Team:Wageningen UR/Results/Detection Device
Detection Device
A major bottleneck is the surveillance of X. fastidiosa occurrence via reliable and easy-to-use detection methods. In this subproject we developed an automated detection tool of X. fastidiosa in insect vectors by Loop-mediated isothermal Amplification (LAMP). The reaction takes place within a plant mimic on which insects feed and thereby transfer the pathogen. In our efforts, we have shown that LAMP can be used to detect X. fastidiosa with a Limit Of Detection (LOD) in the order of 10^3 cells. Furthermore, we have shown that LAMP maintains function at a low concentration of sucrose (0.208%), which is used for insect attraction. Also, we have demonstrated that the LAMP reaction mixture can be encapsulated in paraffin, which is used to ensure separation of feeding medium and LAMP mastermix. Finally, we have succeeded in showing that the vector Cicadella viridis can feed on and transmit bacteria to a prototype plant mimic.
Introduction
One of the central difficulties of combatting Xylella fastidiosa is determining which plants are diseased. This is because plants take long to show symptoms: some species may take up to 2 years, and others never show symptoms at all. These asymptomatic plants can still serve as reservoirs for the spread of X. fastidiosa. This leaves regulators with highly incomplete information on the extend of X. fastidiosa presence. Although several screening tools for X. fastidiosa exist, all are laborious and thus not practical for early detection, which would require large-scale operations. Therefore, we propose detection of X. fastidiosa in vector insects through isothermal DNA amplification in a plant decoy to which insects spread the bacterium by feeding.
Early in the development of this subproject, we considered a complex genetic circuit coding for a biosensor, similar to the DSF sensing circuit to detect X. fastidiosa. However, we soon realized that we were attempting to use synthetic biology for the sake of synthetic biology in a circumstance much better fitted to a traditional approach. We believe that using appropriate rather than fashionable tools in this way helps us better serve our end-users.
-
LAMP Amplification arrow_downward
Loop mediated isothermal amplification (LAMP) is an alternative DNA amplification technique to PCR [1]. As it works isothermally, no expensive thermocyclers are needed for its application, making it easy to automate. It is especially popular in simple screening assays for use in the field. The reaction can be isothermal because the DNA polymerase used has strand-displacement activity. Typically, 6 different primers are used that can anneal and initiate elongation at various sites, resulting in a complex mixture of amplicons of various sizes. This means that LAMP cannot be used for the purposes of cloning, but it is highly suitable for diagnostic purposes. The increase of the concentration of DNA lowers the pH of a solution. This can be used to visualize the result of a LAMP reaction colorimetrically, by including a pH indicator in the mastermix.
In order to realize this detection tool, we first needed to determine whether LAMP has the sensitivity and robustness required for our application. This was done by testing the Limit Of Detection in a model bacterium, testing the reactions resistance to inhibition by a medium used to attract insects, and testing the method on X. fastidiosa infected leaves. Secondly, we needed to find out whether insect vectors could feed on parafilm covered PCR tubes, or what alternative container would be appropriate. Finally, we provided a proof of principle, by showing that vector insects could transmit Escherichia coli to PCR tubes covered by parafilm through feeding.
Suitability of LAMP
To answer the question whether or not LAMP is suitable for our application, we tested LAMP on X. fastidiosa infected leaves, we tested the Limit Of Detection in a model bacterium and the resistance of the reaction to inhibition by a medium used to attract insects. The primers that were used for the detection of X. fastidiosa were derived from literature [2].
Detection of X. fastidiosa in coffee leaves
First, we determined the functionality of the LAMP primers derived from literature to the detection of X. fastidiosa in coffee leaves. This experiment was performed at the NVWA, who provided us access to their X. fastidiosa infected coffeeplants, amongst which a naturally infected plant that was found in the Netherlands. To find out more about this visit, please go here.
Leaves were picked from these infected coffeeplants and extracts were made, which were subsequently used to perform a LAMP essay. Results could not be recorded colorimetrically, as the pH of the solution was heavily influenced by the crudely extracted plant samples. Therefore, results were visualized with gel electrophoresis.
