Seasonal variation of viral infections between the eastern honey bee (Apis cerana) and the western honey bee (Apis mellifera)

Abstract It is a widespread practice in China to keep colonies of both the western honey bee, Apis mellifera, and the eastern honey bee, Apis cerana, in close proximity. However, this practice increases opportunities for spillover of parasites and pathogens between the two host bee species, impacting spatial and temporal patterns in the occurrence and prevalence of the viruses that adversely affect bee health. We conducted a 1‐year large‐scale survey to assess the current status of viral infection in both A. mellifera and A. cerana in China. Our study focused on multiple aspects of viral infections in honey bees, including infection rate, viral load, seasonal variation, regional variation, and phylogenetic relationships of the viruses within the same species found in this study and other parts of the world. The survey showed that the black queen cell virus (BQCV), deformed wing virus (DWV), Israeli acute paralysis virus (IAPV), and sacbrood virus (SBV) were common in both A. mellifera and A. cerana, and infection dynamics of BQCV, DWV, and SBV between bee species or seasons were significantly different. DWV was the most common virus in A. mellifera, and its infection rate and load in A. mellifera were higher than those in A. cerana, which reflects the high susceptibility of A. mellifera to Varroa destructor infestation. The infection rate and viral load of SBV were higher in A. cerana than in A. mellifera, indicating that SBV poses a greater threat to A. cerana than to A. mellifera. Our results also suggested that there was no geographical variation in viral dynamics in A. mellifera and A. cerana. Phylogenetic analyses of BQCV, DWV, IAPV, and SBV suggested the cross‐regional and cross‐species spread of these viruses. This study provides important insights into the complex relationships between viruses and their hosts in different seasons and regions, which will be important for developing effective disease management strategies to improve bee health.


| INTRODUC TI ON
Honey bees (Apis spp.) are the most important insect pollinators of food crops worldwide, playing an irreplaceable role in food security, ecological sustainability, and biodiversity (Gallai et al., 2009;Klein et al., 2007;Potts et al., 2016). Populations of Apis mellifera, the most widely distributed and managed honey bee, have suffered great losses in many countries in recent decades (Brodschneider et al., 2016(Brodschneider et al., , 2018Potts et al., 2010;VanEngelsdorp & Meixner, 2010).
Currently, more than 30 viruses have been identified in honey bee populations (Beaurepaire et al., 2020). The most common and widely spread viruses include the acute bee paralysis virus (ABPV), ABPV and IAPV have been reported to result in rapidly progressing paralysis and death of honey bees after experimental inoculation (Bailey et al., 1963;Maori et al., 2007). KBV can cause great damage to a colony when the levels of the parasitic mite Varroa destructor are high . BQCV is also a member of the family Dicistroviridae and primarily harms queen larvae and pupae, causing the rapid death of queen brood (Bailey & Woods, 1977). CBPV, a currently unclassified virus, causes chronic paralysis in adult bees, leading to the death of adult workers and spreading to other colony members (Bailey & Woods, 1974). DWV and SBV are members of the family Iflaviridae. The typical disease symptom caused by DWV is a wing deformity. The association between DWV and the parasitic mite V. destructor has caused significant damage to colonies (Allen & Ball, 1996;Di Prisco et al., 2016;Martin et al., 2010;Miranda & Genersch, 2010) and has recently been a primary cause of bee colony mortality worldwide (Schroeder & Martin, 2014). SBV mainly infects brood and results in larval death (Nguyen & Le, 2013). It is lethal to Apis cerana but less detrimental to A. mellifera (Gong et al., 2016).
A. mellifera and A. cerana are bee species used in the beekeeping industry in China at a national level and are often kept close to each other. While large-scale epidemiological investigations of honey bee viruses in China have been conducted in previous studies (Ai et al., 2012;Diao et al., 2019;Li et al., 2012;Yang et al., 2013), a broader comparative analysis of viral infections in both host species simultaneously would yield critical insights into the relationship between viruses and their hosts. Furthermore, most viruses infecting honey bees are RNA viruses that have extremely high mutation rates, which could have important epidemiological consequences.
We therefore conducted a 1-year large-scale survey to assess the current status of viral infections in A. mellifera and A. cerana in China.
This study focused on multiple aspects of viral infections in honey bees, including infection rate, viral load, seasonal and regional variation, and phylogenetic relationships with the same viruses in these bee species found in other parts of the world. We anticipate that this study will provide important insights into the evolutionary pattern and risk associated with the epidemic of viral disease in honey bees.

