Edelman, V,P. Mauro, rRNA-comple­tnentarlty in the 5' untranslated region of mRNA spedfylng the Gtx homeodomain protein; evidence that ba>e-- pairing to HIS rRNA affms translational efficiency, Proc, Nat!. Acad, Sci. USA 96 Om) 1.339-1344, 39 Supplementary material Table 2.2. The sequence of the top strand oligonucleotide for the test constructs. SCt SC2 SC3 SC4 SC5 AZ1·wt SC6 SC7 SC8 SC9 SC10 BS-wt AS-wt BS-UUA AS..tJUA stop1 CCCAGCTC TCGACGCCCCAGCTCCAGTGCTGCAG G TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGTGGTGCTCCTGGTGCCCCTCACCCACTGTCGAAGATCCCCGGTGGG CGAGGGGGCGGCAGGGATCCTTCTCTCTCAGCTG TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGTGGTGCTCCTGTTGCCCCTCACCCACTGTCGAAGATCCCCGGTGGGC GAGGGGGCGGCAGGGATCCTTCTCTCTCAGCTG TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGTGGTGCTCCTGCTGCCCCTCACCCACTGTCGAAGATCCCCGGTGGG CGAGGGGGGGGCAGGGATCCTTCTCTCTCAGCTG TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGTGGTGCTCCTCATGCCCCTCACCCACTGTCGAAGATCCCCGGTGGGC GAGGGGGCGGCAGGGATCCTTCTCTCTCAGCTG TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGTGGTGCTCCTTATGCCCCTCACCCACTGTCGAAGATCCCCGGTGGGC GAGGGGGCGGCAGGGATCCTTCTCTCTCAGCTG ~ TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTGATGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTGGTGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTGTTGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCrrCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTGCTGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTCATGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTTATGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGATCCTGATGCCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTGATAGCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGICAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGATCCTTATGCCCCTCACCCACCCCTGAAGATCCCAGGIGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTCCTTATAGCCCTCACCCACCCCTGAAGATCCCAGGTGGG CGAGGGAATAGTCAGAGGGATCACAATCCTTCAGCTG TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGTGGTGCTCCTAGTGCCCCTCACCCACTGICGAAGATCCCCGGTGGG CGAGGGGGCGGCAGGGATCCTTCTCTCTCAGCTG TCGACGCCCCAGCTCCAGTGCTGCAGGCACATTGTTCCAGGGCCTCTGIGGTGCTCCTAATGCCCCTCACCCACTGTCGAAGATCCCCGGTGGGC CTTCTCTCTCAGCTG ACTGTCGAAGATCCCCGGTGGGCGAG G C GCAGGGATCCTTCTCTCTCAGCTG AG C CAACCC TCACCG CHAPTER 3 SCREENING METHODS TO IDENTIFY POLYAMINE ANALOGS WHICH INDUCE ANTIZYME FRAME­SHIFTING, DEPLETE NATURAL POLYAMINES, AND INHIBIT CELL GROWTH Abstract Polyarrlines are polycations that playa vital role in normal cell growth. Nun1erous studies have correlated elevated polyamine levels with abnormal or rapid growth. One therapeutic strategy to treat diseases that have increased cellular proliferation rates, most obviously cancer, has been to identify compounds which lower cellular polyamine levels. An ideal target for this strategy is the protein antizyme; a negative regulator of polyamine biosynthesis and import, and a positive regulator of polyamine export. In this study, two tissue­culture assays have been optimized in 96-well format to allow the rapid high­throughput screening of a 7SD-member polyamine analog library for compounds that induce antizyme expression and inhibit cell growth. After the initial screening steps, six analogs (MQTPA 1-6) were further characterized for uptake efficiency, and effects on natural polyamine levels and enzymes involved in polyamine metabolism. These six analogs were as or more effective franleshifters than spermine at low concentrations making them potential drug candidates. Introduction 41 Polyamines (putrescine, spermidine, and spermine) are essential for cell growth and cell cycle progression (reviewed in Oredsson, 2003), and intracellular levels are highly regulated through complex feedback mechanisms involving synthesis, catabolism/interconversion, import, and export (Figure 3.1, reviewed in Thomas and Thomas, 2003; Wallace et aI., 2003). Based on the observation that increased polyamine levels are correlated with neoplastic growth (reviewed in Pegg, 1988; reviewed in Cohen, 1998), it has been proposed that reducing or depleting intracellular polyamine levels would be an appropriate therapy, or preventative, for diseases involving undesired cell proliferation, such as cancer. Early efforts at modulating polyamine levels focused on reducing synthesis through inhibition of the major polyamine metabolic enzymes, ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (SAMDC). Difluoromethylornithine (DFMO) is an irreversible inhibitor of ODC (Metcalf et aI., 1978) that induces growth arrest and