Vairimorpha Ceranae Alters Honey Bee Microbiota Composition and Sustain the Survival of Adult Honey Bees

Background: Vairimorpha (Nosema) ceranae is the most common eukaryotic gut pathogen in honey bees, Apis mellifera. Infection is typically chronic but may result in mortality. Additional factors may be involved in mortality, including the honey bee gut microbiota. Previous studies of V. ceranae and gut microbiota identied positive associations between core bacteria and V. ceranae infection. These possibly synergistic or mutualistic associations are often disregarded because some core bacteria act as probiotic symbionts. Methods: To clarify the effects caused by the positive associations, we added isomaltooligosaccharide (IMO), a prebiotic also found in honey, to alter the interactions between V. ceranae and gut bacteria. Mortality and sugar consumption of the caged bees were monitored. Infection intensities and gut bacteria were examed after 12 days post inoculation, the plateau phase of infection. The gut bacteria were evaluated using both qPCR and 16S rDNA sequencing. Results: We conrmed that V. ceranae infections alone signicantly enhance several core bacterial populations, including Bidobacterium spp., Snodgrassella alvi, and Gilliamella apicola in the honey bee hindgut microbiota. Moreover, the qPCR results suggested that V. ceranae infected bees had signicantly higher bacterial microbiota populations. In addition to the enhanced core bacteria, Commensalibacter and Bartonella were signicantly increased in the fecal microbiome. Infected bees fed IMO had signicantly higher V. ceranae spore counts but lower mortality; however, infected bees fed IMO did not have signicant changes in gut bacteria populations compared to those fed only sucrose, but feeding IMO further reduced the fecal microbiome alpha-diversity. Conclusions: The microbiota alterations caused by infection were similar to the microbiota differences found between summer bees and winter bees, the latter of which have longer lifespans, and feeding IMO increased this similarity. Our results indicated that the interactions between gut bacteria and V. ceranae not only enhanced both the pathogen and bacteria populations but also sustained the host survival. This mutualistic interaction potentially enhances disease transmission and avoids social immune responses of the honey bee hosts.


Introduction
Microsporidia are a group of intracellular eukaryotic microorganisms that are closely related to the Fungi.
These naturally occurring pathogens often cause disease in laboratory animals that are reared in a dense population [1]. Honey bees are eusocial insects that intimately interact within densely populated colonies, creating a suitable environment for pathogens, including microsporidia, to thrive. There are two common microsporidian species, Varirmorpha (Nosema) apis and Vairimorpha (Nosema) ceranae (genus recently rede ned; [2] identi ed in European honey bees, Apis mellifera. V. ceranae is the most common gut pathogen in adult honey bees [3]. honey bee ileum and rectum harbor abundant gut microbes [20,22]. Among the core bacteria, Lactobacillus Firm-4 and Firm-5, Bi dobacterium spp., Gilliamella apicola, and Snodgrassella alvi, dominate the hindgut environment. These bacteria affect various physiological functions of bees [20,21].
These bacteria are recognized as necessary to maintain homeostasis and as bene cial symbionts in honey bee gut.
Independent studies have shown positive associations between V. ceranae infection and hindgut bacteria populations. G. apicola was signi cantly and positively associated with V. ceranae infections [23]. Our previous studies identi ed positive associations of S. alvi [24] and Bi dobacteria spp. [25]. The associations of bi dobacteria and G. apicola were recently con rmed in another independent study [26], and microsporidian infection was positively associated with Snodgrassella spp. in bumble bees as well [27]. These studies suggested positive associations between V. ceranae and the gut core bacteria species, except for Lactobacillus spp. Such positive associations seem counterintuitive, especially for Bi dobacteria spp. that are generally believed to be probiotics that possibly inhibit pathogenic bacteria [28]. Although most studies of microbiota/pathogen in animals aim to nd agonistic associations, probiotic microbes provide bene ts to the host and may not contradict the needs of a chronic pathogen causing chronic infections. Chronic pathogens rely on the infected host to tolerate the infection and remain active during the time the pathogen reproduces and is disseminated. Therefore, we hypothesized that V. ceranae may have adapted to the gut microbiota and possibly modulated them to maintain host homeostasis, leading to tolerance of the infection and bene ting both the pathogen and the gut bacteria.
To evaluate our hypothesis of the positive associations between V. ceranae and honey bee gut bacteria, we used prebiotics to enhance the gut bacteria populations. Isomaltooligosaccharide (IMO) was selected as a prebiotic because isomaltose (the disaccharide form of IMO) is a common sugar found in honey [29]. We found that feeding IMO to V. ceranae infected bees resulted in signi cantly higher V. ceranae infection intensity but signi cantly lowered mortality. Feeding IMO alone resulted in subtle changes in bacterial populations, but signi cantly reduced microbiome alpha diversity in infected bees. The positive associations between V. ceranae infection and speci c core-bacteria that were previously reported were also con rmed in this study. In addition, we showed that Commensalibacter and Bartonella population ratios were increased in the fecal microbiome. These gut microbiota alterations were similar to those found in winter bees that have long lifespans [30]. We suggest that the positive associations between the pathogen and the gut microbiota serve to sustain the host survival, a bene t for both gut bacteria and V. ceranae. Moreover, these associations may prevent an apoptosis-like social immune response in which infected bees die prematurely to stop pathogen replication.

