Help:Terminators/Measurement/Cassie Huang
< Back to Terminator Measurement Help
| Cassie Huang, as a masters student in Tom Knight's lab, designed, constructed and characterized transcriptional terminators BBa_B1001-BBa_B1010. The following text is largely excerpted from her master's thesis [http://web.mit.edu/huangh/Public/main.pdf Design and Characterization of Artificial Transcriptional Terminators]. |
The characterization system for the artificial terminators uses a GFP/RFP dual fluorescent system with a transcriptional input provided by the arabinose-inducible ParaBAD promoter. The finished constructs were then transformed into E. coli strain CW2553.
Design of terminator characterization system
The transcriptional input for the terminator characterization system is the ParaBAD inducible promoter system. The terminator to be characterized is flanked by two fluorescent proteins, GFP and RFP, which are used to measure the termination efficiency of the terminator. The characterization system is made entirely from BioBrick parts found in the registry. Parts used to construct the characterization system are shown in Table 1. Figure 1 (top) shows the version 1 of the characterization system that uses GFP expression to measure input and RFP expression to measure output. Figure 1 (bottom) shows version 2 of the characterization system that uses the opposite measuring scheme.
Table 1: A list of the BioBrick parts needed to construct the terminator characterization plasmids and a short description of the function of those parts. Data for these parts were found at the Registry of Standard Biological Parts.
| Part Name | Part Type | Description |
|---|---|---|
| BBa_I0500 | promoter | inducible ParaBAD |
| BBa_E0034 | RBS | strong RBS |
| BBa_E0040 | reporter | generates GFP |
| BBa_E1010 | reporter | generates RFP |
| BBa_B0015 | terminator | terminator with high TE |
PoPS input generator
The PoPS generator must be able to vary the input signal to produce a wide range of outputs for device characterization. One possible way to produce a wide range of PoPS inputs is to use an inducible promoter such as the arabinose promoter, ParaBAD. However, inducing the araBAD operon at subsaturation concentrations results a population of cells which exhibit linear behavior in response to chances in inducer concentration but individual cells with either be fully induced or not induced. Decoupling the arabinose transport gene araE from the ParaBAD promoter and putting it under the control of an arabinose independent promoter will remove the all-or-none effects and produce a population of cells that will exhibit linear behavior in ParaBAD expression as a function of arabinose concentration at an individual level with all cells in the population having a similar level of expression as shown in the paper by Khlebnikov et al. Khlebnikov.
Devices under testing (DUT)
The DUT is flanked by two fluorescent proteins, GFP and RFP. The fluorescent protein preceding the DUT measure inputs to the DUT while the fluorescent protein following the DUT measures the output.
The termination efficiency will by measured by the ratio of the first fluorescent protein produced to the second protein produced. If the terminator has a high termination efficiency, very little fo the second protein will be produced. If the terminator has low termination efficiency, there should be no difference in the levels of the first and second proteins. In the off chance that one of the terminators acts as a promoter, more the second protein will be produced than the first protein.
Two sets of the characterization devices were constructed. The first set has GFP flanking on the left of the DUT and RFP flanking on the right. The second set is reversed with RFP on the left, and GFP on the right. This allows calibration of the input and output measurements.
Controls
Controls were needed to calibrate the levels of GFP and RFP expression in the characterization constructs. A list of controls used is shown in Table 2, and all controls are available from the Standard Registry. The controls BBa_I13514 and BBa_I13515 have the same components as the characterization devices, but lack a DUT. These are used to calibrate the input and output between the two sets of characterization devices by showing normal GFP and RFP expression without interference from the DUT, and are shown in Figure 2. Constructs BBa_I13521 and BBa_I13522 each have a fluorescent protein, RFP and GFP respectively, under the control of a constitutive promoter, Ptet. These controls test the maximum levels of GFP and RFP expression and are shown in Figures 3.