decreases the intracellular content of putrescine and spermidine in normal and malignant cells (Marton and Pegg, 1995). Despite its early promise, DFMO had slow uptake and rapid excretion from the body (Grove et aI., 1981), lacked the ability to reduce spermine levels, and led to an up-regUlation of polyamine import (Seiler et aI., 1996). Since extracellular polyamines are readily available from our diet and gut flora, S-adenosylmethionine (SAM) I SAMDC f--. CO2 Methytthioadenosine Decarboxylated SAM Methylthioadenosine 42 L-Ornithine IODCI-@ h-. CO2 ~ Putrescine Spermidine Spermine r----__. Putrescine I \PAO Ntacetylspermidine ~ __ ---, IssAT I Spermidine I '{A 0 N 1_ acetylspermine Iss AT I""--s-pe-r-m-nj -e-"'I Putrescine -+-+-+N 1-acetyl-spermidine Figure 3.1. The polyamine metabolic pathway. SAMDC = S-adenosylmethionione decarboxylase, ODC = ornithine decarboxylase, AZ = antizyme, PAO = polyamine oxidase, SSAT = spermidine/spermine N1-acetyltransferase, and SMO = spermine oxidase. 43 inhibition of polyarrJine synthesis with DFMO alone has shown limited chemopreventive (Alberts et aI., 2000; Fabian et aI., 2002) and therapeutic success (Abeloff et aI., 1986; Pegg et aI., 1995). Likewise, SAM486A (CGP48664), a potent inhibitor of SAMDC (Stanek et aI., 1993; Regenass et aI., 1994), showed rrlixed results in phase II clinical trials (Pless et aI., 2004; Millward et aI., 2005). As with DFMO, growth inhibition by SAM486A can be prevented by exogenous spermidine or spermine (Regenass et aI., 1994). Clinical success with these and other polyamine biosynthetic enzyme inhibitors is likely to be compromised by the up regulation of polyamine transport and the compensatory responses by enzymes involved in the homeostatic control of ce"ular polyamine levels. Combination therapies with more than one enzyme inhibitor have also been developed to affect multiple polyamine maintenance pathways. For example, polyamine transport inhibitors used in conjunction with DFMO have shown promise in cell culture studies and in tumor inhibition studies in nude mice (Devens et aI., 2000; Weeks et aI., 2000). Similarly, polyamine analogs have been designed with the intent of affecting multiple targets in the polyamine pathway. The polyamine analog N1 ,N11 -diethylnorspermine (DENSPM) induces the catabolic enzyme spermidine/spermine N1-acetyltransferase (SSAT), down regulates ODC and SAMDC, suppresses polyamine uptake, and inhibits cell growth (reviewed in Casero and Woster, 2001). While preclinical trials have shown promise for DENSPM as an anticancer drug (Sharma et aI., 1997; Schipper et aI., 2000), clinical trials uncovered gastrointestinal toxicities that compromised the dosing and selectivity (Hahm et aI., 2002; Wolff et aI., 2003). 44 Another strategy to productively alter polyamine homeostasis has been to enhance the cell's innate ability to reduce polyamine levels through induction of antizyme. Antizyme serves as the cell's sensor and regulator of polyamine levels, maintaining homeostasis and preventing toxicity from excess polyamines. Antizyme binds ODC and targets it for proteasomal degradation (Murakami et aI., 1992; Li and Coffino, 1993; Zhang et aI., 2003), inhibits polyamine import (Mitchell et aI., 1994; Suzuki et aI., 1994; Sakata et aI., 1997), and enhances polyamine export (Sakata et aI., 2000). A number of studies have looked at both transient and inducible over-expression of antizyme in cell lines and animal tumor models (Murakami et aI., 1994; Iwata et aI., 1999; Feith et aI., 2001; Tsuji et aI., 2001). The results demonstrate that deliberate antizyme induction can effectively inhibit cell growth and reduce tumor formation in mice. A family of three human antizyme genes has been identified. Antizyme 1 (Miyazaki et aI., 1992) and 2 (Ivanov et aI., 1998; Zhu et aI., 1999) show nearly ubiquitous tissue expression whereas antizyme 3 (Ivanov et aI., 2000b; Tosaka et aI., 2000) is only expressed in developing male germ cells. All three antizymes use the same unique method of protein expression that requires + 1 translational frameshifting (Ivanov et aI., 2000a). mRNA sequences 5' and 3' of the frameshift site (UCC UGA) have been identified as important stinlulators of the frameshift event in both antizyme 1 and 2 (Matsufuji et aI., 1995; Ivanov et aI., 2000a). In addition to these cis-acting mRNA sequences, frameshifting is stimulated by 45 polyamines, thus creating the autoregulatory circuit for controlling intracellular polyamine pools. The mechanism by which polyamines stimulate frameshifting