Team UAlberta’s Beelab focused on performing in vivo experimentation on the efficacy of protoporphyrin IX (PPIX) at inactivating Nosema ceranae spores and the function of our Antifungal Porphyrin-based Intervention System (APIS) prototypes. Our work for our Beelab included collecting and housing honeybees for experimentation under various treatment conditions.

The main motivation behind using live honeybees rather than other animal models, or cell lines, is that N. ceranae is an obligate parasite of honeybees. N. ceranae infections have only been found to infect the Asian honeybee, Apis ceranae, and only recently has it been documented in the Western honeybee, Apis mellifera [1]. Use of cell line studies were also considered but honeybee cell lines are notoriously difficult to acquire and maintain. Published alternatives, such as Lepidoptera-based cell lines have been infected by N. ceranae, but are cited as not suitable for propagating spores [2].

As propagating spores was necessary for our experiments and due to the of the physiology of N. ceranae and limitations in vitro testing, Team UAlberta opted to use live honeybees for our experiments as it would provide the best route for evaluating our designs. Find the full justification for live honeybee experimentation here.

In order to demonstrate the feasibility of our APIS design, we identified four critical outcomes that our technology must achieve:

  • Excess intracellular PPIX must not have any toxic effects on our E. coli chassis as the function of the APIS construct hinges on producing PPIX above natural concentrations. This aspect was focussed on in our Wetlab section.
  • Our E.coli chassis itself must not be harmful to the health of both Nosema-infected or uninfected honeybees treated with it.
  • The PPIX therapy must function as intended so that ingestion of PPIX must decrease Nosema spore loads in vivo or prevent the spread of infection.
  • PPIX must not have a negative effect on the health of both Nosema-infected and healthy honeybees health. We must ensure that bees, both health and unhealthy, do not experience negative health consequences by consuming PPIX.

Achieving these four outcomes individually would provide a set of proof-of-concepts that demonstrate the functionality of APIS, and that our novel strategy of using biosynthesized PPIX as an antifungal therapeutic is effective against N. ceranae infections in vivo.

Establishing a Honeybee Hive

In order to perform our in vivo experimentation, we needed a source for Western honeybees (Apis mellifera). Luckily, we were introduced to Jason McKinnon, a friend of one of our supervisors, who had been looking to start a beehive of his own. With his cooperation, we set up a new honeybee hive in south Edmonton, AB, from which we would collect honeybees!

Our team purchased three hive boxes, a few dozen frames, and the basic accessories for starting a hive. With the hive ready to go, Jason and our team travelled to Hove Apiaries, owned by Alvin and Judy Hove, where we got a five frame nuc of Italian honeybees (Apis mellifera ligustica) which is a subspecies of the Western honeybee [3]. The nuc included brood frames, honey frames, a mated queen honey bees, and thousands of worker bees. Judy and Alvin were also very generous as they donated nine drawn frames for us to use when our hive grows and expands. After transplanting the nuc into our brand new hive, we let the honeybees get settled for a couple of weeks while Jason helped our team tend to the hive in the meanwhile. Watch our hive building process here!

Figure 1: A picture of Team UAlberta visiting our honeybee hive on Jason McKinnon’s property.

Defining Specimens

Before any experimentation took place, our team defined our sample specimens as honeybee hives have different castes that exhibit different physiology. Thus, the subjects of our experiments were to be worker honeybees, as they are the most abundant caste [3] and would be the major carrier of N. ceranae in hives

However, workers are found in a distribution of ages in hives. So, our team had two options of which workers to collect:

  • Newly emerged bees which are adult bees that have just finished pupation and have just emerged from their wax-capped cells. These bees ages can be determined and should have a known pathogen load as they would have not been in contact with other bees. Or,
  • Non-newly emerged bees which are worker bees found in and around the hive. The age of these bees cannot be determined and their pathogen load is unknown as they have been potentially exposed to unhealthy bees. From now on, non-newly emerged bees will be referred simply to as adult bees.

