Nosema ceranae (Microspora: Nosematidae): A Sweet Surprise? Investigating the Viability and Infectivity of N. ceranae Spores Maintained in Honey and on Beeswax

Abstract

Nosema disease is a prominent malady among adult honey bees [Apis mellifera L. (Hymenoptera: Apidae)], caused by the microsporidian parasites, Nosema apis Zander (Microspora: Nosematidae) and N. ceranae Fries et al. 1996. The biology of N. apis is well understood, as this parasite was first described over a century ago. As N. ceranae is an emerging parasite of the honey bee, we do not yet understand how long spores of this parasite survive in honey bee colonies, or all the potential modes of transmission among bees. We investigated the viability and infectivity of N. ceranae spores in honey and on beeswax over time after exposure to 33, 20, −12, and −20°C. Spores in honey maintained viability at freezing temperatures for up to 1 yr and remained viable considerably longer than those on beeswax. Based on this evidence, honey may act as an important reservoir for infective spores to initiate or perpetuate N. ceranae infections in honey bee colonies. This work provides information that may help enhance current management recommendations for apiculturalists.

Nosema ceranae, Apis mellifera, spore, viability, infectivity

Nosema disease is a prominent malady among adult honey bees caused by the microsporidian parasites, Nosema apis Zander and Nosema ceranaeFries et al. 1996. Nosema apis was first described as a parasite of Apis mellifera L. (Hymenoptera: Apidae) over a century ago (Zander 1909), and its epizootiology is well documented (Fries 1997). Unlike its congener, N. ceranae [recently proposed to be reclassified as Vairimorpha ceranae (Tokarev et al. 2020)] was originally described from the Asian honey bee (Apis cerana) in 1996 (Fries et al. 1996). Since its discovery, N. ceranae has been shown to be cross-infective, capable of generating infections in Meliponini and Vespidae (Porrini et al. 2017), several species of Bombus (Plischuk et al. 2009, Fürst et al. 2014), Apis (Chaimanee et al. 2010, Botías et al. 2012), and all castes of A. mellifera (Fries 1997, Alaux et al. 2011, Traver and Fell 2011a, Eiri et al. 2015). Nosema ceranae has a near global distribution (Higes et al. 2006, Holt and Grozinger 2016, Goblirsch 2018) and is now considered to be the dominant Nosema species infecting honey bees in many parts of the world (Chauzat et al. 2007, Paxton et al. 2007, Williams et al. 2008, Invernizzi et al. 2009, Tapaszti et al. 2009, Currie et al. 2010, Stevanovic et al. 2011, Traver and Fell 2011b, Martín-Hernández et al. 2012, Emsen et al. 2016).

The effects of N. ceranae on individual bees have been well documented. Infection with N. ceranae in worker bees decreases lifespan and nursing ability (Higes et al. 2007, Goblirsch et al. 2013), induces precocious foraging (Mayack and Naug 2009, Goblirsch et al. 2013, Li et al. 2019), affects olfactory learning and memory (Gage et al. 2018), and impairs flight (Dussaubat et al. 2013). In addition, infection with N. ceranae suppresses the immune system (Antúnez et al. 2009), degenerates midgut tissues (Dussaubat et al. 2012, Panek et al. 2018), and induces energetic stress (Mayack and Naug 2009, Li et al. 2018). In honey bee queens, infection with N. ceranae can lead to increased production of queen mandibular pheromone, which may serve as a signal for failing or sick queens (Alaux et al. 2011).

Effects of N. ceranae on colonies of A. mellifera appear to vary with study and geographic location. In the United States, colony collapse disorder (CCD)-affected colonies exhibited only slightly higher, nonsignificant, prevalence, and loads of N. ceranae than control colonies (vanEngelsdorp et al. 2009), whereas metagenomic analyses showed that coinfection with both Nosema spp. was one of few pathogenic measures that differentiated colonies with CCD from those that were healthy (Cox-Foster et al. 2007, vanEngelsdorp et al. 2009). Several studies conducted in Spain have demonstrated that infection with N. ceranae is associated with decreases in colony size, honey production, and brood rearing capacity, which may contribute to colony collapse (Higes et al. 2008b, 2009; Botías et al. 2013), whereas another has found that the parasite can be present without causing pathology (Fernández et al. 2012). Additionally, studies in western Europe have shown no clear relationship between N. ceranae presence and colony mortality (Genersch et al. 2010, Gisder et al. 2010).

Currently, Fumagilin-B® (DIN 02231180) is the only effective chemotherapeutic agent for treating infections caused by N. ceranae (Williams et al. 2008, 2011; Higes et al. 2011; van den Heever et al. 2015a). Unfortunately, this product is only effective against active infections, and not spores of the pathogen, which limits N. ceranae management. Fumagilin-based products have been registered for use for over 60 yr in North America, and were originally used to treat infections caused by N. apis (Katznelson and Jamieson 1952, Bailey 1953). At very low concentrations, Fumagilin-B® has been shown to exacerbate N. ceranae infections (Huang et al. 2013), and commercial formulations with the dicyclohexylamine salt may have potential toxic side effects in adult worker honey bees (van den Heever et al. 2015b). As N. ceranae is an emerging parasite of the honey bee, and infection by the parasite does not always generate any immediately obvious symptoms (Faucon 2005, Fries et al. 2006, Higes et al. 2008b, Stevanovic et al. 2013, Horchler et al. 2019), our understanding of how this parasite is transmitted among individuals and colonies is incomplete, thereby precluding effective disease management strategies.