| Sample collection for virus detection
Adult workers were sampled from 244 A. mellifera and 238 A. cerana colonies distributed over 45 apiaries in six provinces (Gansu, Hubei, Zhejiang, Jiangxi, Guangdong, and Yunnan) in China (Table 1; Figure 1).
In each province, A. mellifera and A. cerana colonies were sampled in the same or nearby cities. Sample collections were conducted in autumn (October) and winter (December) of 2017 and spring (April) and summer (July) of 2018. All colonies used in this study were assessed for Source of samples (provinces)

| RNA isolation and cDNA synthesis
Thirty workers (Pirk et al., 2015) from each colony were crushed to a fine powder in liquid nitrogen and used for RNA isolation using the RNApure Total RNA Kit (Aidlab Biotechnologies Co. Ltd.), according to the manufacturer's protocol. cDNA synthesis was conducted using 800 ng RNA and ReverTra Ace qPCR RT Master Mix (Toyobo), according to the manufacturer's instructions.

| qRT-PCR assays for viral load quantification
Informed by our pilot study and previous epidemiological surveys in China (Ai et al., 2012;Diao et al., 2019;Li et al., 2012;Yañez et al., 2015;Yang et al., 2013), we focused on the five most common bee viruses (BQCV, CBPV, DWV-A, IAPV, and SBV). qRT-PCR was performed using 1 μl cDNA as the template in 10 μl reactions, using for 15 s and 60°C for 1 min. A melt curve analysis was used to confirm the specificity of the products. Each PCR amplification included a negative control where 1 μl of RNase-free water was used instead of template cDNA. The primers used for qRT-PCR are shown in Table 2.
Viral loads were quantified using absolute quantification methods. The linear standard curve equation for each virus was based on three linear standard curves obtained through six ten-fold dilutions of known amounts of plasmids (pMD ® 18-T Vector, TaKaRa) containing cloned viral target sequences (Wu et al., 2017). A linear standard curve was used for each qRT-PCR run.

| RT-PCR amplification and sequencing
For each sample, 1 μl cDNA was used for PCR amplification using KOD FX (TOYOBO) according to the manufacturer's instructions.
PCR products were electrophoresed in 2% agarose gels, purified, and sequenced by a commercial company (Sangon Biotch). The sequence specificity of each virus was checked by sequencing analysis using the NCBI BLAST service. The primer pairs used for RT-PCR in this study are shown in Table 3.

| Phylogenetic analysis
The sequences of BQCV, DWV, IAPV, and SBV isolates obtained from this study were individually aligned using Clustal W (Thompson et al., 1997) with other representative homologous sequences retrieved from GenBank. The phylogenetic trees were constructed using MEGA7 software with the Maximum Likelihood method based on the Tamura 3-parameter model (Tamura, 1992) and a bootstrap value of 1000 replicates.

BQCV, DWV, and SBV could be detected in A. mellifera and A. cerana
in each province, while IAPV was absent in samples from Yunnan.
CBPV was detected only in A. mellifera samples from Zhejiang and TA B L E 3 Primer sets used for PCR amplification of BQCV, DWV, IAPV, and SBV

| Virus infection dynamics in different seasons
Seasonally

| Viral infection dynamics in different bee species
There were statistical differences in the infection dynamics of BQCV, DWV, IAPV, and SBV between A. mellifera and A. cerana.
The infection rates of BQCV, DWV, and IAPV in A. mellifera were significantly higher than those in A. cerana in spring (p = 0.003; p = 0.019; p = 0.025) and of DWV in autumn (p = 0.008), while the infection rates of SBV in A. mellifera in spring and winter (p = 0.018; p = 0.037) were significantly lower than those in A. cerana ( Figure 4a).

