Molecular Microbiology (2003)

49

(6), 1627–1637 doi:10.1046/j.1365-2958.2003.03656.x

© 2003 Blackwell Publishing Ltd

Blackwell Science, LtdOxford, UKMMIMolecular Microbiology 1365-2958Blackwell Publishing Ltd, 200349

616271637

Original Article

Tetracycline-aptamer-mediated gene regulationS. Hanson et al.

Accepted 4 June, 2003. *For correspondence. E-mailbsuess@biologie.uni-erlangen.de; Tel. (

+

49) 9131 852 80 85;Fax (

+

49) 9131 852 80 82.

Tetracycline-aptamer-mediated translational regulation in yeast

Shane Hanson,

1

Karine Berthelot,

2

Barbara Fink,

1

John E. G. McCarthy

2

and Beatrix Suess

1

*

1

Lehrstuhl für Mikrobiologie, Friedrich-Alexander Universität Erlangen-Nürnberg, Staudtstrasse 5, 91058 Erlangen, Germany.

2

Posttranscriptional Control Group, Department of Biomolecular Sciences, University of Manchester Institute of Science and Technology (UMIST), PO Box 88, Manchester M60 1QD, UK.

Summary

We describe post-transcriptional gene regulation inyeast based on direct RNA–ligand interaction. Tetra-cycline-dependent translational regulation could beimposed via specific aptamers inserted at two differ-ent positions in the 5

¢¢¢¢

untranslated region (5

¢¢¢¢

UTR).Translation

in vivo

was suppressed up to ninefoldupon addition of tetracycline. Repression via anaptamer located near the start codon (cap-distal) in the5

¢¢¢¢

UTR was more effective than repression via a cap-proximal position. On the other hand, suppression ina cell-free system reached maximally 50-fold and wasmost effective via a cap-proximal aptamer. Examina-tion of the kinetics of tetracycline-dependent transla-tional inhibition

in vitro

revealed that preincubation oftetracycline and mRNA before starting translation lednot only to the fastest onset of inhibition but also themost effective repression. The differences betweenthe behaviour of the regulatory system

in vivo

and

invitro

are likely to be related to distinct properties ofmRNP structure and mRNA accessibility in intact cellsas opposed to cell-extracts. Tetracycline-dependentregulation was also observed after insertion of anuORF sequence upstream of the aptamer, indicatingthat our system also targets reinitiating ribosomes.Polysomal gradient analyses provided insight into themechanism of regulation. Cap-proximal insertioninhibits binding of the 43S complex to the cap struc-ture whereas start-codon-proximal aptamers interferewith formation of the 80S ribosome, probably by block-ing the scanning preinitiation complex.

Introduction

The past decade has seen advances in RNA targeting thathave generated a new field of drug target discovery andRNA therapeutics. Relatively straightforward knock-downtechnologies such as antisense, ribozymes and RNAi areincreasingly being applied in preference to standardknock-out strategies (Opalinska and Gewirtz, 2002; Lori

et al

., 2002). Another promising strategy is the use ofsmall molecule-binding RNA aptamers that exhibit highlyspecific molecular recognition (Sucheck and Wong, 2000).Recently, studies of naturally occurring, small molecule-sensing RNA-dependent gene regulation systems inprokaryotes have been reported. In

E. coli

, the

btuB

5

¢

UTR interacts directly with coenzyme B

12

, resulting in astem–loop structure that includes the Shine-Dalgarnosequence and the start codon, thereby inhibiting ribosomebinding (Nahvi

et al

., 2002). The

btuB

gene encodes thereceptor that transports B

12

into the cell, and the RNA–B

12

interactions function as part of a feedback loop to regulatetransporter synthesis. In another study, the 5

¢

UTRs of

thiM

and

thiC

mRNAs were shown to bind directly to thiamine(vitamin B

1

), thereby functioning as a ‘riboswitch’ thatinhibits synthesis of the thiamine biosynthetic enzymes(Winkler

et al

., 2002a). A similar mechanism has beenproposed for TPP-controlled thiamine synthesis in

Rhizo-bium

spp., in which the TPP–RNA complex interferes withtranslational initiation (Miranda-Rios

et al

., 2001; Stormoand Ji, 2001). However, in

Bacillus

spp., interactionbetween a conserved region in the 5

¢

UTRs of the thiamine(

thi

-box) and riboflavin (

rfn

-box) operons with TPP andFMN, respectively, results in formation of a terminationhairpin that inhibits transcription (Mironov

et al

., 2002;Winkler

et al

., 2002b). Thus, feedback regulation of boththiamine and riboflavin biosynthesis is mediated by smallmolecule-sensing mRNA. Another report describesmodulation of translation by the 5

¢

end of prokaryoticmRNAs with response to environmental stimuli such astemperature shifts or nutrient concentrations (Narberhaus,2002). CspA, an RNA chaperone in

E. coli

, has beenshown to bind RNA without any sequence specificity inorder to facilitate translation at low temperatures by pre-venting the formation of secondary structures (Jiang

et al

.,1997).

A novel approach has been to exploit ligand-bindingRNA aptamers for the design of a small molecule-depen-

1628

S. Hanson

et al.

© 2003 Blackwell Publishing Ltd,

Molecular Microbiology

,

49

, 1627–1637

dent gene regulation system (Werstuck and Green, 1998).Because the stringent selection procedure is based ontight ligand–aptamer association, the 3D structures of theselected aptamers represent optimized targets for ligandbinding (Ellington and Szostak, 1990; Robertson andJoyce, 1990; Tuerk and Gold, 1990). Upon binding, theligand becomes an integral part of the RNA structure, thusadopting a unique conformation (Patel, 1997; Patel

et al

.,1997; Hermann and Patel, 2000). If RNA aptamers areintroduced into the 5

¢

UTR they can result in translationinitiation being rendered subject to control via theaptamer–ligand-complex. Reports of aptamer-basedtranslational control

in vitro

(Werstuck and Green, 1998)and

in vivo

(Werstuck and Green, 1998; Grate and Wilson,2001; Harvey

et al

., 2002) have shown that this type ofcontrol can be imposed on translation. We have generateda tetracycline-binding aptamer by

in vitro

selection whichexhibits affinity to tetracycline in the low micromolar range(Berens

et al

., 2001). Tetracycline and its derivativesserve as excellent modulators for use in gene regulationbecause they are non-toxic and efficiently penetrate mostcell types. Indeed, they have been used successfully in awide range of eukaryotic systems to conditionally controltranscription (Forster

et al

., 1999; Favre

et al

., 2002;Lamartina

et al

., 2002). Interaction of tetracycline withRNA has been studied in detail in the context of studiesof tetracycline resistance in bacteria and ribozyme inhibi-tion (Berens, 2001).