Table 2: This table shows the function and component parts of the control plasmids.
| Part Name | Description | Components |
|---|---|---|
| BBa_I13514 | Calibration of GFP input to RFP output | I0500, E0034, E0040, E0034, E1010, B0015 |
| BBa_I13515 | Calibration of RFP input to GFP output | I0500, E0034, E1010, E0034, E0040, B0015 |
| BBa_I13521 | Maximum RFP output | R0040, E0034, E0040, B0015 |
| BBa_I13522 | Maximum GFP output | R0040, E0034, E1010, B0015 |
E. coli strain CW2553
The E. coli strain CW2553 (araE201, .araFGH::kan) has all arabinose transport genes either deleted or mutated. The araE gene must be under the control of an arabinose independent promoter to ensure homogenous induction of cells. In the study by Khlebnikov et al. Khlebnikov, putting araE under the control of the PCP18 promoter resulted in cells being homogeneously induced by arabinose in the media as well as producing the highest concentrations of fluorescence. The pJAT18 plasmid contains araE under the control of PCP18, and is included in the CW2553 strain obtained from the Endy lab. The pJAT18 plasmid uses gentamicin as a resistance marker, so all transforms will be grown on media containing gentamicin.
Methods
The terminators were characterized by measuring the inputs and outputs of the characterization devices using a protocol developed by Jason Kelly of the Endy lab Kelly. The characterization devices were grown in supplemented M9 media overnight, and then induced with arabinose. The following day, aliquots of the cultures were taken to the MIT Flow Lab, and the induced fluorescence was measured.
Each characterization device and control were first streaked out on plates. One colony from each plate was grown in 5 ml M9 with the appropriate antibiotic was grown for 24 hours at 37°C. The OD600 of each culture was then measured and recorded. Each culture was diluted to an OD600 of approximately 0.07, which contains around 104 CFU. The diluted cultures were then grown from 2 hours at 37°C, and fluorescent protein expression was induced with arabinose. Studies have shown that the best range for arabinose induction is between 0.0001% and 0.01% Khlebnikov. All samples were induced with 5 μl of 0.1% arabinose in 5 ml of culture, creating a final arabinose concentration of 0.001%.
The induced samples were then grown overnight for 12-14 hours to maximize fluorescent protein expression. The following morning, 1ml aliquots of each cultures were placed in Falcon 3026 polypropylene tubes on ice to stop further growth. The aliquots were taken the MIT Flow lab, and GFP and RFP expression were measured.
Results
Controls
Five controls were measured in this experiment. CW2553/pJAT18 was used as the negative control to determine the ranges of background fluorescence. BBa_I13521 and BBa_I13522 constitutively expressed RFP and GFP respectively. BBa_I13514 and BBa_I13515 measured the GFP and RFP expression of the two versions of the characterization plasmids, but lacked the internal terminator under test. The mean GFP and RFP fluorescence and the standard deviations of these controls are shown in Table 3.
Table 3: This table shows the average GFP and RFP expression of the negative control, BBa_I13514, BBa_I13515, I13521 and BBa_I13522. BBa_I13521 and BBa_I13522 constitutively express RFP and GFP, respectively. BBa_I13514 and BBa_I13515 are used to calibrate input and output measurements of the characterization devices. In cases of BBa_I13514 and BBa_I13521, which have two distinct populations of cells, the cells which do not express sufficient fluorescence are discounted from the mean.
| Sample | Description | Mean GFP | Std GFP | Mean RFP | Std RFP |
|---|---|---|---|---|---|
| CW2553 | no fluorescence | 2.86 | 1.32 | 2.18 | 0.81 |
| BBa_I13521 | constitutive RFP only | 3.81 | 1.96 | 83.92 | 77.25 |
| BBa_I13522 | constitutive GFP only | 15.12 | 13.07 | 2.17 | 0.60 |
| BBa_I13514 | inducible GFP/RFP | 20.73 | 23.05 | 17.2 | 14.65 |
| BBa_I13515 | inducible RFP/GFP | 210.30 | 102.48 | 2.22 | 0.62 |
Constitutive expression of RFP and GFP
BBa_I13521 and BBa_I13522 provide a baseline measurement of reasonable ranges of RFP and GFP fluorescence. Figure 4 shows the measured fluorescence of these controls. The sample BBa_I13521 has a mixture of fluorescent and nonfluorescent cells. For purposes of calculating the mean and standard deviation of the cell population in BBa_I13521, only cells expressing sufficient fluorescence, defined as being above 4std of the negative control, were included.
