RESULTS

Introduction

After learning about the issues that beekeepers have faced detecting Nosema ceranae we knew we needed to develop a detection system. But, we also knew that our product needed to be on-site, and cost effective so that beekeepers everywhere can access it. It also needed to be safe. To create the Beetector with all these factors in mind we needed to create our biological solution to deal with all these issues. We broke our project into three separate components that were then put together to create our final Beetector solution:

1. The Spore Ligand- determining the ligand that will bind to N. ceranae spore wall

2. Reporter Phage- creating the conjoined product solution containing amajlime and phage that will attach to the spores of N. ceranae

3. Paper Strip- making the visible detection that allows the beekeeper to see N. ceranae from the elution of the reporter phage solution

Each part of the system plays a vital role in determining the level of N. ceranae infection in Apis mellifera, the European Honey Bee. From here we identified the three major experimental problems that needed to be solved in order to create our overall design:

1. Finding a ligand that would bind specifically to the spore wall of N. ceranae

2. Attaching a reporter tag onto the phage that binds and gives a visual signal

3. Eluting the phages up a paper strip and leaving the phages bound to the N. ceranae spore at the bottom of the strip

Summary of results

After performing these experiments our main findings were:

1. We were able to identify potential ligands for our spore wall as shown by biopanning experiments, in which we performed multiple rounds of biopanning with a phage display library and Nosema spores. This effectively removed phages that bind due to increased infection abilities and phages that were carried along through experiments based on physical proximity.

2. We found that our amajLime expression was able to happen, even though it was not visible to the naked eye but was proven by measuring the OD and seeing that it expressed proteins able to absorb light in the amajLime wavelength. This was done by putting our plasmid into an E. coli BL21 strain that deals better with protein expression.

3. We successfully characterized our amajLime construct by calculating the extinction coefficient to inform us of the fluorescence of the chromoprotein. This allows us to see the absorption of light in our medium and to better analyze and quantify the amajLime chromoprotein.

4. We were able to determine how far the amajLime will elute up the paper strip; moreover, we were able to construct it to elute the same distance on different samples to maintain consistency. This allows beekeepers to maintain control levels and distinguish the differences between our control band and our infection of the spore line.

Our results demonstrate that the Beetector is an effective way to detect N. ceranae, but we still need to assemble the final product. Moreover, our results show where further work could be done to improve our project. Our results, the work that still needs to be completed, our troubleshooting, and much more are described in detail below.

The Spore Ligand

This part of our project aims to determine which phage would bind most effectively to the N. ceranae spore wall. So we decided to go with a chronic infectious phage that is well characterized, M13 (Wang et al. 2019)

Biopanning

We performed biopanning to select for potential candidates to bind to the N. ceranae spore. Biopanning works by selecting potential strong binders and eliminating the weak or non-binders to the N. ceranae spore (Wang et al. 2019). We did this using the commercial Ph.D.^TM-12 library againstN. ceranae spores. The Ph.D.^TM-12 library is composed of random 12-mer peptides fused to a minor coat protein (pIII) of the M13 phage. The library consists of approximately 109 unique sequences that we can use for enriching the most likely candidates of peptide ligands that bind to the spore. The protocol for biopanning can be found on our on our Notebook: Protocols page.

When we added the phage library to the N. ceranae spores, we left them to bind to each other. We then conducted five washes to remove unbound phages from the spores. The final step was acid elution, removing the phages from the spores and then selecting those phages for another round of biopanning (Lakzaei et al. 2018).

Below are the results from our first round of biopanning. When the phages are not washed, the plate is covered in plaques, which can be non-specific. As the washes are introduced, the removal of phages that are non-specific to the spore increases and the acid elution plate only contains the phages that are bound. These results led to further rounds of biopanning, all of these round and their results are listed below.

Table 1: This is the table describing the acid elution output from all our biopanning experiments resulting in more specific binding partners to our spores.