Translational regulation constitutes an important pointof post-transcriptional control of gene expression,enabling the cell to change rapidly the level of gene prod-uct (Mathews

et al

., 2000). Translation is mainly regulatedat the step of initiation (Yoon and Donahue, 1992; Kauf-man, 1994; Gray and Wickens, 1998; McCarthy, 1998).The cap structure at the 5

¢

end of the mRNA facilitates therecruitment of the 43S preinitiation complex (Sonenberg,1994). This complex, along with its associated initiationfactors (eIFs), is thought to scan the entire length of the5

¢

UTR in search of a start codon in a suitable context(Kozak, 2002). Hence the presence of secondary struc-ture elements such as stem–loops or hairpins in the5

¢

UTR negatively affects translation initiation (Cigan

et al

.,1988; Baim and Sherman, 1988; Bettany

et al

., 1989).Moreover, it is known that the extent of inhibition is afunction of the stability and position of the structured ele-ment in the 5

¢

UTR (Oliveira

et al

., 1993b; Vega Laso

et al

.,1993). The presence of short uORFs can also regulatetranslation of the main ORF (Meijer and Thomas, 2002).For example, the uORF located in the 5

¢

UTR of the yeast

CPA1

gene encodes a peptide that represses translationof the downstream ORF (Werner

et al

., 1987). On theother hand, the four uORFs of

S. cerevisiae GCN4

areinvolved in translational regulation of this gene in a mech-anism that does not rely on the specific sequence of an

encoded peptide (Hinnebusch, 1997). The presence ofnonsense mutations in the 5

¢

UTR can induce nonsensemediated decay (NMD) of mRNAs (Oliveira and McCarthy,1995; Gonzalez

et al

., 2001). Both uORFs in the

YAP2

mRNA of

S. cerevisiae

accelerate the decay of the down-stream coding region; this destabilization is potentiated bystable secondary structure 3

¢

of the uORF stop codon(Vilela

et al

., 1999). Thus, a number of types of

cis

-actingelements in the 5

¢

UTR can influence the translation orstability of an mRNA.

Up to now, there has been no mechanistic study of anaptamer-ligand based post-transcriptional regulatory sys-tem in yeast. In the present study, we describe a tetracy-cline-dependent translational control system in

S.cerevisiae

based on ligand binding to an aptamer in the5

¢

UTR. The quantitative degree of regulation that can beachieved by this system exceeds that of

GCN4

. We havecharacterized how the regulatory properties of this systemdepend on the position of the aptamer relative to the startcodon and to the mRNA cap. We have also analysed theeffect of aptamers on reinitiation. Our results suggest ageneral model for tetracycline-aptamer-mediated regula-tion of gene expression.

Results

Effect of aptamer insertion site on regulatory properties and mRNA levels

We inserted the tetracycline-binding aptamer 32sh(Fig. 1D,F), which is a shortened variant of the

in vitro

selected aptamer cb32 (Berens

et al

., 2001), into the5

¢

UTR of a constitutively expressed firefly luciferase genelocated on the plasmid YCp22FL (Fig. 1A; Oliveira

et al

.,1993a). The aptamer was inserted either in a cap-proxi-mal position (9 nucleotides from the cap site using

Afl

II,pWHE711), or near the start codon (5 nucleotides preced-ing the start codon using

Xho

I, pWHE721; Fig. 1B). Theresulting 5

¢

UTRs had a length of 109 nucleotides(Fig. 2A). The control plasmid pWHE702 contains twocopies of a CAA-spacer described to be non-structured(Fig. 1C) to adjust the length of the 5

¢

UTR (108 nucle-otides). The resulting plasmids were transformed into

S.cerevisiae

RS453. Luciferase activity for all constructswas measured in the absence and presence of 100

m

Mtetracycline. In order to rule out the possibility of aptamerinsertion affecting mRNA stability, steady-state mRNA lev-els were determined. The results are summarized inTable 1 and displayed in Fig. 2.

The construct with the cap-distal aptamer insertion(pWHE721) showed a ninefold decrease in luciferaseactivity in the presence of tetracycline. In comparison, thecap-proximal aptamer imposed a threefold decrease inluciferase activity (pWHE711). The luciferase activity of

Tetracycline-aptamer-mediated gene regulation

1629

© 2003 Blackwell Publishing Ltd,

Molecular Microbiology

,

49

, 1627–1637

pWHE711 in the absence of tetracycline was comparableto that of pWHE702. pWHE721, however, showed only35% of the luciferase activity of the control. We attributethis difference in basal activity to the effects of the respec-

tive sequence environments into which the aptamer wasplaced on local structure. We also observed that theaptamer affects the steady-state mRNA level to a limitedextent, depending on the insertion position (Table 1). Theaddition of tetracycline was found to have no influence onmRNA abundance. In conclusion, aptamer-mediatedtranslational inhibition is position-dependent.

Aptamer-mediated inhibition of reinitiation

We then proceeded to address the effect of tetracycline–

Fig. 1.

Expression system to test regulatory properties

in vivo

.A. Schematic view of the vector YCp22FL. The figure shows the

TEF1

promoter (black box), the 5’UTR, the luciferase open reading frame (open box,

LUC

) and the

PGK1

terminator. Unique restriction sites are indicated by capital letters [

Afl

II (A),

Xho

I (X),

Cla

I (C),

Bam

HI (B),

Nde

I (N),

Xba

I (X) and

Hin

dIII (H)]. The arrow indicates the transcriptional start site.B. Sequence of YCp22FL 5’UTR. Beginning from the 5’ end the underlined sequences represent restriction sites

Afl

II,

Xho

I and

Nde

I.C. Sequence of a CAA-repeat spacer. The 35 nt synthetic spacer consisting of CAA repeats is known to generate a non-structured 5’UTR and was used for control constructs to adjust the length of the UTR.D. Sequence of the tetracycline-binding aptamer 32sh.E. Sequence of the

GCN4

uORF1 and uORF4. The uORFs (shown in capital letters) have been inserted into the vector with its original flanking sequence (shown in lower case).F. Secondary structure prediction of the tetracycline-binding aptamer 32sh using the mfold server version 3.0 (http://www.mfold.bur-net.edu.au).

Fig. 2.