As expected, BBa_I13521 has negligible GFP expression and [Part:BBa_I13522|BBa_I13522]] has negligible RFP expression. The mean RFP of BBa_I13521 was 83.92, compared to the negative control of 2.17. The mean GFP of BBa_I13522 was 15.12, compared to a negative control of 2.86. It is not know why constitutive GFP expression was much lower than constitutive RFP expression.
Expression of RFP and GFP from empty characterization plasmids
BBa_I13514 and BBa_I13515 are the empty versions of the characterization plasmids, lacking the terminator under test. BBa_I13514 has GFP followed by RFP under the control of the arabinose promoter ParaBAD. BBa_I13515 is similar, having RFP followed by GFP under the control of the same promoter. Figure 5 shows the measured fluorescence of these controls.
The cell population of BBa_I13514 contains a mixture of nonfluorescent cells, cells only expressing RFP, and cells expressing both RFP and GFP. The majority of the cells express no significant levels of fluorescence. Of the cells producing significant fluorescence, the mean GFP expression was 20.73, and the mean RFP expression was 17.2. As such, the measurements taken from BBa_I13514 cannot be accurately used to calibrate the input (in terms of GFP) to the output (in terms of RFP).
The cell population of BBa_I13515 does not express significant levels of RFP, but expresses high levels of GFP, with a mean GFP expression of 210.30. This may be due to the possible existence of an internal RNAse site in the RFP coding region, which causes fast degradation of RFP mRNA. In effect, the characterization plasmid would only have GFP to measure output of the terminator, instead of RFP to measure input and GFP to measure output.
Terminators
Ten terminators were characterized using two versions of the characterization plasmid. Version 1 contained GFP, followed by the terminator under test and RFP under the control of the arabinose promoter ParaBAD (Figure 1, top). Version 2 switched the locations of the RFP and GFP coding regions, but was otherwise the same (Figure 1, bottom). Under ideal circumstances, if a strong terminator was placed into characterization plasmid 1, the only output should be GFP. Likewise, if a strong terminator was present in characterization plasmid 2, only RFP should be visible as the output.
Results of characterization plasmid 1
Table 4 shows the mean GFP and RFP expression of characterization plasmids BBa_B3101 through BBa_B3110 as compared to both the negative control, CW2553/pJAT18 and the empty characterization plasmid BBa_I13514. The mean RFP expression of all the characterization plasmids was negligible when compared to the negative control. Mean GFP expression ranged from negligible compared to the negative control to close to the maximum indicated by BBa_I13514. The range in mean GFP of the different characterization plasmids was unexpected as the presence of the terminator under test should only affect the coding region downstream from it. The exact fluorescence measurements of each characterization plasmid are shown in Figures 6 through 15.
Table 4: This table shows the mean GFP and RFP expression of characterization plasmids BBa_B3101 through BBa_B3110. The mean GFP and RFP expression of the negative control CW2553/pJAT18 and BBa_I13514 are shown for comparison.
| Sample | Mean GFP | Std GFP | Mean RFP | Std RFP |
|---|---|---|---|---|
| CW2553/pJAT18 | 2.86 | 1.32 | 2.18 | 0.81 |
| BBa_I13514 | 20.73 | 23.05 | 17.2 | 14.65 |
| BBa_B3101 | 5.71 | 9.65 | 2.17 | 0.62 |
| BBa_B3102 | 22.42 | 19.39 | 2.23 | 0.78 |
| BBa_B3103 | 11.18 | 8.92 | 2.17 | 0.61 |
| BBa_B3104 | 22.47 | 26.24 | 2.16 | 0.59 |
| BBa_B3105 | 3.69 | 2.52 | 2.19 | 0.62 |
| BBa_B3106 | 24.86 | 23.88 | 2.16 | 0.60 |
| BBa_B3107 | 14.34 | 11.67 | 2.18 | 0.60 |
| BBa_B3108 | 8.59 | 5.71 | 2.17 | 0.60 |
| BBa_B3109 | 3.57 | 1.73 | 2.18 | 0.62 |
| BBa_B3110 | 25.48 | 33.25 | 2.60 | 1.92 |
Figures 6-15: Flow cytometry data of GFP vs RFP expression for each terminator characterization construct, the negative control E. coli strain CW2553 only, and the empty characterization plasmid BBa_I13514.