Several studies have examined N. ceranae spore viability under laboratory conditions or within honey bees at high temperatures. Most have described the organism as thermotolerant but sensitive to low temperatures (Fenoy et al. 2009; Fries and Forsgren 2009; Martín-Hernández et al. 2009; Gisder et al. 2010, 2017; Higes et al. 2010b; Sánchez Collado et al. 2014), and capable of regeneration after cryogenic storage conditions (McGowan et al. 2016). This apparent lack of environmental cold tolerance is enigmatic given that N. ceranae survives and proliferates in temperate climates as well as warmer climates, putatively displacing N. apis in some regions (Chauzat et al. 2007, Paxton et al. 2007, Williams et al. 2008, Invernizzi et al. 2009, Tapaszti et al. 2009, Stevanovic et al. 2011, Traver and Fell 2011a, Martín-Hernández et al. 2012, Emsen et al. 2016), but not others (Gisder et al. 2010, 2017). Despite infective N. ceranae spores being detected in corbicular pollen (Higes et al. 2008a) and its DNA being identified in honey samples (Giersch et al. 2009), no studies have investigated the patterns of spore viability for this parasite on or within substrates associated with honey bee colonies. Here, we investigate temporal changes in the viability (ability to survive) and infectivity (ability to generate infection) of N. ceranae spores in products commonly associated with honey bee colonies, under differing environmental conditions, in an effort to understand how this parasite survives and may be transmitted. This work has the potential to influence current management recommendations to reduce the spread of the parasite and improve honey bee health.

Materials and Methods

Collection, Enumeration, and Identification of Nosema ceranae Spores

Honey bee colonies from apiaries managed by Agriculture and Agri-Food Canada’s (AAFC) Apiculture Program at Beaverlodge Research Farm (55°11′43.0″N; 119°17′57.3″W) were sampled for the presence of N. ceranae in May of 2014 and 2015, as well as from a cooperating beekeeper’s apiary near Girouxville, Alberta in May 2015. A sample of 60 worker bees from each colony was macerated in 60 ml 1× phosphate-buffered saline (PBS) and used to estimate number of spores/ml by visual examination according to Cantwell (1970) using a Helber Z30000 counting chamber (Cat # Z30000, Hawksley, Sussex, United Kingdom). Nosema spp. were verified using conventional polymerase chain reaction (PCR) techniques outlined in van den Heever et al. (2015b). Samples confirmed to be infected with only N. ceranae were used to create macerates for infecting newly eclosed Nosema spp.-free bees, in order to propagate the large numbers of N. ceranae spores needed for experiments.

Harvesting and Purifying Nosema ceranae Spores from Honey Bee Midguts

A spore purification protocol was adapted from McGowan (2012) to obtain a clean suspension and high number of viable N. ceranae spores without using chemicals. Workers heavily infected with N. ceranae were placed in sterile 50-ml centrifuge tubes and anesthetized on ice prior to dissection. Once anesthetized, digestive tracts were removed with forceps. Using a dissecting microscope, the midgut portions of the tracts were separated from the fore and hindgut using microscissors.

Approximately 50 midguts were macerated in 5 ml of sterile Type I water using a sterile 5-ml tissue grinder (Cat # 0955228, Fisher Scientific, Ottawa, ON, Canada). The macerate was then filtered through a 40-µm cell strainer (Cat # 352340, Fisher Scientific) and the rinsed with 5 ml of water (10× with 500 µl per rinse). This filtrate was then vacuum-filtered through a 10-µm separator (Cat # 60344, Pall Corporation, Ann Arbor, MI). A spore count was performed to estimate the number of spores/ml, as above, then the filtrate centrifuged at 800 × g for 10 min. After centrifugation, the supernatant was removed and a count performed to estimate spore loss. The pellet was resuspended in 2 ml of water. This procedure was repeated until the number of spores required for experiments was achieved—this was ~1.7 × 109 spores for honey and ~7.0 × 10⁸ spores for wax. The final pellet was then resuspended in the appropriate volume of water required for each experiment.