is unknown, although the regulatory 5' sequence and the ORF1 stop codon, but not the 3' sequence, are required (Howard et aI., 2001; Petros et aI., 2005). In the first published assessment of antizyme induction using polyamine analogs, a class of oligoarnines was identified that could induce antizyme up to two-fold better than spermine (Mitchell et aI., 2002). Four classes of analogs (total of 24 analogs) were characterized for antizyme induction and cell growth inhibition. Despite wide variation in cell growth inhibition after 3 days with the analogs, there was a general correlation with level of antizyme induction, with the high antizyme inducers showing potent and rapid growth inhibition. High throughput screens have proven to be a valuable and essential tool for chenlical library screening. We have developed two 96-well assays as a primary screening strategy to identify additional antizyme-inducing molecules in a 750-member polyamine analog library. A cellular antizyme frameshifting reporter assay (Grentzmann et aI., 1998; Howard et aI., 2001) was adapted to a 96-well format providing a method to screen for antizyme frameshift-inducers. A growth­rescue assay, also in a 96-well format, was used to identify those molecules that have limited ability to function like natural polyamines in cell growth. Compounds were identified that were more potent antizyme frameshift inducers than spermidine, with little ability to substitute for polyamines in maintaining cell growth. Cells treated with six of the lead compounds showed reduced polyamine levels, reduced ODC activity, and were growth inhibited without induction of SSAT or inhibition of SAMDC, signifying that endogenous antizyme was being induced. These results illustrate a highly effective library screening strategy for identifying small molecule antizyme inducers for therapeutic purposes. Results Screening for frameshift stimulators 46 Compounds in a 750-member polyamine analog library were screened for their ability to induce antizyme frameshifting using a 96-well dual luciferase reporter assay in cultured mammalian cells (Grentzmann et aI., 1998; Howard et aI., 2001). Cells were transiently transfected with p2Luc reporter plasmids. This reporter contains the Reni/la luciferase gene separated from the firefly luciferase gene by a multiple cloning site. The Renilla luciferase protein acts as an internal control for transfection efficiencies, translation initiation, and mRNA stability (Grentzmann et aI., 1998). The portion of the human antizyme 1 gene known to be important for frameshifting was cloned between the two reporter genes such that the firefly gene is in the +1 frame (to monitor frameshifting) relative to the Renilla luciferase gene (Howard et aI., 2001). The amount of frameshifting is then determined by comparing the ratio of firefly to Renilla luciferase activity in the test (+1) construct normalized to a control construct in which the firefly gene is in the zero frame relative to the Renilla gene. HEK293 cells were incubated with 2.5 mM DFMO for 2 days prior to transfection of reporter plasm ids and addition of polyamine analogs in order to deplete polyamine levels and reduce baseline levels of endogenous antizyme. 47 Previous studies on antizyme induction have shown that the natural polyamines (putrescine, spermidine and spermine) have the ability to induce antizyme frameshifting with a potency order of spermine> spermidine> putrescine (Matsufuji et aI., 1995; Howard et aI., 2001). Using this assay, spermine and spermidine showed a similar result inducing frameshifting to 34% and 29%, respectively (Table 3.1). Spermidine was chosen as the internal control for all screening experiments. Initial screening of compounds was done at 25 IJM, and the percent frameshifting relative to 25 IJM spermidine, % Relative Frameshifting (%RF), was calculated. The well-characterized molecules agmatine and DENSPM were also tested in our screening assays. While agmatine is known to stimulate antizyme frameshifting (Satriano et al., 1998), relative to spermidine it is a poor antizyme frameshifter (290/0 RF) when tested at the same concentration (Table 3.1). The polyamine analog DENSPM is recognized as a potent SSAT inducer as well as an inhibitor of ODC and SAMDC (reviewed in Casero and Woster, 2001). As has been demonstrated previously (Mitchell et aI., 2002), DENSPM also induces antizyme frameshifting (1190/0 RF, Table 3.1). Out of the 750 compounds screened for antizyme frameshifting, 104 (140/0) showed ~80% RF (data not shown). MQTPA1-6 had relative framesbifl:ing values of 106-135% (Table 3.1). These analogs were chosen for further analysis based on the growth rescue screen described below. 