Biopanning Rounds	Acid Elution Plate Plaque Counts (pfu)
1	75
2	378
3	1707

**Figure 1:** Results for first round of biopanning comparing the plaque formation unit with the steps in our second round of biopanning

**Figure 2:** Results for second round of biopanning comparing the plaque formation unit with the steps in our biopanning 1 experiment

**Figure 3:** Biopanning results comparing the plaque formation unit in all 3 acid elution plates

Round One of Biopanning

**Figure 4:** Input plates from the first round of biopanning.

**Figure 5:** Wash 0 plates from the first round of biopanning.

**Figure 6:** Wash 1 plates from the first round of biopanning.

**Figure 7:** Wash 2 plates from the first round of biopanning.

**Figure 8:** Wash 3 plates from the first round of biopanning.

**Figure 9:** Wash 5 plates from the first round of biopanning.

**Figure 10:** Acid elution platefrom the first round of biopanning.

Rounds Two and Three of Biopanning

Performing multiple rounds of biopanning is essential to obtain binders to the N. ceranae spore and eliminate phages that were kept in the washes based on proximity. Therefore, conducting multiple biopanning rounds leads to more accurate ligands that match our spore wall rather than phages that infect faster and outcompete the ligands. We amplified the eluted phage from the first round and used it as the input for the second round and then followed the same procedure for the third round. Below are our results from the second and third round of the biopanning experiment.

Round 2

**Figure 11:** Wash plate from the second round of biopanning.

**Figure 12:** Acid elution plate from the second round of biopanning.

**Figure 13:** LB + Host plate from the second round of biopanning.

Round 3

**Figure 14:** Acid elution plate from the third round of biopanning.

These results indicate that the selection of phages from the acid elution plates are providing us with more specific ligands that match our spores, which is indicated in the table and graph that can be found below. These results led to us doing a selection experiment to determine phages that would bind faster and more efficiently.

After the third round of the biopanning experiment, we selected 10 plaques at random from the acid elution plate. We selected 10 phages in case multiples of the same phage were bound to the spore and not 10 different phages. These phages were then amplified following our protocol which can be found on our on our Notebook: Protocols page. The third round of biopanning confirmed that our phages had more specific ligands for binding.

Negative Controls

Negative Selection on Bee Midgut

In order to verify that our phages would not bind to uninfected bees, we conducted an experiment to verify that our phages from the third round of our biopanning experiment would only bind to N. ceranae and not other microorganisms and the pollen inside the bee midgut. We attained bees from the zoology department (which are not infected with N. ceranae) at the University of Alberta and plated the crushed bee midguts with the phages.

**Figure 15:** Acid elution plate from the third round of the biopanning experiment.

Titering

To complete our phage experiments we used titering to determine how many phages would be there before and after binding. As one plaque on the plate arises from one phage; therefore, with doing multiple dilutions we can determine how many phages are there at the start of our experiment and at the end of our experiment (Gallet et al. 2011).

Test With M13 SVEK Phage

The SVEK peptide phage was one of the negative control conducted to make sure that the phage of a known peptide sequence with a different receptor does not bind to the spore. The biopanning was performed with the SVEK peptide phage against N. ceranae spores. Table 2 presents the results of the M13 control ligand. It shows that there was no binding of the SVEK ligand to the spore wall. This result allowed us to test the spores against the PH.D.^TM-12 phage library for our biopanning experiment as we are aware from the SVEK peptide phage that the phage has no irregular binding to the spore.

**Figure 17:** SVEK peptide ligand M13 phage used for negative control against N. cerenae spore to confirm that the M13 phage does not bind irregularly to an M13 ligand that has no receptor on the spore.

**Figure 18:** Acid elution plates from the second round of the biopanning experiment

Table 2: Description of negative control experiment comparing the phage binding ability of SVEK M13 phage with our N. ceranae spores.

	SVEK M13 Phage (pfu)	Spores
	SVEK M13 Phage (pfu)	Spores
Added	1.19X10^9	3.0X10^6
Supernatant	1.0X10^9	-
Pellet	Confluent	-

Test with T4 phage

This was another one of our negative control experiments for the protocol of selection of phages that we are using. This was done to verify that if the phage was not bound to the spore the phages would be in the supernatant and the spores will pellet down without any phages bound to them. We chose T4 due to it being a lytic phage with no known abilities to bind to N. ceranae since it binds to E. coli (Storms et al. 2014). We also had a known concentration of the T4 phages; therefore, we would be able to calculate the concentration of the phages that would be present in the supernatant or on the spore wall. The results from this experiment show the T4 phages were all found in the supernatant; therefore, there would not be any spores that have phages bound. This indicates that T4 contains no binding affinity to N. ceranae.

**Figure 19:** Biopanning performed with T4 phage as a negative control to test biopanning protocol and get confirmation that the spore is unmodified from the biopanning protocol. The input of the phage and the eluted phage indicate very low loss of phage Titer, the spore count is also unchanged from the biopanning.