Experimental Design, Inoculation of Hive Matrices, and Recovery of Spores

Purified N. ceranae spores were added to two hive matrices (honey and beeswax) and then exposed to one of four temperature treatments (33, 20, −12, and −20°C), generating eight matrix × temperature combinations. These four temperatures were chosen to reflect the broodnest temperature of a colony (33°C), a room temperature benchmark (20°C), and typical fall and winter temperatures experienced in northern temperate climates (−12 and −20°C) to which hives or beekeeping equipment may be exposed. Temperature treatments at 33 and 20°C were maintained ±1.0°C in programmable incubators (models I36NLC8, I36NLC9, Percival Scientific, Perry, IA), whereas −12 and −20°C treatments were maintained ±1.5°C in programmable freezers. Temperature profiles over time were monitored with dataloggers (Hobo TidbiT v2 Temp Logger, Cape Cod, MA). The viability and in vivo infectivity of the treated N. ceranae spores were evaluated over several time intervals: 2, 7, 9, 14, 21, 28, 42, and 365 dpi for honey, and 7, 14, 21, 28, 35, and 365 dpi for beeswax. These time intervals were chosen to represent N. ceranae spore viability and infectivity within colonies during a production or pollination season (short term), and over an entire beekeeping season (long term), to determine whether contaminated equipment could act as a source of spores for infections. Additionally, these time intervals were chosen to be comparable with other N. ceranae viability studies (see Goblirsch 2018 for reference).

Several aliquots of purified N. ceranae spore suspensions were retained for pretreatment viability and infectivity assessment, whereas the remaining volumes were used to inoculate the two hive matrices. At each experimental time point, spores were recovered for each matrix at each temperature × time combination. These spores were used to assess viability and infectivity.

Honey

Two milliliters of spore suspension (containing ~1.7 × 109 N. ceranae spores) were added to 125 ml of liquefied honey determined to be free of Nosema spp. via conventional PCR (Giersch et al. 2009). This quantity of spores was used to ensure enough were available for viability and infectivity assays after recovery steps for all time points. Brassica napus L. honey was chosen as it is the predominant type of honey produced by beekeepers in northern Alberta, where the study was conducted. Spores and honey were mixed thoroughly to ensure even suspension: 1-ml aliquots were transferred into 180 sterile 1.5-ml microcentrifuge tubes (~5.0 × 10⁶ spores/tube, 45 tubes per treatment) and exposed to one of the temperature treatments mentioned in Experimental Design, Inoculation of Hive Matrices, and Recovery of Spores.

To recover spores, a protocol was adapted from McGowan (2012). First, 500 µl of water was added to each aliquot and thoroughly mixed to ensure complete homogenization. The aliquot was then added to a 10-µm separator and vacuum filtered. The separator was washed four times, each time with 1 ml of water. The filtrate (~5 ml) was collected using a 1,000-µl pipette, added to a 15-ml centrifuge tube, and centrifuged at 800 × g for 6 min. After centrifugation, the supernatant was removed and the pellet resuspended in 1 ml water. The number of spores/ml was then estimated, according to Cantwell (1970) using a Helber Z30000 counting chamber. Aliquots of recovered spores were then prepared for viability and infectivity assessments, as described in Assessing Spore Viability.

Beeswax

A sheet of wax foundation (Alberta Honey Producers Cooperative, Spruce Grove, AB) was determined to be Nosema spp. free by washing both sides of the sheet with water, and then following the DNA extraction and Nosema spp. identification protocol described in Collection, Enumeration, and Identification of Nosema ceranae Spores. The wax foundation was cut into 160 – 10 × 10 mm squares, then placed onto 18 × 18 mm glass coverslips, and heated gently on a hotplate to adhere wax to each coverslip.

Purified N. ceranae spore preparations (~7.0 × 10⁸ spores/1.7-ml water) were maintained in a 15-ml centrifuge tube, which was swirled continuously to ensure even suspensions while aliquoting 10 µl of inoculum (several drops/spot, containing ~4.0 × 10⁶ spores) onto each wax square previously cooled to room temperature (RT). Once transferred, squares with suspended spores were air-dried before receiving temperature treatments.

Spore recovery was performed by rinsing each wax square with 500 µl of water into a 50-ml centrifuge tube. The rinse water was then re-drawn into the pipet and used to rinse the wax square another five successive times. An additional 500 µl of water was then used to wash each wax square and was added to the 500 µl in the 50-ml centrifuge tube. The tube was then agitated to evenly suspend the spores and counted to estimate the number of spores/ml. Aliquots were then prepared for viability and infectivity assessments, as described in Assessing Spore Viability and Assessing Spore Infectivity.

Preparation of Recovered Spore Aliquots for Analysis

To assess infectivity, a volume containing ~2.2 × 10⁶ spores was set aside from each temperature × time combination (per matrix). Aliquots destined for viability and infectivity assessments were brought to the same final volume (1 ml) before centrifugation at 800 × g for 6 min. After centrifugation, supernatants were removed and pellets resuspended in an appropriate volume of water for staining (ratio: 2.0 × 10⁶ spores/20 µl of water) or for cage inoculation.

Assessing Spore Viability

The spore viability assessment protocol used in this study was modified from McGowan (2012). The fluorescent stains 4′,6-diamidino-2-phenylindole (DAPI) dilactate and propidium iodide (PI; Life Technologies Inc, Burlington, ON, Canada) were selected for this study. DAPI is a nucleic acid counterstain (Molecular Probes 2006a), whereas PI is a viability stain, as it is membrane impermeant (Molecular Probes 2006b). These stains do not overlap in excitation or emission spectra so this dye combination can reliably differentiate living and dead spores with different colors. Working stocks of both DAPI and PI (both 1 mg/ml in water) were prepared and stored in a dark refrigerator at 4°C. Stains were added simultaneously to each aliquot (2 µl of each dye per 2.0 × 10⁶ spores in 20 µl) destined for a viability assessment. Stained aliquots were incubated in the dark at RT for 20 min.