Position dependence of tetracycline-aptamer-mediated regu-lation

in vivo

.A. Shown is a schematic view of the 5

¢

UTR of the Luc-encoding mRNA indicated as a black line. The cap structure is shown as a black ball and the

LUC

reading frame as a grey box. CAA spacer is dis-played as small open rectangle. The aptamers are drawn at the respective position.B. The plot displays the relative luciferase activity. The closed and open bars correspond to luciferase activity measured in protein lysates in the absence and presence of 100

m

M tetracycline with the activity expressed by the vector pWHE702 set to 100%. The numbers above the bars are the regulatory factors determined as the ratio without and with tetracycline.C. Relative steady-state levels of the luciferase mRNAs detected by Northern blot analysis. The upper band corresponds to the

LUC

and the lower band to the

PGK1

mRNA.

1630

S. Hanson

et al.

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

aptamer interactions on the reinitiation potential of theribosome by placing an uORF in front of an aptamer. Weused GCN4 uORF4 and uORF1 along with 3 and 10nucleotides up- and downstream, respectively, of the nat-ural uORF sequence (Fig. 1E) and inserted them into theAflII site of pWHE721 (Fig. 3A). The resulting plasmidswere pWHE726 and pWHE731. Whereas uORF1 fromGCN4 is known to promote efficient reinitiation, uORF4 isthought to act to enhance post-termination ribosomerelease (Hinnebusch, 1997). As control constructs, weconstructed pWHE727 (DuORF4) and pWHE732(DuORF1), each of which contains an AAG triplet insteadof the uORF start codon. Additionally, we inserted theCAA spacer into the AflII site of pWHE721, resulting inpWHE722.

Yeast cells transformed with the constructs shown inFig. 3A were analysed for luciferase activity and mRNAsteady-state levels in the absence and presence of tetra-cycline. The luciferase levels were normalized to that ofpWHE702 without tetracycline. The results are included inTable 1 and displayed in Fig. 3. Tetracycline-dependentregulation is observed in all cases. Regulation is mosteffective for the constructs containing the CAA spacer(pWHE722) and the inactivated uORFs (pWHE727 andpWHE732). On the other hand, pWHE726 and pWHE731,the GCN4 uORF-containing plasmids, show significanttetracycline-dependent reduction of luciferase activity,although the luciferase level is comparatively low. Indeed,the constructs showed a range of levels of luciferaseexpression in the absence of tetracycline. The insertion ofthe CAA spacer and uORF1 led to the generation ofmarkedly reduced luciferase activities, whereaspWHE727 shows luciferase activity comparable to that ofpWHE721.

Overall, these data indicate that the tetracycline–aptamer interaction still acts to regulate translation whenplaced downstream of either type of uORF, i.e. whetherthis is of the type that promotes reinitiation or not. This

means that reinitiating 40S subunits are also subject tothis mechanism of regulation. Moreover, the control exper-iments using the CAA spacer demonstrate that the regu-latory potential of the tetracycline-aptamer system can beoptimized to exceed the regulatory ratio of GCN4 transla-tional control that is naturally observed under conditionsof amino acid starvation.

Translational inhibition in cell-free extracts

Although it has been shown previously that tetracyclinebinds specifically to the aptamer structures that in thisstudy have been inserted into reporter mRNA, we had yetto investigate the mechanism by which tetracycline inhibitsgene expression. We therefore performed experimentsdesigned to test the hypothesis that the tetracycline–aptamer interaction inhibits translation at the initiationstep. Our strategy was to study the translation of in vitrosynthesized aptamer-containing reporter mRNAs. Titra-tion of the respective mRNAs into the cell-free systemprepared from S. cerevisiae strain 1773 generated typicalsaturation curves for the yield of luciferase activity (datanot shown). The translation rates of the aptamer-contain-ing mRNAs were consistently lower than the controlmRNA translation rate at comparable mRNA concentra-tions. This effect was particularly marked with the start-codon-proximal aptamer (pWHE773). In order to deter-mine the relationship between tetracycline concentrationand the degree of inhibition, we used 0.5 mg RNA tem-plate per 15 ml of extract, which is well below the templatesaturation point. Synthesis of luciferase from this concen-tration of mRNA in the cell-free extract was dose-depen-dent inhibited to differing degrees by the addition oftetracycline, depending on the nature of the mRNA tem-plate. Whereas the control mRNA, lacking the aptamer,was barely affected, translation of the two aptamer-con-taining mRNAs was strongly inhibited. Maximum inhibitionwas reached between 10 and 100 mM tc (data not shown).

Table 1. Relative luciferase activity and mRNA steady state level.

Construct

Relative luciferase activity(%) ± SD

Relative mRNA conc (LUC/PGK1)(%) ± SD

Increase in RNA levela Regulatory factor b-tc +tc -tc +tc

pWHE702 100 ± 6.3 100.0 ± 2.4 100 ± 0 92.0 ± 14.9 0.98 1pWHE711 98.7 ± 8.4 27.2 ± 1.0 111.7 ± 13.9 108.2 ± 8.8 0.97 3pWHE721 33.2 ± 1.8 3.6 ± 0.1 142.2 ± 14.9 144.9 ± 13.9 1.01 9pWHE722 1.9 ± 0.1 0.1 ± 0 111.7 ± 2.2 122.0 ± 3.9 1.09 19pWHE726 2.6 ± 0.2 0.4 ± 0 106.4 ± 7.4 145.5 ± 4.8 1.37 6pWHE727 35.3 ± 1.4 2.3 ± 0 108.0 ± 3.8 117.5 ± 7.5 1.13 14pWHE731 1.1 ± 0.1 0.2 ± 0 77.3 ± 2.6 104.5 ± 1.2 1.35 6pWHE732 10.9 ± 0.7 0.8 ± 0 132.5 ± 9.8 170.0 ± 3.7 1.29 15

a. Indicates ratio of normalized LUC mRNA levels in the presence of tc to that in its absence.b. Regulation factors were determined as ratio of the luciferase activity without and with tetracycline (tc) after correcting for mRNA levels.The relative abundance of the LUC mRNA was normalized to the PGK1 mRNA.

Tetracycline-aptamer-mediated gene regulation 1631

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

The effects of tetracycline at a strongly inhibitory con-centration were investigated over a range of mRNA levels(Fig. 4B). The results show that tetracycline binding to thecap-proximal aptamer (pWHE773) inhibits translation byup to 50-fold. Inhibition is consistently high at differentconcentrations of the added mRNA template. Inhibition ofluciferase production from the mRNA bearing the start-

codon-proximal aptamer (pWHE774) was weaker, maxi-mally sixfold. However, interpretation of this quantitativedifference in the degree of inhibition is potentially compli-cated by the fact that the basal rate of translation of themRNA carrying the start-codon-proximal aptamer struc-ture was already extremely low compared to the othertemplates. This may reflect, for example, an extremeresponse of the in vitro translation system to the presenceof the start codon-proximal structure (see Discussion).Comparison with the degrees of inhibition observed invivo reveals that the start codon proximal location of theaptamer is associated with a more moderate restriction ofthe basal translation rate.