**Figure 20:** Acid elution plates from the second round of the biopanning experiment.

Table 3: Description of negative control experiment comparing the phage binding ability of T4 Phage with our N. ceranae spores.

	SVEK M13 Phage (pfu)	Spores
	T4 Phage (pfu)	Spores
Added	1.75X10^8	4.8X10^7
Supernatant	4.30X10^7	-
Pellet	-	3.15X10^7

Reporter Phage

The reporter phage system is made up of three parts: the tagged M13 Phage, the amajLime reporter tag, and the sortase enzyme which bind them together. To assemble this system we first expressed and purified both our M13 phage and our sortase enzyme, then we purified our sortase enzyme, and performed transformations of the amajLime construct into DH5 alpha cells for protein expression. All this was than combined to create our Reporter Phage.

M13 Tag

In order to work with the M13 phage we needed to start by purifying it after it had been amplified in our E. coli K12 ER2738 culture. We then separated the pellet from the supernatant to get the isolated dsDNA and ssDNA, respectively. Isolated M13 plasmid from infected bacteria due to it having the dsDNA M13, whereas, the supernatant contains the viral phage which is ssDNA. From this, we then had dsDNA for PCR and cloning experiments for further experimentation.

Linearizing M13

In order to run further experiments, we needed to linearize the M13 dsDNA phage, because of M13’s high mutative qualities. When we attempted to digest the phage, the digestion did not work. This could be due to multiple reasons, some including: the cut sites no longer being in range, digestion sites no longer being constant, or the cut sites that should have been there no longer existing. Sequencing is required to determine the genomic structure of the M13 phage.

**Figure 21:** From left to right: 1) 1kb DNA ladder 2) M13 + BspI +KpnI 3) M13 + BspI 4) M13 + KpnI 5) uncut M13

Sortase Enzyme

Once we obtained the sortase enzyme, we transformed it into DH10B and then purified the plasmid before inserting it into a T7 expression strain. We then transformed eSrtA + RFP control onto the plates below. We induced the protein with 0.4 M IPTG and incubated at 15ºC overnight for the protein to express.

The strains were then induced at different conditions: 37ºC for 2h, 37ºC for 4h, 37ºC overnight, and 15ºC overnight. The SDS-PAGE gels showed similar results. From these results we decided to induce 1 L cultures at 37ºC overnight. From this result, we then purified the sortase enzyme.

After receiving the result of the transformation of the sortase enzyme, we then did a column purification and collected fractions of the protein. Elutions 7, 8, and 9 confirm our sortase purification. Unfortunately, when we ran our ladder in the protein gel it did not show up. This could be due to various factors including inadequate digestion of the ladder, insufficient denaturation, or too little protein standard added.

Chromoprotein Tag

We then attempted to transform the amajLime 1.0 complex using NEB Gibson Assembly into pSB1C3. Following this, the new amajLime construct was transformed into DH5-Alpha on a LB-CAM plate and then into an E. coli T7 expression strain on a LB plate. The amajLime construct in DH5-Alpha plated normally with the appearance of green colonies; however, a blue lawn appeared when it was plated in an E. coli T7 expression strain.

**Figure 23:** AmajLime 1.0 expression on chloramphenicol plate.

**Figure 24:** AmajLime 1.0 expression on chloramphenicol plate.

Due to the appearance of E. coli T7 expression strain, we decided to plate 3 different strains on a LB-CAM plate to determine if the E. coli T7 expression strain also had chloramphenicol resistance, like our transformed DH5-Alpha. This test showed that the E. coli T7 expression strain had chloramphenicol resistance. Consequently, we had to transform our amajLime construct into pSB1K3 or pSB1A3. The results from transforming the amajLime 1.0 into those plasmids is shown below.

**Figure 25:** Test plates to identify marker expression for the plasmid on chloramphenicol plate.

**Figure 26:** Positive control of plasmid grown in T7 expression strain on LB plate.

**Figure 27:** Wash 1 from 10⁶ and 10⁴ phage titer experiment.

Sequencing

We then decided to sequence one of the plasmids from DH5-Alpha that was transformed into the cells through Gibson assembly and with verification through Sanger sequencing, only the amajLime complex was present, indicating that the Gibson reaction did not insert into our construct but rather reconstituted the distribution plasmid. From there, we decided to redesign our amajLime construct, thus creating amajLime 2.0.

