Báo cáo khoa học: Gene regulation by tetracyclines Constraints of resistance regulation in bacteria shape TetR for application in eukaryotes Christian Berens and Wolfgang Hillen pptx

The resistance protein TetA, a membrane-spanning H+-[tcÆM]+antiporter, must be sen-sitively regulated because its expression is harmful in the absence of tc, yet it has to be expressed b

R E V I E W A R T I C L E

Gene regulation by tetracyclines

Constraints of resistance regulation in bacteria shape TetR for application

in eukaryotes

Christian Berens and Wolfgang Hillen

Lehrstuhl fu¨r Mikrobiologie, Institut fu¨r Mikrobiologie, Biochemie und Genetik, Friedrich-Alexander Universita¨t Erlangen-Nu¨rnberg; Germany

The Tet repressor protein (TetR) regulates transcription

of a family of tetracycline (tc) resistance determinants in

Gram-negative bacteria The resistance protein TetA, a

membrane-spanning H+-[tcÆM]+antiporter, must be

sen-sitively regulated because its expression is harmful in the

absence of tc, yet it has to be expressed before the drugs’

concentration reaches cytoplasmic levels inhibitory for

protein synthesis Consequently, TetR shows highly

speci-fic tetO binding to reduce basal expression and high affinity

to tc to ensure sensitive induction Tc can cross biological

membranes by diffusion enabling this inducer to penetrate

the majority of cells These regulatory and pharmacological

properties are the basis for application of TetR to

selec-tively control the expression of single genes in lower and

higher eukaryotes TetR can be used for that purpose in

some organisms without further modifications In

mam-mals and in a large variety of other organisms, however,

eukaryotic transcriptional activator or repressor domains

are fused to TetR to turn it into an efficient regulator Mechanistic understanding and the ability to engineer and screen for mutants with specific properties allow tailoring

of the DNA recognition specificity, the response to inducer

tc and the dimerization specificity of TetR-based eukary-otic regulators This review provides an overview of the TetR properties as they evolved in bacteria, the functional modifications necessary to transform it into a convenient, specific and efficient regulator for use in eukaryotes and how the interplay between structure) function studies in bacteria and specific requirements of particular applica-tions in eukaryotes have made it a versatile and highly adaptable regulatory system

Keywords: antibiotic resistance; disease models; fusion pro-tein; inducible gene expression; ligand-binding specificity; mammalian cell lines; protein engineering; structure–activity relationship; Tet repressor; transgenic organism

Properties of bacterial Tet systems

Efflux-mediated tetracycline resistance is always

regulated in Gram-negative bacteria

In Gram-negative bacteria, resistance to tetracyclines (tc)

is mediated by the TetA protein, a proton-[tcÆMg]+

anti-porter embedded in the cytoplasmic membrane [1,2] Eleven

tc resistance determinants (Tet classes A–E, G, H, J, Z, 30,

and 33 [3–5]) share the organization of structural and

regulatory genes (reviewed in [6]) In enteric bacteria, the

efflux-encoding tetA genes are strictly regulated at the level

of transcription by the tc-responsive Tet repressor (TetR)

In the absence of inducer, TetR dimers bind to the operators tetO1 and tetO2, shutting down transcription of its own gene, tetR, and of the resistance gene, tetA Once tc has entered the cell, it binds TetR with high affinity as a [tcÆMg]+complex [7] This induces a conformational change

in TetR [8] resulting in dissociation from tetO [9] The following expression burst of TetA and TetR leads to a rapid reduction of the cytoplasmic tc concentration [10] which, in turn, shuts expression of both genes off again Expression of TetA is fine-tuned in the presence of tc so that export overcomes the slow uptake (compare below)

Regulation of Tc resistance is optimized for tightness and sensitivity

Regulation of tet determinants is subject to strong, opposing selective pressures Expression of the resistance protein TetA is detrimental to the cell [11,12] Overexpression of this integral membrane protein is lethal for Escherichia coli [13], probably due to the collapse of the membrane potential [14] Consequently, expression of TetA must be tightly repressed

in the absence of the drug However, when tc diffuses into the cell the resistance protein must be expressed before the cytoplasmic concentration of tc reaches the micromolar level necessary to inhibit translation This requires: (a) high

Correspondence to W Hillen, Lehrstuhl fu¨r Mikrobiologie, Institut

fu¨r Mikrobiologie, Biochemie und Genetik, Friedrich-Alexander

Universita¨t, Staudtstr 5, D-91058 Erlangen, Germany.

Fax: +49 9131 8528082, Tel.: +49 9131 8528081,

E-mail: whillen@biologie.uni-erlangen.de

Abbreviations: tc, tetracycline; dox, doxycycline; atc,

anhydrotetra-cycline; tTA, tetracycline-dependent transactivator; rtTA, reverse

tetracycline-dependent transactivator; tTS, tetracycline-dependent

trans-silencer; CMV, cytomegalovirus; GFP, green fluorescent

protein.

(Received 8 April 2003, revised 14 May 2003,

accepted 15 May 2003)

affinity of TetR for both tetO and tc to keep the basal

expression level of tetA low and to ensure that its

transcription is initiated at concentrations which are still

subinhibitory for translation; (b) low affinity of the TetR–

[tcÆMg]+complex for DNA; and (c) high-level, but

short-term expression of TetA to initially reduce the internal

concentration of tc A low level of TetR is important for

sensitive induction, since E coli strains expressing high

levels of TetR need high concentrations of tc for full

induction [15] These conflicting requirements are met by

the genetic organization of the resistance determinants

(reviewed in [6]) and by the ligand binding properties of

TetR High sensitivity towards tetracyclines [see Fig 1 for

the structures of tc, doxycycline (dox) and anhydro-tc (atc)]

is achieved by the remarkable binding constant of TetR for

[tcÆMg]+ (Ka 109

M )1), [doxÆMg]+ (Ka 1010

M )1) or [atcÆMg]+(Ka 1011

M )1) [7,9], about 103)105-fold higher than the affinity of the drugs to their intracellular target, the

ribosome [16] Binding of two molecules of [tcÆMg]+to a

TetR dimer diminishes repressor affinity for tetO by about

nine orders of magnitude to the unusually low background

DNA binding level of less than 105M )1[9] This high ratio

of specific over nonspecific DNA binding enables TetR to

bind tetO efficiently, even in larger genomes containing

competing nonspecific DNA to a much higher degree than bacteria Taken together, the evolutionary pressures on tc-dependent gene regulation have led to tight repression in the absence of tc, without compromising sensitivity of induction, so that regulated tc resistance determinants impose no burden on the fitness of E coli in the absence of the antibiotic, but still mediate high levels of resistance to tc

in its presence [12]