Figure 6: BBa_B3101 produces both low levels of GFP and RFP, contrary to the initial belief that it would produce levels of GFP close to that of BBa_I13514, but low RFP. The terminator tested, BBa_B1001, is unlikely to be a strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 7: BBa_B3102 produces a similar amount of GFP compared to the control BBa_I13514, but negligible RFP, and is likely to be a strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 8: BBa_B3103 produces only 54% of GFP of the control BBa_I13514, but negligible RFP, and is unlikely to be strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 9: BBa_B3104 produces a similar amount of GFP compared to the control BBa_I13514, but negligible RFP, and is likely to be a strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 10: BBa_B3105 produces both low levels of GFP and RFP, contrary to the inital belief that it would produce levels of GFP close to that of BBa_I13514, but low RFP. The terminator tested, BBa_B1005, is unlikely to be a strong terminator as it affects input as well as output.
Figure 11: BBa_B3106 produces a similar amount of GFP compared to the control BBa_I13514, but negligible RFP, and is likely to be a strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 12: BBa_B3107 produces only 69% of GFP of the control BBa_I13514, but negligible RFP, and is unlikely to be strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 13: BBa_B3108 produces only 41% of GFP of the control BBa_I13514, but negligible RFP, and is unlikely to be strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Figure 14: BBa_B3109 produces both low levels of GFP and RFP, contrary to the initial belief that it would produce levels of GFP close to that of BBa_I13514, but low RFP. The terminator tested, BBa_B1009, is unlikely to be a strong terminator as it affects input as well as output.
Figure 15: BBa_B3110 produces a similar amount of GFP compared to the control BBa_I13514, but negligible RFP, and is likely to be a strong terminator. The actual termination efficiency cannot be determined as the control BBa_I13514 has a majority of cells that produce no significant fluorescence, and cannot be used to accurately calibrate input measurements to output measurements.
Of the ten terminators tested with this version of the characterization device, four behaved in such a way that indicated high termination efficiency. Devices containing terminators BBa_B1002, BBa_B1004, BBa_B1006, and BBa_B1010 all expressed minimal levels of RFP and high levels of GFP. Devices containing terminators BBa_B1001, BBa_B1005, and BBa_B1009 expressed minimal levels of both GFP and RFP, and as such the termination efficiency of those terminators cannot be accurately judged with these results.
Results of characterization plasmid 2
Table 5 shows the mean GFP and RFP expression of characterization plasmids BBa_B3201 through BBa_B3210. All these characterization plasmids had the same flaw as the control plasmid BBa_I13515. The possible presence of an RNAse site in the RFP coding region made it such that there was limited RFP expression in all the characterization plasmids, and input to the terminators could not be accurately measured. Figures 16 through 25 show the exact fluorescence of these characterization plasmids. For these characterization devices, it is necessary to ignore the RFP measurements, as these devices can only accurately measure the GFP output. A strong terminator characterized by one of these devices will show low levels of GFP output, while a weak terminator will show high levels. The average GFP measured on the empty plasmid BBa_I13515 was 210.30. Of the terminators tested with these characterization plasmids, BBa_B1002 and BBa_B1006 cause the greatest decrease of mean GFP expression to 3.14 and 4.2 respectively.