Transforming Amajlime 2.0

**Figure 28:** AmajLime 2.0 transformed into pUC19 *E. coli* T7 expression strain on LB-Amp plate. White colonies indicate transformation was not successful.

From the amajlime 2.0 complex we attempted to transform it into a pUC19 E. coli T7 expression strain. This was plated on LB-Amp plate and white colonies grew, indicating that our transformation was not successful as the amajLime complex should have made the colonies appear green, not white. For future studies, we would attempt to transform it into E. coli K12 strain; however, from this we did inoculate 10 random colonies for plasmid prep.

Transforming AmajLime 2.0 and 3.0

In order to express our amajLime construct, we needed to transform amajLime 2.0 and 3.0 into chemically competent DH5-Alpha cells. We attempted electroporation; however, the machine was arcing because there was too much salt from the ligation and the machine would send out a pulse without the cuvette being in place properly. So we switched to chemical competency and we realized that our negative control showed the digested plasmid was self ligating, as shown below, even though it was a sticky end digestion. Future work will be done to troubleshoot this to add more insert to the vector in ratios to determine the threshold for proper ligation.

**Figure 29:** AmajLime 1.0 grown on LB plate.

**Figure 30:** Negative control (left) grown on LB and Amp plate and positive control (right) grown on LB and Amp plate (right).

**Figure 31:** AmajLime 1.0 transformation grown on LB and Amp plates (left) and amajLime 1.0 pellet grown on LB and Amp plate (right).

**Figure 32:** AmajLime 2.1 transformation grown on LB and Amp plates (left) and amajLime 2.1 pellet grown on LB and Amp plate (right).

**Figure 33:** AmajLime 3.0 transformation grown on LB and Amp plates (left) and amajLime 3.0 pellet grown on LB and Amp plate (right).

**Figure 34:** AmajLime 2.0 transformation grown on LB and Amp plates (bottom) and amajLime 2.0 pellet grown on LB and Amp plates (top).

**Figure 35:** AmajLime 2.2 transformation grown on LB and Amp plates (left) and amajLime 2.2 pellet grown on LB and Amp plate (right).

AmajLime Purification

For the amajLime purification, we started purification of the standard BBa_K3072010 and BBa_K3072015 which are biobricks that encode for amajLime under the control of constitutive promoters, and strong RBS from the distribution kit, however, this did not contain a purification tag. Please see our Parts pages and Contribution page to read more about these results. In our purification we used BPER to help extract our purified protein. We did this to determine dilution levels for modelling and what would be visible on paper strips. Our next steps, besides running this on our paper, was to transform amajLime 2.0 into a K12 strain to confirm amajLime.

**Figure 37:** Fluorescence spectrum of amajLime.

Extracting Midguts

In order to choose the paper strip we needed to collect spore samples from the bees midgut. This was done by following our protocol, using bees that were known to be infected. To limit the amount of pollen and other bacteria that would be present in the midgut, we were able to attain lab grown bees that were only fed sugar water; moreover, this sample was not filtrated. Under a microscope we can visually detect N. ceranae spores based on literature photos on what they would look like.

**Figure 38:** Bee midguts extracted from lab grown bees.

**Figure 39:** Crushed midguts of lab grown bees in solution.

**Figure 40:** *N. ceranae* spores under microscope.

We also extracted midguts from bees which were from apiaries in the Edmonton area to detect if we could see spores even with background of pollen and other bacteria, following the protocol as mentioned above. When viewed under a microscope it was more challenging to visualize N. ceranae spores due to the higher concentration of spores and microorganisms present. Also, this sample was not filtrated to try to remove the pollen or microorganisms. As shown, it is undetermined if these bees had N. ceranae infections.

Spore Purification of Lab Bees

After we obtained our spores from the midgut of lab grown bees we needed to purify those spores, our protocol for this can be found on our Notebook: Protocols page. Our spore purification sample was stored at 4ºC in 600 uL. In this instance, we could directly see the N. ceranae spores using an optical microscope without any pollen or an abundance of microorganisms. These bees were not filtrated, but since there was no pollen this was not a concern. This sample was easier to purify due to not having pollen in the midgut.

**Figure 41:** Microscopic view of *N. ceranae* spores in the midgut of lab grown bees, therefore without the presence of pollen and other microorganisms.

