A

Tetracycline

Doxycyline

Figure 2 Chemical structure of tetracycline and doxycycline, inducers of the Tet-system. (A) Tetracycline effectively stimulates tTA (Tet-OFF system) but not rtTA (Tet-ON system). (B) Doxycycline activates both Tet-ON and Tet-OFF systems.

Doxycyline

Figure 2 Chemical structure of tetracycline and doxycycline, inducers of the Tet-system. (A) Tetracycline effectively stimulates tTA (Tet-OFF system) but not rtTA (Tet-ON system). (B) Doxycycline activates both Tet-ON and Tet-OFF systems.

as tetO, should be efficiently (i.e., almost fully) occupied by the repressor to adequately inhibit transcription: this implies that when tight repression is required, the repressor itself should be expressed at high and possibly toxic levels within the target cells. Moreover, constitutively high intracellular concentrations of the repressor may severely limit the induc-ibility of the system (24). Finally, studies in cells have demonstrated that efficient silencing of eukaryotic promoters requires proper positioning of tetO according to rules that may vary in accordance to the target promoter (25).

Regulatory systems based on ligand-regulated activators have less severe requirements: in the simplest case, the target promoter is constitutively silent and becomes active only when bound by the cognate activator. Notably, overexpression of the activator is not necessary, provided it is present above a certain threshold level. These considerations prompted H. Bujard and colleagues to convert the TetR into a positively acting factor and use it to regulate gene expression in eukaryotic cells (26).

B. Transcription Factors

TetR was converted to a transcriptional transactivator, called tTA, by fusing the VP16 transactivation domain of H. simplex virus to the C-terminus of TetR (26). tTA retained the DNA-binding and ligand-binding properties of TetR: it was prevented from binding to tetO by Tc or its derivatives, such as doxycycline (Dox), a most potent effector for both TetR and tTA. tTA proved capable of activating expression from tetO-containing promoters in a tetracycline-dependent manner in eukaryotic cells (26). In this ''Tet-OFF'' configuration, therefore, transcription is stimulated by drug withdrawal (Fig. 3). tTA retained an exquisite responsiveness to the effector molecule, in that it was completely inactivated by Dox concentrations as low as 20 ng/ml (26,27). Since its development, the Tet-OFF system has become one of the most widely used systems for modulating gene expression in cells and transgenic animals (28).

As outlined in the introduction, however, a system in which gene expression is induced by drug administration rather than by its withdrawal is highly preferable in the vast majority of gene therapy applications. To generate a Tc-inducible switch, the allosteric behavior of TetR was reversed. By combining chemical mutagenesis and genetic selection in E. coli, a mutant form of TetR was identified that carries four amino acid substitutions and is able to bind tetO only in the presence of tetracycline. Fusion of this mutated TetR to VP16 produced a reverse tTA (rtTA), capable of stimulating transcription only upon adding ligand (Tet-ON system, Fig. 4) (27). However, the same mutations that reversed the phenotype of tTR also affected its affinity for the inducer drug. In fact, rtTA was not induced by Tc and was about 100-fold less responsive to Dox than tTA (27). As a consequence, maximum activation of rtTA is achieved at Dox concentrations as high as 1-3 ^g/ml, which is only about 5-fold lower than the toxic dosage for mammalian cells (29). Moreover, rtTA displayed a significant basal activity that was mainly due to a residual binding of rtTA to

Figure 3 tTA-based, Tet-OFF system. tTA consists of the tetR fused to VP16 or VP16-derived activation domain. In the absence of the inducer drug, tTA binds and activates a promoter consisting of a minimal TATA box containing promoter located downstream of seven tet operator (tetO) sites. Upon addition of Tc or Dox, tTA is released from DNA and the promoter becomes inactive. For the sake of simplicity, only one molecule of tTA is represented.

Figure 3 tTA-based, Tet-OFF system. tTA consists of the tetR fused to VP16 or VP16-derived activation domain. In the absence of the inducer drug, tTA binds and activates a promoter consisting of a minimal TATA box containing promoter located downstream of seven tet operator (tetO) sites. Upon addition of Tc or Dox, tTA is released from DNA and the promoter becomes inactive. For the sake of simplicity, only one molecule of tTA is represented.

tetO in the absence of the inducer Drug (28-30). Both factors limit the potential of tTA for gene therapy applications (see below).