The structural change of TetR associated with induction

by tetracycline is known X-ray crystal structures of free TetR [17], TetR complexed with different tetracyclines [18–21] and with tetO [8] have been determined at resolutions of 1.9–2.5 A˚, revealing the allosteric conformational change leading to induction These results have been reviewed in detail [22] and have been compared to Lac repressor [23] Thus, they are only summarized here (Fig 2) The DNA reading head of TetR (magenta) is connected to the protein core (blue) by the helix a4 (green) Binding of [tcÆMg]+(yellow) to TetR unwinds the C-terminal residues of helix a6 (light blue), which bump into a4 and displace it As the C terminus of a4 is held in place by contacts to tc, the displacement leads to a pendulum-like swing of the a4 N terminus increasing the distance between the recognition helices by 3 A˚, so that they

do not fit into successive major grooves of DNA anymore [24] These conformational changes are consistent with many noninducible TetR mutants [24,25], spectroscopic analysis of TetR in vitro [26], in vivo [27] and in vitro [28] disulfide trapping experiments Furthermore, a movement

of a9 closes the tc binding pocket after the drug has entered [17], and the loop between a8 and a9 is also important for induction [29–31]

Tetracycline penetrates cells by diffusion Tetracyclines (Fig 1) diffuse in their uncharged forms through lipid bilayers without the aid of protein channels [32–36] Measuring the increase in fluorescence intensity of

tc observed upon binding to TetR [7] allows us to determine the cytoplasmic concentration of tc and, thus, to calculate permeation coefficients for tc uptake into liposomes [(2.4 ± 0.6)· 10)9cmÆs)1] and whole E coli cells [(5.6 ± 1.9)· 10)9cmÆs)1] [36] These translate into half-equilibration times of 35 ± 15 min for tc to cross the membranes and are in good agreement with the half-equilibration time of 15 min measured for [3H]tc-uptake in Bacillus subtilis[37], and the slow uptake of tc observed in Staphylococcus aureus[38] Tetracycline diffusion through phospholipid membranes is, thus, slow and appears to be the rate-limiting step of uptake into cells [36] The previously observed rapid uptake of tc [33,39] might rather reflect unspecific adsorption of tc to membrane surfaces [32,36] A detailed model explaining the transport and accumulation

of tc across the Gram-negative cell envelope has been presented by Nikaido and coworkers ([40,41] and references cited therein) In the medium, as well as in the periplasm and cytoplasm, tc is present in one uncharged and several charged or zwitterionic species, due to its three titratable groups (Fig 1) The distribution between these species depends on the pH of the respective compartment [40] The

Fig 1 Structures of tetracyclines used in eukaryotic gene regulation.

(A) Structure of tetracycline with the pK a values of the three titratable

groups (B) Structure of doxycycline (C) Structure of

anhydrotetra-cycline.

uncharged form of tc can penetrate the outer membrane

directly But the major fraction of tc equilibrates as a

[tcÆM]+-complex rapidly through the outer membrane via

porins, with the Donnan potential across the outer

mem-brane leading to a two- to threefold accumulation of this

charged complex in the periplasm Tc then diffuses passively

in its uncharged form through the cytoplasmic membrane

Due to the pH gradient across the cytoplasmic membrane,

a larger fraction of the uncharged tc dissociates in the

cytoplasm than in the periplasm Since equilibrium is

reached when the concentration of uncharged tc is identical

in both compartments, this results in a higher intracellular

concentration of [tcÆM]+, the biologically active compound

Again, accumulation of tc is the product of this passive

equilibration across the inner membrane [40,41]

Tc-based gene regulation functions in

different setups in many eukaryotic systems

The evolved properties of TetR described above combined

with the favorable pharmacokinetics of tetracyclines and

their long record of safe use in clinical practice make the Tet

system a good candidate to fulfill the criteria that are required

for an ideal transcriptional regulator in eukaryotic cells as

given by Saez and others [42,43] Consequently, the past

15 years have seen the broad application of tc-dependent

regulatory systems, mainly in mammalian cell culture, but to

an increasing degree in transgenic organisms like plants,

yeasts, protozoan parasites, slime molds, flies, and rodents

These topics have been extensively reviewed [42–52] The

following section presents an overview of the basic Tet

systems used to regulate gene expression in eukaryotes

Gene regulation by TetR in eukaryotes

The most basic and first published application of

tc-mediated gene regulation in eukaryotes is transcriptional

repression by unmodified TetR [53] Here, TetR most likely acts by interfering sterically with binding of RNA polymerase or auxiliary transcription factors [42,54] To achieve this, one or more tetO elements are placed in proximity of either the TATA box or the transcriptional start site of the respective target gene and TetR is expressed concomitantly by a strong, constitutive promoter Promot-ers of all three eukaryotic RNA polymerases have been targeted in the manner described Unfortunately, as will become evident in the following paragraph, the published, successful approaches do not yet allow formation of a simple strategy for establishing a TetR-repressed system, although they clearly point out that the positioning of the tetOboxes is crucial for efficient regulation

In Leishmania donovani, an RNA polymerase I promo-ter was brought under tc-control by placing a single tetO site 2–24 bp upstream of the transcriptional start site [55], whereas in Trypanosoma brucei at least one tetO element had to be inserted at a position +2 or)2 relativ e to the transcription start site of an RNA polymerase I-like promoter [56] For RNA polymerase III-mediated tran-scription of suppressor tRNA genes, induction factors between two- to fivefold were observed in Saccharomyces cerevisiae, Dictyostelium discoideum and carrot protoplasts when tetO was introduced within 10 bp upstream of the transcriptional start site [57–59] A regulated version of the human U6 snRNA promoter, also transcribed by RNA polymerase III, was developed by replacing sequences between the proximal sequence element and the transcrip-tional start site with tetO [60] Flanking the TATA-box with two operators completely abolished transcriptional activity

In contrast, introduction of a single tetO element affected transcription only slightly, but led to up to 25-fold repression in the presence of TetR A regulated U6 snRNA promoter with a defined expression window [61,62] would

be a very powerful tool as this promoter is used to express the small interfering RNA [63] needed for silencing gene

Fig 2 Structure of the TetR–[tcÆMg] +

complex Tet repressor is shown as a ribbon

diagram with one monomer in gray and the

other monomer color-coded as follows: The

DNA-binding region is in magenta, the helix

connecting it with the protein core is in green.

