Team:MichiganState/Design

Design
 Gene Engineering
 - Secretion System
 - Genome Integration
 Bioinformatics
 - Sequestration vs. Degradation
 - Metabolomics and Transcriptomics
 Device
 - Bee Attraction
 - Biocontainment
 - Prototyping and Future Experiments
 References

Design

Overview

Although Bee-tox research was unable to do much in-person laboratory work this year, the design-build-test cycle still played an integral role in our progress. Below, learn how each of the three subteams utilized these engineering principles in their research!

Gene Engineering

Secretion System

Snodgrassella alvi is a gram-negative, microaerophilic bacterium that resides exclusively in the hindgut of bee species [1]. Since the overall goal of ours was to engineer S. alvi to produce a detoxifying enzyme, we first needed to engineer a way for the enzyme to be secreted from the cell and into the bee hindgut, where a detoxifying enzyme could aid the bee in the breakdown of pesticides.

To accomplish this, we decided to utilize a HlyA secretion system that is native to certain E. coli strains. The HlyA secretion system is a Type 1 secretion system, which spans the entire membrane and secretes specific, tagged proteins into the extracellular space [2]. HlyA is an enzyme that lyses red blood cells, and its secretion system consists of three membrane proteins. HlyB is an ABC transporter located in the inner membrane, HlyD is a membrane fusion protein anchored in the inner membrane that spans across the periplasm, and TolC is an outer membrane protein. The HlyA protein has a C-terminus secretion tag. HlyB interacts with this terminus and recruits the other membrane proteins to form a continuous channel that HlyA will be secreted out of [2].

Figure 1: HlyA secretion system proteins (from iGEM Calgary 2017 wiki)

We decided to engineer the pBTK510 plasmid from the Bee Microbiome Toolkit, which was developed at the University of Texas Austin to effectively transform bee gut symbionts[1]. This plasmid was proven to be effective in transforming S. alvi. We added the entire Hly operon with the genes for HlyB, HlyD, and TolC to this vector so S. alvi will assemble the entire secretion system in its membrane. In order to test this secretion system, we decided to insert mRFP, a fluorescent protein that has been used in secretion experiments due to its slow folding properties. Lastly, we added the C-terminus secretion tag to the mRFP gene so it will be secreted.

First, we created a gblock consisting of mRFP with the secretion tag, as well as TolC. This will be inserted in pBTK510 using Gibson Assembly to produce Plasmid 1 shown below. Then, the linked genes HlyB and HlyD will be amplified out of the iGEM 2019 Distribution Kit and inserted into the Plasmid 1 that was created using Gibson Assembly. A restriction site was introduced at the end of TolC, where the HlyB/HlyD construct will be inserted. This 2-step construction plan was used because the entire DNA sequence containing all of the secretion membrane proteins, as well as mRFP, was too large to have synthesized. Once the plasmid is assembled, we will transform it into donor strain E. coli WM6026, then conjugate it into S. alvi. We plan to test this design by quantifying the fluorescence that occurs extracellularly in the supernatant, to assess whether mRFP was successfully secreted by S. alvi.

Figure 2: This is Plasmid 1, which contains the gblock that consists of mRFP, the secretion tag, and TolC on the pBTK510 backbone.

Figure 3: This is the assembled Plasmid 2, containing all three Hly secretion proteins, as well as mRFP and its secretion tag.

Genome Integration

In the Bee Microbiome Toolkit, it was established that CRISPR could be used to effectively disrupt a gene on the S. alvi chromosome [1]. This was done using a two-step method. First, Cas9 was added to S. alvi via pBTK601, a vector that has a low to medium copy number and lacks a partitioning system, so cells can be readily cured of this plasmid after the gene disruption procedure is complete [1]. Then, a suicide vector pBTK599, which does not replicate in S. alvi, contained the gRNA corresponding to the targeted chromosome location, along with the replacement cassette that contained around 1,000 bp homology flanking an antibiotic resistance gene. The S. alvi gene targeted for insertion of the cassette was an adhesion gene.

Figure 4: This figure from [1] shows the different parts used for CRISPR genome integration in S. alvi.

This method was found to be successful in integrating the replacement cassette into the target location on the chromosome. Using the Cas9 plasmid generated more of the desired S. alvi mutants than when no Cas9 plasmid was added. This method produced many successful double crossover events, in which only the replacement cassette is integrated into the chromosome. A single crossover event is when the vector backbone is also integrated. These results were confirmed by PCR. The success of this experiment inspired us to integrate a functioning gene into the S. alvi chromosome, but into an open location on the chromosome, and then to test the expression of that gene.