After incubation, the aliquots were centrifuged at 800 × g for 6 min. Supernatants were discarded, and the pellets resuspended in 100 µl of water (mixed well by pipetting and expelling). The aliquots were washed successively in 100 µl of water and centrifuged twice at 800 × g for 6 min before being resuspended in a final volume of water that made individual spores readily discernible (~2.0 × 106 spores/50 µl) under the microscope.

All samples were visualized using a Fluoview FV10i fluorescent microscope (Olympus, Tokyo, Japan). Prepared aliquots ready for visualization were mixed well with a 10-µl pipette to ensure homogenization prior to 6 µl being added to a counting chamber. The counting chamber was loaded into the FV10i, the DAPI and PI filter selected, and the total magnification set to 45×. The FV10i acquired the map image of the counting chamber, and all spores in all 16 grids of the counting chamber were counted (phase contrast) and recorded: DAPI only = live spore or DAPI + PI = dead spore. Average number of spores counted per replicate was 132 ± 8 spores.

Calculating N. ceranae Spore Viability

Spore viability for each aliquot was calculated by using the following formula:

$${\displaystyle \begin{array}{l}{\displaystyle \frac{\mathrm{N}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{d}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{D}\mathrm{A}\mathrm{P}\mathrm{I}\mathrm{o}\mathrm{n}\mathrm{l}\mathrm{y}(\mathrm{l}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{s})}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{v}\mathrm{i}\mathrm{e}\mathrm{w}\mathrm{e}\mathrm{d}\mathrm{u}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{r}\mathrm{p}\mathrm{h}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{s}\mathrm{t}}\times 100}\end{array}}$$

Assessing Spore Infectivity

Frames of eclosing worker bees were collected from nonexperimental colonies managed by the Apiculture Program at AAFC Beaverlodge. A minimum of six frames from six different colonies confirmed to be Nosema spp.-free via PCR were maintained at any one time. Bees were collected from these frames daily, so that all newly emerged bees (NEBs) used for the infectivity assays were <24 h old, and free of any Nosema spp. infection.

Infectivity aliquots were prepared from harvested spores (Preparation of Recovered Spore Aliquots for Analysis). After centrifugation, the supernatant was discarded, and the pellet resuspended in an appropriate volume of water such that when sugar syrup was added, each bee would receive ~1.0 × 10⁵ spores in 5 µl of 50% (w/w) sucrose syrup.

NEBs for this assay were collected and placed in individual feeding harnesses adapted from the feeding system developed by Rinderer (1976), and starved for 90 min prior to receiving a 5-µl droplet of inoculum at each time point. The inoculum was homogenized in a 1.5-ml microcentrifuge tube with a 10-µl pipette to ensure even suspension prior to inoculating, and in between doses. After receiving the inoculum, the bees were given 1 h to consume their droplets. After consumption, bees were confined to their harnesses for an additional 30 min to prevent trophallaxis upon caging. If bees did not consume their inoculum, they were not placed in a hoarding cage.

For each matrix, an average of 17 ± 0.5 bees were successfully inoculated per temperature treatment at each time point, except for the honey at 365 d post inoculation (dpi) when 10 were successfully inoculated per treatment. Control bees were given a dose of 5 µl of 50% (w/w) sugar syrup without N. ceranae spores. Once the bees were placed in hoarding cages, they were supplied, ad libidum, with 60% (w/w) sucrose syrup in a gravity feeder in addition to a pollen patty (25% [by weight] irradiated Canadian-collected B. napus pollen, mixed with sucrose, soy flour, yeast, and water). Hoarding cages were maintained in programmable incubators (see Experimental Design, Inoculation of Hive Matrices, and Recovery of Spores) for 14 d at 33 ± 1.0°C, with the diet being replenished every 72 h. After 14 dpi, all living bees were collected and frozen at −20°C until they could be examined for spores. We chose to examine living bees at 14 dpi because infections are considered to be fully developed at this time (Paxton et al. 2007, Forsgren and Fries 2010, Huang and Solter 2013, Huang et al. 2013), and because low mortality (<10%) across both matrices was observed. Additionally, much of this mortality was attributed to two instances in which sugar syrup feeders became clogged during initial assays.

