Wetlab
Figure 1: Engineering-Design-Cycle with details how each step was or should be tackled within the wet lab work.
Biological engineering is a major part of iGEM’s objectives, already contained in the acronym itself.
Thus, this creative and systematic process, necessary for great projects and outcomes, is supposed to
receive greater attention this year. Figure 1 gives an overview of the Engineering-Design-Cycle
and how
we fulfilled each step.
Research: During project finding and planning, we researched literature on heavy metal pollution
and on the possibilities of binding heavy metal ions. Environmental heavy metal contamination is often
man-made, stemming from our agricultural and industrial activities [1, 2]. This is not only the case in
developing countries [3, 4]; areas with heavy metal contaminated soil can be found in Europe [5].
Furthermore, it was shown that rising ground water temperatures are connected to higher manganese
concentrations in Germany [6]. Thus, rising temperatures caused by climate change could possibly enhance
problems with heavy metal pollution. This local connection and the fact that not many iGEM projects have
worked with this metal so far, led to our decision to focus on manganese. Plants, animals, and human
health can be affected by manganese exposure. In plants, exposure to manganese pollution causes biomass
loss. In humans, its inhalation causes neurotoxicity [7–10], and recent studies suggest that
manganese-contaminated drinking water poses health risks [11]. Interestingly, plants possess a
remarkable disposal mechanism for heavy metals; they produce cysteine-rich oligopeptides named
phytochelatins, which form complexes with heavy metal ions, especially cadmium. The complexes are then
stored in the vacuole, preventing interference with cellular biochemical processes [12]. To address
heavy metal contamination, an overexpression of synthetic phytochelatins on the surface of E.
coli was
attempted and resulted in more accumulation of cadmium in comparison to wild-type E. coli [13].
Imagine: This led us to our project PacMn: Phytochelatin-actuated complexation of Manganese. To
combat heavy metal contamination in water, methods for detection and capture of these ions are
crucial. Therefore, we want to create a bifunctional biosensor that fulfills both tasks using the
synthetic phytochelatin EC20 already included in the parts registry (BBa_K1321005), which has been
shown to increase manganese adsorption when expressed on C. metallidurans (14). although to a
lower
extent than other metals. Thus, future adaptations of phytochelatin may be necessary to increase its
complexing efficiency for manganese.
One of our primary aims is to have a bifunctional biosensor with a high specificity for manganese ions.
Accordingly, we hypothesize that by using a riboswitch-based sensing mechanism, we will gain stronger
specificity than via standard repressor / activator proteins of bacterial operons [15]. The effects that
manganese-complexing will have on the functionality of the biosensor are unknown. At this point, we can
only assume that complexed manganese ions in the phytochelatin will not be sensed by the riboswitch due
to its high specificity for free dissolved ions [16], creating a self-regulated system in the form of a
negative feedback loop. In contrast, if riboswitches could be stimulated by complexed manganese ions, an
intracellular accumulation of manganese would lead to a positive feedback loop and, ultimately, a false
positive signal for extracellular manganese concentrations.
Because E. coli is a model organism for cloning work, we chose it as our biosensor host. However, other
microbial organisms might be more suitable for distinct extreme conditions. Thus, at a later stage, we
may consider switching hosts. Naturally, we wish to create a product that adheres to biosafety
regulations. Considering our work will entail the usage of genetically modified organisms (GMOs), a
potential safety measurement for the application of our product would be the implementation of a filter
system permeable to manganese ions and impermeable to cells.
Design: To address our questions and reduce unknown variables which would arise in a more complex
system, we decided to simplify our design as much as possible. In order to study the individual
components of each construct, we designed four similar constructs for comparative analysis (Table 1).
Although only one type of phytochelatin was included in our constructs, other variants were also
modelled by our dry lab.
Table 1: Description of all components (BioBricks) of our PacMn constructs.
| Name | Part ID | Function |
|---|---|---|
| Manganese- promoter | BBa_K902073 | A manganese-inducible promoter for transcriptional regulation |
| Anderson-promoter | BBa_J23102 | A constitutive promoter to evaluate the effect of having and not having transcriptional regulation |
| Manganese- riboswitch | BBa_K902074 | A manganese-inducible riboswitch for translational regulation |
| FAST2-tag | BBa_K3510000 | A fluorogen-activating protein for interchangeable fluorescence |
| Phytochelatin | BBa_K1321005 | A chelator for manganese-decontamination |
| Chromoprotein | BBa_K864401 | A blue chromoprotein to evaluate the effect of phytochelatin on the sensing mechanism of the construct and to provide a direct color change for the naked eye |
| Double terminator | BBa_B0015 | A reliable double terminator |
Different combinations of the seven components (Table 1) resulted in four PacMn constructs (Fig.
2). For
details regarding design and function of these parts please visit the
corresponding wiki page.
