Part:BBa_M50053:Experience
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_M50053
Device Design
In an effort to research and develop less invasive methods for the testing and diagnosis of diabetes, we designed a protein-based sensor that employs FRET to assess glucose levels. For FRET to be effective, there must be a sufficient overlap between the emission and absorption spectra of the donor and acceptor fluorophores respectively, but enough separation in emission spectra to avoid. Based on construct length constraints, we chose to use cometGFP as our donor and mRuby3 as our acceptor fluorophore (Fig. 1). cometGFP has an maximum excitation wavelength of 403 nm and maximum emission of 511 nm while mRuby3 has a maximum excitation of 558 nm and maximum emission of 592 nm. Ideally, when the GBP is in an open state, the two fluorophores would be relatively far from each other and we would be able to see independent emission - green fluorescence when excited at 403 nm and red fluorescence when excited at 558 nm. When the GBP is in a closed state, the two fluorophores would be close enough together such that some of the cometGFP emission would be transmitted to the mRuby3 protein, producing red fluorescent emission through FRET in addition to the expected green when excited at 403 nm. Since the conformational shifts of the GBP are likely not drastic enough to eliminate all FRET in the open state, we did not expect binary fluorescent readings. However, by assessing the relative levels of green versus red fluorescence we aimed to characterize the extent of glucose-binding.
In our plasmid, these two fluorophores were joined on each end of a unique GBP isolated from Thermus thermophilus bacteria, optimized for transformation in E. coli. This GBP was selected for its reasonably short sequence length (1242bp) and stable glucose binding ability (Kd value of 0.08(+/- 0.03) mM).10 mRuby3 was connected to the N-terminus of the GBP and cometGFP followed the C-terminus of the GFP. A linker sequence was inserted between each fluorophore and the protein in order to maintain linearity of the gene and prevent protein folding. We selected the T5 IPTG-inducible promoter which is suitable for any E. coli transformation and is normally repressed by LacI expression. Since it is also repressible with 2% glucose, we induced protein expression first with IPTG and used dilute glucose concentrations when testing glucose binding. Toxicity of our protein in E. coli was not a concern so a strong RBS and an origin of replication with a high copy number were selected to increase the amount of GBP produced. The construct also included a selective marker for ampicillin. All of these generic elements were derived from the pd444-cc plasmid from DNA2.0 (Fig. 2).
To create our modified glucose-sensing E. coli, we first transformed competent E. coli cells with our plasmid DNA according to BioE 44 - Practical 3. We performed three transformations: the first with our plasmid DNA, pKAT, the second with a known pColi plasmid that confers ampicillin resistance to serve as our positive control, and a mock transformation with no plasmid to ensure the validity of the experiment. We then plated the all of these cells on ampicillin plates; we also plated an additional aliquot of our pKAT-transformed cells on a kanamycin plate to confirm our plasmid insert only conferred ampicillin resistance. As expected, we saw colonies on the plates with pKAT and pColi and ampicillin; interestingly, our pKAT colonies on the ampicillin plate appeared pink in color. We figured that the nutrients in the media and on the plate may have induced our protein expression such that the presence of cometGFP caused the colonies to appear pink to the naked eye. Thus, in following experiments, we used a minimal media without these nutrients, such as lactose and glucose, so that we could better control promoter induction and protein binding.
We then selected a single colony from our experimental plate and streaked the cells on an LB + ampicillin plate in order to further isolate potential contaminants and serve as a source of clonal populations of cells for each experiment.
Assay 1: Inducing gene expression with IPTG
We wanted to test what level of IPTG would be sufficient to induce expression of our protein as the promoter region of our plasmid insert is IPTG inducible. A dynamic range of IPTG concentrations were used - from 1.0 mM to 50.0 mM - to ensure some measurable amount of induction would occur between these two extrema. LB + ampicillin media, IPTG, and transformed cells were added to each well for a total volume of 250 ul to achieve the desired IPTG concentration and cell concentration (measured as OD600) of 0.01. In order to evaluate induction efficiency, needed to measure both mRuby3 and cometGFP peak expression using their respective excitation/emission wavelengths of 558/592 nm and 403/511 nm. Furthermore, since IPTG induction can take anywhere from 4-8 hours, we measured mRuby3 and cometGFP levels every hour after the 4th hour until induction levels were sufficient.
