Results
The InflammatoryToxinSensor can detect and report bacterial toxins
In order to generate our cell-based sensor for the early and minimally-invasive detection of bacterial infection, we designed and cloned many different genetic constructs. These constructs are composed of different combinations of the single parts we use for our sensor and therefore had to be characterized and validated. As proof of concept experiments, we performed in vitro studies in which we transfected HeLa cells with either genetic construct and subsequently analyzed the cells’ protein expression by appropriate methods.
Expression of MagA and Gaussia Luciferase in HeLa cells
In a first set of experiments, we aimed to show that the expression of our two reporters – MagAcreated part:
BBa_K3338000 and
Gaussia Luciferasecreated part:
BBa_K3338001 – was possible in HeLa cells.
For this, we designed two genetic constructs with either reporter gene under the control of the
constitutively active CMV promoterused part:
BBa_I712004. In the case of MagAcreated part:
BBa_K3338000, we further genetically introduced a
fluorescent tag (eGFP) to the protein to make expression analysis via fluorescent microscopy
possible. For the generation of both plasmids, NEB HiFi DNA cloning kit was used. The DNA sequence
of the respective reporter gene was amplified by PCR using appropriate primers for subsequent
introduction into the previously linearized plasmid backbone pEGFP-C2created part:
BBa_K3338020.
Successful cloning was
verified by DNA sequencing.
Transfection of HeLa cells with either plasmid, respectively, was performed by lipofection (ViaFect transfection reagent) or electroporation.
The constitutive expression of the eGFP-MagAcreated
part:
BBa_K3338000 fusion protein was analyzed by fluorescent microscopy 24
hours after transfection. Figure 1 shows that transfected HeLa cells synthesized MagAcreated part:
BBa_K3338000 and that the
protein was not localized in the area of the nucleus or the cytoplasm, but rather in membrane
regions. This finding is in good agreement with our expectations, as in magnetotactic bacteria,
MagAcreated part:
BBa_K3338000
is a transmembrane protein serving as an iron transporter.
BBa_K3338000 expressing HeLa cells. Both fluorescence (left) and brightfield channel (middle) as well as a merge (right) are shown. Scale bar: 10 µm.
The constitutive expression of Gaussia Luciferasecreated part:
BBa_K3338001 was analyzed by performing a luminescence assay. In
this, the enzyme’s substrate coelenterazine was added to the cell culture supernatant, which is
expected to contain the Gaussia Luciferasecreated
part:
BBa_K3338001 that was secreted by the transfected HeLa cells. Since
luminescence is produced in the reaction of Gaussia Luciferasecreated part:
BBa_K3338001 with its substrate, we qualitatively
detected the enzyme’s expression via camera. Figure 2 shows that luminescence was only observed when
supernatant from previously transfected HeLa cells was utilized, while usage of negative controls
such as cell culture media or supernatant from non-transfected HeLa cells did not lead to a
generation of light. On the one hand, these results demonstrate the expression of Gaussia Luciferasecreated part:
BBa_K3338001
by transfected HeLa cells. On the other hand, it also shows that the protein is secreted by the
cells. Furthermore, this study proves the functionality of the Gaussia
Luciferasecreated part:
BBa_K3338001 by showing that it
actually exhibits the expected enzymatic activity.
BBa_K3338001 in supernatant of previously transfected HeLa cells. Negative controls: cell culture media or supernatant from non-transfected HeLa cells.
Taken together, the results we achieved by these first experiments suggest that both our chosen
reporters can be expressed by human epithelial cells such as HeLa. While we showed the correct
intracellular localization of MagAcreated
part:
BBa_K3338000, we further demonstrated the functionality of Gaussia Luciferasecreated part:
BBa_K3338001,
making both proteins possibly well suited for our intended sensor concept.