Kinetics of tetracycline binding to the aptamer-containing mRNA

The above results indicated that the binding of tetracyclineto its specific aptamer located in the 5¢¢¢¢UTR of an mRNAleads to strong inhibition of translation. We then askedwhether inhibition in vitro was a fast event that followedon rapidly after addition of the ligand. We were promptedto examine the kinetics of this process by the fact thatinhibition of the targeted mRNA in whole yeast cells fol-lows only nine hours after addition of tetracycline. Strik-ingly, we found that addition of tetracycline to the templatemRNA 5 min before the cell free extract resulted inreduced translation of the aptamer-containing mRNAsfrom the earliest points of measurement (Fig. 4B).

We wondered whether equally rapid inhibitory kineticswould be observed if tetracycline is added subsequent tothe onset of translation. We therefore added tetracycline,to a final concentration of 10 mM, 5 min subsequent toaddition of cell-free extract at time-point 0 (Fig. 5). Again,the tetracycline imposed its inhibitory effect within a fewminutes of addition. However, in a further experiment (notshown), we found that when tetracycline was added20 min after initiation of translation, a reduction in thetranslation rate only became evident after the 40th min ofthe time-course. This suggests that once the in vitro syn-thesized mRNA has been packaged into functional RNPcomplexes, the addition of tetracycline becomes lesseffective.

Polysomal gradient analysis of tetracycline-aptamer-bearing mRNAs

We wanted to obtain further information about the mech-anism by which the translation process is affected byligand binding to the aptamer sequence. In order to dothis, we studied the distribution of 32P-labelled mRNAs onsucrose gradients prepared from the in vitro translationextracts (Fig. 6). As expected, the control mRNA lackingthe aptamer structure was found to be associated with the

Fig. 3. Influence of different cis-acting elements in the 5¢UTR on tetracycline-aptamer-mediated translational inhibition.A. Shown is a schematic view of 5¢UTRs containing different cis-acting elements. The CAA spacer is displayed as small open box. The GCN4 uORF1 and uORF4 are shown as a black box along with its AT/GC-rich context (shown as a white box). The inactivated uORFs which contain an AAG instead of the AUG start codon are shown as grey box. All other denotations are as in Fig. 2.B. The plot displays the relative luciferase activity. The closed and open bars correspond to luciferase activity measured in protein lysates in the absence and presence of 100 mM tetracycline with the activity expressed by the vector pWHE702 set to 100%. The numbers above the bars are the regulatory factors determined as the ratio without and with tetracycline.C. Shown is a Northern blot analysis to determine relative steady-state levels of the luciferase mRNAs. The upper band corresponds to LUC and the lower band to PGK1 mRNA.

1632 S. Hanson et al.

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

40S and 80S fractions, with some of it associated withlarger polysome peaks (Fig. 6B). This is consistent withactive binding of 40S subunits and subsequent formationof translation-competent 80S complexes on the addedmRNA. The addition of tetracycline had no effect on thisdistribution in the case of the control mRNA. In contrast,the presence of tetracycline almost completely eliminatedthe cap-proximal aptamer-containing mRNA from the ribo-

somal peaks (Fig. 6C). This result suggests that the bind-ing of tetracycline to the aptamer binding site preventsstable association of 40S subunits with the mRNA. In thecase of the start codon-proximal aptamer, the presenceof tetracycline drastically reduces the 80S peak, therebysuggesting the inhibition of large subunit binding to thescanning 40S (Fig. 6D). Control experiments using theinhibitors m7G cap analogue and GMP-PNP demonstratedthe changes in 32P-labelled mRNA localization in responseto inhibition of 43S binding to the cap (Fig. 6E) and of 60S-40S joining at the start codon (Fig. 6F) respectively.

Taken together, these data demonstrate that transla-tional inhibition by aptamer insertion at distinct positionsin the 5¢UTR is targeted to different stages of the initiationprocess.

Discussion

We have described a regulatory mechanism of translation

Fig. 4. In vitro translation using cell free extracts of S. cerevisiae strain 1773.A. Schematic view of the vector pHSLUCS4 and the analysed con-structs. The upper diagram shows the T7 promoter (black box), the 5¢UTR and the luciferase open reading frame (open box, LUC). A stretch of 30 adenine residues constitutes the poly (A) tail that ends in the NsiI (Ns) restriction site. Unique restriction sites are indicated by capital letters [BglII (B), NcoI (N), BsiWI (Bs) and XbaI (X)]. In the lower diagrams the aptamers are drawn at the respective position of the 5¢UTR of the Luc-encoding mRNA indicated as a black line. All denotations are as in Figs 2 and 3.B. Dependence of in vitro translation inhibition on mRNA concentra-tions. In vitro synthesized capped LUC mRNA was translated in the absence (filled bars) and presence (open bars) of 10 mM tetracycline using a range of mRNA concentration. mRNA was incubated at RT for 5 min with tetracycline in the translational buffer before the addition of the cell free extract.

Fig. 5. Kinetics of tetracycline binding to the aptamer-containing mRNA. In vitro synthesized capped LUC mRNA (pWHE773) was translated in the absence (filled circle) and presence (open circle) of 10 mM tetracycline. Luciferase activity was measured every 20 min The arrow indicates the time-point of addition of 10 mM tetracycline 5 min after the start of translation.

Tetracycline-aptamer-mediated gene regulation 1633

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

initiation in yeast based on the binding of tetracycline toRNA aptamer sites within the 5¢UTR. Binding between theligand and the aptamer results in a conformational changein the latter (Berens et al., 2001). We have been able tocharacterize the influence of insertion site and cis-actingsequences on the degree of tetracycline-dependent inhi-bition of reporter gene expression and to analyse themechanism underlying this regulatory phenomenon.

Whereas insertion of the aptamer sequence into the5¢UTR near the cap site has no effect on reporter expres-sion in the absence of tetracycline, locating the aptamerclose to the start codon does result in a reduced basallevel of expression. We suspect that the inhibitory effectseen at the start codon-proximal site is attributable tointeractions between the aptamer and the local nucle-otide sequence that stabilize inhibitory secondary struc-ture. Previous work showed that there is little differencein the inhibitory capacity of a stem–loop structure ineither of these positions in yeast (Oliveira et al., 1993b;Vega Laso et al., 1993). This contrasts with the situationin mammalian systems, in which a given stem–loopstructure is more inhibitory when located in a cap-proxi-mal position (Kozak, 1986). In the yeast system studiedhere, it is however, notable that aptamer-mediated trans-lational regulation imposed by ligand binding is strong ineither position.