Calculating N. ceranae Spore Infectivity

Individual frozen bees were placed into 1.5-ml microcentrifuge tubes. One milliliter of 70% ethanol was added to each tube, and each bee was macerated with a sterile micropestle. The macerate of each tube was vortexed and a spore count performed to determine number of spores/ml (individual bee) using a Helber Z30000 counting chamber (see Collection, Enumeration, and Identification of Nosema ceranae Spores) according to Cantwell (1970). Bees were classified as infected if at least two spores, representing 1.5 × 10⁵ spores/ml were present. To calculate spore infectivity for each treatment group at a given time point, the following equation was used:

$${\displaystyle \begin{array}{l}{\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{e}\mathrm{d},\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{v}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{b}\mathrm{e}\mathrm{e}\mathrm{s}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{v}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{b}\mathrm{e}\mathrm{e}\mathrm{s}}\times 100}\end{array}}$$

Statistical Analyses

Statistical analyses were performed in ‘R’studio v. 1.1.463 for Mac OS X (R Core Team 2014). We determined the impact of temperature on N. ceranae spore viability in honey at eight time points. One-way ANOVAs followed by Tukey’s HSD comparisons (TukeyHSD, v. 1.4–10, multcomp) were used at 7 and 21 dpi, whereas Kruskal–Wallis tests followed by Dunn’s multiple comparisons (dunn.test, v. 1.3.5., dunn.test) were employed for data at the remaining six time points (Zar 2010). The latter nonparametric techniques were also used to evaluate the effect of temperature on N. ceranae spore viability on beeswax at all five time points. All data were assessed for normality using the Shapiro–Wilk test, with homogeneity of variances being assessed using Bartlett’s test for normal data, and Fligner-Killeen’s test for non-normal data (Crawley 2013).

To estimate 50% viability (when spores are considered to be uninfective; Undeen et al. 1993) for N. ceranae spores under each temperature × substrate treatment combination, we first employed Prism’s curve-fitting function using Prism 6 (Graphpad, San Diego, CA) for Mac OS X in order to evaluate the appropriateness of linear versus nonlinear models for each treatment (Graphpad Software Inc. 2014). All treatments were best described using nonlinear models (curves). After curve-fitting, replicates tests were performed to assess the fit of curves to the various datasets (Draper and Smith 1998). One-phase decay curves had the lowest discrepancy values (F-values close to one) for all treatments and were selected to model all data sets.

The binary-response infectivity data were analyzed per matrix and by time point, using Fisher’s exact test for equality of proportions followed by multiple pairwise comparisons when necessary (fisher.multcomp, v. 0.9–73, RVAideMemoire), to determine whether there was an effect of treatment on infection status (Crawley 2013). To determine whether substrate had an effect on spore number, we compared overall spore levels from surviving infected bees from each of the two substrates using Wilcoxon rank-sum tests (wilcox.test, R Stats Package; Crawley 2013).

Results

Viability of N. ceranae Spores Maintained in Honey

Differences in spore viability among temperature treatments for N. ceranae stored in honey were not detected until 21 dpi (F_3,12 = 5.18, P = 0.01), when spores maintained at 33°C experienced greater reductions in viability than those at 20, −12, and −20°C (Fig. 1, Supp Table 1 [online only]). This trend was maintained throughout 28 dpi (χ ² = 12.18, df = 3, P = 0.01) and 42 dpi (χ ² = 8.33, df = 3, P = 0.04). Spores maintained at 33 and 20°C reached 50% viability at 31 and 149 dpi, respectively, whereas spores maintained at −12 and −20°C failed to reach 50% viability over the 365-d study period (Table 1).

Table 1.

Fifty percent viability estimates of Nosema ceranae spores in honey or on beeswax after exposure to 33, 20, −12, and −20°C for 365 d

Treatment	50% viability (d)	95% CI
		LCL	UCL
Honey
33°C	31a	22.9	49.8
20°C	149a	82.9	748.4
−12°C	>365b
−20°C	>365b
Wax
33°C	2a	1.9	3.1
20°C	<90a
−12°C	2a	0.8	6.2
−20°C	<1a

CI, confidence interval; LCL, lower confidence limit; UCL, upper confidence limit.

^aEstimated using a one-phase decay equation (Y = Y₀ − plateau × e^−kχ + plateau).

^bInestimable due to low decay.

Fig. 1.

Nosema ceranae spore viability for spores stored in honey at 33, 20, −12, and −20°C, modeled using one-phase decay curves. Each point represents mean spore viability (±SE; n = 2–4 honey replicate samples per time point; also refer to Supp Table 1 [online only] for statistical comparisons).

Overall, viability decreased at a greater rate for spores maintained at 33 and 20°C, and was significantly lower than the viability maintained by spores at −12 and −20°C at 365 dpi (χ ² = 12.71, df = 3, P = 0.01; Fig. 1, Supp Table 1 [online only]).

Viability of N. ceranae Spores Stored on Beeswax

Nosema ceranae spores experienced a rapid decrease in viability on beeswax at 33, −12, and −20°C, whereas viability was largely maintained for spores at 20°C for up to 21 dpi (Fig. 2, Supp Table 2 [online only]). Any effect of treatment on spore viability was lost after 35 dpi (χ ² = 0.69, df = 1, P = 0.71) when viability for all treatments ranged from 1% to 3.2% (Fig. 2, Supp Table 2 [online only]).

Fig. 2.

Nosema ceranae spore viability for spores stored on beeswax at 33, 20, −12, and −20°C, modeled using one-phase decay curves. Each point represents mean spore viability (±SE; n = 2–4 beeswax replicate samples per time point; also refer to Supp Table 2 [online only] for statistical comparisons).