The protein core is dark blue, with the helix a6

in light blue Tetracycline is displayed as

space-filling CPK model in yellow For clarity,

the helices a1–a10 of one monomer are

num-bered and the N and C termini of both

sub-units are indicated The coordinates were

taken from the PDB entry 2TRT [18].

expression by RNA interference [64] Repression of RNA

polymerase II promoters exerted by TetR is strongest in

plants [65,66], mammalian cells [67] and fungi like

Schizo-saccharomyces pombe [68,69] when multiple tet operators

are positioned within a region from 5 bp upstream to 35 bp

downstream of the TATA-element In contrast, placement

of one to four tet operators immediately downstream of the

transcription initiation site has been shown to be most

effective in the parasitic protozoa Entamoeba histolytica

[70,71], Toxoplasma gondii [72] and Giardia lamblia [73]

Gene regulation by TetR-based transregulators

While unmodified TetR acts as a transcriptional repressor in

plants and lower eukaryotes, it can be, but not always is

efficient in mammalian cells [67,74] A consistently

func-tional version for yeasts, flies and mammalian cell lines is

TetR fused to an eukaryotic regulatory domain, such as an

acidic activation domain (Fig 3A; tTA or Tet-Off) [75–80]

or a repression domain (Fig 3B; tTS) [81–83] The

trans-activator tTA directs expression from a tc-dependent

promoter that contains seven repeats of a tetO2 element

from the transposon Tn10 The palindromic centers of two

adjacent operators are separated by 41 bp This element is

fused to a minimal promoter, typically derived from the

human cytomegalovirus (CMV) immediate early promoter

[75] When both components are stably integrated into

proper chromosomal loci of mammalian cell lines,

tran-scription from the hybrid promoter is silent in the presence

of more than 10 ngÆmL)1 dox Removal of dox leads to

binding of tTA to tetO and subsequent activation of

transcription Regulatory factors of up to five orders of

magnitude can be reached with sensitive reporter genes like firefly luciferase [75] Luciferase activity is expressed within

4 h of removal of tc and about 20% of the steady-state level

is reached after 12 h While the use of a strong, constitutive promoter (CMV IE, EF-1a, Ubiquitin C) is common in cell culture applications, the use of tissue-specific promoters in transgenic animals provides spatial control to the Tet system, restricting expression of the Tet transregulator and, subsequently, the transgene to the desired tissue [84,85] In Drosophila, usage of the Gal4-UAS system to control Tet transregulator expression allows the generation of spatially delimited expression patterns by simple crossing with one of the many Gal4 driver lines available in the Drosophila research community [86]

One concern has been the expression levels of Tet transregulators as influenced by a potentially low mRNA stability or efficiency of translation This was recently addressed by generating a synthetic coding sequence for tetR Potential splice donor and acceptor sites identified by sequence analysis, several potential endonuclease cleavage sites, and potential stable hairpin structures in the mRNA were eliminated and human codon usage was used [87–90] The consequence of this optimization protocol is a higher protein level in Drosophila, HeLa and HEK293 cells Another concern voiced was that the CMV-derived minimal promoter was not transcriptionally silent under all experimental conditions [91–93] This promoter leakiness can be caused by promoter-dependent or integration site-dependent effects and has been discussed in detail [94] Promoter-dependent leakiness has been addressed by the use of alternative minimal promoters [75,95,96] In transient transfection experiments, these show lower basal activities

Fig 3 Regulation of gene expression by Tet transregulators The promoter proximal tetO boxes are represented by black boxes The transregulators are shown as follows: the DNA reading heads are in light gray, the inducer-binding and dimerization domain is in dark gray, activation domains are black boxes, and the silencing domain is stippled The conformational change leading to the loss of DNA-binding activity is pictured as a light gray box High-level activated transcription is displayed by a bold arrow, low-level basal transcription by a dotted arrow (A) tTA (B) tTS (C) rtTA.

than Ptet-1, but also do not reach its maximal activation

level Thus, the regulatory window for target gene

expres-sion is shifted and expanded due to the stronger reduction of

the basal activity Integration site-dependent leakiness has

been attributed to enhancers located close to the integration

site of the target gene construct Besides screening additional

clones until one harboring the desired properties is found,

the problem has been approached by insulating Ptet-1 from

external activating signals through insertion of a chicken

lysozyme matrix attachment region just upstream of Ptet-1

[87] or by flanking the target gene expression unit with either

chicken b-globin insulators [90] or SCS and SCS’ boundary

elements from Drosophila [86]

A different strategy was adopted by engineering a

tc-controlled trans-silencer protein [81] Fusion of the

KRAB domain of Kox1 [97] to TetR yielded a hybrid

protein called tTS, that not only substantially repressed

basal transcription from Ptet-1 even if the tet operators were

located 3 kbp distant from the minimal promoter, but also

efficiently down-regulated gene expression from a CMV

enhancer-driven Ptet-1 [83] This strategy therefore appears

to be more versatile in coping with unwanted target gene

expression than the promoter adaptation proposed above

In addition and in contrast to tTA, the tc-dependent

silencing of complex promoters offers the unique possibility

of reversibly down-regulating the expression of cellular

genes on top of their normal regulation The KRAB

domain is inactive in S cerevisiae and Drosophila where it

was replaced with repression domains from the proteins

SSN6 [82], knirps, giant or dCtBP [83]

The expression of transfected genes can be rapidly

repressed in mammalian cells by epigenetic mechanisms

[98] Although this transgene silencing is not specific for

the Tet system, it is often observed for genes under tc

control due to its frequent usage as conditional expression

system Approaches to achieve stable gene expression have

been to: (a) screen many transfected clones; (b) the use of

lentiviral vectors [99]; (c) replace the viral promoters that

direct expression of the transregulators with promoters of

human origin [100]; (d) use chromatin insulator sequences

to protect transgene expression [98]; or (e) couple transgene

expression to a selectable marker via an IRES element

[101] or by fusion of the transregulator with green

fluorescent protein (GFP) [102] Note that in this fusion

protein GFP is connected to the DNA-binding domain of

TetR which can interfere with nonspecific DNA-binding

activity of TetR at low levels of dox (see Fig 2A in [102]

and [78]) The few published examples make it impossible

to recommend one of the strategies for use in establishing

homogenous expression of transgenes, but silencing of

transregulator expression is not completely suppressed by

the use of lentiviral vectors [103] or insulator sequences

[101]

Modifications of the Tet transregulators

The TetR–VP16 fusion works very well in many cases, but

may not be optimal for all applications Structure–function

studies based on powerful selection and screening systems in

E coli [104,105] and in S cerevisiae [88] have lead to a

profound understanding of how DNA binding, inducer

binding and dimerization function in TetR This

informa-tion can be used to find soluinforma-tions to some of the problems and limitations that arise for Tet system applications in eukaryotes