Spore Purification of Bees From Beekeepers

When working with bees we collected from beekeepers there was a lot more background contaminants than with the lab bees. Due to the presence of pollen, microorganisms and potentiallyN. ceranae spores, purifying this solution took more time and we were unable to determine if we removed all of the background contaminants including the spore, or just the background itself due to the inability to determine if this bee was originally infected with higher levels of N. ceranae.

**Figure 42:** Microscopic view of purified solution of extracted midguts of bees from apiaries.

**Figure 43:** Microscopic view of purified solution of extracted midguts of bees from apiaries.

Spore Dye

We obtained a spore dye called Chromotrope 2R from to determine if it binds to our N. ceranae spores. If it did bind we would be able to determine the spore count in our solutions more accurately and be able to do trials with our paper strip. When following the protocol adding spore dye to bees, there was need to do multiple dilutions as the dye bled from the pellet during centrifugation into the supernatant.

The results below show what happened when we added the spore dye to a bee with a N. ceranae infection. In this case, adding the spore dye created more visual distractions under a microscope than a bee without spore dye. This dye could be binding to N. ceranae; however, the shapes that the dye is producing do not match our microscope photos of N. ceranae. This means that the spore dye could be either in very high quantity or binding unspecifically.

**Figure 45:** Addition of spore dye, Chromotrope 2R, to a bee solution with *N. ceranae* infection.

**Figure 46:** Bee midgut infected with *N. ceranae* without spore dye.

Paper Strip

To start our testing with paper strips the first thing we had to do was determine the proper paper and shape to provide the best elution and maintain a separation of spores from the result of the dye. Below are different shapes of P5 filter paper with the above spore dye that was placed in a beaker filled with water. The results (shown below), which indicate that the spore dye bound to the bee did not create a solid line at the bottom and top as we needed regardless of shape. This result could be due to the chemical nature of Chromotrope 2R being an acidic polar compound interacting with other polar compounds, and due to the inconsistencies with the spore dye and paper we had to directly use the purified amajLime dye to determine if it would elute with a band at the top.

**Figure 47:** Whatman P5 Filter paper placed in solution of spore dye, Chromotrope 2R, and water. The absence of a solid line on the paper indicates the spore dye is unable to elute on the paper.

AmajLime and Paper Strips

The last part of our paper strip experiments was testing to see if it would bind to our amajLime and elute up the paper. We did this by using filter paper P5 and putting around 500 uL -1 mL of amajLime with BPER solution into a 50 mL falcon tube. The P5 strip was cut to be around 2.2 cm; however, regardless of the width of the paper strip, the amajLime always eluted 1 cm up the paper strip. This result led to being able to use smaller falcon tubes when necessary.

Since we are eluting the amajLime and BPER pure construct up our filter paper there will be an appropriate rf value for the amajLime unbound phages. We also tested if at different times the amajLime and BPER solution would elute higher, as seen on the 2c m paper strip. This showed that regardless of larger time intervals, unbound phages did not elute higher. We then compared the rates of elution between the different sizes of paper strip.

rf(phage) = amajlime/ solvent front

This lead to our calculated result.

rf 1cm = 1/3.5cm = 0.29

rf 2cm = 1/2.75cm = 0.36

rf 2cm-30min = 1/5.6cm = 0.17

**Figure 48:** Whatman P5 filter paper eluted with 2 mL of purified amajLime in B-PERII.

These results lead to the decision not to have a longer time interval for our paper strip, indicating that the amajlime is not as soluble as the BPER; therefore not needing as much time to elute the paper strip. This also tells us that regardless of time BPER does not influence a higher rf value for amajlime.

REFERENCES:

[1] Wang X-Y, Yang T, Wang S-Y, Du K-D, Chen M-L, Wang J-H. 2019. M13 phage as network frame for the quantification of Pb2 based on the Pb2 -induced in-situ growth of gold nanoparticles. Analytica Chimica Acta 1073:72–78.

[2] Lakzaei M, Rasaee MJ, Fazaeli AA, Aminian M. 2018. A comparison of three strategies for biopanning of phage‐scFv library against diphtheria toxin. Journal of Cellular Physiology 234:9486–9494.

[3] Gallet R, Kannoly S, Wang I-N. 2011. Effects of bacteriophage traits on plaque formation. BMC Microbiology 11:181.

[4] Storms ZJ, Brown T, Cooper DG, Sauvageau D, Leask RL. 2014. Impact of the cell life-cycle on bacteriophage T4 infection. FEMS Microbiology Letters 353:63–68.

Team:UAlberta/Results