RESULTS

The first step toward a good characterization in terms of statistics, reliability, repeatability and feasibility was to develop a protocol that fulfills each one of these points. Exploring a large window of different parameters and validating the data collected through statistic analysis we obtained a good and easy protocol for transcriptional terminators characterization in vivo, that gave ratiometric results affected by an acceptable standard deviation (as shown in the terminators characterization charts below). Theoretically, this protocol should work for any platform that exploits a bicistronic message encoding two fluorescent proteins (of course it would have to be tested on them and some minor changes could be necessary). Its fundamental points are probably the maintaining of the same optical density between the samples during induction, the sonication of the samples and the incubation over-night that allows fluorescence stabilization. The protocol and further details can be found here.

Figure 5: Transcriptional termination efficiencies of T7 and E. coli terminators with T7 and E. coli RNA polymerases (RNAP).

While collecting the above data, we noticed that the raw fluorescence intensities for mCherry seemed to increase in the presence of the T7 transcriptional terminator when using the T7 RNA polymerase. This effect was previously described for some terminators by H. Abe and H. Aiba [2]. To further explore this potential influence, we had to compare different cultures, with and without intervening terminator in the transformed plasmid. Thus, we repeated the experiment ensuring that all measurements were taken from cultures that were induced at the same optical density (O.D.), more information can be found here. The mCherry fluorescence was then analyzed with the following equation, that represent the Relative Increase in the upstream gene expression, where Cs is the intensity of mCherry when followed by the terminator of interest, and Cc is the intensity of mCherry of the control construct that did not contain the terminator.

The data showed that the combination of a T7 transcriptional promoter and a T7 transcriptional terminator increased mCherry levels two-fold with respect the same construct in the absence of an intervening terminator. This effect was not observed for the remaining three combinations tested.

Figure 6: Relative increase in upstream gene expression of constructs with T7 and E. coli terminators, T7 and E. coli RNA polymerases (RNAP).

Since the apparent transcriptional termination efficiencies were calculated as a ratio of mVenus over mCherry levels, an increase in mCherry fluorescence would result in an over estimation of the termination efficiency. To investigate whether our values were, indeed, over estimated, we calculated raw termination efficiencies from the same samples used for the mCherry measurements. The fluorescence intensities of mVenus were evaluated without considering mCherry emissions with the following equation: where Vs is the intensity of mVenus when preceded by the terminator of interest, and Vc is the intensity of mVenus control, i.e. in the absence of the preceding terminator.
When the contribution of mCherry was removed, the data suggested that the RNA polymerase did not influence the transcriptional termination efficiency (Figure 7).

Figure 7: The raw and apparent termination efficiencies of T7 and E. coli terminators with T7 and E. coli RNA polymerases.

One potential reason for the increased mCherry levels could be that the terminator protects the transcript from the activity of cellular RNases. Although RNases would not explain why the effect was only observed for the T7 promoter - T7 terminator combination, we decided to test some of the constructs in a purified system that lacked the presence of nucleases. Therefore, we exploited the PURExpress kit from New England BioLabs, which consists of purified transcription - translation machinery, including T7 RNA polymerase and E. coli ribosomes. First, the data showed that the transcriptional termination efficiencies, both apparent and raw, were significantly lower in the purified, in vitro system for both T7 and E. coli terminators than the data obtained from the in vivo experiments (Figure 8). This is consistent with the fact that T7 lysozyme, which was not present in our in vitro experiments but was present in vivo, aids in transcriptional termination [3]. Second, the stabilization effect was significantly diminished in vitro, suggesting that RNases or other unidentified cellular factors were involved. Some kinetic profiles of the reaction can be found below (Figure 9).

Figure 8: The activity of T7 and E. coli terminators in vitro with purified T7 RNA polymerase and E. coli translation machinery. Left, right and bottom panels show apparent termination efficiencies, raw termination efficiencies, and normalized mCherry expression.

Figure 9: in vitro kinetics measurements. Left panel shows the raw emission peaks intensities obtained with T7 RNA polymerase and T7 terminator. Right panel shows the raw termination efficiencies of both terminators with the same platform.