If our team ended up collecting drone bees, or even the queen bee, our protocol was to return them to the hive.

Bee Collection

After letting our hive establish its new home, we began collection procedures for our various Beelab experiments. First, we consulted with Paul Greidanus, a commercial honey producer, about the effects of collecting honeybees on hive health. He told us that hives regularly lose hundreds of workers daily without adverse effects as healthy queens are consistently laying new brood (eggs). With this information, our team was reassured that we could collect enough bees for our experiments without harming our hive.

Our team’s first choice was to use newly emerged bees as more variables can be accounted for. To collect newly emerged bees, we placed frame cages around capped brood. Our frame cages were enclosed nets which would trap any workers that had just emerged. However, we quickly realized that bees did not emerge consistently in large numbers for our experiments which reduced the number of bees we could collect for a given day.

Due to this limitation, our team changed our approach and collected adult bees instead. Our decision was further justified by the idea that collecting a distribution of workers would be a closer approximation of the age variation within actual hive, thus providing a more realistic dataset for demonstration. To collect adult bees, we removed frames from the hive and ran our container across it. This caused multiple bees to gently fall into our container, which was faster than trying to individually collect bees. Paul Gredanius showed us this method of collecting bees and it drastically streamlined the process.

Bee Containment

To house our bees in our laboratory, we created custom bee cages according to previous methods [4]. Our design is depicted below. These cages had the appropriate openings for both feed inlets and removal of honeybees. Our mesh walls and flooring were designed to provide our honeybees with necessary ventilation and drainage.

Our standard feed for the honeybees was filter-sterilized 2M sucrose solution. These were poured into our feeding tubes which were 15 mL disposable centrifuge tubes with one to three holes drilled at the ends. When filled and capped, a vacuum is created which prevents the feed from pouring out while still allowing the solution to drip out slowly so that the honeybees are able to feed.

Figure 3: Depicted are several of our custom bee cages with our vacuum feeding tubes, filled with sugar-water, inserted.

Check-in Procedures

For all our experiments, our members performed daily check-ins on the bees. Each day, the number of dead bees was counted and then removed from the cage. If the feeding tubes had a volume of less than 5 mL, they were refilled with the appropriate feed. The number of dead bees per cage and the initial and final feeding tube volumes were recorded.

Honeybee Termination

Our termination protocols prioritized the welfare of the honeybees. Through our conversations with bee-researchers, we found out that honeybees are classified to have no experience of pain, only a reflex to pain stimuli. However, our team still decided to create termination procedures that both reduce any pain stimuli the bees might experience and are consistent with established methods. Termination, by instantaneous decapitation/physical destruction, was conducted swiftly in order to eliminate the possibility that the bees might experience a pain stimulus. Our termination protocols were only used when the honey bees had fulfilled their intended purpose for our experiments, meaning that we plan on terminating no bees unless we can obtain quality measurements from them.

Nosema ceranae Spore Counts

After dead honeybees were collected, or live bee sacrificed, N. ceranae, spore counts were conducted when appropriate. Individual or multiple bees were homogenized in 1 mL of PBS per bee and depending on the spore load, the liquid fraction was diluted further with PBS. 10 uL of each sample were then loaded onto a hemocytometer and Nosema spore were counted using a phase contrast microscope at 400x magnification.

With our bees specimens collected and with our four ultimate goals in mind, we designed several experiments to answer the following questions:

Experiment 1: Is PPIX toxic to E. coli?

Given that APIS is based on increased production of PPIX above natural levels in E. coli, it was necessary to test whether high concentrations of PPIX would have adverse effects on our chassis. Therefore, before commencing on live bees, we conducted an PPIX minimum inhibitory concentration assay for E. coli which is further detailed in Wetlab.

PPIX Source

For our various Beelab investigations, unless explicitly referred to as biosynthesized PPIX, the PPIX we used was a HPLC-grade reagent sourced from Enzo Life Sciences. Stock solutions were prepared by dissolving solid PPIX in a 1:1 Tris-base:10 mg/mL EtOH solution in the dark and further diluted into experimental solutions.

