Engineering success can be defined by following the engineering design cycle:
Design → Build → Test → Learn
In order to develop a test that could be used to determine the concentration of bioavailable phosphate in soil, we needed to go through the steps of design, build, test, learn, retry. We have been working through these steps to design a testing procedure that is simple, cost-effective, and accurate.
For our project last year, we designed a construct that used the Lac repressor to control the expression of the bla gene in response to phosphate levels. The bla gene codes for Beta-lactamase which is an enzyme that can produce a red color when a substrate called nitrocefin is added. One significant problem that we found was that there was a lot of background red color suggesting that there is Beta-lactamase being produced independently of the phosphate-controlled regulation.
It is known from studying the lac operon that the Lac repressor allows leaky expression of the genes it is regulating so we wanted to try using a different repressor that may provide more tight regulation of the bla gene. This led to us trying TrpR which is a negative regulator of the trp operon. This repressor binds more tightly to the trpL operator sequence so we thought this might help remove some of the background red color that makes our data difficult to interpret.
In the design stage of our project we needed to decide which regulators would potentially work the best to control the expression of the bla gene in response to phosphate concentrations. We knew that we were going to use the phoA promoter because it is sensitive to phosphate levels and there are very few alternatives currently known. The phoA promoter is tricky, however, because it works the opposite of how we would like it to work. The phoA promoter turns expression off when there are high levels of phosphate and turns expression on when there are low levels of phosphate. To account for this mechanism, we put the phoA promoter in control of a negative regulator. This way, if there are high levels of phosphate, it would reduce the amount of the negative regulator which would turn on the expression of whatever the negative regulator was regulating.
As mentioned above, we initially used the Lac repressor as our negative regulator but this proved to be too leaky. We considered using AraC from the arabinose operon since this is a tightly regulated operon but AraC can work as both a positive and negative regulator so it would lead to a more complicated system. We settled on TrpR from the trp operon so we placed the trpR gene under the control of the phoA promoter and placed the trpL operator sequence and the trp promoter in front of the bla gene. In theory, this would put the expression of the Beta-lactamase enzyme under the control of phosphate levels.
When building our constructs, we had to make sure all of the components we were trying to put together were exactly right. The phoA promoter contains a phoB box that is necessary for the promoter to be regulated properly so the complete sequence of the promoter including the phoB box and -10 and -35 sequences needed to be accounted for. This had to be aligned properly with the transcription start site of the trpR gene so the entire transcript would be transcribed and subsequently translated. The trpL promoter containing the trpL operator sequence had to be determined and placed ahead of the bla gene. The bla gene also had the trpL transcription start site, with the trpL ribosome binding site, ahead of the bla open reading frame.
We made two additional constructs that had ssrA degradation signal sequences attached to either the trpR or the bla genes. This is a specific DNA sequence that is added to the 3’ end of the gene that encodes a short peptide that leads to a shorter half-life of the protein. In both cases, the signal sequence had to be place in the correct reading frame at the end of the gene but before the stop codon.
After we had these constructs synthesized by Twist Biosciences, we transformed them into E. coli and began to test them. Initially, we exposed the E. coli strains containing the constructs to high concentrations of phosphate buffers to determine if we would get Beta-lactamase expression. We would determine if Beta-lactamase was being expressed by whether the red color was being produced when nitrocefin was added. We had a positive control that was a strain of E. coli with a plasmid that constitutively expressed the bla gene so there was a high amount of Beta-lactamase present and a deep red color after nitrocefin was added and allowed to sit for one hour. We also had a negative control which was a strain of E. coli that did not contain the plasmid construct so no Beta-lactamase should be produced. In this case, there was no significant red color observed when nitrocefin was added.
When exposed to high concentrations of phosphate buffer, we saw that the samples mostly turned red after incubation with nitrocefin. The constructs that either did not have the ssrA degradation tag or had the tag on the trpR gene produced a large amount of red color. The strain containing the construct with the ssrA tag on the bla gene produced considerably less red color suggesting that the Beta-lactamase protein was being degraded more rapidly so it could not cleave the nitrocefin as much.
After this experiment, we decided to expose the E. coli strains containing the constructs to lower concentrations of phosphate that were more similar to what would be found in soil (approximately 30 ppm). We saw similar results as the first experiments with high phosphate concentrations except that the strain containing the construct with the ssrA tag on the trpR gene was the only one that showed a positive correlation between the concentration of phosphate and the amount of red color produced.
The last set of experiments we tried was to actually test soil samples for which we knew the phosphate concentrations as determined previously through chemical tests. We only used the construct with the ssrA tag on the trpR gene. While the results were not conclusive due to controls not working as well as we would have hoped, we were able to show that we could see a difference between the phosphate levels in the two soil samples that were similar to the known difference between the two. This suggested that with more experimentation and fine tuning, it is possible to accurately measure bioavailable phosphate concentrations using the system we developed.
We learned that we need to make our plasmids as specific as possible to make the most efficient sensor. Understanding what can cause background noise and how we can limit it in our test samples was a consistent obstacle we had to troubleshoot throughout our experimentation. We have learned to be conscious of everything you use in your experiment, as even buffers can have major effects on your results. We learned a lot about the biology of operons and chemical cleaving through this project which our members can use in their future careers in biochemical research.