First, we chose a location on the S. alvi chromosome to integrate a new gene. We chose a location that was unlikely to disrupt the function of another gene. Much of the S. alvi genome is not annotated, so we chose an open spot close to two other genes since there was a smaller chance of an unannotated gene being found in a small space between two genes than a large space.

We decided to use GFP to integrate into this location on the genome. We designed a gblock that contains the gRNA targeting the PAM site on the chromosome where Cas9 will make a cut, as well as the homology regions that correspond to the chromosome where we will insert the GFP. The gblock contains a mutated PAM site so the Cas9 will not re-target the mutated genome. In between the homology regions of the gblock, there is a restriction cut site into which we will insert GFP to have a complete replacement cassette of gRNA and homology regions flanking GFP that will be inserted. This glblock will be added to the suicide vector pBTK599 via Gibson Assembly.

Figure 5: This is the assembled suicide vector pBTK599 containing the gRNA, and 1,000 bp homology flanking superfolder GFP. We decided to use a plasmid called pX2-cas9 that is easily cured from cells to introduce the Cas9 enzyme into S. alvi.

Figure 6: The pX2-Cas9 plasmid.

To test the genome integration, we plan to first transform pX2-Cas9 into donor strain E. coli WM6026, then conjugate it into S. alvi. We will then transform the suicide vector pBT599 into the donor strain, and perform a second round of conjugation to add this plasmid into our recombinant S. alvi. We will test the success of the genome integration by measuring GFP fluorescence emitted from the recombinant S. alvi.


Bioinformatics

Sequestration vs. Degradation

During the first stages of project Bee-tox, we discussed two different methods to control imidacloprid exposure in bees. The first idea was to implement an enzyme into the bee gut microbiome that would break down/transform imidacloprid into less toxic molecules that would be excreted from the bee. The second idea that we discussed was the usage of a sequestration protein that would bind to imidacloprid and render it unable to bind to nAChR target receptors, allowing it to be excreted out of the bee through the digestive tract.

Both ideas had their benefits and drawbacks. For the degradation pathway, we understood that an enzyme would be able to perform many degradation reactions per protein. In addition, several different enzymes had been previously identified that played a role in imidacloprid resistance in insects[1-10]. However, these enzymes have specific optimized conditions that are difficult to control while in the bee gut. In addition, the overall workflow of our project would be very demanding as modeling and possible modification of enzymes would need to occur.

Following degradation, we then shifted our discussion on the advantages and disadvantages of using a sequestration technique. This method has the benefit of simplicity in terms of protein modeling as sequestering proteins would be smaller. Also, there would be a possibility of designing a protein ab initio, a technique that we were excited to learn about. However, we raised the concern about the metabolic stress a large sequestration system would put on our engineered microbe Snodgrasella alvi. To address this concern, we estimated the total proportion of active proteins our sequestration system would require in each S. alvi cell (Table 1). With the calculation, we concluded that only about 2% of the total amount of proteins would be required, a percentage that we considered unsubstantial. Still, there still existed the disadvantage of creating such a large number of sequestering proteins and trying to contain these proteins to the bee gut.

Name Value Units
Number of S. alvi for Production 1.00E+08 cells
Imidacloprid Forager Exposure 2.49E+12 molecules
Protein Production Needed (Imidacloprid Exposure) 2.49E+04 proteins/S. alvi cell
Volume of S. alvi (Cylindrical Approx.) 0.50 μm3
Protein density of B.subtillis (Gut microbe with cell volume of 0.62 μm3) 2.10E+06 proteins/μm3
Number of proteins in S. alvi cell 1.06E+06 molecules
Percent of proteins needed for sequestering 2.36 %

Table 1. Protein Requirement Calculations

Between degradation and sequestration, it was decided that the degradation technique would be more feasible. The largest deciding factor was the amount of available research on current degradation pathways and breakdown of imidacloprid. The sequestering concept, although seeming to have more benefits, did not have the research backing that was needed to move it forward into the experimental phase, especially with our limited lab space access.