Infectivity of N. ceranae Spores Maintained in Honey

Nosema ceranae spores maintained at 33 and −12°C in honey underwent a sharp reduction in infectivity at 14 dpi and experienced a further reduction from 21 dpi onwards (Fig. 3, Supp Table 3 [online only]). For spores maintained in honey at 33°C, infectivity decreased to zero at 21 dpi, whereas infectivity for spores kept at −12°C remained constant at 11–13% for the remainder of the experiment. From 14 dpi onward, spores stored at 20°C maintained moderate infectivity until 42 dpi. Nevertheless, at some point between 42 and 365 dpi, infectivity was reduced to zero for this treatment group. Spores stored at −20°C maintained high to moderate infectivity as well as high viability for the entirety of the experiment. Although N. ceranae spores stored in honey at −12°C maintained high viability over the 1-yr study period, infectivity observed at 365 dpi was low, but also decreased significantly after 14 dpi.

Fig. 3.

Nosema ceranae infectivity for spores stored in honey at 33, 20, −12, and −20°C for up to 1 yr. Each point reflects the percentage of surviving, infected bees at the end of a 14-d incubation period (n = 6–20 replicate bees per time point; see Supp Table 3 [online only] for statistical comparisons).

Infectivity of N. ceranae Spores Stored on Beeswax

Nosema ceranae spores stored on beeswax at 20°C maintained moderate infectivity for 28 dpi (Fig. 4, Supp Table 4 [online only]). Spores at 20°C were significantly more infective than those at all other temperatures, except at 35 dpi (P = 0.14, df = 3). In contrast, spores maintained at 33, −12, and −20°C experienced a rapid decline in infectivity after 7 d across all three treatments. Spores at −12 and −20°C did not generate any infections after 7 dpi, whereas spores stored at 33°C had infectivity levels that varied between 7% and 17% after 7 dpi, and spores stored at 20°C had infectivity levels that varied between 10% and 83% after 7 dpi.

Fig. 4.

Nosema ceranae infectivity for spores stored on beeswax at 33, 20, −12, and −20°C for up to 1 yr. Each point reflects the percentage of surviving, infected bees at the end of a 14-d incubation period (n = 11–20 replicate bees per time point; see Supp Table 4 [online only] for statistical comparisons).

Infectivity Experiment Mean Infection Intensities

Mean N. ceranae spore levels for bees used in infectivity experiments were 7.67 × 10⁶ ± 5 × 10⁵ spores/bee at 14 dpi. Bees infected with spores maintained in honey had significantly higher spore levels than those infected with spores maintained on beeswax (11.3 × 10⁶ ± 6 × 10⁵ and 4.1 × 10⁶ ± 4 × 10⁵, respectively, W = 14,545, P < 0.0001).

Discussion

This is the first study to examine the viability and in vivo infectivity of N. ceranae spores maintained in honey bee-associated matrices under semifield realistic environmental conditions for northern temperate climates. Long-term storage of N. ceranae spores in honey at low temperatures (−12 and −20°C) proved effective at maintaining high spore viability and low to moderate infectivity, whereas storage at room and broodnest temperature (20 and 33°C respectively) had a negative effect on both N. ceranae spore viability and infectivity. With the exception of cryopreservation studies (McGowan 2012, McGowan et al. 2016), our findings contrast with existing knowledge about N. ceranae spore biology. Much of the literature describes N. ceranae as a cold-intolerant and thermotolerant organism, capable of surviving for extended periods of time (>80 d in one instance) at high (33–60°C) temperatures in PBS, worker honey bees, and water (Fenoy et al. 2009, Fries and Forsgren 2009, Martín-Hernández et al. 2009, Higes et al. 2010b, Sánchez Collado et al. 2014). Nevertheless, the bulk of this previous work has examined N. ceranae spore viability in suspensions, under laboratory conditions—rather than field-realistic conditions, which undoubtedly contributes to differences in spore viability and infectivity.

In many host-microsporidia systems, cold tolerance is the rule rather than the exception (Weiser 1956, Thomson 1958, Revell 1960, Oshima 1964, Kramer 1970). For example, Perezia fumiferanae Thomson, a microsporidian of Choristoneura fumierferana Clemens, can survive 4–6 mo at −5°C (Weiser 1956). Encephalitozoon cuniculi, a parasite of mammalian hosts, also remains infective after exposure to low (4, −12, and −24°C), but not high (60 and 70°C) temperatures (Koudela et al. 1999). Given the high sugar and low moisture content of honey, the long-term viability of N. ceranae spores in honey at −12 and −20°C is not particularly surprising. The high sugar:water ratio likely inhibits the formation of ice crystals, acting as a cryoprotectant (similar to glycerol) preventing damage to the spores. This finding is in accordance with McGowan et al. (2016) who found that N. ceranae spores exhibited high viability and infectivity after storage at −70°C in 10% glycerol.

