Can terminators be used as insulators into yeast synthetic gene circuits?

Abstract

Background: In bacteria, transcription units can be insulated by placing a terminator in front of a promoter. In this way promoter leakage due to the read-through from an upstream gene or RNA polymerase unspecific binding to the DNA is, in principle, removed. Differently from bacterial terminators, yeast S. cerevisiae terminators contain a hexamer sequence, the efficiency element, that strongly resembles the eukaryotic TATA box i.e. the promoter sequence recognized and bound by RNA polymerase II.

Results: By placing different yeast terminators (natural and synthetic) in front of the CYC1 yeast constitutive promoter stripped of every upstream activating sequences and TATA boxes, we verified that the efficiency element is able to bind RNA polymerase II, hence working as a TATA box. Moreover, terminators put in front of strong and medium-strength constitutive yeast promoters cause a non-negligible decrease in the promoter transcriptional activity.

Conclusions: Our data suggests that RNA polymerase II molecules upon binding the insulator efficiency element interfere with protein expression by competing either with activator proteins at the promoter enhancers or other RNA polymerase II molecules targeting the TATA box. Hence, it seems preferable to avoid the insulation of non-weak promoters when building synthetic gene circuit in yeast S. cerevisiae.

Keywords: Efficiency element; Insulation; S. cerevisiae; TATA box; Terminator.

Fig. 1

CYC1 promoters. a Scheme of the yeast CYC1 promoter as reported in [12, 13]. b Scheme and sequence of the pCYC1noTATA promoter used in this work. The thymine in green (+1 position) represents the promoter transcription start site (TSS)

Fig. 2

Terminators used to construct synthetic promoters. Sequences in red represent the efficiency elements, the ones in blue are the positioning elements (DEG1 positioning element is not known). Sequences underlined in purple are strong TATA-box motifs, the ones underlined in orange are weak TATA-box motifs

Fig. 3

Characterization of synthetic promoters made of a terminator placed in front of pCYC1noTATA. Fluorescence levels are normalized with respect to the fluorescent signal produced by the bare pCYC1noTATA. The negative control (background relative fluorescence) is also shown. The strong TATA boxes into DEG1t and Tsynth8 cause, respectively, a 6.6 and 5.1 fold increase in pCYC1noTATA fluorescence level. The weaker TATA box belonging to the genCYC1t has a lower effect (2.2 folds increase). The * symbol indicates a statistically significant difference with respect to the fluorescence level of the negative control

Fig. 4

Mutational study on DEG1t-pCYCnoTATA and Tsynth8-pCYC1noTATA. Mutations on the efficiency element of DEG1t (mut_DEG1t-pCYC1noTATA) and Tsynth8 (mut_Tsynth8-pCYC1noTATA) make it no longer recognized by RNA polymerase II molecules as a TATA-box with a consequent decrease in fluorescent expression. The insertion of a 100-nucleotide-long spacer between Tsynth8 and pCYC1noTATA extends the distance between the TATA-box-like efficiency element and the TSS up to 157 nucletides. This prevent any TSS activation from the TATAAA motifs along Tsynth8. Fluorescence level are normalized to the one corresponding to pCYC1noTATA. The ∙ symbol indicates no statistically significant difference with respect to pCYC1noTATA fluorescence level, whereas the ∘ symbol indicates no statistically significant difference with respect to shortADH1t-pCYC1noTATA fluorescence level

Fig. 5

Possible scenarios for promoter competition induced by the presence of a terminator-insulator. a A strong insulator efficiency element (EE) is placed in proximity of the promoter upstream activating sequence. DNA steric occupancy by RNA polymerase II at the EE prevents activator binding at the UAS and recruitment of other RNA polymerase II molecules to the promoter TATA box. b DNA bending–here due to the presence of an activator at its UAS–puts the EE spatially close to the TATA box causing a competition among RNA polymerase II molecules to get access to the EE and the TATA box themselves. DNA bending might also put the EE near the UAS provoking a competition (or even a collision) between RNA polymerase II and activator molecules

Fig. 6

GPD promoter. a Schematic of pGPD structure. b Wild-type, consensus, and mutated sequence of the two components of the strong pGPD UAS. Notice that the UAS can be placed on either DNA strand. Low-case letters indicate mutated nucleotides just outside the GRF1 binding site

Fig. 7

Fluorescence levels of synthetic promoters made of different terminators preceding the strong GPD promoter. All fluorescence levels are normalized with respect to the one of wild-type GPD promoter. The * symbol indicates a statistically significant difference with respect to pGPD fluorescence level. Among the four terminators, DEG1t causes the most considerable reduction in pGPD fluorescence level

Fig. 8

Mutational study on DEG1t-pGPD and DEG1t-mut_pGPD. Similarly to pCYC1noTATA, the fluorescence level of the wild-type pGPD is restored by removing, via double mutations, the TATA-box-like motif from the DEG1t efficiency element. Moreover, DEG1t is proved to be able to reduce the already rather low fluorescence expressed by mut_pGPD, where the sequence of the strong bipartite UAS has been deeply modified to prevent GRF1 binding. Therefore, RNA polymerase II molecules recruited by DEG1t are able to interfere also with far downstream transcription activation processes. By substituting DEG1t with mut_DEG1t, fluorescence grows back to the original level of mut_pGPD. All fluorescence levels are normalized with respect to the wild-type pGPD one. The ∙ symbol indicates no statistically significant difference with respect to pGPD fluorescence level, whereas the ∘ symbol indicates no statistically significant difference with respect to mut_pGPD fluorescence level

Fig. 9

DEG1t as insulator of five different yeast constitutive promoters. DEG1t drives a decrease in the fluorescence level of strong and medium-strength promoters, whereas it seems not to spoil the transcriptional activity of the rather weak ACT1 promoter. Values shown in this Figure are the ratio between the fluorescence levels of the insulated promoters with respect to the ones of the non-insulated promoters. The ** symbol indicates a statistically significant difference between the fluorescence expressed by an insulated promoter with respect to the one corresponding to its wild-type configuration

Fig. 10

Green fluorescent protein expression by two-transcription-unit systems. DEG1t placed between two transcription units does not alter, in a considerable way, the production of GPF by the downstream transcription unit. This scenario is quite different from the one illustrated in Fig. 7. We think that two adjacent transcription units integrated into the same S. cereviaese genomic locus reduce DNA bending and, as a consequence, the competition between the RNA polymerase II molecules recruited by the insulator efficiency and other molecules targeting nearby DNA sequences. Each construct is labeled as tu1(CYC1) (i.e. the leftmost transcription unit that encodes for yomKate2 and ends with CYC1t) and the name of the insulator (if present) placed in front of the GPD promoter that drives yEGFP expression. Fluorescence levels are normalized with respect to the one expressed by pGPD on a single transcription unit. The * symbol indicates a statistically significant difference with respect to the wild-type pGPD fluorescence level