Generator

Part:BBa_K3174006:Experience

Designed by: iGEM19_Austin_UTexas   Group: iGEM19_Austin_UTexas   (2019-08-08)


This experience page is provided so that any user may enter their experience using this part.
Please enter how you used this part and how it worked out.

Applications of BBa_K3174006

User Reviews

UNIQa64c184678e4e134-partinfo-00000000-QINU

No review score entered. University of Texas at Austin iGEM 2019

Creating calibration strains for the University of Texas at Austins 2019 iGEM team project

The 2019 UT Austin iGEM team [1] created a 'burden monitor' that used percent reduction in growth rate to quantify the metabolic burden of synthetic constructs from the iGEM kits. As part of the model, we had to create 'controls' that represented a diverse range of promoter and RBS strengths. These 5 promoter/RBS combinations were transformed into the burden strain of DH10B E.coli and used for this purpose.

In order to visualize the growth rate of these parts, the 5 glycerol stocks were streaked out onto LB + CAM plates and grown in a 37 degree incubator for a period of 24 hours. Fig.1 below shows the results.

Fig.1: Visualization of growth rates for the 5 BFP plasmids in TOP10 cells streaked onto LB + CAM plates from glycerol stocks, and viewed under a UV light: (left to right) BBa_K3174002, BBa_K3174003, BBa_K3174004, BBa_K3174006, and BBa_K3174007. BBa_K3174002 and BBa_K3174003 have very small fluorescent colonies, BBa_K3174004 has bigger colonies, and BBa_K3174006 and BBa_K3174007 have the largest colonies. The disparity in these growth rates originates from the variation in metabolic burden imposed on each strain due to their relative promoter and RBS strengths.

Once we had observed growth rates of each of these 5 parts, we wanted to see if the BFP expression levels matched our expectations based on what we know about the promoter strength of each part. We did this by selecting 3 colonies from each of these 5 plates and growing them in 5mL overnight LB + CAM cultures. We then placed 4 replicates of each isolate into a 96 -well plate (using 200uL of culture for each well) and took a single BFP measurement for each well. The resulting data is depicted in the Fig.2 bar plot below.

We expected BBa_K3174002 and BBa_K3174003 to have the highest BFP expression levels because these parts have the strongest promoter/RBS combinations. Surprisingly, we found that they had the lowest level of BFP expression. We hypothesized that the cells with the intact genetic devices must have been under a high level of metabolic stress, resulting in loss-of-function mutations that contained a 'broken' genetic device (making these cells express lower levels of BFP). The new mutated cells, with their fitness advantage over the highly burdened intact cells, must have then quickly overtaken the population.

We expected BBa_K3174004 to have a BFP expression level that was lower than BBa_K3174002 and BBa_K3174003 but higher than BBa_K3174006 and BBa_K3174007. However, it instead had the highest level of BFP expression out of all 5 parts. This is probably due to the fact that it had a strong enough promoter to cause BFP to be expressed strongly but a weak RBS that prevented considerable stress to the cell. It is important to note that although BBa_K3174004 expressed BFP at the highest level, it lacked consistency between each of the four replicates for each isolate.

BBa_K3174006 and BBa_K3174007 expressed BFP as expected. They had the second and third highest BFP expression levels respectively. Both these parts have relatively weaker promoters that are unlikely to cause burden to the cell. They seemed to have the highest consistency between replicates.

Fig.2: Single BFP expression measurements for the 3 isolates of each of the 5 BFP parts. This experiment was carried out by selecting 3 colonies from the plates in Fig.1, growing them in 5mL O/N cultures, and then loading them in replicates of 4 onto a 96-well plate. The error bars represent the standard deviation between the 4 replicates of each isolate. Blanks (only LB) and a non-BFP expressing part ('BB_0') were also included as controls for this reading. The average OD600 measurement values of the LB blanks were subtracted from all the other samples prior to calculating BFP expression (to account for background emissions). BFP expression of each sample was calculated by dividing BFP by the adjusted OD600 value.

These 5 constructs had their sequences confirmed via Sanger Sequencing.

The major takeaways from this measurement data are:

  • BBa_K3174002 and BBa_K3174003 should be used with caution because they break almost immediately due to a high level of metabolic stress to the cell that causes loss of function mutations, leading to a population of cells that no longer express BFP at a high level.
  • BBa_K3174004 expresses BFP at the highest level but probably does not have very high stability based on the inconsistency in BFP expression measurements between replicates of each of the three isolates depicted in the bar graph above.
  • BBa_K3174006 and BBa_K3174007 express BFP at low levels but appear to be the most stable based on the consistency in BFP expression measurements between replicates of each of the three isolates depicted in the bar graph above. These parts could be useful for experiments that do not require high levels of BFP expression but do rely on evolutionary stability.
UNIQa64c184678e4e134-partinfo-00000004-QINU