Do they have the expected effect on E. coli growth?
We created constructs containing either the toxin or antitoxin in the pCDFDuet vector under control of an IPTG inducible T7 promoter.
We grew these strains in standard LB media at 37C, with a starting OD600 of 0.05. After two hours we added 0.5ul/ml of 1M IPTG and looked for effects on growth. Measurements of OD600 were taken every hour and the results (below) show that the production of the toxin does have a negative effect on growth and the antitoxin does not.
We also performed a Ni++ bead his tag purification of our cultures and were able to see both proteins on a gel post-induction (data not shown).
Although we cloned three toxin/antitoxin pairs, the results were best with the YoeB/YefM pair. The data shown above is for this pair and the rest of our experiments were done using this set. As a side note, we found that the toxicity of the MazE/MazF pair was very strong and had a hard time even growing the cells!
Can the toxin/antitoxin system be used to affect growth between two strains of E. coli?
We tried several different experiments to determine if the toxin and antitoxin pair could move between cells to cause an effect on growth. The final experiment we performed utilized protein induction at 30C which we believe helped maintain the stability of these small proteins.
A flask with 150ml LB+Spec was inoculated with either – Toxin strain, Antitoxin strain, or BL21 (empty vector) control cells.
Each flask was started at an initial OD600nm of 0.05 and was grown for 2 hours at 37C before being induced with 0.5 ul/ml of 1M IPTG.
After the addition of IPTG, the flasks were transferred to 30C to promote stable protein production. These cultures were grown overnight, ~16 hours.
The next morning the cultures were centrifuged and the supernatant was filtered over a 0.2um filter to remove any cell debris, but retain any small proteins or molecules in the supernatant.
The filtered supernatants were then diluted (25ml supernatant + 75ml fresh LB) to provide a fresh supply of nutrients.
Then toxin, antitoxin, and control cells were inoculated into these filtered, diluted supernatants and induced with IPTG after 2 hours of growth at 37C.
The effect of supernatants potentially containing toxin or antitoxin proteins was observed in the growth of each strain and is shown in the graphs below.
The graphs shown below demonstrate growth of toxin (YoeB), antitoxin (YefM), and control cells in the media produced after step 5 above.
This first graph shows control cells that were grown in filtered supernatant from either Toxin, Antitoxin, or control cell growth. The inoculated control cells grow well in the supernatant from control and antitoxin cells but do not grow well in the presence of toxin supernatant. This indicates that the toxin supernatant actually contained toxins that had a growth effect on the control cells.
This graph shows YefM, the antitoxin, being grown in filtered supernatant from control, toxin, and antitoxin cells. It is observed that the antitoxin grows well in each condition, except when inoculated into the supernatant where toxin cells had previously grown (green line).
This graph shows YoeB, the toxin, being grown in filtered supernatant from control, toxin, and antitoxin cells. It is observed that the YoeB toxin does not grow as well as any of the strains in the above graphs. However, there is a slight rescue of growth only in the case when it is exposed to the antitoxin supernatant (pink line).
Summary: The toxin-antitoxin pair of YoeB/YefM is potentially useful as a communication and tuning system between two cells. Previously, toxin-antitoxin pairs have only been known to function within cells and we show there is potential for them to be used in a variety of applications, but in particular to tune population ratios.