48 Table 3.1. Structure, frameshifting, and growth rescue of the natural polyamines and select analogs. °/oFS is the percent frameshifting, %RF is the percent relative frameshifting to 25 IJM spermidine, and GRR is the growth rescue ratio of a 6-day assay with 2.5 mM DFMO. Name Spermidine Spermine Agmatine DENS PM MTQPA1 MTQPA2 MTQPA3 MTQPA4 MTQPA5 MTQPA6 Structure H H.. N............-.N~N"H H H H t;t I;i 'N~N~~N .. H H R 25 uM Screening %FS GRR 290/0 3.4 340/0 1.68 1.61 1.28 1.30 0.89 0.85 0.95 1.29 49 Screening for growth rescue The structural features of the natural polyamines that are essential for growth and antizyme frameshift induction are unknown, but are likely different. The target of polyamines in antizyme induction could be the mRNA, the ribosome, or other molecules. Polyamines most likely playa role in growth regulation through multiple targets, affecting transcription, translation, mRNA splicing, chromatin condensation, and apoptosis (Cohen, 1998; Coffino, 2001; Childs et aI., 2003). To be therapeutically useful, a polyamine analog must activate antizyme induction without functionally replacing the natural polyamines in growth regulation. Compounds were screened for the inability to rescue HEK293 cells that were growth inhibited by the polyamine synthesis inhibitor, DFMO. DFMO shows a growth inhibition IC50 of 150 IJM in a 3-day assay on HEK293 cells (Figure 3.2A). When 1 J.JM spermidine was included in the media, growth inhibition due to polyamine reduction was completely prevented because of uptake of exogenous spermidine (Figure 3.28). A half-maximum rescue for spermidine was found to be 0.37 J.JM. To screen compounds for their inability to substitute for the natural polyamines in growth, HEK293 cells were incubated with 2.5 mM DFMO and 25 J.JM compound for 6 days. A growth rescue ratio (GRR), the number of living cells with a given compound and DFMO compared to the number of living cells with DFMO alone, was determined. Growth with DFMO alone was given a GRR of 1.0. Cell growth with DFMO plus 25 J.JM spermidine gave a GRR of A 8 ....r...: :-e.- ~c ,-0 (!)O tot- =0 (1) u?f-!. 100 75 50 25 0 0 • (+) SPO, 1 J..1'M" ...... ". ...... .. ....... (-) SPO 3 10 30 100 300 10003000 OFMO Cone. (J.1M) 125m ; .... .r::. ~ 100 .. · .. · .. ~ ~ 75; "- 0 (!) 0 = '0 50 Q) u?f_!. 25 o '. -..... • (+) DFMO, 2.5 mM ...•... (-) OFMO o 0,1 0.3 1 3 10 30 100 SPO Cone. (J.1M) Figure 3.2. Three-day growth assays on HEK293 cells. (A) HEK293 cells were incubated for 3 days with 1 J.1M spermidine (SPO), 1 mM AG, and a titration of OFMO. (8) HEK293 cells were incubated for 3 days with 2.5 mM OFMO, 1 mM AG, and a titration of SPO. Cell growth was determined by colormetric MTS assays. 00 values are the averages from triplicate wells. The figure is representative of two experiments. 50 51 approximately 3.4 (Table 3.1). If the drug had the desired effect of no rescue, then it had a value close to DFMO alone (1.0). If the value was less than one, then the compound was growth inhibitory. If the value was zero, the compound was cytotoxic. Compounds that rescued cells from DFMO-induced growth inhibition had a GRR greater than 1.0 and were less desirable. Of the 104 compounds that had shown the desired >80% RF, only 18 (170/0) had acceptable GRRs of 0.5-1.5 (data not shown). This screening step was critical in identifying compounds that were not simply mimics of natural polyamines. Out of 18 lead compounds, six compounds with similar structures (MQTPA 1-6) were selected for further characterization. MQTPA 1, MQTPA2, and MOTPA6 had GRR of about 1.3 (Table 3.1). In contrast, MOTPA3-5 had GRR values less than 1.0, suggesting not only are these compounds unable to substitute for the natural polyamines, but they also further inhibit cell growth. Both agmatine and DENSPM could minimally growth-rescue DFMO-treated HEK293 cells (GRRs of 1.68 and 1.61, respectively). These GRRs are greater than those of MOTPA1-6, but much less than the GRR of spermidine. Frameshifting potency and growth rescue profile MQTPA 1-6 were tested at lower concentrations to determine frameshifting potency. Figure 3.3 compares the %RF of the six lead compounds and spermidine at 0.1-100 jJM. Three of the six compounds (MQTPA1-3) showed frameshifting greater than spermidine at 0.1-1 jJM. The other three compounds (MQTPA4-6) showed a concentration-dependent frameshifting pattern very