DESIGN

The Problem

One of the main challenges when treating Nosema ceranae infections is detecting the infection in the first place. This delays proper treatment and makes this parasitic fungus an issue that plagues all beekeepers. There are currently two common ways of detecting N. ceranae in honey bees:

(1) Visual Inspection: Beekeepers told us that they were instructed to look for visible infection symptoms, including declining populations, poor honey production, dysentery, and bees with swollen or greasy-looking abdomens [1]. This is the only on-site detection method and is a very inadequate diagnostic method, as these symptoms only appear once the infection has gone past the point where treatment would be effective.

(2) Sending a sample to a lab for microscopy inspection. This is the gold standard for N. ceranae diagnosis [1], but it also poses problems as beekeepers find this process costly and time-consuming, with a local beekeeper saying it can take upwards of a month to receive results. See our Integrated Human Practices page to learn more about this.

Since the current detection methods are not convenient, effective, nor field-ready, Team UAlberta decided to develop a simple and affordable detection system with beekeepers in mind.

The Beetector

We wanted our diagnostic system to be simple, affordable, and familiar enough for commercial and hobbyist beekeepers alike to use on their hives in the field. We learned through multi-level consultations with beekeepers that they were familiar with using paper-based detection methods, like pH strips for nectar and honey, and are comfortable with the idea of crushing up a handful of bees into a solution, as bees reproduce rapidly and this is required for current standards of detection as well. The Beetector was designed with all of this in mind: simplicity, familiarity, timeliness, and most importantly, ready for on-site use.

With the Beetector system, the beekeeper initially collects a sample of bees from a hive, crushes them, and then mixes them with the first Beetector solution, which is composed of amajLime-tagged phages. A second solution, composed of a non-ionic detergent called Tween 20, is then added. Our paper strip, made of Whatman P5 filter paper, is dipped into the mixture, and the second solution helps elute the components, with phages traveling further along the paper. A positive sample for N. ceranae would show colour at different intensities and regions on the paper compared to a negative sample, with the colour intensity directly corresponding to the severity of the infection. The signal is given off by a chromoprotein, amajLime, attached to an engineered M13 phage, which itself displays a ligand specific to N. ceranae spores. The Diagnostic Procedure and the Beetector prototype that we had a beekeeper test is available for viewing.

**Figure 2:** A schematic of our engineered phage displaying the chromoprotein as well as a ligand specific for *Nosema ceranae* spores.

Our approach to developing a diagnostic system allowed us to isolate the individual components of the Beetector and work on them separately. These parts are: the spore ligand, the reporter phage, and the paper strip.

The Spore Ligand

The first objective for our system was to find a ligand that could selectively bind to N. ceranae spores. Regular spore dyes would not work however, as they bind to the chitin on the spore walls, which are also present on the bee exoskeleton [2, 3]. More sophisticated binders such as antibodies have been found in the literature [4], but these are too unstable and impractical for field use. This prompted us to find our own ligand.

To find our own spore ligand, we used the NEB Ph.D.TM-12 Phage Display Peptide Library to screen against N. ceranae spores. Phage display libraries consist of M13 phages with different peptides on one of its proteins, allowing for 10⁹ unique ligands to be tested simultaneously.

**Figure 3:** A schematic of the phage display workflow.

To select peptide binders in the library, we first mixed the phage library with the N. ceranae spores that we have obtained from dissected, purified honey bee midguts, allowing for the phages with good ligands to bind onto the spore. The unbound phages are then washed out afterwards, leaving just the spores and the bound phages. Glycine-hydrochloride, which is our acid treatment, is then used to separate the phages from the spores and use these phages as the input for the next round of experiments. In doing so, we are able to enrich for better ligands with each round of phage display. For more information about our phage display protocol, click here.

**Figure 4:** A serial dilution schematic for our biopanning experiments. Each serial dilution consists of 10 μl of the previous tube and then adding 990 μl of lysogeny broth (LB), resulting in a 10-2 dilution.

**Figure 5:** A plating schematic for our biopanning experiments.

The Reporter Phage

Once the phage binds onto the spore, our next objective was to use a chromoprotein to give off a visual signal, which for our system was amajLime. AmajLime is a chromoprotein that is also a green fluorescent protein (GFP) homologue isolated from corals, displaying a strong yellow-green color in the visible spectra (BBa_K1033916). For more information on amajLime specifically, click here.