Figure 2: Schematic overview of the PacMn constructs. Each Gibson Assembly consisted of two fragments and a pUC19 vector backbone. The dotted line indicates the ends of the two insert fragments. Here they are going to get connected by Gibson Assembly.
Build: To assemble our constructs, we chose the Gibson Assembly® method, which is an efficient
and seamless method for assembly of multiple linear DNA fragments [17]. This method relies on the
addition of complementary overhangs to each DNA fragment by PCR. The overhangs of a fragment must be
complementary to its adjacent fragments, and careful primer design is required, in which we succeeded
with the support of the SnapGene software. The extended fragments were assembled together along with
pUC19 as the vector backbone. Upon successful cloning, our constructs were subjected to measurement
experiments and others.
As neither the initial PCRs nor the Gibson Assembly had worked on the first try, we had to iterate
through the Build-Test-Learn-Improve cycle multiple times, before we finally obtained the correct
constructs.
Test: Initial issues occurred with the PCRs to create the overhangs for Gibson Assembly. When
analyzing PCR products by gel electrophoresis, we encountered several problematic results ranging from
no bands, over faint bands, to many unspecific bands.
Learn: Therefore, we had to investigate ways to troubleshoot and optimize PCR. A summary can be
found
in
Contributions.
Improve: With this knowledge we then edited the PCR protocol for each fragment and decided to
include
gel-extraction in our process.
Build: Once all PCRs worked and we had all fragments with the necessary overhangs, we approached
Gibson Assembly.
Test: To assess successful assembly and transformation, we performed colony-PCR as an initial
detection
method. Plasmid DNA was then isolated from potential candidates and subjected to analytical restriction
digestion with the restriction enzymes EcoRI and HindIII. Only the GA1 construct (Fig. 1) was
successfully cloned on the first try.
Learn & Improve: Following the failed attempt to clone the other three constructs of the
manganese
project, we adjusted the vector-insert molar ratios based on the analysis of the gel images from
colony-PCR.
Build & Test: With this adjustment, we were able to clone the other three main constructs (GA2-4)
the
second time we attempted Gibson Assembly. All positive clones were then sequenced to check for
mutations. Unfortunately, we noticed that an ordering mistake was made for the FAST2I-tag. The
original FAST2-tag sequence contains the restriction site for PstI, which is incompatible with the
accepted iGEM standards BioBrick RFC[10] and Type IIS RFC[1000]. This issue was initially excluded
with a silent mutation in the restriction site. However, the original FAST2-tag sequence was
accidentally ordered instead of our modified version.
Learn: To solve the compatibility issue, we considered two options: Ordering the correct
sequences
and repeating the cloning work with our reliable protocols, or ordering primers for site-directed
mutagenesis. Due to the late detection of the problem and the differing shipping times of primers
and longer DNA fragments, we decided to order both and execute the method that was more readily
available.
Build: To limit invested time and money, and considering we already had a polymerase appropriate
for site-directed mutagenesis, we did not order a special kit for this. Since both orders arrived on the
same day, we chose to repeat PCR and Gibson Assembly with the new FAST2-tag sequences, which was the
more familiar and therefore reliable cloning method for us.
Test: To assess successful assembly and transformation, we applied the same test procedure as
previously: Colony PCR, analytical restriction digestion of isolated plasmid DNA, and sequencing.
Correct transformants were obtained on the first try for all four constructs. For assessment of the
site-directed mutagenesis, we had planned to digest isolated plasmid DNA of potential transformants
(after analysis of colony PCR) with PstI to verify if the restriction site was successfully changed
or not. The samples which were not cut would have then been sequenced for final confirmation of
successful mutagenesis. For our proposal on how to successfully assemble GA5 and GA6 from our
vitamin B12 side project, which are composed of one more fragment than GA1-4, see
Outlook.
Despite our encountered problem with the iGEM BioBrick compatibility, we continued experimenting with
the already older constructs to gain a first impression of what results to expect, and to optimize the
experimental set-up. By changing one nucleotide at a wobble position in the PstI restriction site, the
resulting amino acid sequence is identical. Of course, codon usage is irregular, but this only provides
more value by allowing comparative analysis of two constructs that are different in their nucleotide
sequence yet identical in their amino acid sequence. The information collected could then serve future
iGEM teams when deciding if a particular illegal sequence can be changed in a certain way that makes it
compatible with the standards and comparibly functional to its original form.
The experimental results obtained from the older and also from the newly modified constructs indicated
that there was a functional issue: No color change nor obvious fluorescence was observed in any case,
contrary to our anticipated results. The fluorescence intensity measured from E. coli containing FAST2
was similar to the negative control, as only autofluorescence was observed. Furthermore, no amount of
manganese was able to induce a color change in E. coli containing the chromoprotein.