Materials And Methods
Honey bee rearing and inoculation Honeybee colonies, commercially available hybrids with phenotypes of Apis mellifera ligustica, were locally purchased from an apiary in Fuzhou. The rearing method and conditions were identical to those of our previous work (Zhang et al., 2019;. Newly emerged workers were collected within 24 h and transferred into cages, 100 bees per cage. Six cages were prepared and randomly selected for inoculation with one of two sugar-water types, IMO and sucrose. The IMO solution contained 10% isomaltooligosaccharide (IMO), 45% sucrose, and 45% ddH2O, and the sucrose solution contained 50% sucrose and 50% ddH2O. IMO, purchased from Yuanye Bio (Shanghai), was food-grade fermented from starches and mainly resulted in disaccharides with the degree of polymerization up to ve [31].
Arti cial pollen patties and the sugar water solutions were fed ad libitum. The bees were held in a 34˚C growth chamber for ve days until inoculation with V. ceranae.
Vairimorpha ceranae spores were freshly isolated from foragers collected in the eld. Spores were isolated using the same method reported in our previous work (Zhang et al., 2019; Huang and Solter 2013) and used immediately to inoculate bees. Five-day-old bees from the same cage were divided into treated and control groups after anesthesia on ice, and then the treated groups were inoculated individually by feeding a dosage of 10 5 spores in 2 µl sugar water as previously published (Zhang et al., 2019;. At 30 min post-inoculation, bees were transferred to new cages, 45-50 bees per cage, and given the same sugar-water type and pollen patties that were fed before inoculation. Twelve cages were generated, three cages each for bees inoculated with V. ceranae and fed either IMO or sucrose, and two control groups of three cages each for bees fed IMO or sucrose without inoculation with the pathogen. The cages were held in a 30˚C growth chamber, 24 h dark, 70% relative humidity.