We chose to tag the M13 phage protein P8, as this has the most copies in the viral particle, allowing for a larger signal per phage. However, we could not directly express the amajLime onto the phage as it would compromise the phage’s structural integrity and assembly. Chung, Lee, and Yoo demonstrated this when they tried to express GFP directly onto the P8 protein on the M13 phage [5]. To circumvent this, they added a Sortase-A tag onto GFP and a second Sortase-A tag on the M13 phage. Inspired by their work, we decided to express our tagged components separately and then react them with the sortase enzyme. Sortase-A is a transpeptidase from Staphylococcus aureus that covalently binds two peptide residues, allowing for proteins to conjugate — in our case, amajLime with the phage displaying the N. ceranae ligand [6].

**Figure 6:** A schematic of the gene design for amajLime with the Sortase-A tag.

**Figure 7:** A schematic depicting the use of sortase to conjugate the phage and the amajLime chromoprotein by covalently binding their attached peptide residues.

The Paper Strip

The paper strip, made of Whatman P5 filter paper (with a pore size of 2.5 μm), aims to separate the bound phages from the unbound phages and other honey bee debris. A separation mechanism for the Beetector is needed as the unbound phages can still produce a visible signal due to being conjugated with the amajLime chromoprotein. Additionally, honey bee debris — such as tissue, pollen, and other microorganisms — can hinder the proper visualization of a N. ceranae infection by the Beetector.

**Figure 8:** A microscopic view of crushed honey bee midguts showing various debris.

The paper strip allows for the different components of the mixture to elute at different rates. Because they are much lighter than the spore, the unbound phages will elute to the top of the paper and appear as a dim-coloured band. As a comparison, the N. ceranae spore dimensions are 4.4 μm x 2.2 μm [7], whereas the M13 phage has a size of 880 nm x 6 nm [8]. If spores are present in the bee sample, however, the phages bound to the spores would elute less and create a darker band at the bottom, whereas the unbound phages would continue to elute through the paper. A very dark band being visible at the bottom with no dim band at the top would indicate a stronger N. ceranae infection, with the intensity of the colour directly correlating with the severity of the infection.

**Figure 9:** A schematic of different paper strips showing bands at different regions and intensities on the paper. A darker band at the bottom indicates a more severe *Nosema ceranae* infection.

References

[1] “Established Pests: Nosema,” Bee Aware. [Online]. Available: https://beeaware.org.au/archive-pest/nosema/#ad-image-0. [Accessed: 16-Oct-2019].

[2] H. R. Qin, J. L. Li, and J. Wu, “Detection and identification of Nosema ceranae by dual fluorescent staining with Calcofluor White M2R and Sytox Green,” Journal of Applied Entomology, vol. 49, no. 5, pp. 1392–1396, 2012.

[3] M. Elias-Neto, M. P. M. Soares, Z. L. P. Simões, K. Hartfelder, and M. M. Bitondi, “Developmental characterization, function and regulation of a Laccase2 encoding gene in the honey bee, Apis mellifera (Hymenoptera, Apinae),” Insect Biochemistry and Molecular Biology, vol. 40, no. 3, pp. 241–251, 2010.

[4] C. Del Aguila De La Puente, F. Izquierdo Arias, C. Fernandez Vadillo, S. Fenoy Rodriguez, M. y Higes Pascual, and R. Martin Hernandez, “Specific anti-exospora monoclonal antibody 7D2 Nosema ceranae for diagnosis by immunochemical techniques of bee's nosemosis”, Spain Patent 2548424A1, 15 April 2014.

[5] S. Y. Yoo, W.-J. Chung, and D.-Y. Lee, “Chemical modulation of M13 bacteriophage and its functional opportunities for nanomedicine,” International Journal of Nanomedicine, p. 5825, Dec. 2014.

[6] A. W. Jacobitz, M. D. Kattke, J. Wereszczynski, and R. T. Clubb, “Sortase Transpeptidases: Structural Biology and Catalytic Mechanism,” Structural and Mechanistic Enzymology Advances in Protein Chemistry and Structural Biology, pp. 223–264, Jun. 2017.

[7] M. L. Smith, “The Honey Bee Parasite Nosema ceranae: Transmissible via Food Exchange?,” PLoS ONE, vol. 7, no. 8, Aug. 2012.

[8] G. T. Hess, J. J. Cragnolini, M. W. Popp, M. A. Allen, S. K. Dougan, E. Spooner, H. L. Ploegh, A. M. Belcher, and C. P. Guimaraes, “M13 Bacteriophage Display Framework That Allows Sortase-Mediated Modification of Surface-Accessible Phage Proteins,” Bioconjugate Chemistry, vol. 23, no. 7, pp. 1478–1487, Jul. 2012.