Spores maintained at higher temperatures (20 and 33°C) in honey exhibited lower viability toward the end of the experiment (28 dpi onwards) than spores stored at −12 and −20°C, with all matrices experiencing some reductions in infectivity during this period. These effects may be explained by several hypotheses regarding spore germination, an event governed by combinations of stimuli (including pH shift, ion concentration, enzyme activity, membrane hydration state, osmolarity, and redox potential) that appear to vary with species (Ishihara 1967; Weidner and Byrd 1982; Pleshinger and Weidner 1985; Undeen et al. 1987; Undeen and Frixione 1990; Undeen and Vander Meer 1990, 1999; De Graaf et al. 1993; Frixione et al. 1997). In several species of microsporidia-infecting aquatic hosts (namely N. algerae Vávra and Undeen), the catabolism of trehalose, the main carbohydrate storage of many microsporidian species (Wood et al. 1970, Undeen and Vander Meer 1999, Keeling and Fast 2002), into subunits of glucose by the enzyme trehalase is considered an important step in the germination process (Undeen et al. 1987, Undeen 1990, Undeen and Vander Meer 1999).

It is not documented whether N. ceranae possesses a trehalose store. If this species requires trehalose degradation for germination, it is possible that some of the warmer storage conditions used in these experiments may present some but not all the stimuli (suboptimal stimuli) required for germination, leading to ineffective germination and spore death. Possibly, the broodnest temperature and acidic pH of honey may trigger a gradual breakdown of trehalose by trehalase, resulting in lower pressure buildup within the posterior vacuole than is necessary to effectively fire the polar filament during the germination process, and thereby incomplete germination. This mechanism of slow trehalose catabolism has been suggested previously for spores of Edhazardia aedis Kudo (Undeen et al. 1993), and it is through this mechanism that spores of N. algerae lose viability after exposure to irradiation (Undeen et al. 1984, Undeen and Vander Meer 1990). Although less is known about this mechanism for terrestrial spores, it would explain the variation between percent infectivity and percent viability in the present study; spores that had not germinated but were inefficiently breaking down trehalose would appear as viable, but may be incapable of infecting a host. Nosema apis possesses trehalose, and it is interesting to note that the trehalase of N. apis is active between 35 and 45°C (Vandermeer and Gochnauer 1971), and most active at pH 6.0–7.5 (Vandermeer and Gochnauer 1971, De Graaf et al. 1993), which is the pH maintained by the honey bee ventriculus (Hoskins and Harrison 1934). Although the catabolism of trehalose does not appear to play a major role in the germination of N. apis (De Graaf et al. 1993, Undeen and Vander Meer 1999), it is conceivable that it may play a role in N. ceranae germination.

If N. ceranae is similar to N. apis, and a large degradation of trehalose does not precede germination, a mechanism involving the breakdown of another storage carbohydrate, or change in osmolarity, may be influencing the viability and infectivity of the spores. For spores of N. apis, De Graaf et al. (1993) suggest that germination is stimulated by monovalent anions. Although this may also be possible for N. ceranae spores, the possibility also exists that the reduction in viability and infectivity predominantly observed at high temperatures in honey is not due to ineffective germination, but rather damage to the membrane by some other mechanism. Malone et al. (2001) indicate that caged bees inoculated with 10² or 10³ spores maintained in honey had lower spore levels than bees inoculated with the same dosage of spores stored in sugar syrup. They suggest that the observed difference in infectivity may be attributed to the high pH or peroxide activity of honey, but do not propose a mechanism for how these properties affect spore viability. It may be possible that high pH and peroxide activity contribute to the thinning of the apical portion of the spore membrane, resulting in spore death, and thereby lower infectivity, or that they damage transmembrane pathways, preventing or reducing the required amount of water for effective germination.

The results of our study suggest that honey may act as an important route of transmission for N. ceranae. Nosema apis, a cold-tolerant parasite of the honey bee, is frequently referred to as a wet nosema, as dysentery is a common symptom and transmission is primarily via a fecal-oral route (Fries 1993). The phenology of a N. apis infection is characterized by low spore levels over the summer when high host population turnover occurs and bees can defecate outside the colony. An increase in spore numbers occurs in the fall and peaks in the winter when population turnover is low and the hosts are confined to, and defecate inside, the colony (Fries 1993). In an effort to remove the fecal deposits from within the colony over the winter, the bees ingest N. apis spores in fecal material and perpetuate the infection. Although N. ceranae is often referred to as a dry nosema (Higes et al. 2010a; dysentery is not always a common symptom, suggesting a route of transmission other than fecal-oral exists), and is traditionally thought of as a heat-tolerant and cold-intolerant parasite, the potential for infection of bees with spores from honey may be similar for both parasites.

Honey bee populations are greatest during the summer. For N. ceranae spores, displaying moderate viability and infectivity for over 42 d in honey at brood nest temperature is likely sufficient to ensure the parasite population is sustained due to the high number of available hosts during this time. During winter months, with fewer hosts available, low bee turnover, and greater N. ceranae spore viability and infectivity in honey at low temperatures, the parasite’s chance of reproduction and proliferation within the colony increases, without the need for a fecal-oral route of transmission. The phenology of infection resulting from this mechanism of transmission most likely results in increased levels of N. ceranae in the spring followed by reduced levels over summer and early fall, with a peak in the winter. This pattern is consistent with infections already observed in temperate climates (Williams et al. 2008, 2011; Gisder et al. 2010; Copley and Jabaji 2012; Ibrahim et al. 2012; McGowan 2012; Mulholland et al. 2012; Traver et al. 2012) and is similar to the phenology of N. apis (see Fries 1993). In addition, N. ceranae-infected honey bees have also been shown to self-medicate with honey (Gherman et al. 2014). From the perspective of the parasite, this is beneficial, as transmission is further enhanced. This mechanism would also support the aforementioned phenology. The next step in evaluating contaminated honey as a route of transmission for N. ceranae would be to determine whether the honey in N. ceranae-infected colonies contains sufficient spore loads to generate new infections.