Examination of caged bees
The cages were checked daily after inoculation. The dead bees were removed and recorded for survival analysis (Kaplan-Meier method using SPSS23, IBM). Sugar-water consumption was recorded every 3 d when the feeders were replaced. To evaluate the infection intensity and bacterial microbiota, ve bees were randomly collected from the cages at 12 d post-inoculation (dpi) and then every other day. Collected bees were dissected using the same method as previously described (Zhang et al., 2019). We separated the midgut, the hindgut and the feces for individual storage. The dissected hindgut was transferred into 50 µl TE buffer, and then the balloon-like rectum with the accumulated feces of the entire caged period was broken. The hindgut, ileum and broken rectum, were washed twice in TE buffer before storing.
Midgut tissues were stored in 100 µl PBS and macerated for microscopic examination. The infected bees with midgut spore count < 10 7 , the approximate plateau phase level of V. ceranae infection [32,33] were not included in the following qPCR and 16S rDNA sequencing. Primers for quantifying the core bacteria [34,35] are listed in the supplementary table. Honey bee actin primer set [36] and the universal qPCR primer set for all bacteria [37] were used to normalize the bacteria qPCR results. The signi cance of the differences in results was calculated using the One-way ANOVA method (SPSS 23, IBM).
We noted that qPCR may not generate robust results from fecal samples due to the complex composition and effects of food debris [25]. To investigate if the prebiotic and V. ceranae infection affect the fecal microbiota, fecal samples were submitted for 16S rDNA sequencing using Illumina MisEq. We selected infected bees that had a fully developed infection (total spore count > 10 7 ); DNA samples from six bees of the same group were pooled and ve pooled samples of each group were submitted for sequencing. The samples were PCR ampli ed using a universal primer set, 338f/806r, with adaptor sequence for the Miseq library and then sequenced on Miseq (Illumina) with the assistance from Majorbio (Shanghai, China). Brie y, the PCR was performed using Fastpfu DNA polymerase (TransGen, Beijing) in 30 cycles. PCR products were revealed in 2% gel-electrophoresis, and products in the expected size range were extracted using AxyPrepDNA kit (Axygen) and quanti ed using QuantiFluor™ (Promega). The PCR products were then processed using TruSeqTM DNA Sample Prep Kit (Illumina) before sequencing. The obtained data were processed using preset settings and analyzed on the Majorbio Cloud Platform (www.majorbio.com). Single reads were excluded from the results, and the operational taxonomic units (OTU) sharing more than 97% identity were recognized as the same species. Silva database (Release132, http://www.arbsilva.de) was used for the species annotations.

Results
The effects of V. ceranae infection and the prebiotic isomaltooligosaccharide (IMO) on the honey bee hindgut microbiota were investigated in this study. IMO alone (no infection) did not signi cantly affect mortality and sugar water consumption. However, mortality in infected bees fed IMO was not different from the controls while mortality in infected bees fed sucrose was higher. Feeding IMO signi cantly increased V. ceranae infection intensities but did not signi cantly alter bacteria population ratios in either qPCR or microbiome analyses; nonetheless, feeding IMO signi cantly reduced microbiome alpha diversity and unevenness in the fecal microbiome of infected bees. These subtle changes in IMO-fed bees appeared to reduce mortality caused by V. ceranae infection. These results support our hypothesis that the speci c gut bacteria were enhanced by the infection to maintain the host homeostasis.
Sugar consumption and mortality of the caged bees No signi cant difference in consumption of sugar water was found between IMO and sucrose fed groups but V. ceranae-inoculated groups consumed signi cantly more sugar-water, both IMO and sucrose (P = 0.034, Fig. 1), than the uninfected controls. There was no difference in mortality between control groups fed the two sugars, however, mortality was signi cantly lower (P < 0.001) in the infected group fed IMO compared to the infected group fed only sucrose (Fig. 2). Mortality was not different for V. ceranaeinoculated bees fed IMO and uninfected control bees fed either solution.

Differences in V. ceranae infection intensities
To evaluate infection intensities in the midgut, ve bees were analyzed by microscopy from each cage at 12 dpi and then every other day. The trial was terminated at 20 dpi because fewer than ve bees survived in the cages. The inoculation resulted in a 100% infection rate in the collected bees, and the uninoculated bees were free of spores. Midguts of the same group had similar infection intensities collected in 12-20 dpi. V. ceranae infection intensities were slightly, but signi cantly, higher in IMO groups (Fig. 3). Individual differences and variation were found among the samples, and the bees fed IMO tended to have higher spore counts in the dissected midguts. Because we were investigating fully developed infections, infected bees with less than 10 7 spores were not included in the following analyses.