Alterations of the activator domain of tTA Especially for gene therapy, concern about a viral protein is often voiced, as humoral as well as cellular immune response against the VP16 protein has been found in herpes simplex infected humans [106–108] Thus, immune res-ponses against transactivators containing the VP16 domain cannot be rigorously excluded, although they have not been observed so far in a mouse model using reverse tTA (rtTA; Tet-On) [109] Two solutions circumventing this concern have been developed: (a) the VP16 domain has been replaced by three repeats of a minimal activation domain derived from a 12-amino acid activating patch of the VP16 protein (tTA2 [76]); and (b) a variety of human activator domains from the acidic, glutamine-rich, serine/threonine-rich and proline-serine/threonine-rich functional groups were tested for their ability to replace the VP16 domain When fused to TetR, only acidic activation domains were highly active [78–80] Minor activation was observed with the serine/threonine-rich domains from the transcription factors ITF-1, ITF-2, and MTF-1 Transactivators with activation potentials spanning more than three orders of magnitude have been generated by combination of various minimal activation domains (see above; [76]) They are attractive for combined

knock-in/knock-out strategies to convey tissue-specific expression of the transactivator, while at the same time inactivating expression of the genomic copy of the target gene Expression of the regulatory protein is then an invariant function of the genomic locus and, if too high, can lead to squelching [110] This can be addressed by employing a transactivator with reduced activation poten-tial as these are tolerated in the cell at higher concentrations [76]

Conversion of TetR to reverse TetR Eukaryotic gene regulation by tTA shows a high dynamic range and works consistently well, but has several practical drawbacks Tc has to be continually present to keep expression of the gene of interest downregulated Although

tc is not toxic at the levels utilized in gene regulation, prolonged exposure to the antibiotic is not always desirable

in transgenic animals nor is it possible in gene therapy Furthermore, induction of target genes is mostly slow as it requires removal of the drug from the culture or organism

To be able to control the time point of induction more precisely, and since organisms are more easily saturated with an effector than depleted of it [111,112], reverse TetR variants which bind tetO only in the presence of tc were searched for and found (Fig 3C) Screening in E coli [113] and in S cerevisiae [88] revealed that a small number of mutations in TetR can lead to that phenotype [113] Once this was discovered, intensive screening led to rtTA alleles in which the initial disadvantages of occasional background expression and low sensitivity for dox were eliminated [88] The rtTA-S2 allele was obtained by screening for reduced background expression and rtTA-M2 was the result of screening for higher sensitivity towards dox starting from

the alterations in rtTA-S2 that are responsible for the

reverse phenotype [88] None of the exchanges found in

these new alleles were present in the original rtTA The

mutations leading to the reverse phenotype are located at

the interface between the DNA reading head and the

protein core or in the last turn of helix a6 that undergoes a

conformational change upon inducer binding Structural

analysis of the DNA-bound form of TetR has led to the

proposal that the mutations present in rtTA [113] restrict

the repressor to a noninducible conformation and lock the

DNA-binding domains in the position necessary for

oper-ator binding [8] Taken together, the phenotype of rtTA can

be improved and designed by using appropriate screens

Tet transregulators vary in their sensitivity towards

tetracyclinic inducers

The tTA and rtTA variants presently employed in

eukary-otic gene regulation display differential sensitivity towards

tc and its derivatives While tTA can be induced by tc,

dox and atc [114], reverse transactivators respond only to

dox and atc [113] and tTSGis about twofold less sensitive to

dox than rtTA [115] The response range of tTA to dox (0.1–

10 ngÆmL)1) is clearly lower and, more importantly,

non-overlapping with that of rtTA to dox (100–3000 ngÆmL)1)

[114], but slightly overlapping with that of the more sensitive

rtTA2s-M2 allele (2–200 ngÆmL)1) [88] The molecular

mechanisms responsible for these different sensitivities are

presently unknown The isolation of a tc-like antagonist for

TetR [116] and the demonstration of its activity in

transgenic plants [117] make it seem likely that alternative

inducers for TetR can be identified by screening

The DNA binding specificity of Tet-transregulators

can be changed

Structure–function analyses of TetR–tetO interactions had

shown that only few changes (shown in Fig 4) in the DNA

binding helix–turn–helix motif of TetR suffice to switch the

recognition specificity from the 19-base pair wild-type tetO

to variants containing symmetric exchanges of bases at

position 4 (tetO-4C [118]) and position 6 (tetO-6C [119])

The TetR mutants were converted into the transactivators

tTA24Cor tTA26Cand minimal promoters Ptet4and Ptet6

were constructed with the respective tetO variants [114]

DNA binding of the modified transactivators is efficient;

in transient transfections in HeLa cells, they specifically achieve induction factors between 2000 and 8000 and are, thus, as active as wild-type tTA2 Moreover, they are also highly specific, as they induce the converse operator less than twofold [114] Modulation of the DNA-binding specificity is not confined to tTA Alleles specific for the 4C- [120] and 6C-tet operators [114] have been constructed with rtTA and also regulate tc-dependent expression units efficiently This now leaves us with different tTA- and rtTA-operator combinations capable of controlling gene expres-sion tightly over a wide range of inducer concentrations Mastering subunit recognition of TetR

Comparison of the TetR primary structures reveals 38–90% identical amino acids overall, but only 18% in the four-helix bundle involved in dimerization Detailed structural infor-mation [19] of the dimerization interface [121] suggested that TetR proteins from individual classes would not readily form heterodimers The modular architecture of TetR allows the combination of a class B DNA-binding domain with the inducer-binding and/or dimerization domains of Tet repressors from other classes [121] Fusion to the reading head from TetR(B) increases activity of Tet repressors from several other classes [121] and ensures tight binding to the tetO boxes from Tn10 [122] Class B TetR does not form heterodimers with Tet repressors from classes

D [121], E [93,114,120], or G [115] The fusion points can be chosen with some flexibility; functional chimeras have been obtained either by connecting the entire protein core from TetR(D) or TetR(E) to a TetR(B) DNA-reading head [114,120,121] or by replacing the four-helix bundle formed

by the helices a8 and a10 from both subunits (see Fig 2), with the respective region from TetR(G) [115] The resulting transrepressors or transactivators regulate gene expression efficiently and do not form heterodimers as demonstrated in DNA-retardation assays [114], immunoprecipitation and FACS analysis [115] opening up the possibility to introduce two or more TetR-based regulatory proteins into the same cell without having to cope with the disadvantages of heterodimer formation [114,115]

Combinatorial Tet regulation solves special problems and allows sophisticated

applications

The previous section has shown that DNA-binding speci-ficity, subunit recognition and response to the inducer can

be altered in TetR Fig 5 gives an overview of the present state of the Tet modules that are available for use and the following section presents a few principles of how the modular nature of the transregulators can be exploited to address specific experimental requirements and open up new applications for conditional regulation