In the cell-free translation system, both the cap-proxi-mal and the cap-distal aptamer structure are highly inhib-itory (see Fig. 5), even in the absence of tetracycline. It ispossible that the aptamer (and its environment) assumesa different conformation to the one in vivo because the invitro synthesized mRNA does not pass through the cellu-lar synthesis, processing and transport pathway and is notassociated with the normal complement of RNA-bindingproteins. Alternatively, this effect may be attributed to thegeneral instability of the association between 40S sub-units and mRNAs in the cell-free system. Previous work(K. Berthelot and J. E. G. McCarthy, unpubl. data) hasalready demonstrated that the 43S-mRNP structuresformed in yeast cell-free extracts are less stable than thepolysomes formed in vivo.

The tetracycline-dependent inhibitory effect also showssignificant quantitative differences in vitro as compared toin vivo. Whereas the degree of tetracycline-dependentsuppression of the construct with the cap-distal aptamerinsertion is comparable between the two systems, thecap-proximal aptamer shows threefold inhibition in vivocompared to 50-fold in vitro. Thus, in the cell-free system,the cap-proximal location of the aptamer allowed strongtranslational inhibition. Studying the kinetics of tetracy-cline-dependent translational inhibition revealed that pre-mixing of tetracycline and mRNA before incubation in the

Fig. 6. In vitro polysomal gradient profiles of tetracycline-aptamer containing mRNAs 32 P-labelled mRNA of pWHE771 (B), pWHE773 (C) and pWHE774 (D) were transcribed in cell free extracts in the absence (filled circles) and pres-ence (open circles) of 10 mM tetracycline. The profiles show radioactivity counted in the frac-tions of a 5–30% linear sucrose gradients. Plot A shows the absorption profile of the gradient. Positions of the 40S, 60S and the 80S riboso-mal particles are indicated. pWHE771 mRNA was analysed in the presence of GMP-PNP (E) and cap analogue (F).

1634 S. Hanson et al.

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

cell-free extract led to the highest degree of inhibition.However, addition of tetracycline either 5 or 20 min aftertranslation had been initiated markedly reduced tetracy-cline-dependent inhibition. One explanation of these datais that tetracycline added after starting in vitro translationhas reduced access to the target aptamer because thelatter is now bound up in RNP complexes. In this contextwe should note that tetracycline achieves its full inhibitoryeffect on translation in vivo only 16 h after being added toyeast cells. This may be partially due to the kinetics oftetracycline transport into the cells. Alternatively, it mayalso reflect the influence of mRNP structure on accessi-bility at various stages of the process leading from tran-scription in the nucleus via nuclear export to incorporationinto polysomes. In conclusion, although the quantitativedetails of tetracycline-aptamer function differ in vitro, wehave found that this regulatory mechanism is reproducedin the cell-free system.

We employed two types of uORF for studying the effectof tetracycline-aptamer mediated translational regulationwhen the aptamer is positioned between the upstreamand the main ORF. Comparison of the tetracycline-depen-dent inhibition obtained with the two uORFs indicates thatalso the post-termination scanning ribosome is subject toregulation via the aptamer-tetracycline system.

Polysomal gradient analysis provided further insight intothe mechanisms underlying tetracycline-aptamer medi-ated translational inhibition as a function of the position ofthe aptamer in the 5¢UTR. The control transcript lackingan aptamer showed no change in the polysome patternin response to the addition of tetracycline. In the case ofthe cap-proximal aptamer transcript, tetracycline causedreductions in both the 40S peak and the 80S peak. Thisis consistent with the tetracycline-aptamer preventingrecruitment of the small ribosomal subunit to the 5¢¢¢¢ endof the mRNA. The low level of translation of the startcodon-proximal aptamer transcript, and thus of activeengagement between the ribosome and the mRNA, madeit difficult to obtain a clear sucrose gradient profile againstthe background of general protein–mRNA association.However, the sharp decrease in the size of the 80S peaksuggests that the tetracycline-aptamer complex causesmarked inhibition of scanning by the 40S subunit, thusreducing binding of the large ribosomal subunit at the startcodon. This can be expected to have a knock-on effect onrecruitment of 40S subunits at the 5¢ end of the mRNA.

On the basis of the above results we propose a modelto explain tetracycline-aptamer-mediated regulation oftranslation initiation (Fig. 7). This form of regulation workswith the target aptamer in both the cap-proximal and thecap-distal positions, albeit via somewhat different mecha-nisms. A tetracycline–aptamer complex near the cap siteinhibits recruitment of the 43S preinitiation complex in thepresence of tetracycline. On the other hand, a cap-distal

aptamer essentially functions as an efficient roadblock forthe scanning preinitiation complex and hinders 80S for-mation. At least in yeast, therefore, this type of regulatorysystem is effective irrespective of aptamer position withinthe 5¢UTR, which means that it provides a relatively flex-ible method of imposing post-transcriptional control ondiverse genes. The observation that translation can beregulated in a tc-dependent manner without the need foradditional components provided demonstrates the poten-tial of a small molecule-modulated riboswitch for use inbasic research. Such a translationally regulated systemwill find various applications as a novel experimental toolfor studies requiring inhibition of translation either in vitroor in vivo.

Experimental procedures

Strains and media

Saccharomyces cerevisiae strain RS453 (MATa ADE2-1TRP1-1 CAN1-100 LEU2-3 LEU2-112 HIS3-1 URA3-52)

Fig. 7. Model to explain tetracycline-aptamer-mediated translational control. The addition of tetracycline facilitates the formation of a tetracycline-aptamer complex which interferes with different steps of translation initiation dependent on the site of aptamer insertionA. Cap-proximal aptamer insertion interferes with binding of the 43S preinitiation complex.B. Insertion of the aptamer close to the start-codon prevents 80S formation by either hindering successful scanning of the 40S subunit or binding of the large ribosomal subunit.C. Aptamers located between an uORF and the main ORF hinder scanning or reinitiation of a terminating ribosome. Relative luciferase activity and mRNA steady state level.

