KULeuven iGEM 2011



1. Promotor testing of inducible promotors

pBAD + GFP generator (Part:BBa_K584000)

To test the usefulness of the arabinose-inducible promoter I13453 in our 2011 iGEM project, we fused the promoter to a GFP reporter, and assayed the promoter’s activity after addition of different amounts of arabinose. We tested this activity both in a TOP10F’ (figure 1) as well as a MG1655 (figure 2) E. coli strain background. For more information on E. coli strain descriptions, we recommend the following website.

As can be seen in figure 1, addition of arabinose had only minor, if any, effect on the growth of TOP10F’ cells, most likely because these cells cannot metabolize the sugar. However, to our surprise, we did not observe an arabinose-induced increase of fluorescence compared to the control without arabinose for the first 6 hours. The observed increase in fluorescence after arabinose addition at the 8 hour time point could not be confirmed when the experiment was repeated.

Figure 2 demonstrates that in MG1655 cells, the addition of 0.2% and 2% arabinose results in a growth defect after about 4 hours of growth, which may relate to the metabolism of the sugar in this strain. In contrast to the TOP10F’ cells, we see a clear induction of promoter activity after addition of arabinose. The addition of only 0.02% resulted in the greatest induction. Increasing arabinose concentration did not increase fluorescence, probably due to the observed growth defect. As an additional control, we checked the activity of a constitutive promoter under the same conditions as described here. For results on these experiments, check out our BBa_K584001 page.

Constitutive promotor + GFP generator (Part:BBa_K584001)

We fused the constitutive promoter to a GFP reporter, and assayed the promoter’s activity after addition of different amounts of arabinose and IPTG. As such these results can serve as a control for the results obtained for the arabinose-inducible BBa_K584000 and IPTG-inducible BBa_K584002 bricks. We tested the activity both in a TOP10F’ (figure 1 & 3) as well as a MG1655 (figure 2 & 4) E. coli strain background.

Apart from a small growth defect and somewhat lower fluorescence at the 8 hour time point after the addition of 2% arabinose, the activity of the constitutive promoter in TOP10F’ cells is unaffected by arabinose addition, as seen in Figure1. For the MG1655 cells (Figure 2), however, we see clear effects of adding arabinose at concentrations of 0.2% or 2%, with promoter activities significantly lower than the control condition. This points to interference of high inducer levels with the activity of the promoter and the importance of metabolism of the inducer, as arabinose can be metabolized by MG1655 cells, but not by TOP10F’ cells.

For both TOP10F’ and MG1655 cells, the activity of the constitutive promoter is unaffected by the addition of IPTG (Figures 3 and 4).

pLac-Lux hybrid + GFP generator (Part:BBa_K584002)

To gain insights into the usefulness of the lactose-inducible promoter K091100 in our 2011 iGEM project, we fused the promoter to a GFP reporter, and assayed the promoter’s activity after addition of different amounts of IPTG. IPTG is a molecular mimic of allolactose, a lactose metabolite that triggers transcription of the lac operon. Unlike lactose, IPTG cannot be metabolized by wild-type E. coli cells, leading to a constitutive high presence of inductor. For this reason, IPTG is often used instead of lactose to induce the lac operon. We tested the activity both in a TOP10F’ (figure 1) as well as a MG1655 (figure 2) E. coli strain background.

Addition of IPTG to TOP10F’ cells results in a minor growth defect, as can be seen in Figure 1A. However, this does not seem to inhibit the induction of the promoter, as Figure 1B clearly demonstrates that IPTG results in a clear induction of fluorescence, while without IPTG no such induction is seen.

In MG1655 E. coli cells, IPTG at a concentration of 1.5mM leads to a minor growth defect (Figure 2A). Looking at the fluorescence measurements (Figure 2B), it is clear that the promoter already displays high activity in the absence of inducer, and that this activity is not increased by adding IPTG. More still, adding IPTG at concentrations of 1mM or 1.5mM even results in a lower fluorescence, and hence promoter activity, than the situation without IPTG addition. We think that this is due to the fact that we express our construct on a multicopy vector, which may outcompete the inhibitory activity of the LacI repressor under these conditions. In contrast to the MG1655 cells, the TOP10F’ strain contains a high copy LacI repressor, making the investigated promoter activity not leaky. These experiments highlight the importance of checking the strain background for compatibility with a desired system. The investigated promoter, pLac-Lux (K091100), should also be repressed by the luxR repressor bound to the corepressor CO6HSL. Due to time limitations, we could not investigate this aspect of the promoter. As an additional control, we checked the activity of a constitutive promoter under the same conditions as described here. For results on these experiments, check out our BBa_K584001 page.

HybB promotor + GFP generator (Part:BBa_K584004 )

To test the usefulness of the cold shock-inducible promoter J45503 in our 2011 iGEM project, we fused the promoter to a GFP reporter, and assayed the promoter’s activity after a temperature shift from 37°C to 25°C or 4°C. We tested this activity both in a TOP10F’ (figure 1) as well as a MG1655 (figure 2) E. coli strain background.