Expression can be switched between two alleles

of one gene The expression of two genes or of two alleles of one gene can

be controlled in a mutually exclusive manner by combining different dimerization domains, different operator-binding

Fig 4 Operator specificity combinations for the Tet system The

pri-mary structure of the TetR(B) recognition helix a3 and the flanking

loops is given in standard one-letter abbreviations The entire sequence

of tetO 2 is shown with the palindromic center marked by an asterisk

and the base numbering shown above one operator half-side The

exchanges in TetR and tetO are highlighted in inverse print for each

matching pair (wt, 4C, 6C).

specificities and by exploiting the differential sensitivity of

Tet transregulators towards tetracyclines [114] Interference

between the two expression units is excluded by using a tTA

allele with an alternative class E or G dimerization domain

and by furnishing rtTA with a modified DNA-binding

domain that contacts the tetO-4C operator in Ptet4

speci-fically Expression of the wild-type allele, for example, is

placed under tTA control and represents the normal state of

the cell A knockout situation can be generated by adding

either tc or, alternatively, atc or dox at concentrations

between 10 and 100 ngÆmL)1which dissociates tTA from

the promoter but does not lead to DNA binding by rtTA

[114] Maintaining the intermediate concentration of dox

needed to shut down expression of both alleles will be

feasible in cell culture applications In transgenic animals,

however, the necessary fine-tuning of a dox or atc

concen-tration may prove impossible suggesting instead the use of tc

to shut down tTA-dependent gene expression without

interfering with regulation by rtTA To switch to the

expression of the mutant allele requires atc or dox

concen-trations of 1 lgÆmL)1or more

Such a dual control system can provide valuable

insights into developmental and pathogenic processes

One can imagine shutting down expression of a tumor

suppressor while inducing expression of an oncogene to

study cancerogenesis Switching off expression of the

oncogene after tumor formation can establish whether the

respective protein is a valid target for therapeutic

inter-vention One could also switch from a wild-type to a

mutant allele at a defined developmental state of the

organism and then return to wild-type expression at a later

stage This type of regulatory circuit can also deliver an

additional degree of freedom to gene therapeutic

strat-egies) one regulatory circuit may be used to control a

therapeutic gene, while the other may be exploited to serve

as a suicide switch to terminate the treatment once the

therapeutic goal has been reached or, if necessary, in case

of emergency

One gene can be regulated stringently by conversely acting transrepressor and transactivator

Detectable levels of transgene expression in animals or cells

in which the transactivator is not active can limit the usefulness of any conditional expression system for mode-ling complex biological processes or evaluating the effects of

a gene product For the Tet system, this transgene leakage has been attributed either to basal activity of the respective tetO-based minimal promoter used (see above; [115]); or, in systems with rtTA, to residual binding of the reverse transactivator to tetO in the absence of dox [123,124] A stringently controlled regulatory system can now be accomplished by combining a trans-silencer with a reverse transactivator, since heterodimer formation and concomit-ant phenotype blurring will be prevented if the trans-silencer

is equipped with a dimerization domain from the TetR classes E or G Thus, both transregulators bind in a mutually exclusive manner Gene expression is actively repressed in the absence of dox by the binding of tTSE/tTSG

to the minimal promoter Upon addition of dox, tTSE/tTSG dissociates from tetO, allowing the reverse transactivator to bind and activate transcription This setup efficiently reduces background expression in yeast [82], in mammalian cell lines [93,115,120] and in transgenic animals [125–127], while affecting the maximal expression level only slightly [128] or not at all [93]

Transgenes can be expressed in a graded

or in a binary manner Transcriptional control has generally been assumed to operate as a binary switch with on/off characteristics [129,130], but several examples displaying graded changes

in gene expression have recently been published [131,132] The manner of gene expression might well be a key factor in programs of cell differentiation or stimulus response Different regulatory setups of the Tet system allow a transgene to be expressed in one or the other manner [133– 135], enabling not only an analysis of a gene’s function, but also of its mode of expression When tTA and rtTA are expressed constitutively in mammalian cells and also in

S cerevisiae, they drive transgene expression in a dose-dependent, graded manner [133,135] However, when rtTA was expressed in S cerevisiae under conditions of positive feedback using an autoregulatory circuit, the cell population was clearly divided into regulator-expressing and nonex-pressing cell pools [135] In mammalian cells, the combined usage of tTSGand rtTA also led to bimodal expression of the GFP reporter (see Fig 3 in [134]) Although not formally proven, we assume that a bimodal expression pattern will not be observed for all repressor/activator combinations, but only for those in which the sensitivity of tTS for the inducer is lower than that of the rtTA allele used,

as is the case for tTSG(compare the dose–response curve of tTSEand rtTA of Fig 4 in [93] with the one for tTSGand rtTA of Fig 2 of [134]) This will ensure that rtTA is preloaded with inducer and ready to activate transcription the moment the dox concentrations needed for binding to

Fig 5 The Tet toolbox TetR modules and regulatory domains are

displayed with the possible combinations The different binding

func-tions of TetR were coded in different shades of gray and placed at their

approximate position in the protein, but not drawn to scale The TetR

variants characterized were classified in the corresponding module.

The regulatory domains that can be fused to TetR are coded in

dif-ferent shades of gray according to their viral, human, insect or fungal

origin Note that not every possible combination of modules need

result in a transregulator with acceptable regulatory properties.

tTSG are reached and tetO is subsequently released In

principle, only two regulatory states are observed: either the

tetOsites are fully occupied with tTSGand gene expression

is shut off, or they are saturated with the rtTA variant,

resulting in full transcriptional activation The consequence

is a binary expression pattern of the target gene While this

setup already works with rtTA, the effect should be even

more pronounced with rtTA2s-M2, as its inducer response

range overlaps completely with that of tTSG

Highlighting the regulatory potential

and looking into the future

The properties and the adaptability of Tet regulation as

presented in the previous sections allow its use in many

different applications We would like to demonstrate this

enormous variability by referring to a few key studies that,

in our opinion, highlight the potential of Tet regulation

Regulation by tetracyclines is sensitive and efficient

enough to control target gene expression in pathogenic

organisms even when they have been injected into a

mammalian host The role of individual genes in infection

and pathogenesis can, thus, be probed and their validity as

targets for therapeutic intervention determined in an in vivo

disease model [136] This has not only been demonstrated

for trypanosomes [137,138], but also for common human

pathogens like Staphylococcus aureus [139] and Candida

glabrata[140] In the fungus, squalene synthase [136] and

sterol 14a-demethylase [141] were, thus, shown not to be

ideal targets for antifungal development

The successful expression of the diphtheria toxin A

subunit by tTA/Ptet-1 in transgenic mice has demonstrated

the stringency of regulation that can be reached with the Tet

system [142] Although mouse lines that carried the target

transgene were obtained at an approximately 10-fold lower

frequency than normal, those that were established

regula-ted the transgene efficiently Induction of toxin expression

led to cell death and development of cardiomyopathies

Stringent control of transgene expression using rtTA has

also been achieved in HeLa cells for the Shiga toxin B

subunit [143], for the proapoptotic gene PUMA in SAOS-2

and H1299 cell lines [144] and, using rtTA2s-S2 in transgenic

mice, for Cre-recombinase [145]