Tetracycline-aptamer-mediated gene regulation 1635

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

(Sauer and Stadler, 1993), S. cerevisiae strain 1773 (MATaARG9) and E. coli strain DH5a [supE44, DlacU109(f80lacZDM15M) hsdR17 recA1 endA1 gqrA96 thi-1 relA1](Ausubel et al., 1994) were used. Yeast strains were grownin YPD (Sherman, 1991) and YNB (2 g l-1 yeast nitrogen base(DIFCO), 5.5 g l-1 ammonium sulphate, 20 g l-1 glucose, ade-nine 12 mg l-1, arginine 40 mg l-1, histidine 40 mg l-1, leucine60 mg l-1 and uracil 40 mg l-1, pH 5.6) in the absence or pres-ence of 100 mM tetracycline. Yeast cells were transformedusing the Frozen EASY yeast transformation II kit (Zymoresearch, Orange, CA).

Plasmid constructions

We used the plasmid YCp22FL (Oliveira et al., 1993a), whichconstitutively express the LUC gene from a TEF1 promoter,for all in vivo analysis. The aptamer 32sh (Fig. 1B) was ampli-fied by PCR using primer pairs A5AflF (5¢-ATACTTAAGGCCTAAAACATACCAG-3¢)/A6AflR (5¢-ATACTTAAGGCCTAGGTGGTCG-3¢) and 32 k_xho_F (5¢-ATATCACTCGAGGCCTAAAACATACC-3¢)/32 k_xho_R (5¢-ATATCACTCGAGGCCTAGGTGGTCG-3¢). The restriction sites have been under-lined. The resulting fragments were inserted into the 5¢ UTRof a LUC reporter gene using AflII and XhoI restriction sitesof YCp22FL, respectively, by standard cloning procedures(Sambrook et al., 1989). A CAA-spacer (Fig. 1C) and theGCN4 uORF1 and uORF4 fragment (Fig. 1E) were gener-ated using synthetic oligonucleotides (MWG Biotech, Ger-many) and inserted into the respective vectors. Syntheticoligonucleotides with AAG instead of AUG were used togenerate DuORF1 and DuORF4.

The plasmid pSLUCS4, in which the T7 promoter precedesthe LUC gene, was used for in vitro transcription. The entireaptamer containing 5¢UTRs from YCp22FL derivatives wereamplified by PCR using primer pairs synutr1_in_bgl (5¢TATATAGATCTAATTATCTACTTAAGAAAAC)/synutr_rev_nco (5¢ATATACCATGGTTCTCGAGTTTTGTGTTCTTAAG), 32 k_afl_in_bgl (5¢TATATAGATCTAATTATCTACTTAAGGCC)/synutr_rev_nco and synutr1_in_bgl/32 k_xho_out_nco (5¢ATATACCATGGTTCTCGAGGCCTAGG) respectively. The fragmentswere cloned into pSLUCS4 using BglII and NcoI to displacethe original UTR.

Luciferase assay

Yeast cells transformed with the respective plasmids weregrown overnight in YNB medium in the absence of tetracy-cline. They were then diluted in YNB (± 100 mM tetracycline)to give OD600 = 0.05 and grown until OD600 = 0.9–1.0 wasreached. Cells (1 ml) were harvested and washed twice withice-cold 50 mM Tris HCl buffer pH 7.5. Yeast cell lysates wereprepared using the Yeast Protein Kit (Zymo research,Orange, CA). Ten ml of lysates or 15 ml of the in vitro transla-tion assay were added to 350 ml of reaction buffer (25 mMglycylglycine pH 7.8, 5 mM ATP, 15 mM MgSO4) andluciferase activity was measured using a luminometer(Lumat-LB9501, Berthold). The reaction was initiated by theinjection of 100 ml of 0.2 mM luciferin (in 25 mM glycylglycinepH 7.8). Five ml of yeast cell lysate was used for proteinestimation using the BCA protein assay kit (Pierce, Rockford,

IL, USA). The assay was carried out according to the manu-facturers instructions in 96-well plates. Absorbance was mea-sured at 550 nm using a Spectra Fluor Plus (Tecan,Crailsheim).

RNA isolation and Northern blots

Total RNA was isolated from 20 ml of yeast culture harvestedat OD600 = 0.9–1.0 using the hot phenol method (Köhrer andDorndey, 1991). Northern blot analysis was carried out asdescribed (Homuth et al., 1997). In vitro synthesis and label-ling of RNA probes, hybridization and detection of signalswas done using the digoxigenin (DIG) RNA labelling kit anddetection chemicals (Roche Diagnostics, Mannheim). PGK1mRNA was used as the internal control. Yeast chromosomalDNA was prepared using YeaStar Genomic DNA kit (Zymoresearch, Orange, CA). DIG-labelled RNA probes wereobtained by in vitro transcription with T7 RNA polymerase(Roche Diagnostics, Mannheim) using PCR fragments ampli-fied from respective YCp22FL derivatives (LUC) or yeastchromosomal DNA (PGK1) (primer pairs specific for LUC:Luc_F_NH (5¢-GAAAGGCCCGGCGCCATTCTATCC-3¢)/Luc_R_NH (5¢-CCAAGTAATACGACTCACTATAGGGAAACAAACACTACGGTAGGCTGC-3¢) and PGK1: Pgk_F_NH (5¢-ATGTCTTTATCTTCAAAGTTGTCTGTCC-3¢)/Pgk_R_NH (5¢-CCAAGTAATACGACTCACTATAGGGACCTTTCTGGAACCTTCTTCTTCG-3¢). The T7 RNA polymerase recognitionsequence is underlined. The blots were visualized and quan-tified using Lumi-imager F1 and Lumi Analyst 3.1 (RocheDiagnostics, Mannhein).

Preparation of yeast cell-free extracts

Cell-free extracts were prepared essentially as describedpreviously (Iizuka and Sarnow, 1997). Yeast cells of the strain1773 were harvested in log phase (OD600 = 3–4), cooled for20 min on ice, washed twice with cold water and then washedthrice in freshly prepared chilled buffer A (20 mM Hepes/KOHpH 7.4, 100 mM potassium acetate, 2 mM magnesium ace-tate, 2 mM dithiothreitol) with 8.5 g l-1 mannitol. Cell pellet (4–5 g) was resuspended in buffer A with mannitol (1.5 ml g-1)and treated with 1 mM PMSF. The cells were disrupted man-ually using glass beads (3 g g-1) six times with 1 minute inter-vals on ice. Cellular debris and glass beads were removedby centrifugation at 30 000 g for 5 min followed by a 10 mincentrifugation step of the supernatant. The supernatant(4 ml) was loaded on a 100-ml GF-25 column equilibratedwith buffer A treated with 1 mM PMSF. 2 ml fractions werecollected and protein content estimated at A260/280. Transla-tional activity was routinely checked using a LUC mRNA astemplate, and active fractions were pooled, aliquoted, andfrozen in liquid nitrogen before storing at -80∞C.