Metabolomics and Transcriptomics

With the degradation workflow decided, we began work on identifying a plausible enzyme candidate. Early into the research phase, it was clear that not all pathways were well characterized and it was difficult to pinpoint what exact intermediates were being produced. In light of this, we began to design a two-part experiment in an attempt to “pathway crack” the degradation of imidacloprid (Figure 7). The first component of the experiment would be a metabolomics approach where we would attempt to capture the intermediates of the imidacloprid degradation pathway using tandem mass spectroscopy (LC-MS/MS). The second component of our pathway cracking would take a look into the enzymes produced by cells after imidacloprid exposure using different transcriptomic methods. The first transcriptomic method would be RT-qPCR of biotransformed cells to glimpse into the regulation of genes known to be associated with imidacloprid resistance in other insects such as GSTs[11-15], Aldehyde oxidases [16-17], and Cytochrome p450s [2-8,16]. This would allow us to glimpse into the transcriptomic profile of exposed cells. If the RT-qPCR is successful and we see upregulation of specific genes, we would move on to the second transcriptomic method RNA-Seq. This method would allow us to capture the full transcriptomic profile of a cell undergoing exposure to imidacloprid and to determine all genes that are regulated during exposure. At first glance, it seems redundant to do a RT-qPCR then an RNA-Seq; however, due to the high costs of RNA-Seq, we agreed a smaller transcriptomic profile should be established.

Figure 7. Conceptual overview of “Pathway Cracking” Method.


With the plan outlined, we directed our focus towards studying soil microbes known to degrade imidacloprid as Snodgrassella alvi has not been previously documented as having imidacloprid degradation capabilities. Through literature research, we focused 5 possible microbes that have been known to degrade imidacloprid (Figure 8). From this list, we selected Pseudomonas putida KT2440 as we had access to the appropriate lab equipment and P. putida KT2440 had an annotated genome. However, we decided to use the KT2440 variant P. putida EM371 as we had greater accessibility to the strain. Because only phage elements were deleted from EM371, we made an assumption that EM371 still contained the same degradation enzymes as KT2440 [18]. With this, P. putida EM371 would be used in all biotransformation and metabolomic/transcriptomic experiments.

Figure 8. Imidacloprid Degradation pathway within the Soil [28].


Device

Bee Attraction

One important aspect of designing the bee feeding device was ensuring our feeder was attractive to bees. To begin this process, we performed a literature review on elements of bee attraction. We then used this literature review, as well as research into conventional bee feeders, to select certain variables to test for (color, feeding angle, etc.). Next, we drafted a basic feeder (“mini feeder”) for 3D printing that we could use to test the effect of different variables on attractiveness. An image of the 3D design is pictured on the right.

Initially, our plan was to conduct trials with these mini feeders to determine the ideal color, feeding angle, and other feeder parameters. Unfortunately, due to several setbacks in the approval process, both from iGEM and our university, there was not enough time to conduct these trials. Instead, we used existing literature we had previously researched to select for these parameters.





Biocontainment

Another important goal for the device team was to incorporate a physical method of biocontainment that would prevent non-target organisms from accessing the probiotic. We knew we wanted to incorporate image recognition software to accomplish this, but we had not decided on what type of physical barrier would be used. To begin the design process, we researched conventional bee feeders to understand basic designs. Next, we brainstormed initial designs that could incorporate biocontainment mechanisms and created CAD models, pictured below:

Design A



Design B



Design C


Then, we selected the feeder that would be easiest to 3D print, as well as most effective in keeping other organisms out of the feeding solution. Device A from above was selected. Once the design was selected, our team received feedback from our graduate student advisor Shaylynn Miller and modified the design to make it printable in resin. In order to print the device in our resin printer, the design needed to be resized and some of the design elements were modified. The finalized design, including the separate motor component piece and the removable acrylic cover, is shown below:

Prototyping and Future Experiments

For future prototype testing, we propose a modified version of our “mini feeder” testing protocol. The two primary design attributes to be tested are image recognition gating system functionality and bee attraction.

Attraction

In order for the feeder to effectively spread the probiotic amongst a wild bee population, it needs to be accessible enough to bees to consistently attract them and successfully administer the treatment. In order to test relative attractiveness, the feeder prototype will be set out in an area with a high wild bee population during peak foraging times (approximately midday) alongside an open control such as a dish. Both the control and prototype will contain equal volumes of a 1:1 sucrose-water solution acting as a nectar surrogate. Stationary cameras will be placed two feet from the feeders and will record from multiple angles for four hours. Team members will then analyze footage, recording the number of bee visits per hour for each feeder as well as qualitative observations regarding bee behavior.
Literature indicates that placement of the feeding site holes level with the “herb layer” (or at the same elevation as the vegetation where bees tend to forage) would be most effective for data collection and bee attraction [1]. Aromatic attractants such as lemongrass oil placed in equal volumes on the feeders may aid in increasing bee attraction during trials [2].