The strength of a true conditional system) the

possibi-lity to switch gene expression on and off at leisure and

repeatedly) represents a powerful method with which to

explore the relationship between mutant protein expression

and disease progression This has become evident upon

studying transgenic mouse models for cancer and

neuro-logical disorders Here, the use of tTA and rtTA to control

expression of an oncogene revealed for solid tumors

[146,147] and for leukemias [148,149] that the oncogene is

not only necessary for tumor formation but also for tumor

maintenance, suggesting pharmacological inactivation of

oncogenes as a possible therapeutic strategy for cancer This

assumption has been substantiated by the unexpected

observation that, after having gone through one cycle of

MYC-gene expression and silencing, reactivation of the

oncogene does not lead to tumor regrowth, but rather to

apoptosis [150] Similar effects have also been found for

neurological disorders In a conditional model of

Hunting-ton’s disease, mice expressing a mutated huntingtin

fragment in the brain demonstrated that its continuous supply was needed to maintain the characteristic neuro-pathology and behavioral phenotype, raising the possibility that the disease may be reversible by targeting the causative agent [151]

Regulation by the Tet system has also had a significant impact on behavioral studies Expression of constitutively active mutant forms of the calcium/calmodulin dependent kinase II or calcineurin in the brain of adult mice resulted in altered synaptic plasticity and impairments in spatial memory storage and retrieval, but these deficits were fully reversed when transgene expression was suppressed [84,152] Because expression of the transgene was limited

to the hippocampus, this structure was additionally proven

to be the site responsible for the behavioral effects In a different example, knockout mice lacking the serotonin 1A receptor show increased anxiety-like behavior which could

be rescued by conditional expression, but only if the receptor was synthesized during the early postnatal period

in the hippocampus and cortex [153]

Nevertheless, improvement and additions to the Tet system, among them the regulatory components, are still possible and necessary Promoter development has not received the same degree of attention as the transregulators The number of tetO elements and their spacing [154], as well

as the linker sequence separating the operators [155] have not been optimized yet It remains to be seen if an ideal minimal promoter with no intrinsic leakiness supporting very high-level activation can be identified or designed Fortunately, screens for regulators with improved prop-erties can now be performed in eukaryotic systems [88] and, as an example, the isolation of novel Tet regulators which recognize nonantibiotically active tetracyclines or even nontetracyclinic inducers, would be of great benefit They would not only facilitate gene therapy applications which, at the moment, can be impaired by the use of tetracyclines in anti-infective therapy or their misuse as growth promoting additives to animal food If these novel inducers are not only ecologically safe, but also easy and nonexpensive to manufacture, the inducer–regulator pair-ing could also be useful in insect population control uspair-ing dominant, repressible, lethal genetic systems [156,157] and might even introduce regulation by the Tet system to crop plants They would add to the repertoire of transregulators and finally, since multiple dimerization and DNA-binding specificities are already present, allow fully independent expression control of more than one gene by the Tet system

A major experimental challenge will be to express a target gene within its physiological window, which might depend

on environmental stimuli and even change during develop-ment, since over- or underexpression often results in altered phenotypes [131] or pathologies While tc-controlled expres-sion can mimic the natural level [146], this must not always

be the case A solution might be precise promoter targeting

by tetO elements, to minimally interfere with gene expres-sion This will be difficult and will require extensive knowledge about the influence of chromatin structure on gene expression and its sensitivity to perturbation, partic-ularly when regulatory regions are modified [158] But, if successful, this approach will provide an additional degree

of freedom to manipulate gene expression, as the existing

transregulators can be used to activate or silence gene

expression, in addition to and independent of the

promo-ter’s natural expression pattern

Conclusion

The Tet system is the most widely used regulatory system

for conditional gene expression at the moment The

increasing number of: (a) cell lines stably transfected with

tTA and rtTA; (b) cell lines harboring tTA or rtTA that

have been derived from transgenic mice; and (c) transgenic

mice expressing either the transregulators via cell-type

specific promoters or a target gene under Ptet-1 control will

greatly facilitate genetic studies by allowing combination of

the existing components instead of having to generate all cell

and mouse lines, a costly and time-consuming process

Ongoing improvement of the existing components as well as

the continuous addition of new components to extend its

applicability have turned the Tet system into a highly

versatile and flexible regulatory system that can be adapted

to many different applications Starting from an extensive

knowledge-base of TetR structure–activity relationships

and the strength of the genetic screening and selection

systems in both pro- and eukaryotes, the Tet system is

becoming more and more capable of modeling the

sophis-ticated regulatory setups needed [48,51] to analyze complex

and multifactor biological processes in development and

disease, thereby not only improving our understanding of

living organisms, but also revealing novel and innovative

approaches to treat maladies

Acknowledgements

This work was supported by the Bayerische Forschungsstiftung

through their FORGEN initiative, by the Deutsche

Forschungsgeme-inschaft through SFB473 and the Fonds der Chemischen Industrie

Deutschlands We would also like to thank Dr Anja Knott and Felix

Kuphal for critical reading of the manuscript.

References

1 Yamaguchi, A., Udagawa, T & Sawai, T (1990) Transport of

divalent cations with tetracycline as mediated by the transposon

Tn10-encoded tetracycline resistance protein J Biol Chem 265,

4809–4813.

2 Yamaguchi, A., Iwasaki-Ohba, Y., Ono, N., Kaneko-Ohdera,

M & Sawai, T (1991) Stoichiometry of metal-tetracycline/H+

antiport mediated by transposon Tn10-encoded tetracycline

resistance protein in Escherichia coli FEBS Lett 282, 415–418.

3 Levy, S.B., McMurry, L.M., Barbosa, T.M., Burdett, V.,

Courvalin, P., Hillen, W., Roberts, M.C., Rood, J.I & Taylor,

D.E (1999) Nomenclature for new tetracycline resistance

determinants Antimicrob Agents Chemother 43, 1523–1524.