We can see clearly that transferring cells (both TOP10F’ and MG1655) to lower temperatures (4°C and even 25°C) results in a growth arrest between the 1 and 4 hour time points of our experiment (Figures 1A and 2A). We see that promoter activity is induced when cells are transferred to 25°C and even when they are put in and ice bath (4°C) (Figures 1B and 2B). Unfortunately, however, cells that are kept at 37°C also display an increase in promoter activity, indicating leakiness in the system.

2. Qualitative measurements of INP and AFP activity

First of all, four different types of water were tested to see which one showed the least fluctuations in ice nucleation temperature and thus could be used for an optimal initial characterisation of INP and AFP. To do this, we measured the temperature at which ice nucleation in 25mL of tap water, Milli-Q, deionized water and filtered Milli-Q was induced. An ethanol bath was set at 0°C and falcons containing 25mlLof the different types of water were added. The temperature was then gradually cooled to -30°C. This experiment was performed in triple and results showed that deionized water had the smallest fluctuations in induction of ice nucleation (Figure 1). Deionized water will thus be used for all further experiments. It should be noted that this is no quantitative approach as the temperature in the falcons was not monitored, but only the temperature of the ethanol bath.

To see whether the time necessary for the induction of ice nucleation was influenced by AFP and INP, an ethanol bath was set at - 30°C and falcons containing 25mL of deionized water with no cells, cells expressing a GFP, the INP or the AFP construct were added. The time necessary for ice nucleation was measured and the experiment was done in triple and repeated two times. Figure 2 shows that the time necessary to induce ice nucleation is vastly reduced when cells expressing INP are added, whereas the time increased when cells expressing AFP were added. The speed of ice nucleation of deionized water without cells or with cells expressing a GFP construct was only marginally different.

Another experiment was carried out to learn if the cells expressing AFP or INP could be used multiple times. Ice nucleation was induced in 25mL of deionized water containing cells expressing either AFP or INP, the formed ice was then thawed and the same water was used to re-induce ice nucleation. The time necessary to re-induce ice nucleation was not strikingly different from the first ice nucleation induction. The experiment was performed in triple and the data are shown in figure 2. In conclusion, both AFP and INP can be re-used without loss of activity. Finally, we checked if INP or AFP expression influences the speed of ice thawing. This was performed by freezing 25mL of deionized water at -30°C containing no cells, cells expressing a GFP, the INP or AFP construct. Subsequently the falcons were put in a water bath at 21°C and the time necessary for thawing was measured. This experiment was done in triple and, as can be seen in figure 3, all the different solutions thawed at a similar speed. Hence, we conclude that neither AFP nor INP has an effect on the thawing speed.

3. Quantitative measurements: Differential scanning calorimetry


Through qualitative analysis of E.D. Frosti coated with Ice Nucleating Protein (INP) and the Antifreeze Protein (AFP), we were able to prove that these proteins work. But to know ‘how good’ these proteins work, we decided to do quantitative measurements. Questions that needed to be answered were: What is the exact difference in temperature on which the ice crystallization occurs when we add E.D. Frosti with INP in water compared to only water? Could ice rinks save energy if they use E.D. Frosti? On which temperature does ice crystallization happen when we add E.D. Frosti coated with AFP to water? Is this temperature lower than that of salt? To answer these questions we did Differential scanning calorimetry measurements.

Differential Scanning Calorimeter

A differential scanning calorimeter (DSC) (DSC 7 Perkin Elmer) measures the difference in the amount of heat required to increase the temperature of a sample and reference, so they can both be maintained at the same temperature. During a phase transition heat is released (exothermic reaction) or absorbed (endothermic reaction). Depending on which process is happening, less or more heat will flow in to the sample than in the reference.

A typical curve from cooling a water sample with DSC is shown in Fig. 1. Such a DCS-curve makes it possible to determine some absolute values for phase transition characteristics: the onset temperature is considered as the nucleation temperature, the area below the curve is a measure for the difference in fusion enthalpy. Fusion enthalpy is the enthalpy for the liquid-solid phase transition. Crystallization is an exothermic process, which means that during crystallization, heat will be released and less heat will be needed to keep the sample at the same temperature as the reference. We can also determine freezing points out of heating DSC-curves. A typical heating DCS-curve is shown in Fig. 2 and here the heat flow is reversed because melting is an endothermic process. However, for all our different conditions, we only looked at the effect on the ice nucleating temperature.

The difference between the nucleation point and the freezing point is shown in Fig. 3. Nucleation point is the temperature where crystallization is initiated and freezing/melting point is the temperature where crystal growth/melting takes place. This difference is crucial to understand the comparison between AFP and salt later on.

Experimental procedures

We dissolved the various cells till OD600nm of 0.1 (10^8 cells/mL) using different kinds of water. Each time we pipetted 10µL (10^6 cells) in the sample cell and in the reference cell we put nothing. Afterwards, we programmed the DSC to decrease in temperature at 10°C/min, so we can measure on which temperature crystallization occurs. Different kinds of water and cells where used, in total we tested 9 combinations of 3 kinds of water and 3 kinds of additives.