In vitro translation and polysomal gradient analysis

DNA templates for in vitro translation assays and polysomalgradient analyses were prepared by discontinuous caesiumchloride gradient (Sambrook, 1989) and linearized with NsiI(after poly(A) stretch) and BsiWI (155 nt downstream of LUCstart codon) respectively. RNAs were transcribed in vitro

1636 S. Hanson et al.

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

using T7 RNA polymerase (New England Biolabs) accordingto the manufacturer’s instructions in the presence of capanalogue m7GpppG. [a-32P] UTP (ICN) was added to thetranscription reaction for synthesis of radiolabelled RNAs.The transcripts were purified by phenol and phenol-chloro-form-isoamylalcohol extractions, ammonium acetate, sodiumacetate precipitations and purification on NAP-5 columns(Amersham Biosciences). RNAs were denatured at 42∞C for15 min followed by renaturation at room temperature for10 min before the start of each translation experiment. In vitrotranslation assays were performed using 0.5 mg RNA intranslation buffer (22 mM Hepes/KOH pH 7.4, 120 mM potas-sium acetate, 2 mM magnesium acetate, 0.75 mM ATP,0.1 mM GTP, 25 mM creatine phosphate and 1.7 mM DTT),1.2 mg ml-1 creatine phosphokinase, 32 U RNasin, 5.56 mMamino acid complete mixture (Promega), the respective con-centration of tetracycline and 15 ml of cell-free extract. Tetra-cycline was preincubated with the RNA in the translationbuffer for 5 min at room temperature before addition of thecell-free extract. Translation was carried out at 25∞C for 1 h.In vitro translation for the polysomal gradient analysis wascarried out using 5 ng of [a-32P] radiolabelled and purifiedRNAs and 100 ml of cell-free extract at 25∞C for 10 min in thepresence or absence of 10 mM tetracycline. The translationmix was then loaded onto 5–30% DEPC-treated linearsucrose gradient and recorded on an ISCO UA-5 densitygradient fractionator. Fractions (200 ml) were collected andradioactivity was counted using a WALLAC 1409 liquid scin-tillation counter.

Acknowledgements

A part of these studies were carried out in the laboratory ofWolfgang Hillen, whose support is greatly appreciated. Weare grateful to the Volkswagenstiftung and the Fonds derChemischen Industrie. J.E.G.M. thanks the Biotechnologyand Biological Sciences Research Council (UK) and theWellcome Trust (UK) for support. B.S. was a recipient of apersonal grant from Bayerischer Habilitationsförderpreis andS.H. from the Boehringer Ingelheim Fonds.

References

Ausubel, F.A., Brent, R., Kingston, R.E., Moore, D.D.,Seidmann, J.G., Smith, J.A., and Struhl, K. (1994) CurrentProtocols in Molecular Biology. New York: John Wiley &Sons.

Baim, S.B., and Sherman, F. (1988) mRNA structures influ-encing translation in the yeast Saccharomyces cerevisiae.Mol Cell Biol 8: 1591–1601.

Berens, C. (2001) Tetracyclines and RNA. In RNA-bindingAntibiotics. Schroeder, R., and Wallis, M.G. (eds). Austin,Texas: R. G. Landes Company, pp. 73–88.

Berens, C., Thain, A., and Schroeder, R. (2001) A tetracycline-binding RNA aptamer. Bioorg Med Chem 9: 2549–2556.

Bettany, A.J., Moore, P.A., Cafferkey, R., Bell, L.D., Goodey,A.R., Carter, B.L., and Brown, A.J. (1989) 5¢-secondarystructure formation, in contrast to a short string of non-preferred codons, inhibits the translation of the pyruvatekinase mRNA in yeast. Yeast 5: 187–198.

Cigan, A.M., Pabich, E.K., and Donahue, T.F. (1988) Muta-tional analysis of the HIS4 translational initiator region inSaccharomyces cerevisiae. Mol Cell Biol 8: 2964–2975.

Ellington, A.D., and Szostak, J.W. (1990) In vitro selection ofRNA molecules that bind specific ligands. Nature 346:818–822.

Favre, D., Blouin, V., Provost, N., Spisek, R., Porrot, F.,Bohl, D., et al. (2002) Lack of an immune responseagainst the tetracycline-dependent transactivator corre-lates with long-term doxycycline-regulated transgeneexpression in nonhuman primates after intramuscularinjection of recombinant adeno-associated virus. J Virol76: 11605–11611.

Forster, K., Helbl, V., Lederer, T., Urlinger, S., Wittenburg,N., and Hillen, W. (1999) Tetracycline-inducible expressionsystems with reduced basal activity in mammalian cells.Nucleic Acids Res 27: 708–710.

Gonzalez, C.I., Bhattacharya, A., Wang, W., and Peltz, S.W.(2001) Nonsense-mediated mRNA decay in Saccharomy-ces cerevisiae. Gene 274: 15–25.

Grate, D., and Wilson, C. (2001) Inducible regulation of theS. cerevisiae cell cycle mediated by an RNA aptamer-ligand complex. Bioorg Med Chem 9: 2565–2570.

Gray, N.K., and Wickens, M. (1998) Control of translationinitiation in animals. Annu Rev Cell Dev Biol 14: 399–458.

Harvey, I., Garneau, P., and Pelletier, J. (2002) Inhibition oftranslation by RNA–small molecule interactions. RNA 8:452–463.

Hermann, T., and Patel, D.J. (2000) Adaptive recognition bynucleic acid aptamers. Science 287: 820–825.

Hinnebusch, A.G. (1997) Translational regulation of yeastGCN4. A window on factors that control initiator-tRNA bind-ing to the ribosome. J Biol Chem 272: 21661–21664.

Homuth, G., Masuda, S., Mogk, A., Kobayashi, Y., and Schu-mann, W. (1997) The dnaK operon of Bacillus subtilis isheptacistronic. J Bacteriol 179: 1153–1164.

Iizuka, N., and Sarnow, P. (1997) Translation-competentextracts from Saccharomyces cerevisiae: effects of L-ARNA, 5¢ cap, and 3¢ poly (A) tail on translational efficiencyof mRNAs. Methods 4: 353–360.

Jiang, W., Hou, Y., and Inouye, M. (1997) CspA, the majorcold-shock protein of Escherichia coli, is an RNA chaper-one. J Biol Chem 272: 196–202.

Kaufman, R.J. (1994) Control of gene expression at thelevel of translation initiation. Curr Opin Biotechnol 5:550–557.