More recently, improved versions of rtTA have been isolated by functional selection in yeast (30). These novel trans-activators, called rtTA2S-S2 and rtTA2S-M2, carry different mutations compared to rtTA but retain Dox-dependency and display considerably lower activity in the off state (30-33). These latest versions of the rtTA also carry an optimized activation domain (VP16-F3) consisting of three tandem repeats of a 12 amino acid peptide derived from the VP16 activation domain (34). Finally, the cDNAs for these novel transactiva-tors have been optimized for expression in human cells. Because of their lower leakiness, both rtTA2S-S2 and rtTA2S-M2 were better inducers than rtTA and enabled hundreds-fold stimulation of gene expression in transiently and stably transfected cells (30-32). rtTA2S-S2 displays the lower unin-duced activity between these two novel versions, thus enabling more stringent control. rtTA2S-M2 is more responsive to Dox, as it is about 10-fold more sensitive to Dox than rtTA and is fully induced by Dox concentrations as low as 100-200 ng/ ml (30,33).

A modified version of the tet system has been used to repress transcription in a ligand-dependent manner. Dox-de-pendent silencing was achieved by fusing the transrepressor KRAB (Kruppel-associated box) domain of the human Kid protein to the tetR (35,36). In the absence of Dox, this transre-pressor, called tTS, was able to silence a CMV promoter engineered to contain seven tetO sequences in its 5' region (36). In the presence of the effector molecule, tTS is released from DNA and gene expression is induced. For the reasons outlined above, systems based on repression are probably not appropriate for gene therapy.

More interestingly, tTS has now become a standard technology in combination with rtTA to repress basal activity in uninduced conditions, thus widening the regulatory window of the Tet-ON system. A tTS unable to heterodimerize with rtTA was designed, which contains a dimerization domain of different class specificity compared to rtTA (37,38). As indicated in Fig. 5, in the absence of Dox, tTS binds to tetO and inhibits basal transcription: as Dox is added, tTS dissociates from the target DNA while rtTA becomes active and triggers transcription (37,38). Coexpressing tTS and rtTA (rtTA/tTS system) reduces the basal expression to almost un-

Figure 4 rtTA-based, Tet-ON system. rtTA consists of rTetR fused to an activation domain, usually derived from VP16. rtTA is inactive in the absence of the drug: following Dox administration, it binds and activates the target promoter. See text for additional details.

detectable levels in vivo (39-43). Despite these attractive features, one drawback of the rtTA/tTS system is that it is fully active only at high (^g/ml) Dox concentrations, because of the low responsiveness of rtTA to Dox (33,37,44). From this viewpoint, the use of the most advanced version of rtTA in combination with tTS proved useful. The properties of rtTA/ tTS, rtTA2S-S2/tTS, and rtTA2S-M2/tTS combined systems have been compared in a recent study (33). Both rtTA2S-S2/ tTS and rtTA2S-M2/tTS displayed a lower baseline activity and a 10-fold expanded dynamic range of induction than rtTA/ tTS, with thousands-fold induction measured in Dox-treated cells in the absence of detectable leakiness. Moreover, the rtTA2S-M2/tTS system displayed the highest Dox sensitivity, as it became fully active at concentrations as low as 100-200 ng/ml of Dox (33). Of course, the greater Dox-responsiveness of rtTA2S-M2/tTS makes this system more suitable for gene therapy applications in humans.

C. Tet-responsive Promoter

The Tet-responsive promoter in its more classical configuration (PhCMV*-1, also called Ptet-1) consists of a minimal version of the human cytomegalovirus immediate-early promoter fused 5' to a heptameric tetO repeat (tetO7) (29). In this configuration, the promoter may display a relatively high leaki-ness because of the intrinsic basal activity of the CMV minimal promoter (29). The use of different minimal promoter sequences, such as a modified mouse mammary tumor virus (MMTV) promoter or a core promoter derived from the plant viral 35S promoter, decreases the basal activity of the Tet-responsive promoter by one order of magnitude (45,46). However, such promoters are also considerably less potent than PhcMV*-1 upon induction with Dox: therefore, it may be particularly appropriate to use them in those cases where stringency of control is the most critical factor. It is worth mentioning that the leakiness of PhCMV*-1 is effectively suppressed by tTS also when coexpressed with the various activators (see above).

The heptameric version of the tetO was initially evaluated as the most potent operator, and therefore little or no effort was dedicated to modifying this architecture (26). In point of fact, the heptameric sequence is the standard target sequence for tTA, tTS, and the various versions of rtTA. It has been recently reported that IFNa-stimulated response elements (ISREs) are located in the linker regions between the heptam-

Figure 5 rtTA/tTS-based system, Tet-ON system. rtTA consists of rTetR fused to the VP16 activation domain and tTS consists of TetR fused to the KRAB repressor domain. rtTA and tTR recognize the same DNA sequence but cannot dimerize because they carry dimerization domains of different class specificity. In the absence of Dox, tTS binds and represses the target promoter. Following drug administration, tTS is released from DNA while rtTA interacts with the target promoter, thus triggering gene expression.