4 Tauch, A., Pu¨hler, A., Kalinowski, J & Thierbach, G (2000)

TetZ, a new tetracycline resistance determinant discovered in

Gram-positive bacteria, shows high homology to Gram-negative

regulated efflux systems Plasmid 44, 285–291.

5 Tauch, A., Go¨tker, S., Pu¨hler, A., Kalinowski, J & Thierbach, G.

(2002) The 27.8-kb R-plasmid pTET3 from Corynebacterium

glutamicum encodes the aminoglycoside adenyltransferase gene

cassette aadA9 and the regulated tetracycline efflux system Tet 33

flanked by active copies of the widespread insertion sequence

IS6100 Plasmid 48, 117–129.

6 Hillen, W & Berens, C (1994) Mechanisms underlying expres-sion of Tn10 encoded tetracycline resistance Annu Rev Micro-biol 48, 345–369.

7 Takahashi, M., Altschmied, L & Hillen, W (1986) Kinetic and equilibrium characterization of the Tet repressor-tetracycline complex by fluorescence measurements Evidence for divalent metal ion requirement and energy transfer J Mol Biol 187, 341–348.

8 Orth, P., Schnappinger, D., Hillen, W., Saenger, W & Hinrichs,

W (2000) Structural basis of gene regulation by the tetracycline inducible Tet repressor-operator system Nature Struct Biol 7, 215–219.

9 Lederer, T., Kintrup, M., Takahashi, M., Sum, P.E., Ellestad, G.A & Hillen, W (1996) Tetracycline analogs affecting binding

to Tn10-encoded Tet repressor trigger the same mechanism of induction Biochemistry 35, 7439–7446.

10 McMurry, L., Petrucci, R.E Jr & Levy, S.B (1980) Active efflux

of tetracycline encoded by four genetically different tetracycline resistance determinants in Escherichia coli Proc Natl Acad Sci USA 77, 3974–3977.

11 Moyed, H.S., Nguyen, T.T & Bertrand, K.P (1983) Multicopy Tn10 tet plasmids confer sensitivity to induction of tet gene expression J Bacteriol 155, 549–556.

12 Nguyen, T.N., Phan, Q.G., Duong, L.P., Bertrand, K.P & Lenski, R.E (1989) Effects of carriage and expression of the Tn10 tetracycline-resistance operon on the fitness of Escherichia coli K12 Mol Biol Evol 6, 213–225.

13 Berg, C.M., Liu, L., Wang, B & Wang, M.D (1988) Rapid identification of bacterial genes that are lethal when cloned on multicopy plasmids J Bacteriol 170, 468–470.

14 Eckert, B & Beck, C.F (1989) Overproduction of transposon Tn10-encoded tetracycline resistance protein results in cell death and loss of membrane potential J Bacteriol 171, 3557–3559.

15 Bertrand, K.P., Postle, K., Wray, L.V Jr & Reznikoff, W.S (1984) Construction of a single-copy promoter vector and its use

in analysis of regulation of the transposon Tn10 tetracycline resistance determinant J Bacteriol 158, 910–919.

16 Epe, B & Woolley, P (1984) The binding of 6-demethyl-chlortetracycline to 70S, 50S and 30S ribosomal particles: a quantitative study by fluorescence anisotropy EMBO J 3, 121–126.

17 Orth, P., Cordes, F., Schnappinger, D., Hillen, W., Saenger, W & Hinrichs, W (1998) Conformational changes of the Tet repressor induced by tetracycline trapping J Mol Biol 279, 439–447.

18 Hinrichs, W., Kisker, C., Du¨vel, M., Mu¨ller, A., Tovar, K., Hillen, W & Saenger, W (1994) Structure of the Tet repressor-tetracycline complex and regulation of antibiotic resistance Science 264, 418–420.

19 Kisker, C., Hinrichs, W., Tovar, K., Hillen, W & Saenger, W (1995) The complex formed between Tet repressor and tetra-cycline-Mg2+reveals mechanism of antibiotic resistance J Mol Biol 247, 260–280.

20 Orth, P., Saenger, W & Hinrichs, W (1999) Tetracycline-chelated Mg2+ ion initiates helix unwinding in Tet repressor induction Biochemistry 38, 191–198.

21 Orth, P., Schnappinger, D., Sum, P.E., Ellestad, G.A., Hillen, W., Saenger, W & Hinrichs, W (1999) Crystal structure of the Tet repressor in complex with a novel tetracycline, 9-(N,N-dimethylglycylamido)-6-demethyl-6-deoxy-tetracycline J Mol Biol 285, 455–461.

22 Saenger, W., Orth, P., Kisker, C., Hillen, W & Hinrichs, W (2000) The tetracycline repressor-A paradigm for a biological switch Angew Chem Int Ed Engl 39, 2042–2052.

23 Matthews, K.S., Falcon, C.M & Swint-Kruse, L (2000) Relieving repression Nature Struct Biol 7, 184–187.

24 Mu¨ller, G., Hecht, B., Helbl, V., Hinrichs, W., Saenger, W &

Hillen, W (1995) Characterization of non-inducible Tet repressor

mutants suggests conformational changes necessary for

induc-tion Nature Struct Biol 2, 693–703.

25 Hecht, B., Mu¨ller, G & Hillen, W (1993) Noninducible Tet

repressor mutations map from the operator binding motif to the

C terminus J Bacteriol 175, 1206–1210.

26 Tiebel, B., Radzwill, N., Aung-Hilbrich, L.M., Helbl, V.,

Steinhoff, H.J & Hillen, W (1999) Domain motions

accom-panying Tet repressor induction defined by changes of interspin

distances at selectively labeled sites J Mol Biol 290, 229–240.

27 Tiebel, B., Garke, K & Hillen, W (2000) Observing

conforma-tional and activity changes of Tet repressor in vivo Nature Struct.

Biol 7, 479–481.

28 Tiebel, B., Aung-Hilbrich, L.M., Schnappinger, D & Hillen, W.

(1998) Conformational changes necessary for gene regulation by

Tet repressor assayed by reversible disulfide bond formation.

EMBO J 17, 5112–5119.

29 Berens, C., Schnappinger, D & Hillen, W (1997) The role of the

variable region in Tet repressor for inducibility by tetracycline.

J Biol Chem 272, 6936–6942.

30 Kintrup, M., Schubert, P., Kunz, M., Chabbert, M., Alberti, P.,

Bombarda, E., Schneider, S & Hillen, W (2000) Trp scanning

analysis of Tet repressor reveals conformational changes

asso-ciated with operator and anhydrotetracycline binding Eur J.

Biochem 267, 821–829.

31 Scholz, O., Kintrup, M., Reich, M & Hillen, W (2001)

Mechanism of Tet repressor induction by tetracyclines: length

compensates for sequence in the a8-a9 loop J Mol Biol 310,

979–986.