Köhrer, K., and Dorndey, H. (1991) Preparation of highmolecular weight RNA. Methods Enzymol 194: pp. 398–405.

Kozak, M. (1986) Influences of mRNA secondary structureon initiation by eukaryotic ribosomes. Proc Natl Acad SciUSA 83: 2850–2854.

Kozak, M. (2002) Pushing the limits of the scanning mecha-nism for initiation of translation. Gene 299: 1–34.

Lamartina, S., Roscilli, G., Rinaudo, C.D., Sporeno, E., Silvi,L., Hillen, W., et al. (2002) Stringent control of gene expres-sion in vivo by using novel doxycycline-dependent trans-activators. Hum Gene Ther 13: 199–210.

Lori, F., Guallini, P., Galluzzi, L., and Lisziewicz, J. (2002)Gene Therapy Approaches to HIV Infection. Am J Pharma-cogenomics 2: 245–252.

Tetracycline-aptamer-mediated gene regulation 1637

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 49, 1627–1637

Mathews, M., Sonenberg, N., and Hershey, J.W.B. (2000)Origins and principles of translational control. In Transla-tional Control of Gene Expression Cold Spring Harbour,New York: Cold Spring Harbour Laboratory Press, pp. 1–31.

McCarthy, J.E. (1998) Posttranscriptional control of geneexpression in yeast. Microbiol Mol Biol Rev 62: 1492–1553.

Meijer, H.A., and Thomas, A.A. (2002) Control of eukaryoticprotein synthesis by upstream open reading frames in the5¢-untranslated region of an mRNA. Biochem J 367: 1–11.

Miranda-Rios, J., Navarro, M., and Soberon, M. (2001) Aconserved RNA structure (thi box) is involved in regulationof thiamin biosynthetic gene expression in bacteria. ProcNatl Acad Sci USA 98: 9736–9741.

Mironov, A.S., Gusarov, I., Rafikov, R., Lopez, L.E., Shatalin,K., Kreneva, R.A., et al. (2002) Sensing small moleculesby nascent RNA. A mechanism to control transcription inbacteria. Cell 111: 747–756.

Nahvi, A., Sudarsan, N., Ebert, M.S., Zou, X., Brown, K.L.,and Breaker, R.R. (2002) Genetic control by a metabolitebinding mRNA. Chem Biol 9: 1043.

Narberhaus, F. (2002) mRNA-mediated detection of environ-mental conditions. Arch Microbiol 178: 404–410.

Oliveira, C.C., and McCarthy, J.E. (1995) The relationshipbetween eukaryotic translation and mRNA stability. A shortupstream open reading frame strongly inhibits translationalinitiation and greatly accelerates mRNA degradation in theyeast Saccharomyces cerevisiae. J Biol Chem 270: 8936–8943.

Oliveira, C.C., Goossen, B., Zanchin, N.I., McCarthy, J.E.,Hentze, M.W., and Stripecke, R. (1993a) Translationalrepression by the human iron-regulatory factor (IRF) inSaccharomyces cerevisiae. Nucleic Acids Res 21: 5316–5322.

Oliveira, C.C., van den Heuvel, J.J., and McCarthy, J.E.(1993b) Inhibition of translational initiation in Saccharomy-ces cerevisiae by secondary structure: the roles of thestability and position of stem-loops in the mRNA leader.Mol Microbiol 9: 521–532.

Opalinska, J.B., and Gewirtz, A.M. (2002) Nucleic-acid ther-apeutics: basic principles and recent applications. Nat RevDrug Discov 1: 503–514.

Patel, D.J. (1997) Structural analysis of nucleic acid aptam-ers. Curr Opin Chem Biol 1: 32–46.

Patel, D.J., Suri, A.K., Jiang, F., Jiang, L., Fan, P., Kumar,R.A., and Nonin, S. (1997) Structure, recognition andadaptive binding in RNA aptamer complexes. J Mol Biol272: 645–664.

Robertson, D.L., and Joyce, G.F. (1990) Selection in vitro ofan RNA enzyme that specifically cleaves single-strandedDNA. Nature 344: 467–468.

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecu-lar Cloning: a Laboratory Manual. Cold Spring Harbour,New York: Cold Spring Harbour Laboratory Press.

Sauer, N., and Stadler, R. (1993) A sink-specific H+/monosaccharide co-transporter from Nicotiana tabacum:cloning and heterologous expression in baker’s yeast.Plant J 4: 601–610.

Sherman, F. (1991) Getting started with yeast. MethodsEnzymol 194: 3–21.

Sonenberg, N. (1994) mRNA translation: influence of the 5¢and 3¢ untranslated regions. Curr Opin Genet Dev 4: 310–315.

Stormo, G.D., and Ji, Y. (2001) Do mRNAs act as directsensors of small molecules to control their expression?Proc Natl Acad Sci USA 98: 9465–9467.

Sucheck, S.J., and Wong, C.H. (2000) RNA as a target forsmall molecules. Curr Opin Chem Biol 4: 678–686.

Tuerk, C., and Gold, L. (1990) Systematic evolution of ligandsby exponential enrichment: RNA ligands to bacteriophageT4 DNA polymerase. Science 249: 505–510.

Vega Laso, M.R., Zhu, D., Sagliocco, F., Brown, A.J., Tuite,M.F., and McCarthy, J.E. (1993) Inhibition of translationalinitiation in the yeast Saccharomyces cerevisiae as a func-tion of the stability and position of hairpin structures in themRNA leader. J Biol Chem 268: 6453–6462.

Vilela, C., Ramirez, C.V., Linz, B., Rodrigues-Pousada, C.,and McCarthy, J.E. (1999) Post-termination ribosomeinteractions with the 5¢UTR modulate yeast mRNA stability.EMBO J 18: 3139–3152.

Werner, M., Feller, A., Messenguy, F., and Pierard, A. (1987)The leader peptide of yeast gene CPA1 is essential for thetranslational repression of its expression. Cell 49: 805–813.

Werstuck, G., and Green, M.R. (1998) Controlling geneexpression in living cells through small molecule–RNAinteractions. Science 282: 296–298.

Winkler, W., Nahvi, A., and Breaker, R.R. (2002a) Thiaminederivatives bind messenger RNAs directly to regulate bac-terial gene expression. Nature 419: 952–956.

Winkler, W.C., Cohen-Chalamish, S., and Breaker, R.R.(2002b) An mRNA structure that controls gene expressionby binding FMN. Proc Natl Acad Sci USA 99: 15908–15913.

Yoon, H., and Donahue, T.F. (1992) Control of translationinitiation in Saccharomyces cerevisiae. Mol Microbiol 6:1413–1419.