Quanti cation PCR results
To investigate bacterial microbiota alterations of hindgut linings and feces, we used qPCR to quantify the core-bacteria species with higher populations. Lactobacillus spp., Bi dobacterium spp., Snodgrassella alvi, and Gilliamella apicola, were included in the qPCR. We attempted to normalize the results using the honey bee Beta-actin gene and the universal bacteria primer set; however, both primer sets were problematic as the references for the bacteria qPCR results. Fecal samples yielded low (> 35) or no Ct values using the actin gene primer set. The universal bacteria primer set yielded slightly but signi cantly lower (P < 0.001) values in V. ceranae infected bees, both hindgut and fecal samples, whereas IMO feeding did not alter the values. The results suggested V. ceranae infection signi cantly increases bacteria populations, but the prebiotic, IMO, did not. Because this study aimed to investigate the microbiota alterations caused by the infection, we decided to present the hindgut results (Fig. 4A) normalized by the host Beta-actin to show the bacterial population alterations and the fecal results (Fig. 4B) normalized by the universal bacteria primer set to show the bacteria ratio alterations. Hindgut sample results normalized with the universal bacteria primer set are provided in supplementary materials.
In addition, we noted that not all fecal samples generated detectable Ct values within 40 cycles, especially in the qPCR of S. alvi and G. apicola. The correlation analysis suggested that detectable S. alvi and G. apicola Ct values in fecal samples were signi cantly correlated with V. ceranae infection (P < 0.001). To further clarify the alterations in fecal samples, we submitted randomly selected fecal samples for 16S rDNA sequencing.
Vairimorpha ceranae infection alone enhanced all quanti ed core-bacteria populations and ratios in hindgut and feces ( Fig. 4A and B), except Lactobacillus. The hindgut samples used for DNA extraction and qPCR included no visible feces or food debris; therefore, Snodgrassellla and Gilliamella that attached to the ileum generated much lower values as expected (Fig. 4A). Overall, Bi dobacterium, Snodgrassella and Gilliamella were higher in infected bees; all alterations are signi cant in hindgut samples (Fig. 4A), and some in fecal samples (Fig. 4B).
Feeding IMO alone signi cantly increased the Snodgrassella population in hindgut samples and signi cantly decreased the Lactobacillus population ratio in fecal samples, which resulted in a marginally signi cant difference (P = 0.053) between the control and the infected groups fed IMO. For infected groups, IMO did not signi cantly alter bacteria populations in hindgut and fecal samples according to the qPCR results.

16S rDNA sequencing of fecal samples
The sequencing results of the fecal samples from the four groups showed signi cant differences. In the alpha diversity analyses (Fig. 5), the infected groups showed higher Shannon index values and lower Simpson index values, which suggested that V. ceranae infection signi cantly reduces the alpha diversity of the fecal microbiota. In addition, the infected group fed IMO showed the most distinctive changes in the analyses (Fig. 5). Feeding IMO alone did not result in any signi cant alteration in the microbiota (alpha diversity analyses), but the within-group differences were much higher than those in other groups (Fig. 5). Infected bees fed IMO showed the opposite results, the lowest within-group differences. Beta diversity analyses suggested that all the infected samples were grouped together in hierarchical clustering and Principal Co-ordinates Analysis (PCoA). In permutational multivariate analysis of variance (PERMANOVA) with Bray-Curtis method, V. ceranae infection alone signi cantly affected the microbiota composition (P = 0.002), and IMO feeding alone did not (P = 0.488).
Bacterial population analysis at the genus level indicated that Lactobacillus, Commensalibacter, Snodgrassella, and Bartonella populations were signi cantly different among the four groups; Bi dobacterium and Gilliamella were marginally signi cantly different (Fig. 6). Comparing the infected and control groups fed IMO, all the core bacteria species were signi cantly altered. Similar to the qPCR results, there were no signi cantly altered bacterial populations between the infected groups fed sucrose and IMO. Neither was there a signi cant alteration in bacterial population ratios between control groups fed sucrose and IMO, although the mean value was different (Fig. 6), possibly because of the high diversity and deviation of the control group fed IMO, shown in Fig. 5.