Table 5: This table shows the mean GFP and mean RFP expression of characterization plasmids BBa_B3201 through BBa_B3210. The mean GFP and RFP expression of the negative control CW2553/pJAT18 and BBa_I13515 are shown also for comparison.
| Sample | Mean GFP | Std GFP | Mean RFP | Std RFP |
|---|---|---|---|---|
| CW2553/pJAT18 | 2.86 | 1.32 | 2.18 | 0.81 |
| BBa_I13515 | 210.30 | 102.48 | 2.22 | 0.62 |
| BBa_B3201 | 37.95 | 40.89 | 2.82 | 1.46 |
| BBa_B3202 | 3.14 | 1.61 | 2.17 | 0.60 |
| BBa_B3203 | 33.87 | 41.70 | 2.34 | 0.81 |
| BBa_B3204 | 13.28 | 14.71 | 2.59 | 1.17 |
| BBa_B3205 | 29.70 | 32.85 | 3.16 | 2.02 |
| BBa_B3205 | 4.20 | 4.98 | 2.18 | 0.67 |
| BBa_B3207 | 35.85 | 29.08 | 3.02 | 1.55 |
| BBa_B3208 | 9.90 | 8.04 | 2.19 | 0.62 |
| BBa_B3209 | 12.50 | 9.56 | 2.20 | 0.62 |
| BBa_B3210 | 9.01 | 6.17 | 2.18 | 0.60 |
Figures 16-25: Flow cytometry data of RFP vs GFP expression for each terminator characterization construct, the negative control E. coli strain CW2553 only, and the empty characterization plasmid BBa_I13515. However, due to the presence of an RNAse cut site in the RFP coding region in the terminator characterization construct, only the GFP fluorescence is measurable.
Figure 16: Terminator BBa_B1001 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 37.95, decreasing the output by 81%
Figure 17: Terminator BBa_B1002 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 3.14, decreasing the output by 99%
Figure 18: Terminator BBa_B1003 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 33.87, decreasing the output by 83%
Figure 19: Terminator BBa_B1004 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 13.28, decreasing the output by 94%
Figure 22: Terminator BBa_B1007 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 35.85, decreasing the output by 83%
Figure 23: Terminator BBa_B1008 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 9.90, decreasing the output by 95%
Figure 24: Terminator BBa_B1009 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 12.50, decreasing the output by 94%
Figure 25: Terminator BBa_B1009 reduces the mean GFP output from 210.30 as measured by the control BBa_I13515 to 9.01, decreasing the output by 96%
Termination Efficiency
Only the results from the second set of characterization devices were used to calculate termination efficiency. The control for the first set, BBa_I13514, did not have enough cells with significant GFP or RFP expression to accurately measure input and output of the terminator under test. Calculations of termination efficiency can be performed with only the output of the terminators, measured by the second set of characterization devices. Termination efficiency would be measured by the ratio of the mean GFP of a characterization device to the mean GFP of control BBa_I13515. The mean TE was calculated by average the TE of each cell in the sample population. Termination efficiency was calculated by the following formula.
Table 6 shows the average termination efficiency of the artificial BioBrick terminators BBa_B1001 through BBa_B1010 while Figures 26 through 35 show the histograms of the TE of the terminators as measured by the second set of characterization devices. Terminators BBa_B1002 and BBa_B1006 are the strongest terminators with mean % TE of 0.99 and 0.98 respectively. Other strong terminator with a % TE above 0.9 are BBa_B1004, BBa_B1008, BBa_B1009, and BBa_B1010. The remaining four terminators, BBa_B1001, BBa_B1003, BBa_B1005 and BBa_B1007 are all weaker, with % TE under 0.86. As these termination efficiencies were only calculated with the data from the second set of characterization plasmids, no final conclusions can be made until the behavior of the terminators is verified with the first set of characterization plasmids.
Table 6: This table shows the termination efficiencies of the new BioBrick terminators BBa_B1001 through BBa_B1010. The strongest terminators are BBa_B1002 and BBa_B1006. The weakest terminator is BBa_B1001.
| Terminator | TE |
|---|---|
| BBa_B1001 | 0.81 |
| BBa_B1002 | 0.99 |
| BBa_B1003 | 0.83 |
| BBa_B1004 | 0.93 |
| BBa_B1005 | 0.86 |
| BBa_B1005 | 0.98 |
| BBa_B1007 | 0.83 |
| BBa_B1008 | 0.95 |
| BBa_B1009 | 0.94 |
| BBa_B1010 | 0.95 |
Figures 26-35: Histogram of termination efficiencies for each terminator.