Experiment 2: Will E. coli make our bees sick?

To demonstrate the feasibility of APIS, it was also necessary to test whether feeding bees E. coli, which is not native to the bee microbiome, has an adverse effect on their health. To test this, we designed a bacterial toxicity assay (Figure 6). We introduced varying concentrations of our expression strain, E. coli BL21 (DE3), to the bees’ food source. These tests were conducted in both healthy adult bees and Nosema-infected bees. Groups were observed over the course of a week, during which we monitored their behaviour and observed the overall survival trend. Given the success of previous iGEM teams, UBC 2015’s Probeeotics and NYMU-Taipei’s Bee.coli, at introducing E. coli into bee intestines via introduction into their food source [5][6], we hypothesized that the introduction of E. coli into Nosema-infected bees would not negatively affect bee mortality.

Figure 4: A schematic of our design for our E. coli toxicity experiment. For each (A) healthy bees and (B) Nosema-infected bees, there was a sugar-water only control, and a treatment group for E. coli concentrations of 0.0025625g/mL, 0.01025g/mL, 0.0205g/mL, 0.041g/mL, 0.164g/mL.


Figure 6: The percent survival of healthy honeybees of that were fed a range of E. coli concentrations.

A note on the above plot, first looking at the healthy, uninfected bees, we found that only one group fed with E. coli statistically differed from the healthy bees fed only with sugar water. While this initially caused alarm, the statistically different group was the one fed the lowest concentration of bacteria. If this effect was due to toxic effects from the bacteria, we’d expect the groups with higher concentrations to have the same or lower survival rates, which we did not see. Thus, while we’re not sure of the exact cause, we believe that the significant difference is not a result of bacteria toxicity, but rather an anomaly.

Figure 7: The percent survival of Nosema-infected honeybees of that were fed a range of E. coli concentrations.

Interestingly, when we look at the Nosema-infected bees, we don’t see any significant differences between any of the bacteria-fed groups against the infected bees that were just fed sugar water, making us comfortable with the conclusion that the bacteria did not cause increased bee mortality. Even the highest tested E. coli concentration did not lead to a significant difference in bee mortality over time or in the overall survival trend. Bee mortality over the time course is therefore expected to a function of natural causes in the variably-aged bee population.

With no pronounced toxic effects induced by ingestion of E. coli, results demonstrate a valuable proof-of-concept that E. coli could be a feasible delivery vehicle of our therapeutic treatment.

Experiment 3: Does PPIX lead to a reduction in spore load?

Of course, we needed address our primary hypothesis: Will ingested PPIX, the same chemical that APIS will produce, decrease the N. ceranae spore load in honeybees in vivo? Both healthy and infected bees were fed with sugar-water supplemented with PPIX at a concentration of 80 uM, which is around the expected production levels enabled by our APIS construct.. In parallel, healthy and uninfected controls were fed regular sugar-water. PPIX was introduced to the feed 24 hours after Nosema infection, then this concentration was held constant for the duration of the treatment. PPIX tubes were wrapped in tin foil to minimize exposure to light and mitigate photosensitive degradation of PPIX. Altogether, there were six conditions tested:

  • Healthy bees fed regular sugar-water
  • Healthy bees fed PPIX treatment
  • Healthy bees fed Tris-base:Ethanol solvent system (solvent control)
  • Nosema-infected bees fed regular sugar-water
  • Nosema-infected bees fed PPIX treatment
  • Healthy bees fed Tris-base:Ethanol solvent system (solvent control)
Figure 8: A schematic for the design of our PPIX and solvent system toxicity experiment. For both (A) healthy bees and (B) Nosema-infected bees, a sugar-water only control was used, along with treatment groups for PPIX+Solvent System (80 uM) and Solvent System in sugar-water without PPIX.

We controlled for the solvent system in which our store-bought PPIX was prepared by introducing it into the sugar water solution, such as not to mistakenly attribute the possible action of the solvent system to PPIX’s therapeutic capability, as this might skew our results.