Simultaneous expression of MagA and Gaussia Luciferase in HeLa cells – Introduction of P2A
In our final sensor, we aim to use both MagAcreated
part:
BBa_K3338000 and Gaussia Luciferasecreated part:
BBa_K3338001 as reporters which are expressed
under control of the same, single promoter. To achieve this goal, it needs a special part in our
genetic construct which connects both reporters and enables the simultaneous synthesis of both
proteins. For this, we tested two options: IREScreated part:
BBa_K3338004
and P2Acreated part:
BBa_K3338003
site. IREScreated part:
BBa_K3338004
stands for Internal Ribosomal
Entry Site and P2Acreated part:
BBa_K3338003
represents a 2A peptide which encodes for a self-cleaving peptide. The difference
between these two “connections” is that they control/regulate the formation of the two reporter
proteins at different stages of the expression. The use of IREScreated part:
BBa_K3338004
enables a second ribosome to bind to the mRNA independently from the 5’-cap, thereby leading to the generation of the two individual reporter proteins:
MagAcreated part:
BBa_K3338000
and Gaussia Luciferasecreated part:
BBa_K3338001. In contrast, the presence of a P2Acreated
part:
BBa_K3338003
site first leads to a long, single mRNA
transcript containing MagAcreated part:
BBa_K3338000, P2Acreated part:
BBa_K3338003
and Gaussia Luciferasecreated part:
BBa_K3338001, which is cleaved into its individual
components only during translation. In the end, both IREScreated part:
BBa_K3338004
and P2Acreated part:
BBa_K3338003
can lead to the expression of our
two reporters, but the stoichiometry of both proteins might be different in either case.
Since our goal is to let both reporters be expressed in a stoichiometrically equal manner, we
designed two genetic constructs that carry the fluorescent proteins eGFP and mCherry as well as
either IREScreated part:
BBa_K3338004
or P2Acreated part:
BBa_K3338003
as connecting part. In both cases, we used the constitutively active CMV
promoterused part:
BBa_I712004
to control the expression. For the generation of both plasmids, NEB HiFi DNA cloning kit was used.
The DNA sequence of the single parts was amplified by PCR using appropriate primers for subsequent
introduction into the previously linearized plasmid backbone pEGFP-C2created part:
BBa_K3338020.
Successful cloning was
verified by DNA sequencing (please see Parts.).
To compare the stoichiometry of simultaneous expression of both fluorescent proteins in the case of
IREScreated part:
BBa_K3338004
and P2Acreated part:
BBa_K3338003,
respectively, transfection of HeLa cells with either plasmid was performed by
lipofection (ViaFect transfection reagent) or electroporation and fluorescence microscopy was used
for subsequent analysis. As demonstrated in Figure 3, we observed a large difference in fluorescence
intensity – meaning expression strength – of eGFP and mCherry in the case of the IREScreated part:
BBa_K3338004
construct.
Further, many eGFP positive cells did not show an mCherry signal. In contrast, in the case of the
P2Acreated part:
BBa_K3338003
construct, nearly all cells that expressed eGFP were mCherry positive as well. The mCherry
fluorescence intensity is higher than observed for the IREScreated part:
BBa_K3338004
construct and it is roughly comparable
with the respective eGFP signal.
BBa_K3338004-mCherry (top) or CMV-eGFP-P2Acreated part:
BBa_K3338003-mCherry (bottom). Images of three channels are shown: green fluorescence (left), red fluorescence (middle) and brightfield (right). Scale bar: 100 µm.
As a result, we decided to implement the P2Acreated
part:
BBa_K3338003
site as connecting part for simultaneous expression of
both reporters in our sensor concept.
How to make our construct “sensing”
Until this point, we have constantly utilized the constitutively active CMV
promoterused part:
BBa_I712004 to analyze the
gene expression for characterization of the sensor’s reporters. For our sensor concept though, a
promoter with a low basal expression and an activity that is inducible by inflammatory signals is
needed to actually make our construct sensing.
A broad variety of bacterial components, including for example the bacterial endotoxin lipopolysaccharide (LPS), act as such inflammatory signals and are detected by human cells via binding to extracellular Toll-like receptor domains. In the course of the signal transmission, these receptors often induce the intracellular NF-κB pathway, in which the transcription factor NF-κB is released. After translocation to the nucleus, NF-κB can bind to specific DNA promoter sequences thereby activating gene expression. Hence, for our sensor concept, we aimed to decipher the best conditional promoter design for NF-κB sensing.
For this, we tested four different inducible promoters, which are described in more detail below. For
every promoter, we designed a genetic construct composed of the promoter and Gaussia Luciferasecreated part:
BBa_K3338001 as
reporter for quantitative readout of expression strength. For the generation of all plasmids, NEB
HiFi DNA cloning kit was used. The DNA sequence of the single parts was amplified by PCR using
appropriate primers for subsequent introduction into the previously linearized plasmid backbone
pEGFP-C2created part:
BBa_K3338020.
Successful cloning was verified by DNA sequencing. For analysis, HeLa cells were
transfected with either plasmid. The plasmid carrying Gaussia Luciferasecreated part:
BBa_K3338001 under control of the CMV
promoter served as reference. On the day after transfection, cells were treated with different
concentrations of LPS (0, 0.5, 1 or 2 µg/mL) for 3 hours in order to stimulate the NF-κB pathway.