Discussion
Differences in data from the two different characterization devices need to be resolved. Several terminators, when tested with characterization plasmid 1, reduced the expression of the upstream GFP as well as the downstream RFP but testing with characterization plasmid 2 shows that they have high termination efficiency. In addition, the actual termination efficiencies were very different from the predicted values.
Effects of mRNA stability
The presence of an RNAse site in the RFP coding region would destabilize the mRNA for both proteins and result in minimal RFP expression in all samples. A strong hairpin of a terminator would help stabilize the mRNA after it has been cut, and slow the rate of degradation. When using the version of the characterization plasmid with GFP upstream of the terminator and RFP downstream, the terminator under test would have the job of stabilizing the remaining mRNA. A strong terminator would be able to slow the degradation of the remaining mRNA, so the resulting system would have high levels of GFP expression. A weak terminator would be unable to protect the remaining mRNA, causing the GFP coding region to be degraded as well. The resulting systems would then produce neither GFP nor RFP.
Of the terminators tested, BBa_B1002, BBa_B1004, BBa_B1006, and BBa_B1010 proved to be strong enough to prevent degradation of the GFP coding region. Terminators BBa_B1001, BBa_B1005, and BBa_B1009 proved to be poor at mRNA stabilization as shown by the fact that they have the lowest levels of GFP expression compared to the control BBa_I13514. The remaining terminators BBa_B1003, BBa_B1007 and BBa_B1008 provided a moderate amount of protection and allowed approximately 50% GFP expression as compared to the control.
Table 7 shows a summary of the BioBrick terminators and their termination efficiencies, as well as their ability to stabilize mRNA. With the exception of BBa_B1008 and BBa_B1009, the stong terminators were able to prevent degradation of the GFP mRNA in the first characterization plasmid. Conclusions of the termination efficiency of BBa_B1008 and BBa_B1009 cannot be made as their behavior in the two characterization plasmids contradict each other. The four best terminators are BBa_B1002, BBa_B1004, BBa_B1006, and BBa_B1010.
Table 7: This table shows the termination efficiencies of the new BioBrick terminators BBa_B1001 through BBa_B1010 as well as the mRNA stabilization ability. mRNA stabilization is based on much GFP was produced a terminator was tested with version 1 of the characterization plasmid as compared to control BBa_I13514. Strong terminators should also be able to stabilize mRNA better than weak terminators. BBa_B1008 and BBa_B1009 have high % TE, but are unable to stabilize mRNA. As the data from the two different characterization plasmids conflict for these two terminators, no conclusions can be made about them. The best terminators are BBa_B1002, BBa_B1004, BBa_B1006 and BBa_B1010.
| Terminator | TE | GFP produced |
|---|---|---|
| BBa_B1001 | 0.81 | 0.28 |
| BBa_B1002 | 0.99 | 1.08 |
| BBa_B1003 | 0.83 | 0.54 |
| BBa_B1004 | 0.93 | 1.08 |
| BBa_B1005 | 0.86 | 0.18 |
| BBa_B1006 | 0.98 | 1.20 |
| BBa_B1007 | 0.83 | 0.69 |
| BBa_B1008 | 0.95 | 0.41 |
| BBa_B1009 | 0.94 | 0.18 |
| BBa_B1010 | 0.95 | 1.23 |
Accuracy of predicted termination efficiencies
The most accurate predictions of termination efficiency occurred when the terminator in question had a poly(T) tail of approximately 6nt. The formula was least accurate when predicting termination efficiencies of terminators with tails less than 5nt long. Predicted termination efficiencies of the new BioBrick terminators are shown in Table 8.
