The most attractive feature of the mazEF system is that cytotoxity of a toxin protein (MazF) is determined by the stoichiometric ratio of a toxin protein to a corresponding antitoxin protein (MazE) in cells. The story of Snow White involves sleep and resuscitate from sleep, and thus, we think that this feature is useful for our project.

Transformants shown below were prepared.

E. coli A: carrying the Pcon ‐ rbs ‐ gfp (pSB6A1) , Plac ‐ rbs (pSB3K3)

E. coli C: carrying the PBAD ‐ rbs (pSB6A1) , Plac ‐ rbs (pSB3K3)

E. coli B: carrying the PBAD ‐ rbs ‐ mazF ‐ tt ‐ Pcon ‐ rbs ‐ gfp (pSB6A1) , Plac ‐ rbs (pSB3K3)

E. coli D: carrying the PBAD ‐ rbs ‐ mazF ‐ tt ‐ Pcon ‐ rbs ‐ gfp (pSB6A1) , Plac ‐ rbs ‐ mazE (pSB3K3)

It is well known that the expression from Plac is usually stronger than that from PBAD. Therefore, we expected that the expression level of mazE would be finally higher than that of mazF. Samples were incubated with vigorous shaking at 37℃. When turbidity reached 0.03, 0.02% arabinose was added to induce mazF expression. Two hours after the addition of arabinose, 2 mM IPTG was added to induce mazE expression. The turbidity and Relative Fluorescence Units (RFU) of GFP were measured at several time points.

As shown in Fig. 3-1-2-1-3-1, when the mazE expression was induced 2 h after the mazF expression, cell growth resumed gradually. Similarly, the RFU of GFP also increased mazF expression, and about 8 h later, cell growth resumed. Similarly, the RFU of GFP was also resumed(Fig. 3-1-2-1-3-2). Before starting the experiments, we anticipated that recovery from the growth inhibition by mazF would be observed immediately after the induction of mazE expression. However, it took 8 h for us to observe the recovery. The reason is unclear, but the prior expression of mazF might damage the cells and it might take time to repair. Alternatively, the mazF expression might lead cells to the dormant state from which cells can hardly escape. We assume that this feature is advantageous for our Snow White project, because we can prolong the sleeping time of Snow White and take sufficient time for Prince to find Snow White. As a conclusion, the “Stop and Go” experiment was successful.

In Experiment 1-2-1, we found that a toxin inhibits cell growth, and an antitoxin resuscitates it. However, what will happen when a toxin is expressed after the antitoxin constitutive expression? Therefore, we conducted the experiment. Since cell growth was resuscitated after cells had grown, we named this experiment, "Go & Stop".

Transformants as shown below was prepared. E. coli A: carrying the Pcon - rbs - gfp (pSB6A1) , Plac - rbs (pSB3K3)

E. coli C: carrying the PBAD - rbs (pSB6A1) , Plac - rbs (pSB3K3)

E. coli E: carrying the PBAD ‐ rbs ‐ mazF ‐ tt ‐ Pcon ‐ rbs ‐ gfp (pSB6A1) , vector (pSB3K3)

E. coli F: carrying the PBAD ‐ rbs ‐ mazF ‐ tt ‐ Pcon ‐ rbs ‐ gfp (pSB6A1) , Pcon ‐ rbs ‐ mazE (pSB3K3)

E. coli G: carrying the PBAD ‐ rbs ‐ mazF ‐ tt ‐ Pcon ‐ rbs ‐ gfp (pSB6A1) , Pcon ‐ rbs(weak) ‐ mazE (pSB3K3)

In order to adjust the amount of expression of mazE, two types of RBS (BBa_B0034 and BBa_J61117: J61117 is weaker than B0034) was used. Using these plasmids, we tried clarifying stoichiometric relationship between A and B by changing the expression of mazE. Samples were incubated with vigorous shaking at 37℃. When turbidity reached 0.03, 0.02% arabinose was added to induce mazF expression. The turbidity and Relative Fluorescece Units (RFU) of GFP were measured at several time points.

Increase in turbidity and RFU of E. coli G stopped earlier than that of E. coli F(Fig. 3-1-2-2-3-1 and Fig. 3-1-2-2-3-2). As a result, turbidity and the RFU of E. coli F in the stationary phase was smaller than that of E. coli G.

Fig. 3-1-2-2-3-1 Time vs Turbidity (Go &Stop)

Fig. 3-1-2-2-3-2 Time vs RFU of GFP (Go &Stop)

Calculation of the change of RFU of GFP / Turbidity per unit time (translation efficiency) indicates that the expression level of mazE correlated with the translation efficiency(Fig 3-1-2-2-3-3).

Fig. 3-1-2-2-3-3 Translation efficiency of each E. coli.

We found that MazF inhibited cell growth and translation even when there was MazE. In addition,these results suggested that there exsited relation between the resuscitation and mazE expression.

   From results of Experiment1.2.1. and Experiment1.2.2., it was expected that mazEF can repeatedly control cell growth.

　The control of cell growth by the mazEF system has been shown until the previous sections. In this section, we analyzed whether the “stop & go” experiment can be repeated many times.

A pSB6A1-based plasmid containing both the PBAD(BBa_I0050) - rbs(BBa_B0034) - mazF (BBa_K1096002) cassette was constructed. Furthermore, a pSB3K3-based plasmid containing the Plac(BBa_R0010) - rbs(BBa_B0034) - mazE (BBa_K1096001) cassette was constructed. These plasmids were co-introduced into E. coli. We prepared LB medium containing arabinose (Ara(+)), IPTG (IPTG (+)), both inducers (Ara(+), IPTG(+)), and no inducers (Ara (-), IPTG (-)). We tested whether the E. coli cells formed colonies on above plates by controlling the mazEF system. The procedures are shown below.

(i) Day1: Each transformant was streaked onto the Ara(+)-, IPTG(+)-, and
                  Ara(-)_IPTG(-)-plates.

(ii) Day2: The colonies on the above plates were re-streaked onto the Ara(+)_IPTG(+)-plate.                   Even when no colonies were found on the plates, the streaked areas on day 1 were
                  scratched by toothpicks and re-streaked.

(iii) Day3: The same procedure was conducted as day 2 except that Ara(+)-plate was used.

(iv) Day4: The same procedure was conducted as day 2.

From the result, it was clarified that growth of E. coli cells was repeatedly controlled by expression of mazE(Fig. 3-1-2-3-3-1 and Table 3-1-2-3-3-1). The results of this experiment are very useful in our project, and it is expected to lead to new biotechnological application.

Table 3-1-2-3-3-1 Control of colonization by mazEF system on the LB Agar Plate