Discussion
Vairimorpha (Nosema) ceranae is the most common eukaryotic pathogen of A. mellifera. A. mellifera foragers can support tens of millions V. ceranae spores and still perform daily activities. Such tolerance for V. ceranae infection is surprising because the spores occupy an enormous part of the midgut cells. Positive associations between V. ceranae and several core bacteria symbionts, including Bi dobacterium spp., Snodgrassella alvi, and Gilliamella apicola that are generally considered to be probiotics, led to the hypothesis that these enhanced symbiotic bacteria may positively affect host homeostasis and longevity. Probiotics are expected to add health bene ts to the hosts, including protection from pathogens. But probiotics that have no pathogen protection effects may be able to form synergistic relationships with pathogens. To test this hypothesis, we added IMO as a prebiotic to alter the interactions. Although results suggested that infected bees have increased microbiota populations and reduced microbiome alpha diversity, and feeding IMO may intensify these alterations, only alpha diversity was signi cantly impacted by IMO. However, the intensi cations seemed to be enough to diminish the mortality effect caused by V. ceranae infections.
Modulated bacterial populations may have distinct functions that could enhance V. ceranae infection, host tolerance, or both, but the current knowledge about the enhanced bacteria and physiological changes in bees provides only associations, not causation. For example, the population numbers of Bi dobacterium and Snodgrassella spp. in the rectum were associated with longevity in comparison of worker and queen bees [38]. The increase of Commensalibacter and Bartonella population ratios and the reduced microbiome alpha diversity in V. ceranae-infected bees coincided with results found in the comparisons between summer and winter foragers [30]. Winter foragers that have longer lifespans had overall higher bacterial populations [30], which is similar to our ndings in V. ceranae-infected bees. Bi dobacterium spp. affects juvenile hormone titer, an effect that also has been reported in V. ceranae infection [39]. Interestingly, Lactobacillus spp., the only core bacteria that showed a negative association with V. ceranae infection, has no known speci c physiological function in the bees [21], but the metabolites they produced could reduce V. ceranae infection intensity [40].
Similar results to ours were reported in V. ceranae studies evaluating the effects of feeding pollen; infection intensity increased [17,41] and mortality was reduced [17], but the possible microbiota alterations were not analyzed in these studies. Because pollen is the main polysaccharide source in the honey bee diet, we assumed that pollen feeding and V. ceranae infections altered the bee microbiota in those studies. Similar results were obtained when indigenous gut bacteria were gavaged into bees with microsporidian infection [42], but neither did this study evaluate microbiota alterations. Feeding antibiotics and V. ceranae to bees showed the opposite results in that mortality increased [43]. Overall, these independent studies generated similar results that support the concept that gut microbiota can reduce the negative impacts caused by V. ceranae infection. However, only live bees were analyzed in the studies, including the work described here and there is no evidence to suggest that bees dying prematurely from the V. ceranae infection have unaltered microbiota. One study suggested that the modulations and the positive associations may not exist in early infection stages [44].
Vairimorpha ceranae infection may modulate the host's core bacteria by affecting polysaccharide digestion, in addition to the hypothesized immune modulations [45]. Polysaccharides are digested in the honey bee midgut in varying degrees. The isomaltooligosaccharides (IMO) we used are mostly small oligos, and honey bees have enzymes to digest such small oligos [46]. We found pronounced prebiotic effects only in V. ceranae infected bees, and IMO appeared only to decrease diversity and unevenness in the microbiome of infected bees. V. ceranae infection may have impaired or altered polysaccharide digestion in the midgut, but the expression of some related genes was upregulated [47] and proteomic levels were differentially regulated [48]. Interestingly, fumagillin, the antibiotic treatment for Nosema disease in bees, also affected modi cation of polysaccharide digestion enzymes and enhanced intensity of V. ceranae infection at low levels [49]. Nonetheless, digestion impairments caused by V. ceranae cannot fully explain all of the microbiota alterations. Both lactobacilli and bi dobacteria have the genes needed for metabolizing IMO [29], but lactobacilli populations were still lower in infected bees fed IMO in our fecal microbiome analysis. Although niche competition among the core bacteria might explain the decrease of Lactobacillus spp., additional modulations are probably involved. Immune response or other physiological functions modulated by V. ceranae infection may involve the core bacteria, and inhibiting immune responses could bene t some bacteria species, especially S. alvi and G. apicola that are not immune neutral [45].
Establishing mutualistic interactions with the gut microbiota appears to be a reasonable function for V. ceranae, which has a broad host-range among apid bees. Because the microbiota is a signi cant part of the gut environment to which V. ceranae needs to adapt, establishing mutualistic interactions with gut bacteria may facilitate V. ceranae infection in other host species that share similar gut microbiota. The gut microbiota in corbiculate bees are nearly identical [22,50], and V. ceranae infects most corbiculate bees, including bumble bees [7,8] and different species of honey bees. The mutualistic interactions between V. ceranae and the gut microbiota may have contributed to the broader host range and allowed host-switching in the eld [51]. Although not all bees developed the high infection intensities found in honey bees, the possibility that the pathogen can use wild bees as reservoirs has possibly increased the di culty of managing the disease in managed honey bees.
The results of our study suggest a synergistic and potentially mutualistic interaction between honey bee gut bacteria and V. ceranae: In infected hosts, the populations of both V. ceranae and the associated bacteria are higher and host lifespan is not affected. Preserving the host lifespan could be a strategy to avoid social immune reactions of the honey bee colony as a super-organism [52,53]. A social immune response includes the premature death of infected individuals to stop the spread of disease, similar to apoptosis by infected or damaged cells. Because social immune response in honey bees may be e cient enough to drastically reduce immune-related genes in honey bee genome [54], it could also be the selection force that leads to the synergistic associations between V. ceranae and gut bacteria.