Table 8: This table shows sequences, predicted % TE, measured % TE, and the error in the prediction. The strongest terminators are BBa_B1002, BBa_B1006, BBa_B1010, and BBa_B1004. The formula used to predict termination efficiency was most accurate when the terminator had approximately 6 thymine residues in the tail. The most accurately predicted terminators were BBa_B1002, BBa_B1003, BBa_B1006 and BBa_B1007.
| Terminator | Predicted TE | Measured TE | Error |
|---|---|---|---|
| BBa_B1001 | 0.95 | 0.81 | 0.17 |
| BBa_B1002 | 0.90 | 0.99 | 0.09 |
| BBa_B1002 | 0.80 | 0.83 | 0.04 |
| BBa_B1004 | 0.55 | 0.93 | 0.40 |
| BBa_B1005 | 0.25 | 0.86 | 0.70 |
| BBa_B1006 | 0.95 | 0.98 | 0.03 |
| BBa_B1007 | 0.80 | 0.83 | 0.04 |
| BBa_B1008 | 0.70 | N/A | N/A |
| BBa_B1009 | 0.40 | N/A | N/A |
| BBa_B1010 | 0.10 | 0.95 | 0.89 |
The most surprising result was that BBa_B1010 proved to be one of the most effective terminators while BBa_B1001 had the lowest termination efficiency. Since B1010 only had a 3nt T tail and a long stem loop structure, its predicted termination efficiency was only around 0.1 but its actual termination efficiency was measured to be 0.95. The terminator BBa_B1004 was another terminator that proved to be much more successful than predicted, with a predicted % TE of 0.55 but an actual % TE of 0.93. BBa_B1005, while not as effective as BBa_B1010 or BBa_B1004, had a measured % TE of 0.86, but was predicted to have a % TE of .25. BBa_B1001, contrary to initial expectations, was a poor terminator despite a poly(T) tail of 9nt and a high t score. It is not known at this time why these terminators behaved in this manner.
Of the remaining terminators, BBa_B1002, BBa_B1003, BBa_B1006 and BBa_B1007 all behaved as expected, with the predicted termination efficiency coming within 10% of the measured termination efficiency. Terminators BBa_B1008 and BBa_B1009 could not be accurately characterized, and as such, no comparisons could be made between the measured termination efficiency and the predicted. This leads to the conclusion that the formula used to predict termination efficiency is most accurate when applied to terminators with moderate poly(T) tails of approximately 6nt in length.
Future work
Several things can be done to improve the terminators in the future. The first thing to do would be to verify and retest the control plasmid BBa_I13514 to determine why only a minority of the cells carrying that plasmid produce sufficient GFP and RFP. This may be due to mRNA instability, but further tests should be performed, and, if necessary, a new control designed. The control plasmid BBa_I13515 should also be updated to prevent mRNA stability by removing mRNA3 cut sites from the RFP coding region or use hairpins to stabilize the remaining mRNA if the cut site cannot be moved. The constitutive controls of BBa_I13521 and BBa_I13522 should also be reviewed to see if the fluorescence of these controls could be increased.
In future experiments, calibration beads will be run on the flow cytometer, so data from different days of measurements can be compared. Three sets of measurements were taken during when the terminators were characterized, but only one set could be analyzed. While the results were overall consistent across multiple days of measurements, minor differences in the setup of the machine made combining those measurements ill advised. Running calibration beads at the start of each session of flow cytometry will provide a baseline to compare the performance of the flow cytometer across different days.
Further studies into designing new terminators will include using the device characterization plasmid designed by Endy Lab to characterize any new terminators. Using the Endy Lab plasmid will result in reducing the number of constructions necessary to prepare a BioBrick part for characterization. New terminators will have more varied stem loops in an attempt to vary termination efficiency, and will have thymine tails of approximately 6nt to maximize the effectiveness of the formula used to predict termination efficiency.
References
<biblio>
- Khlebnikov pmid=12080425
- Smolke pmid=12402322
- Choe pmid=15767274
- Nicholson pmid=10371039
- Nojima pmid=17150951
- Kelly Kelly J. Design and Evolution of Engineered Biological Systems. MIT SBWG technical reports 2005. doi: http://hdl.handle.net/1721.1/21169
</biblio>