Figure 5 rtTA/tTS-based system, Tet-ON system. rtTA consists of rTetR fused to the VP16 activation domain and tTS consists of TetR fused to the KRAB repressor domain. rtTA and tTR recognize the same DNA sequence but cannot dimerize because they carry dimerization domains of different class specificity. In the absence of Dox, tTS binds and represses the target promoter. Following drug administration, tTS is released from DNA while rtTA interacts with the target promoter, thus triggering gene expression.

eric tetO sequences (47). Consequently, IFNa can stimulate tet promoter activity, thus interfering with the stringency of gene control (47). To avoid undesired induction, these ISREs would probably have to be deleted before this system can be used in humans.

D. Tet System for Gene Therapy Applications

The various versions of the Tet systems have been delivered in animal models by a variety of vectors and in several different tissues to regulate expression of reporter as well as therapeutic genes (29).

The tTA system has been and continues to be used to regulate transgene transcription in vivo in a variety of animal models (48-54). However, because of the unfavorable properties of an OFF-system in vivo, these reports can be considered as interesting proof-of-concept studies, which are nonetheless unlikely to move from the preclinical to the clinical setting.

In contrast, results obtained with the various configurations of the Tet-ON system are of greater interest and forecast future outcomes in humans. A few reports, listed in Table 1, have indeed demonstrated that the original rtTA-based system, delivered in mice and nonhuman primates using both viral and nonviral vectors, can be used for long-term regulation of gene expression in a Dox-dependent manner (55-58). However, a certain leakiness was observed, especially in those cases where rtTA was used for modulating expression of high-potency protein hormones, such as erythropoietin (Epo), the major regulator of erythropoiesis in mammals (59). In this case, even the low amount of protein produced in the unin-duced state stimulated a hematocrit (Hct) increase in untreated mice (31,55,57).

In vitro and in vivo leakiness was more pronounced in those cases where the rtTA cDNA was cloned downstream of strong enhancer/promoter elements and inserted with the tet-responsive gene into the same vector (55,60). This type of configuration probably results in very high intracellular concentrations of rtTA, which may thus unspecifically activate transcription. Moreover, the basal activity may be increased by the interference between the regulatory elements driving rtTA expression and the intrinsically ''leaky'' PhcMv*-1 promoter. As a matter of fact, more efficient control could be achieved by splitting the system into two vectors, thus allowing to cells to be transduced with substoichiometric amount of transactivator with respect to the regulated gene (56,57,61).

As expected on the basis of in vitro results, the novel versions of rtTA enabled tighter control when used in vivo (Table 1). Better results were obtained with the less leaky rtTA2S-S2. In a mouse model of hepatitis, intrahepatic delivery of a gutless-Ad vector carrying an rtTA2S-S2-regulated IFNa protected mice from the disease: protection was only observed in mice treated with Dox, which stimulated IFNa production in a Dox-dose dependent manner (62). rtTA2S-S2 also enabled tight control of Epo gene when delivered intramuscularly by electroinjection as plasmid DNA (31). Epo production and Hct levels could be modulated up to 300 days postinjection in response to Dox delivery and withdrawal, in the absence of any background expression in untreated mice. Notably, this result was achieved by using a single plasmid containing a CMV-rtTA2S-S2 expression cassette and the tet-responsive Epo gene (31). In the same experimental setting, rtTA2S-M2 displayed a certain degree of leakiness, which was, however, much lower than that of rtTA (31). The leakiness of rtTA2S-M2 was also evident in another study, in which a gutless-Ad vector containing rtTA2S-M2-regulated secreted alkaline phosphatase (SEAP) gene was delivered into mouse muscles (63). In vivo experiments thus confirmed that rtTA2S-S2 represented a major improvement over rtTA in terms of basal activity. Moreover, Dox-dose response experiments demonstrated that the system is activated in mice by oral dosages of Dox comparable to those normally used in clinical practice in humans (31). Nonetheless, vector-dose response experiments demonstrated that the tightness of control is partially lost at higher vector doses, thus imposing an upper limit to the amount of vector that can be delivered (31,33,62). Use of this activator may be appropriate in those cases in which a certain degree of basal expression level is acceptable.

Coexpression of tTS has been adopted as an alternative strategy to reduce the basal activity of rtTA following in vivo delivery. For historical reasons, initial studies involved the use of tTA with rtTA (Table 1). In a first study, the system

Table 1 Tet-on System: In Vivo Studies in Animal Models of Gene Therapy

Transactivator

Transgenea

Vector

Tissue

Species

Refs.

rtTA

Was this article helpful?

0 0

Post a comment