32 Argast, M & Beck, C.F (1984) Tetracycline diffusion through

phospholipid bilayers and binding to phospholipids Antimicrob.

Agents Chemother 26, 263–265.

33 Argast, M & Beck, C.F (1985) Tetracycline uptake by

suscep-tible Escherichia coli cells Arch Microbiol 141, 260–265.

34 Katiyar, S.K & Edlind, T.D (1991) Enhanced antiparasitic

activity of lipophilic tetracyclines: role of uptake Antimicrob.

Agents Chemother 35, 2198–2202.

35 Pezeshk, A., Pezeshk, V., Firlej, A., Wojas, J & Subczynski,

W.K (1993) Transport of spin-labeled tetracycline across model

and biological membranes Life Sci 52, 1071–1078.

36 Sigler, A., Schubert, P., Hillen, W & Niederweis, M (2000)

Permeation of tetracyclines through membranes of liposomes and

Escherichia coli Eur J Biochem 267, 527–534.

37 Sumita, Y & Shishido, K (1985) Regulation of tetracycline

accumulation in Bacillus subtilis bearing B subtilis plasmid

pNS1981 FEMS Microbiol Lett 30, 403–406.

38 Hutchings, B.L (1969) Tetracycline transport in Staphylococcus

aureus H Biochim Biophys Acta 174, 734–748.

39 Yamaguchi, A., Ohmori, H., Kaneko-Ohdera, M., Nomura,

T & Sawai, T (1991) DpH-dependent accumulation of

tetracycline in Escherichia coli Antimicrob Agents Chemother 35,

53–56.

40 Nikaido, H & Thanassi, D.G (1993) Penetration of lipophilic

agents with multiple protonation sites into bacterial cells:

tetra-cyclines and fluoroquinolones as examples Antimicrob Agents

Chemother 37, 1393–1399.

41 Thanassi, D.G., Suh, G.S & Nikaido, H (1995) Role of outer

membrane barrier in efflux-mediated tetracycline resistance of

Escherichia coli J Bacteriol 177, 998–1007.

42 Gossen, M., Bonin, A.L & Bujard, H (1993) Control of gene

activity in higher eukaryotic cells by prokaryotic regulatory

ele-ments Trends Biochem Sci 18, 471–475.

43 Saez, E., No, D., West, A & Evans, R.M (1997) Inducible gene

expression in mammalian cells and transgenic mice Curr Opin.

Biotechnol 8, 608–616.

44 Shockett, P.E & Schatz, D.G (1996) Diverse strategies for tet-racycline-regulated inducible gene expression Proc Natl Acad Sci USA 93, 5173–5176.

45 Blau, H.M & Rossi, F.M.V (1999) Tet B or not tet B: advances

in tetracycline-inducible gene expression Proc Natl Acad Sci USA 96, 797–799.

46 Baron, U & Bujard, H (2000) Tet repressor-based system for regulated gene expression in eukaryotic cells: principles and advances Methods Enzymol 327, 401–421.

47 Fussenegger, M (2001) The impact of mammalian gene regula-tion concepts on funcregula-tional genomic research, metabolic engine-ering, and advanced gene therapies Biotechnol Prog 17, 1–51.

48 Lewandoski, M (2001) Conditional control of gene expression in the mouse Nature Rev Genet 2, 743–755.

49 Yamamoto, A., Hen, R & Dauer, W.T (2001) The ons and offs

of inducible transgenic technology: a review Neurobiol Dis 8, 923–932.

50 Gossen, M & Bujard, H (2002) Studying gene function in eukaryotes by conditional gene inactivation Annu Rev Genet.

36, 153–173.

51 Jonkers, J & Berns, A (2002) Conditional mouse models of sporadic cancer Nature Rev Cancer 2, 251–265.

52 Zhu, Z., Zheng, T., Lee, C.G., Homer, R.J & Elias, J.A (2002) Tetracycline-controlled transcriptional regulation systems: advances and application in transgenic animal modeling Semin Cell Dev Biol 13, 121–128.

53 Gatz, C & Quail, P.H (1988) Tn10-encoded tet repressor can regulate an operator-containing plant promoter Proc Natl Acad Sci USA 85, 1394–1397.

54 Gatz, C., Kaiser, A & Wendenburg, R (1991) Regulation of a modified CaMV 35S promoter by the Tn10-encoded Tet repressor in transgenic tobacco Mol Gen Genet 227, 229–237.

55 Yan, S., Myler, P.J & Stuart, K (2001) Tetracycline regulated gene expression in Leishmania donovani Mol Biochem Parasitol.

112, 61–69.

56 Wirtz, E & Clayton, C (1995) Inducible gene expression in trypanosomes mediated by a prokaryotic repressor Science 268, 1179–1183.

57 Dingermann, T., Frank-Stoll, U., Werner, H., Wissmann, A., Hillen, W., Jacquet, M & Marschalek, R (1992) RNA poly-merase III catalysed transcription can be regulated in Saccharo-myces cerevisiae by the bacterial tetracycline repressor-operator system EMBO J 11, 1487–1492.

58 Dingermann, T., Werner, H., Schu¨tz, A., Zu¨ndorf, I., Nerke, K., Knecht, D & Marschalek, R (1992) Establishment of a system for conditional gene expression using an inducible tRNA sup-pressor gene Mol Cell Biol 12, 4038–4045.

59 Ulmasov, B., Capone, J & Folk, W (1997) Regulated expression

of plant tRNA genes by the prokaryotic tet and lac repressors Plant Mol Biol 35, 417–424.

60 Ohkawa, J & Taira, K (2000) Control of the functional activity

of an antisense RNA by a tetracycline-responsive derivative of the human U6 snRNA promoter Hum Gene Ther 11, 577–585.

61 Weinberg, R.A & Penman, S (1968) Small molecular weight monodisperse nuclear RNA J Mol Biol 38, 289–304.

62 Hutva´gner, G & Zamore, P.D (2002) A microRNA in a mul-tiple-turnover RNAi enzyme complex Science 297, 2056–2060.

63 Yu, J.Y., DeRuiter, S.L & Turner, D.L (2002) RNA interference

by expression of short-interfering RNAs and hairpin RNAs in mammalian cells Proc Natl Acad Sci USA 99, 6047–6052.

64 McManus, M.T & Sharp, P.A (2002) Gene silencing in mammals by small interfering RNAs Nature Rev Genet 3, 737–747.

65 Frohberg, C., Heins, L & Gatz, C (1991) Characterization of the interaction of plant transcription factors using a bacterial repressor protein Proc Natl Acad Sci USA 88, 10470–10474.