X. fastidiosa DNA was found in the positive controls made of pure X. fastidiosa cultures. Furthermore, DNA was found in previously extracted coffee plant material, but not in the sampled leaves extracted in this experiment. As the pathogen is spread in an irregular fashion throughout the plant xylem, this may be explained by only leaves not containing X. fastiosa having been sampled. Only two leaves were sampled.
Construction of a plasmid containing primer-specific gene
To further determine properties of this LAMP amplification reaction, a plasmid was constructed using the gene for where specific X. fastidiosa primers were designed. To do this, part of the RimM gene, corresponding to biobrick BBa_K3286230, was amplified from X. fastidiosa genomic DNA. Subsequently, Golden Gate Assembly was used to construct a pACYC184 plasmid with this insert. This plasmid was transformed into E. coli DH5a.
Limit of Detection
To determine whether LAMP is sensitive enough to detect the levels of bacteria brought into the feeding container, we have determined its Limit Of Detection (LOD) on whole bacteria. The LOD of this reaction has earlier been found to be 250 copies of DNA [2]. By performing LAMP assays on a serial dilution of an overnight culture in triplo, of whom the cell density was determined by OD600 measurement, we determined that the amount of cells that can reliably be detected is in the order of 10^2 . Taking into account that the copy number of pACYC184 is 10-12, the amount of bacteria that can reliably be detected is of the order of 10^3. See Figure 2.
Combining LAMP with feeding solution
To attract insects, the option of using feeding media previously used to artificially feed spittlebugs (Philaenus spumarius) was explored [3]. Most feeding media consisted mainly of amino acids, which would have a significant effect on pH and thus interfere with the colorimetric output of our LAMP reaction through a pH indicator. Therefore, a sucrose medium was chosen. In order to use this medium, it was first necessary to determine whether this medium would interfere with the LAMP reaction itself. It was found that LAMP is inhibited by higher concentrations of sucrose (from 1.04%), but that lower levels of sucrose do not inhibit the reaction (up to .208%). See figure 3.
Encapsulating LAMP reagents in paraffin
The detection method as outlined above would require a feeding medium to attract insects, and another medium for the LAMP polymerase to function. To incorporate both these functions into a one-pot reaction, using paraffin to separate the LAMP reagents from the feeding medium was explored. Paraffin is used because it has a melting temperature that allows it to be solid at room temperature and melt at the LAMP reaction temperature (65 ºC). Also, it was previously found to work in encapsulating PCR mastermix and primers [4].
Subsequent to encapsulation, the enzyme was found to work, but it’s function was reduced, as yellowing was significantly less than expected for the amount of DNA that was added. In literature, PCR reagents have previously be encapsulated in paraffin, and it was noted that drying at 37 ºC is a mayor cause for loss of activity. It is likely that using lyophilisation could significantly reduce this loss of activity. [4]
-
Encapsulation parafin arrow_downward
Figure A1: LAMP solution encapsulated in paraffin. 
Figure A2: LAMP assay after encapsulation in paraffin. The positive control is slightly more yellow.
Suitability of Plant Mimic
Our detection method works by detecting the presence of X. fastidiosa in insects that feed on our plant mimic. It is therefore crucial for the functionality of our tool to determine that vector insects can feed on this mimic and transmit bacteria to it. It was determined here that Cicadella viridis, a common and abundant potential X. fastidiosa vector in Europe [5], feeds on PCR tubes covered by parafilm and can take up bacteria by this feeding.
The design of the feeding container was determined by presenting the spittlebug (Philaenus spumarius) and C. viridis with feeding containers of different shapes and sizes. Feeding on PCR tubes was observed directly for C. viridis
C. viridis was subsequently fed a bacterial solution through parafilm covered PCR tubes. To detect bacteria in these insects, they were crushed in LB, part of which was subsequently plated.
It was hypothesized that the highest number of alive bacteria would be present in the mouthparts of the insects, as they may not survive the conditions in the insect digestive system. Also, crushing and plating of the rest of the insect body may expose bacteria still alive in the mouthparts to digestive enzymes and acids, which may kill them or hamper their growth.
It was found in the case of C. viridis that colonies were much more distinct and numerous if the head of C. viridis was dissected by cutting through the thorax, and only the crushed head was plated. This technique was subsequently used in the transmission assay outlined in the section below, as described in the protocol.