After 4 hours, results showed that mRuby3 and cometGFP fluorescence was very high across all concentrations. While there seemed to be baseline protein expression in the negative control, expression in those induced with IPTG was far higher. However, increasing the IPTG concentration did not seem to significantly improve protein expression implying that perhaps even our lowest concentration of IPTG (1 mM) was already an excessive concentration. Moreover, analysis of cell concentrations across this time course revealed that cell growth with all our levels of IPTG was severely stunted. After 6 hours, cells in the negative control, which had no IPTG in solution, had an OD600 of 0.27 whereas cells grown in the presence of 1 mM IPTG had an OD600 of 0.06 - without IPTG cells grew almost 5 times better (Fig. 4). After consulting with our professor, we learned that too much IPTG is toxic to cells and that our lowest concentration of 1.0 mM was already an excess amount. Since our promoter seemed to be highly induced by IPTG, we figured that lowering future IPTG induction concentrations to 0.1 mM would not only still be able to induce protein expression but be low enough such that cell growth would not be inhibited; finally, IPTG induction seemed to be sufficient after 6 hours, so we implemented this time scale in following experiments.
Assay 2: Assessing glucose binding with FRET analysis
After developing our methods for bacterial culture and protein expression, we conducted a fluorescence-based assay of glucose binding. 6 hours after 0.1 mM IPTG induction, we added glucose, minimal media (EZ Rich Media and 2% glycerol), and cells for a total volume of 250 µL. Since the Kd of our GBP is 0.08 mM10, we used five glucose concentrations from 0.001 mM to 100 mM; the negative control had no glucose. We first diluted cells to an OD600 of 0.01, but in subsequent trials we diluted to 0.1 to facilitate a shorter growth period. Plates were incubated for 24 hr. in a 37˚C shaker then for 12 hr. at 25˚C to allow the fluorophores to properly mature. Fluorescence readouts included OD600 absorbance, FRET fluorescence spectrum, and mRuby3 fluorescence spectrum.
After normalizing the FRET spectra readings to the GFP emission at 510 nm, we found that the highest emission of mRuby3 (peak around 615 nm) was with a glucose concentration of 0.001 µM, the next highest was with a glucose concentration of 0.01 µM, and so on. As expected, our lowest emission of mRuby3 was our negative control with a glucose concentration of 0.0µM (Fig. 5). This indicated that when glucose was present, cometGFP and mRuby3 were physically close enough to cause FRET - cometGFP emission excited mRuby3 to increase its emission.
Our next trial helped us determine the ideal time frame to produce best FRET results. Again, we normalized the spectra readings to the GFP emission at 510 nm and collected data 24hr. and 36hr. after introducing glucose. After 24hr., we found that the highest mRuby3 emission was with the highest glucose concentration (0.1 µM) (Fig. 6). While this was unexpected given that 0.001 µM was most effective in producing FRET in our first trial, we believe the low amounts of mRuby3 emission for 0.001 µM in this trial could be due to pipetting errors. After 36hr. we found consistent results that the highest mRuby3 emission was with 0.1 µM glucose. However, we also found an increase in mRuby3 emission in the negative control (Fig. 7). We believe this could be due to the mRuby3 protein fully maturing after 36hr. in an environment with 0.0 µM glucose and maturing at a faster rate when there is glucose present for the cells to consume. So, after 24hr. we are seeing FRET happen, but after 36hr. there is no significant amount of FRET occurring in the experimental wells compared to the negative control. We believe that the protein stability is not aligning with the mRuby3 maturation time, which is therefore giving us some conflicting results.
mRuby3 spectrum readouts provided important information to verify mRuby3 was being expressed, to identify the wavelength of mRuby3 emission when excited at the normal excitation wavelength of 558 nm, and to assess the maturation time of mRuby3. Given our inconsistent data with FRET, we went on to analyze mRuby3 emission over time when excited at 530nm. After normalizing the spectra readings to the negative control, we see different results for the different concentrations of glucose. We found that at the lowest glucose concentration (0.001 µM) there is higher mRuby3 emission after 36hr. than emission after 24hr. However, when looking at the highest glucose concentration (0.1 µM), there is higher mRuby3 emission after 24hr. than emission after 36hr. (Fig. 8). This may explain why at 0.1 µM glucose concentration, the mRuby3 emission in the FRET analysis is higher than the negative control at 24hr., but not at 36hr. This finding is consistent with our interpretation that mRuby3 needs a maturation period of 36hr. to see full emission, especially when the glucose concentration is so low, such as 0.001 µM or 0.0 µM (Neg. Ctrl.).
Stanford Location
Plasmid Name: pKAT
DNA2.0 Gene #: pD444-CC
Organism: E. coli
Device Type: Sensor
Box Label: BIOE44 F16
Barcode Number: 0133026155
User Reviews
UNIQe5956653713143fc-partinfo-00000000-QINU UNIQe5956653713143fc-partinfo-00000001-QINU