Gating System Functionality

We also require field testing to determine the accuracy rate for our image recognition system. We propose a similar setup to that which is described above, where the prototype is positioned level with the herb layer during peak foraging time and footage is recorded by multiple cameras. At the conclusion of each trial, video footage should be reviewed and the following information recorded:

  • Total number of bee visits
  • Number of bee visits in which the image recognition system successfully recognizes the bee and opens the gate to the feeder
  • Number of false gate openings (in the presence of non-bee organisms)

References

Gene Engineering

  1. Kwong, W. K., and Moran, N. A. (2013) Cultivation and characterization of the gut symbionts of honey bees and bumble bees: description of Snodgrassella alvi gen. nov., sp. nov., a member of the family Neisseriaceae of the Betaproteobacteria, and Gilliamella apicola gen. nov., sp. nov., a member of Orbaceae fam. nov., Orbales ord. nov., a sister taxon to the order “Enterobacteriales” of the Gammaproteobac- teria.
  2. Gentschev, I., Dietrich, G., & Goebel, W. (2002). The E. coli α-hemolysin secretion system and its use in vaccine development. Trends in microbiology, 10(1), 39-45.
  3. Leonard, S. P., Perutka, J., Powell, J. E., Geng, P., Richhart, D. D., Byrom, M., ... & Barrick, J. E. (2018). “Genetic engineering of bee gut microbiome bacteria with a toolkit for modular assembly of broad-host-range plasmids”. ACS synthetic biology, 7(5), 1279-1290.,
  4. Kwong, W. K., Engel, P., Koch, H., & Moran, N. A. (2014). Genomics and host specialization of honey bee and bumble bee gut symbionts. Proceedings of the National Academy of Sciences, 111(31), 11509–11514. https://doi.org/10.1073/pnas.1405838111
  5. Marcelo C. Bassalo, Andrew D. Garst, Andrea L. Halweg-Edwards, William C. Grau, Dylan W. Domaille, Vivek K. Mutalik, Adam P. Arkin, and Ryan T. Gill. ACS Synthetic Biology 2016 5 (7), 561-568. DOI: 10.1021/acssynbio.5b00187