After ten days, honeybees were collected for spore counting according to the procedure described below. Spore load in healthy and infected bees treated with PPIX were compared to that of the sugar water-only control groups. It was hypothesized that the PPIX treatment would lead to a significant reduction in spore load, ultimately protecting the infected from the detrimental effects of Nosema infections.


Figure 9: Displayed are the spore counts per honeybee for each treatment group determined using a hemocytometer. PPIX-treated Nosema-infected bees show a statistically significant difference in final spore load as compared to its untreated counterpart.

As our results will show, the PPIX-treated group experienced a significant decrease in the spore load. Our results indicate that the spore load was unaffected by the solvent system alone, and thus the significant decrease in spore load is due to the action of PPIX alone. In fact, when comparing treated Nosema-infected groups treated with PPIX to the healthy, untreated controls, there was no significant difference in spore count. These results, within the scope of this preliminary experiment, demonstrate a definitive reduction in spore count as a function of our novel porphyrin treatment.

Experiment 4: Will PPIX negatively affect the health or longevity of the bees?

After finding that our therapy was capable of reducing spore load, we had to make sure that extended treatment was fully helping the bees in the long term as opposed to ultimately hurting them. To test whether increased concentrations of PPIX affects bee longevity or health, we designed a PPIX toxicity time course study for the duration of 18 days. We fed Nosema-infected and healthy honeybees with sugar-water supplemented with PPIX 80 uM. Bee death was recorded daily, such that a mortality trend could be observed. It was hypothesized that PPIX would not have a negative health effect on infected bees or healthy bees. This hypothesis was based on a previous study which used a chemically synthesized PPIX derivative and demonstrated that porphyrin-based treatments had no negative effects on bee longevity or behaviour [8].


This experiment has recently been completed, though the data analysis remains to be completed.

That said, given the therapeutic effects observed in Experiment 3, we anticipate that our findings will be consistent with that of literature, and thus, by extension, that our hypothesis will confirm that PPIX is a safe therapeutic which does not exhibit any adverse health risks to the target population. Validation of this outcome, again, is still forthcoming.


Upon answering our questions, through experiments 1 to 3, Team UAlberta obtained promising results that suggest APIS is a feasible method to treat Nosema ceranae infections in vivo. Our team anticipates that these experiments provide an experimental foundation which justifies continued experimentation with the finalized genetic construct. As shown in our Wetlab results here, we are close to demonstrating that our genetically engineered device successfully functions as a dedicated PPIX-producing machine. In subsequent work, we anticipate that the completion of Experiment 4 and the other methods described above will serve as a successful framework in which we can test the finalized construct’s therapeutic ability, and hopefully demonstrate that this technology is a promising solution to the threat of Nosema ceranae, whether via direct introduction into the bee diet, or as a tailored biosynthetic means of high-level production PPIX production. The latter feature may be made possible via incorporation of our design into a large-scale bioreactor to fill the gaping hole in apicultural antifungal therapies.


[1] M. L. Smith, "The Honey Bee Parasite Nosema ceranae: Transmissible via Food Exchange?," PLoS ONE , vol. 8, no. 8 , pp. 1-6, 2012.

[2] Gisder S, "A cell culture model for Nosema ceranae and Nosema apis allows new insights into the life cycle of these important honey bee-pathogenic microsporidia,”, 2018. [Online]. Available: [Accessed: 16-Oct- 2018].

[3] “About Honey Bees”,, [Online]. Available: [Accessed 16-Oct-2018]

[4] Williams et al. “Standard methods for maintaining adult Apis mellifera in cages under in vitro laboratory conditions,” Journal of Apicultural Research, vol. 52, no. 1, pp. 1-35, 2013

[5] NYMU-Taipei, "Entering of Bee. coli," 2013. [Online]. Available: [Accessed 28 5 2018].

[6] University of British Columbi iGEM. , "Probeeotic Prototype Testing Details," 2015. [Online]. Available: [Accessed 16 10 2018].