Subsequent analysis of expression strength was performed by luminescence assay. Therefore, cell
culture supernatant was collected 24 hours as well as 48 hours after LPS treatment and luminescence
was quantitatively read out. A schematic representation of the experimental procedure is shown in
Figure 4.
CMV promoter
As mentioned before, the CMV promoterused part:
BBa_I712004 is characterized by its constitutive activity with rather high
basal expression levels. Due to this, we did not expect to observe any LPS induced changes in the
promoter’s activity, making it an ideal reference. The results, which are shown in Figure 5, are in
good agreement with this and demonstrate the importance of choosing the right promoter to make our
sensor concept work.
BBa_I712004 in HeLa cells 24 hours (A) and 48 hours (B) after treatment with different concentrations of LPS for 3 hours, normalized to untreated control (0 µg/mL LPS). Data shown represents mean ± SEM of n=4 biological replicates. Statistical analysis was performed by unpaired t-test in comparison to untreated control, significance level: 10 %, significance is indicated by asterisk.
Interleukin-6 promoter
The Interleukin-6 (IL6-promotercreated part:
BBa_K3338008),
which plays an important role in (patho)physiological inflammatory
processes, features such a desired NF-κB inducible activity. Thus, we chose this promoter for our
sensor concept.
Figure 6 shows that LPS treatment of cells indeed led to an increased expression of Gaussia luciferase. While the relative enhancement is rather small, it is still significant.
BBa_K3338008 in HeLa cells 24 hours (A) and 48 hours (B) after treatment with different concentrations of LPS for 3 hours, normalized to untreated control (0 µg/mL LPS). Data shown represents mean ± SEM of n=4 biological replicates. Statistical analysis was performed by unpaired t-test in comparison to untreated control, significance level: 10 %, significance is indicated by asterisk.
As the sequence of the IL6-promotercreated
part:
BBa_K3338008
contains a restriction site which is not compatible with the iGEM
BioBrick system though, we introduced a point mutation at this site. We named the resulting part
IL6mutcreated part:
BBa_K3338005
promoter.
Figure 7 shows that, in contrast to our expectations and to the behavior of the native IL6-promotercreated part:
BBa_K3338008,
LPS stimulation of HeLa cells carrying the plasmid with Gaussia
Luciferasecreated part:
BBa_K3338001 under control of the
IL6mutcreated part:
BBa_K3338005
promoter did not lead to an enhanced expression within 24 or 48 hours,
respectively, after
treatment. While it seems as if LPS treatment negatively influenced the promoter activity, no
significant changes could be detected in comparison to untreated control.
These results indicate that the introduced point mutation might have led to an unpredicted negative effect on the promoter’s activity in terms of its NF-κB sensing capabilities. Unfortunately, this makes the IL6mut promoter unsuitable for our sensor concept.
NF-κB-AP1-minimal-CMV promoters
Besides the IL6-promotercreated part:
BBa_K3338008
and the CMV promoterused part:
BBa_I712004, we further tested two self-designed promoters (NF-κB-AP1-minCMV1created part:
BBa_K3338007
and
NF-κB-AP1-minCMV2created part:
BBa_K3338002).
These
were based on a minimal CMV promoter with AP1 and NF-κB transcription factor binding sites. Using
Matlab, we generated random sequences containing three AP1 and NF-κB binding sites each and the
minimal CMV promoter. For generation of the construct, we used synthesized sequences.
As seen in Figure 9 and Figure 10, both synthetic NF-κB-AP1-minCMV promoters exhibited a significant upregulation of luciferase expression upon LPS stimulation. This behavior was observed 24 hours as well as 48 hours after stimulation. In the case of NF-κB-AP1-minCMV1 promoter, significance increased after 48 hours compared to 24 hours. These results suggest that both self-designed promoters could serve as potential candidate parts for our final sensor construct.
Conclusion
Taken together, we showed that the IL6-promotercreated part:
BBa_K3338008‘s
as well as our two self-designed NF-κB-AP1-minCMV
promoters‘ activities can be induced by LPS stimulation. Since this feature is essential for our
sensor’s capability to detect a bacterial biofilm, we concluded that these promoters pose potential
candidate parts for our final sensor construct.
A low basal expression of the reporter proteins is important for our sensor concept
Besides the sensing capability, what else do we need to consider? To obtain a high sensitivity of our sensor, it needs a low basal expression level. Moreover, we do not want constitutive expression of both our reporter proteins in the human body. Consequently, we compared all basal expression levels based on luminescence.