Research: Therefore, we returned to literature research and took a closer look at our design to
figure out
where a problem might have occured. We quickly detected a misconception during our project
design that led to these results. Originally, we thought that due to the nature of the
relationship between a riboswitch and an RBS, they would have to be at a specific distance from
each other and, therefore, a RBS would be included in the riboswitch sequence from the parts
registry. In accordance with our presumption, we did not add an RBS in front of the FAST2
sequence, as this would have overridden the regulatory function of the riboswitch. This turned
out to be an expensive mistake leading to no expression of our genes of interest.
Design & Test: At the time of realization, we were a few days away from wiki freeze. Since there
was
no time to order new sequences, and repeat our mastered Gibson Assembly method, we came up with a
plan to at least test the functionality of our manganese-inducible promoter. Our plan was to analyze
transcriptional regulation, since this was independent of the RBS. The only method available to us
on such short notice was RT-PCR to analyze mRNA synthesis. We designed a qualitative experiment to
determine if mRNA synthesis was being promoted by the manganese-inducible promoter of GA1 and GA2 in
media containing 10 µM manganese(II) chloride. We even thought of a creative way of lysing our cells
for RNA extraction (see Troubleshooting). Unfortunately, the experiment was inconclusive (see
Results).
Research & Imagine: While conducting our troubleshooting literature research, we also gathered
more knowledge about manganese homeostasis in E. coli. It was found that the cells reduce the expression
of the manganese importer MntH in the presence of high manganese levels [18, 19]. This may interfere
with our biosensor, as it would prevent a proportional response of the sensor towards extracellular
manganese concentrations. In contrast, it was shown that manganese import is increased under oxidative
stress [19]. Strains lacking catalase/peroxidase (Hpx-) cannot degrade hydrogen peroxide (H2O2) and
accumulate H2O2 when cultured in oxic media [20]. As presented, there are multiple options for
optimization of our system (see Outlook).
Conclusion: Looking back at all the roadblocks that we were able to overcome, we consider our
engineering experience as successful. We experienced first-hand how unpredictable and sensitive
laboratory work can be, and how beautiful it is when different minds cooperate to find solutions for our
human mistakes. Like for many others, the SARS-CoV-2 pandemic was our largest roadblock. Its
consequences included difficulties finding a lab and consequently our late start there in the middle of
Summer, transferring to a second lab in Autumn, the limited time and space available in these labs to
ensure safety in the workplace, and not being able to meet in real life to discuss early design and
planning steps.
In spite of the difficult situation, we managed to clone our main composite parts by iterating through
the Engineering-Design-Cycle (Fig. 1). Using the extensive toolbox synthetic biology has to offer, we
overcame many issues occurring in the wet lab, especially for preparing our individual fragments for
Gibson Assembly, which took time to perfect. For cloning of our composite parts, we managed to
troubleshoot efficiently and can now provide reliable protocols to replicate this section of our work.
By applying the Build-Test-Learn-Improve cycle throughout the course of our work, we detected the
sequence mistake responsible for the shortcomings of our composite parts, and can now proceed to fix
this mistake and test them again. This proves how important it is to follow this procedure in a
meticulous way to implement a new system in synthetic biology.
Although we realized that our systems require a sequence adjustment to work as expected, and we lacked
the time to correct this before this year’s wiki freeze, all of us are satisfied with the development of
our project. Each and every one of us learned new skills and became a better scientist in a matter of
months. We could not have asked for a more realistic experience to teach us how complicated and fun
science can be.
Drylab
Figure 3: Engineering-Design-Cycle with details how each step was approached within the dry lab work.
Research: In order to find a way to contribute to the project from a perspective of modelling,
the drylab members conducted literature and database research on their own, primarily about
riboswitches, phytochelatin, and their binding mechanisms to manganese and other ions. Based on this
research, we assumed that the amino acid cysteine, which appears often in our phytochelatin, binds to
the manganese ions through its thiol groups.
Imagine: We identified the great potential in the improvement of the binding of phytochelatin to
manganese, so we decided to tackle this challenge. Additionally, we realized that the three-dimensional
structure of the phytochelatin we are using is unknown. This led us to wonder how replacing cysteine
with other amino acids would affect the structure of our phytochelatin and its binding to manganese.
Design: We exchanged all cysteine codons in the phytochelatin sequence, once for histidine and
once
for aspartic acid codons. The decision was based on their higher electronegativity in comparison to
cysteine, which might strengthen the binding to manganese.
Build: The structures of the original sequence and the two modified sequences were predicted using a
mixture of homology and ab initio modelling (see Model).
Test: To test the stability of our newfound structures, we needed to employ molecular dynamics
modelling over 50 ns.
Learn: While the structure of the original sequence and the histidine variant were stable over
the whole simulation time span, the aspartic acid variant was not. We then had to evaluate whether to
discard the structure or change the model parameters.
From our collaboration with Waterloo, we learned about the potential of quantum mechanical force fields
such as CHARMM for better estimation of sequence stability.
Improve: After changing the forcefield, the structure of the aspartic acid variant was stable for
a
relatively long time of 48 ns. Through this modification, we were able to validate this structure as
well.