Deionized water
Fig. 4 shows that the onset temperature of deionized water equals -22,930°C. Comparing this to the onset temperature of deionized water with AFP -26,053°C (Fig. 5), it is clear that E.D. Frosti coated with AFP decreases the ice nucleation point. In contrast with AFP, INP increases the temperature at which ice crystallization starts. In Fig. 6 we see that INP has a much bigger effect, the onset equals -10,101°C. AFP causes a difference of 3,123°C, while INP changes the ice nucleation temperature with 12,829°C. Given we added the same amount of cells, we can conclude that both proteins work and INP is much more active than AFP.

Ice rink water
The Belgian Ice Hockey Federation provided us with water they use for their ice rinks. They asked us to compare it with the effect of E.D. Frosti with INP. Fig. 7 shows that the water of the Belgian Ice Hockey Federation has a higher ice nucleation temperature (-17,406°C) than deionized water (-22,930°C) (Fig. 4). The extra filtering they do certainly has an effect (from -22,930°C to -17,406), but the addition of INPs has a much stronger effect (from -22,930°C to -10,101°C) (Fig. 6). As we can see in Fig. 9, if we use ice rink water and E.D. Frosti with INP on its membrane, we get even better results. The onset temperature is raised to -5,686°C. Fig. 8 proves that AFP has approximately the same effect on ice rink water as on deionized water.

NaCl water
One of our applications is using AFP to scatter on the roads during the winter. Comparing the ice nucleation point of salty water (-25,794°C) (Fig. 10) and AFP (-26,053°C) (Fig. 5), we see that AFP decreases the onset temperature more than 1M NaCl. But the combination of salty water and AFP is additive, as in Fig. 11 we can see that the ice nucleation occurs at -27,094°C. Hence, it would be better to scatter with a mixture of AFP and NaCl water. To evaluate whether INP would still work if NaCl ions are present, we also tested NaCl water with INP. As shown in Fig. 12 INP still works, but a little bit less than when used in combination with deionized water. The difference in the degree of supercooling drops from 12,829°C to 10,394°C (Fig. 6, 8, 10, 12).

Protein activity
To know how INP would react to high temperature, we raised the temperature till the average denaturation temperature of proteins. Each time we increased the maximum heat temperature at a rate of 200°C/min and returned to -40°C with a rate of 10°C/min.

Table 1 and Figure 13 show that the nucleation temperature decreases when we raise the temperature of our sample with the DSC before cooling down again. We can conclude that loss of activity due to denaturation of the INP increases with higher temperatures and starts to occur when our samples are heated to a temperature between 50-60°C.


Nucleation points
When we look at Fig. 14, the effect of INP and AFP on the nucleation points is very clear. Addition of INP to any of the 3 kinds of water results in a major decrease in the absolute value of the nucleation point. When we add AFP to any of the 3 kinds of water, we always observe an increase of the absolute value of the nucleation point, although to a lesser extent.

Besides the effect of INP and AFP, also a difference in nucleation points is visible for the different kinds of water. Ice rink water freezes at a lower degree of supercooling in comparison with deionized water. On the other hand, salty water has a higher degree of supercooling, although this degree is lower than when AFP is added to water.

We also observed differences in enthalpy (Fig. 15) for which no direct explanation can be given. Addition of INP increases the enthalpy in deionized water, where it decreases in ice rink water. For AFP, lower enthalpies are observed under both conditions. However, since the major amounts of energy that can be saved during the initial freezing process is due to the higher ice nucleation temperature, these differences are of minor importance.


The enormous difference in nucleation temperature for water with and without INP opens up possibilities to save energy in refrigeration costs. Here, we suggest the implementation of INP in ice rink water, but every process where freezing is an important step could possibly benefit from using INP.

The difference in nucleation temperature for water with and without AFP seems to be small compared to the differences with INP. However this effect may not be underestimated: when the comparison is made between deionized water with AFP (first green balk, Fig. 14) and salty water without cells (third red balk, Fig. 14), they seem to be almost equal. This fact does not implicate that AFP could be an alternative for salt. Salt has another activity in preventing freezing, it can depress the freezing point. This phenomenon is colligative in that it is directly proportional to the concentration of the solute species regardless of its chemical nature. The antifreeze action by AFP is definitely a non-colligative phenomenon, as proven by early studies.[1,2] The essential difference between salt and AFP activity, is that salt decreases the freezing and nucleation temperatures and AFP only decreases the nucleation temperatures (see Fig. 3). A result is that melting ice is impossible with AFP.[3] We could, however, use AFP in addition to salt, because the nucleation is made more difficult, referring to the decrease of nucleation temperature when adding AFP.


[1] Yeh Y., Feeny R.E., Antifreeze proteins: Structures and Mechanisms of Function. Chemical Reviews 96 (1996) 601-618
[3] Chattopadhyay M. K; (2007); Antifreeze Proteins of Bacteria; Resonance p 25-30
[2] DeVries, A.L., Wohlshlag, D.E. Science(1969) 163, 1073