Bioinformatics

  1. Whitefly-active metabolites of imidacloprid: biological efficacy and translocation in cotton plants Link: https://onlinelibrary.wiley.com/doi/pdf/10.1002/%28SICI%291096-9063%28199903%2955%3A3%3C265%3A%3AAID-PS891%3E3.0.CO%3B2-C?casa_token=TndI4TQbXx4AAAAA:S4P3UHunyJEPs8z7uoERNH7KznCirTwRexWfttX7WTIV4xphCv1F8ZMsMMJ7Grj6Uw1sRwSbUpxB
  2. Metabolism of imidacloprid in Apis mellifera Link: https://onlinelibrary.wiley.com/doi/full/10.1002/ps.772?casa_token=vOLYjbyBdO4AAAAA%3AD-hrS4MuVONj1FsFtcTrbOhlqNKEihPYVKpAh-NeoroJZobmgwsOCDyTjgLID_7tIx5jHqVCw6lVvNY
  3. Structural model and functional characterization of the Bemisia tabaci CYP6CM1vQ, a cytochrome P450 associated with high levels of imidacloprid resistance Link: https://www.sciencedirect.com/science/article/pii/S0965174809001258
  4. Metabolism of imidacloprid and DDT by P450 CYP6G1 expressed in cell cultures of Nicotiana tabacum suggests detoxification of these insecticides in Cyp6g1-overexpressing strains of Drosophila melanogaster, leading to resistance. Link: https://onlinelibrary.wiley.com/doi/10.1002/ps.1472
  5. Soluble and membrane-bound Drosophila melanogaster CYP6G1 expressed in Escherichia coli: Purification, activity, and binding properties toward multiple pesticides Link: https://www.sciencedirect.com/science/article/pii/S0965174813000271?casa_token=D8-7W3QYntgAAAAA:BfA9Kb7g_MEcqH-iVXhjW2wDdMzhUTpTqu7StaeNXilqtzNECdqhwOpBGqU2JhDcZD5YI8zR
  6. Soluble and membrane-bound Drosophila melanogaster CYP6G1 expressed in Escherichia coli: Purification, activity, and binding properties toward multiple pesticides Link: https://www.sciencedirect.com/science/article/pii/S0965174813000271?casa_token=D8-7W3QYntgAAAAA:BfA9Kb7g_MEcqH-iVXhjW2wDdMzhUTpTqu7StaeNXilqtzNECdqhwOpBGqU2JhDcZD5YI8zR
  7. Global transcriptome profiling and functional analysis reveal that tissue-specific constitutive overexpression of cytochrome P450s confers tolerance to imidacloprid in palm weevils in date palm fields Link: https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-019-5837-4
  8. Cytochrome P450 monooxygenase-mediated neonicotinoid resistance in the house fly Musca domestica L Link: https://www.sciencedirect.com/science/article/abs/pii/S0048357510000696
  9. Microbial degradation of imidacloprid and toxicological analysis of its biodegradation metabolites in silkworm (Bombyx mori) Link: https://www.sciencedirect.com/science/article/pii/S1385894713008152?casa_token=jCQ69nvXLz0AAAAA:FmpFmfjENdznQmFRszUgb3pwnVKnhRuI879f6upUj_Zrcdko5lCInfdRFG-RN7J0oTW_FO0U
  10. Bacterial biodegradation of neonicotinoid pesticides in soil and water systems Link: https://academic.oup.com/femsle/article/363/23/fnw252/2726779
  11. Insect glutathione transferases and insecticide resistance Link: https://onlinelibrary.wiley.com/doi/full/10.1111/j.1365-2583.2004.00529.x
  12. Identification of a diazinon-metabolizing glutathione S-transferase in the silkworm, Bombyx mori Link: https://www.nature.com/articles/srep30073
  13. Effects of imidacloprid on detoxifying enzyme glutathione S-transferase on Folsomia candida (Collembola) Link: https://link.springer.com/article/10.1007/s11356-016-6686-1
  14. Overexpression of a Glutathione S-transferase (Mdgst) and a Galactosyltransferase-Like Gene (Mdgt1) Is Responsible for Imidacloprid Resistance in House Flies Link: https://onlinelibrary-wiley-com.proxy1.cl.msu.edu/doi/full/10.1002/ps.5125
  15. Oxidative Stress and DNA Damage Induced by Imidacloprid in Zebrafish (Danio rerio) Link: https://pubs-acs-org.proxy1.cl.msu.edu/doi/10.1021/jf504895h
  16. Neonicotinoid metabolic activation and inactivation established with coupled nicotinic receptor-CYP3A4 and -aldehyde oxidase systems Link: https://www.sciencedirect.com/science/article/pii/S037842740500247X#bib4
  17. Regulation of Hydroxylation and Nitroreduction Pathways during Metabolism of the Neonicotinoid Insecticide Imidacloprid by Pseudomonas putida Link: https://pubs.acs.org/doi/abs/10.1021/acs.jafc.6b01376
  18. Peng J, Miao L, Chen X, Liu P. 2018. Comparative transcriptome analysis of Pseudomonas putida KT2440 revealed its response mechanisms to elevated levels of zinc stress. Front Microbiol. 9(JUL):1669.
  19. Sue M, Mikawa T, Ueda T, Nomoto Y, Miyamoto T. 2006. A novel function of housefly glutathione S-transferase 6B—Its effect on the retention and increase of insecticidal activity of the insecticide prothiofos. J Pestic Sci. 31(2):139–145.
  20. Nothias, L.F. et al Feature-based Molecular Networking in the GNPS Analysis Environment bioRxiv 812404 (2019).
  21. Wang, M. et al. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nat. Biotechnol. 34, 828–837 (2016).
  22. Katajamaa, M., Miettinen, J. & Oresic, M. MZmine: toolbox for processing and visualization of mass spectrometry based molecular profile data. Bioinformatics 22, 634–636 (2006).
  23. Pluskal, T., Castillo, S., Villar-Briones, A. & Oresic, M. MZmine 2: modular framework for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics 11, 395 (2010).
  24. FAO, Imidacloprid (206) First draft prepared by Usula Banasiak, Federal Biological Research Centre for Agriculture and Forestry (BBA), Kleinmachnow, Germany.

Device

  1. Plowright, R. C., and Candace Galen. “Landmarks or Obstacles: The Effects of Spatial Heterogeneity on Bumble Bee Foraging Behavior.” Oikos, vol. 44, no. 3, 1985, p. 459., doi:10.2307/3565787.
  2. Malerbo-Souza, D. T., et al. “Honey Bee Attractants and Pollination in Sweet Orange, Citrus Sinensis (L.) Osbeck, Var. Pera-Rio.” Journal of Venomous Animals and Toxins Including Tropical Diseases, vol. 10, no. 2, 2004, doi:10.1590/s1678-91992004000200004.