Our results (Figure 11) show that, indeed, the native IL6-promotercreated part:
BBa_K3338008
proved to be the best choice as it
resulted in a very low basal expression. Only the BioBrick system adapted IL6mut promoter
showed an
even lower basal expression, but unfortunately, this is possibly due to the fact that the promoter
was not really active and inducible anymore in general. Interestingly, both self-designed promoters,
which are based on the minimal CMV promoter, exhibited a strong enhancement of basal expression
compared to the CMV promoterused part:
BBa_I712004. One reason for this observation might be an overall increased
transfection efficiency due to the lesser size of the plasmid. Further experiments are needed to
clarify.
BBa_I712004 which served as reference. Data shown represents mean ± SEM of n=4 biological replicates.
Taken together, we showed that the native IL6-promotercreated part:
BBa_K3338008’s
properties suit best for our sensor concept.
While the synthetic promoters proved to be very good for LPS detection, based on their high basal
expression, they could not fully fulfil our purpose though. So we had our Inflammatory Toxin Sensor ready:
BBa_K3338010.
Can we characterize magnetic cells?
Our idea was to build and operate a measuring chamber, which will allow future iGEM Teams to characterize their own magnetic cells or particles in a fast and efficient way.
We used magnetic beads to proof our measuring concept and identify possible problems with our chamber. Iron microparticles with a diameter of 10 µm to 100 µm and iron-oxide beads with a diameter of 1 µm were used. The particles were suspended in water in the particle solutions.
During our first trials we had problems with the entrapment of air bubbles in the chamber. Air bubbles were in general bigger than our fluid channels, therefore shutting it down completely by clogging the micro fluid channels. To resolve this problem we added soap to our solution, reducing the surface tension of the water and making the formation of bigger bubbles impossible. The magnetic field was measured using a Hall-effect sensor.
The measuring chamber is connected to 200 µl pipette tips. These are used as injector, as well as collection vials. The inputs are fed by two syringe pumps with B Braun Original Perfusor®. We choose 10 mm /s as a valid flow velocity for observation and particle acceleration. Also we determined that at this velocity the impact of diffusion would be minimal. We calculated that at this speed the flow would be laminar using the Reynolds number:
\[ R={\mu L \over \nu} \]Therefore the volumetric flow rate of the mixing chamber had to be: \[\dot{V} = 40 ml/h\] To guaranty separated output streams we adjusted the input volumetric flow rate, the rate for the particle solution has to be slightly lower than for the buffer solution. Thus no particles end up in the output if not attracted by the magnet. Thus we used \[\dot{V} = 23 ml/h \] for the buffer solution and \[\dot{V} = 17 ml/h \] for the particle solution.
For counting, we used a brightfield microscope. We placed a droplet of our output solution on a glass coverslip and took an image with the microscope. Afterwards we used the Particle Analysis plugin of the software ImageJ [1].
We were able to observe an influence of the magnet on the particles within our measuring chamber, which verified our concept of a magnetophoresis-micro-fluidic-system. Figure 14 and figure 15 show the results of our trails with two different kind of magnetic particles. The particles were counted by the ImageJ software and then the ratio of the moved particles was calculated. All values were then displayed in a boxplot diagram. With an increase of the magnetic field, the ration of moved particles is increased, for each the iron micro particles as well as the magnetic beads.
However, some particles were moved into a different channel even without an eternal magnetic field. Diffusion and most likely convection are the reason for this phenomenon, due to design and manufacturing errors in our measuring chamber.
For the iron micro particles with a diameter of 100 µm to 10 µm even small magnetic fields with 8.7 mT showed a drastic increase of movement of the particles. This increase in movement compared to the smaller 1 µm magnetic beads can be contributed to their bigger volume and therefore increased magnetic momentum. On the other hand, the drag of the bigger surface is smaller than the magnetic force moving the particles in the surrounding fluid. For the magnetic beads, a defined increase of movement is detectable at 45 mT and above.
The measuring chamber is valid for characterizing magnetic particles, for the cellular experiments
with our MagAcreated part:
BBa_K3338000 expressing cells some adjustments are needed. The magnetic momentum in the MagAcreated part:
BBa_K3338000 cells
is
located in the magentosomes in nanosized magnetite crystals. Therefore the magnetic field for the
measurement has to be at least 50 mT following the experiments with the magnetic beads. Also the
drag
of the cell will be increased compared to the beads. This is analyzed in ongoing experiments.