The results from the experiment conducted to examine the effect of temperature on the viability and infectivity of N. ceranae spores dried onto beeswax indicate that desiccation has a substantial negative impact on both N. ceranae spore viability and infectivity. This is consistent with much of the literature, which suggests that spores dried onto a surface are much more susceptible to desiccation than those suspended in liquids or dead hosts (White 1919, Allen 1954, Weiser 1956, Malone et al. 2001). Although Gisder et al. (2010) did find that freshly isolated N. ceranae spores air dried onto microscope slides were able to successfully germinate, this germination rate decreased to <10% after 4 d of storage at 4°C.

Fenoy et al. (2009) examined the viability of N. ceranae spores dried onto glass slides using DAPI and Sytox Green as fluorescent dyes; however, they did not observe a dramatic decrease in spore viability after 1 wk of exposure at room temperature. An alternative explanation for these results could be that Fenoy et al. (2009) may have experienced significant autofluorescence of N. ceranae spores under the green filter (excitation 470–490 nm) that is used for viewing Sytox Green, contributing to increased apparent viability. This phenomenon was recognized by McGowan (2012) when viewing fresh, unstained spores under the same wavelengths.

For the current experiment, N. ceranae spores maintained on beeswax were wetted to prepare them for viability and infectivity assays. Kramer (1970) found that dried feces-bound spores of Octosporea muscadomesticae Flu re-wetted with water before being delivered to susceptible hosts for infection assays resulted in fewer infections than spores given to hosts dry. He suggested that the environmental moisture provided by the water may have triggered a futile germination, resulting in reduced infectivity (Kramer 1970). For the current (beeswax) experiment, it is possible that futile germination was triggered by preparing samples for assays, leading subsequently to poor viability and infectivity. Nevertheless, no extruded polar filaments were observed during the viability assays, suggesting that futile germination was not a confounding factor.

This study has illustrated that N. ceranae spore viability and infectivity are dependent on storage substrate and temperature, and it has provided evidence for cold tolerance, which makes N. ceranae’s prevalence and persistence in temperate climates much less mysterious. The high viability and infectivity of N. ceranae spores in honey at low temperatures suggest that honey may act as a main route of transmission for the parasite.

Based on the results of these experiments, economical, chemical-free methods for reducing N. ceranae spore loads within beekeeping operations that are nontoxic to honey bees are possible. When spore viability is <50%, spores can be considered noninfectious (Undeen et al. 1993). To ensure <50% viability for N. ceranae-contaminated honeycomb, our results suggest that beekeepers could maintain comb at −12°C or colder for 7 d to reduce the viability of any spores present on the surface of beeswax. In addition, comb exposed to 33°C for 50 d will reduce the viability of spores in honey to <50%.

Conclusions

We have verified the long-term viability of N. ceranae spores in honey, suggesting this food source as a potential route of transmission within and among colonies. This movement could occur through the consumption of honey, the robbing of weak hives or, given the lack of external symptoms associated with the parasite, movement of contaminated honeycomb by beekeepers. It also suggests that N. ceranae may be transferred more readily among hives than N. apis, the latter species relying heavily on a fecal-oral route of transmission, often through the cleaning of contaminated comb by bees within a colony. Given that N. apis is often associated with dysentery and sick/crawling bees (visible symptoms), beekeepers may be less likely to reuse comb from N. apis-contaminated colonies without first decontaminating it, thereby reducing the spread of the parasite among colonies.

Given the way in which honey bee colonies are managed in temperate climates (i.e., overwintered), and the excellent viability of N. ceranae spores in honey at low temperatures, the phenology of N. ceranae will probably continue to mimic that of N. apis. The lack of overt symptoms associated with N. ceranae, and the fact that contaminated honey could act an effective route of transmission for the parasite, also suggests that N. ceranae will continue to displace N. apis in areas where the parasite is present.

Acknowledgments

The authors wish to thank technicians Abdullah Ibrahim and Michael Peirson as well as all other employees and students in the Pernal lab, past and present, for their assistance with field work. Jamie Malbeuf assisted with Nosema ceranae spore propagation and individual bee inoculation. Grande Prairie Regional College’s National Bee Diagnostic Centre provided us with access to laboratory equipment, including their confocal microscope which was used for all viability assessments in this study. Thanks also go to beekeepers in the Peace Region of Alberta that allowed us to sample their colonies for N. ceranae. This research was financially supported by Agriculture and Agri-Food Canada (project ID J-000112), and a Queen Elizabeth II Graduate Scholarship and Project Apis m.-Costco Fellowship Scholarship to C.I.M. Funding sources had no role in study design; in the collection, analysis, or interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.