-
Keeping Insects arrow_downward
A series of tests preceded the transmission assay to create an environment that is suitable for P. spumarius and C. viridis to survive in. The first test on P. spumarius resulted in insect death within 5 hours. This was first interpreted as indicating that P. spumarius was unable to feed on this container. Later, it was found however, that the lack of humidity in these containers may have contributed to the fast death of the insects.
It was also attempted to feed P. spumarius on a tube-like artificial feeder to replicate existing literature [3]. This experiment failed also, resulting in rapid insect death (within 5 hours). This was interpreted to be attributable to a lack of humidity in containers due to the use of moisture-absorbent cardboard and no source of humidity. We improved this in our next design.
After showing that both P. spumarius and C. viridis could survive in containers with moist tissue paper as source of humidity and plant material as a source of food for up to 7 days, C. viridis was exposed to PCR tubes filled to the brim with feeding medium covered with parafilm. Feeding behavior was subsequently observed directly.
Proof of Principle
Finally, it was shown that C. viridis could transmit E. colito PCR tubes covered by parafilm through feeding. This was done in a transmission assay.
For the transmission assay, two lanes of PCR tubes are placed in a plastic containers, of which one is spiked with E. coli transformed with a plasmid that makes it constitutively express GFP and has a Chloramphenicol resistance marker. The other one is unspiked, indicating that it does not contain any bacteria. Insects are introduced in these containers, and allowed to feed for 2 days. The content of the non-spiked PCR lane and the insects are subsequently plated. This was done in sixfold with 5 insects per container.
In one of the containers, an insect was found to contain the bacterium, as verified by growing on a chloramphenicol resistant plate and by detecting GFP expression. In this same container, plating verified that the non-spiked PCR lane had also been infected with the bacterium. Negative controls without insects did not show growth, and we therefore conclude that at least one of these insects has fed on the spiked container and infected himself, and subsequently fed on the unspiked container, infecting that container (Figure 8). Therefore, C. viridis can feed on PCR tubes, take up E. coli while feeding and transmit that E. coli to unspiked PCR tubes. In literature, amount of cells transmitted for related species, namely Homalodisca vitripennis and Graphocephala atropunctata were found to be around 300 cells and 200 cells respectively [6].
Conclusion
In this subproject, we first determined that LAMP could be used to detect X. fastidiosa with a Limit Of Detection in the order of 10^3 cells / reaction. Secondly, LAMP was shown to be functional for concentrations of a sucrose medium used to attract insects. Thirdly, we found that the vector insect C. viridis could feed on parafilm covered PCR tubes, showing that it could feed on similar structures in our plant mimic. Finally, we provided a proof of principle, by showing that vector insects could transmit E. coli to PCR tubes covered by parafilm through feeding.
-
References arrow_downward
- Notomi, T., Okayama, H., Masubuchi, H., Yonekawa, T., Watanabe, K., Amino, N., & Hase, T. (2000). Loop-mediated isothermal amplification of DNA. Nucleic Acids Res.
- Harper, S. J., Ward, L. I., & Clover, G. R. G. (2010). Development of LAMP and real-time PCR methods for the rapid detection of Xylella fastidiosa for quarantine and field.
- Cornara, D., Ripamonti, M., Morente, M., Garzo, E., Bosco, D., Moreno, A., & Fereres, A. (2019). Artificial diet delivery system for Philaenus spumarius, the European vector of Xylella fastidiosa. Journal of Applied Entomology.
- Qiu, X., Mauk, M. G., Chen, D., Liu, C., & Bau, H. H. (2010). A large volume, portable, real-time PCR reactor. Lab on a Chip, 10(22), 3170-3177.
- Jeger, M., Caffier, D., Candresse, T., Chatzivassiliou, E., Dehnen‐Schmutz, K., Gilioli, G., ... & Niere, B. (2018). Updated pest categorisation of Xylella fastidiosa. EFSA Journal, 16(7).
- Rashed, A., Killiny, N., Kwan, J., & Almeida, R. P. (2011). Background matching behaviour and pathogen acquisition: plant site preference does not predict the bacterial acquisition efficiency of vectors. Arthropod-Plant Interactions, 5(2), 97-106.