Conclusion
We demonstrated that the interaction of symbiotic bacteria and pathogens is not unidirectional, only favoring the hosts. Symbiotic bacteria and pathogens can establish mutualistic associations if there are no obvious harmful effects for the hosts. In this case, the V. ceranae infected honey bees with enhanced symbiotic gut bacteria had mortality rates similar to the uninfected controls, and populations of speci c gut bacteria and V. ceranae were all increased. Honey bees are important pollinators and some reports have suggested pathogens, including V. ceranae, are having serious negative impacts on honey bee populations. An understanding of the success (high infection rates and intensities) of V. ceranae in honey bees is a step toward better control of the disease. The mutualistic interactions between gut microbiota and V. ceranae may lead to nding solutions for nosema disease, e.g. nding the optimal gut bacteria for gene-modi cations [55] as treatments. We believe this interaction is not a unique case. V. ceranae is not the rst microsporidium identi ed that can affect gut bacteria [56] and microsporidia are not the only pathogens that can destroy or affect the host microbiota [57]. Honey bees have a simple gut microbiota that provides an opportunity to identify these interesting mutualistic associations.

Declarations
♣ Ethics approval and consent to participate Not applicable.
♣ Consent for publication Not applicable.
♣ Availability of data and material Please contact author for data requests.

♣ Competing interests
The authors declare that they have no competing interests.   Accumulated sugar water consumption of the four honey bee treatment groups. Y-axis indicates the accumulated sugar-water consumption in grams. The four treatment groups were generated by two variables: IMO/sucrose and V. ceranae inoculation/control.   Survival analysis using the Kaplan-Meier method of the four treatment groups. Only the group inoculated with sucrose showed a signi cant increase in mortality.

Figure 3
Infection intensities of the inoculated groups fed sucrose and IMO sugar water. The Y-axis indicates the spore counts. Each bee midgut was individually processed and spores counted under a phase contrast microscope. Control bees were free of spores.

Figure 3
Infection intensities of the inoculated groups fed sucrose and IMO sugar water. The Y-axis indicates the spore counts. Each bee midgut was individually processed and spores counted under a phase contrast microscope. Control bees were free of spores. Similar to the hindgut results, but Snodgrassella and Gilliamella were not signi cantly altered in sucrose feeding groups. IMO also caused no signi cant alterations in bacterial population ratios in infected groups.

Figure 4
Relative qPCR results of the hindgut and the fecal samples. The lower ΔCq values indicate higher bacteria populations. (A) Hindgut samples qPCR results, the relative quanti cation results normalized by the honey bee Beta-actin gene. V. ceranae infection signi cantly increased bacterial populations. Feeding IMO signi cantly increased Snodgrassella (P<0.001) and marginally increased Bi dobacterium (P=0.058) populations in control bees; however, IMO caused no signi cant bacteria population differences in infected bees. (B) Fecal sample qPCR results normalized by the universal bacteria primer set results. Similar to the hindgut results, but Snodgrassella and Gilliamella were not signi cantly altered in sucrose feeding groups. IMO also caused no signi cant alterations in bacterial population ratios in infected groups.