| Phylogenetic relationship
All BQCV, IAPV, and DWV isolates from A. mellifera and A. cerana in different regions of China were phylogenetically clustered together and did not form distinct clades based on the species of the host. All of the BQCV isolates from Asia and the United States were clustered into the same clade, which was isolated from another clade mainly composed of European isolates (Figure 7). DWV isolates from China were clustered into several clades with isolates from South Korea, Brazil, and the UK (Figure 8). The IAPV isolates from China formed a common clade with isolates from South Korea, Australia, Japan, the United States, and the Czech Republic ( Figure 9).
The SBV phylogenetic tree consisted of two branches. One branch consisted of Asian SBV isolates from both A. mellifera and A. cerana, which further split into two subbranches. The first one was dominated by A. cerana SBV isolates but also contained six A. mellifera SBV isolates from Vietnam and China. Among the six isolates, four were identified in this study, which accounted for 14.28% (4/28) of the total sequenced A. mellifera SBV isolates. Another subbranch consisted of six A. mellifera SBV isolates from China. The second branch consisted of SBV isolates from A. mellifera from Oceania, Europe, Asia, and North America ( Figure 10).

| DISCUSS ION
Our work provided the first seasonal data revealing common virus  (Ai et al., 2012;Ding & Shi, 2015). Simultaneous multiple infections of honey bee viruses have been reported (Chen et al., 2005;Chen, Zhao, et al., 2004), and our results also showed that multiple infections were common in A. mellifera and A. cerana in China. The high infection rate and load of DWV in A. mellifera, which has been regarded as potentially responsible for A. mellifera colony losses (Nordström et al., 1999), suggested that it is an important factor affecting the health of A. mellifera in China. Moreover, the impact of the widespread prevalence of BQCV on A. mellifera in China is noteworthy because of its high pathogenicity to queen larvae (Anderson, 1993) and the large royal jelly production industry in China (Zheng et al., 2018). SBV was first identified in A. mellifera in the United States in 1913 and was subsequently found in A. cerana in Thailand (Thai sacbrood viruses, TSBV) and China (Chinese sacbrood viruses, CSBV; Allen & Ball, 1996;Bailey et al., 1982;Zhi & Chou, 2008).

A. cerana
Zero One Two Three Four mellifera and A. cerana in certain seasons were significantly different, which suggested that the infection pattern and dynamics of the same virus in A. mellifera and A. cerana vary.
Our study also allowed us to compare the prevalence and titers of the five common honey bee viruses in A. mellifera and A. cerana colonies throughout six provinces of China. The results indicated that there was no regional variation in the infection status of BQCV, DWV, IAPV, and SBV among different provinces in A. mellifera and A. cerana. This can be explained by migratory beekeeping, which facilitated the transfer of bee viruses between different provinces (Zheng et al., 2018).
Previous studies have reported seasonal changes in bee viral infections in A. mellifera (Natsopoulou et al., 2017;Runckel et al., 2011;Tentcheva et al., 2004). Our results also showed that seasonality impacted the prevalence of BQCV and DWV in A. mellifera and that of SBV in A. cerana. The infection rate of BQCV in A. mellifera was found to be high in spring, then decreased in summer, and remained stable in summer, autumn, and winter. However, a national investigation conducted in France showed that the infection rate of BQCV in summer was significantly higher than that in spring and autumn, and an investigation conducted in southwest Germany showed that BQCV frequencies and titers in A. mellifera were significantly lower in spring than in autumn (Natsopoulou et al., 2017;Tentcheva et al., 2004), which was different from our results. The discrepancies between these studies might be due to climate differences among the different countries or the genetic differences in the honey bees or viruses. The infection rate of DWV in A. mellifera increased during summer, which agreed with previous surveys conducted in France and China (Diao et al., 2019;Tentcheva et al., 2004) and could be related to the increase in V. destructor density from spring to summer (Gloria et al., 2017).
However, due to a lack of data on the parasitism rate of V. destructor in this study, a more comprehensive epidemiological investigation is warranted.
Phylogenetic analysis of BQCV, DWV, and IAPV did not show distinct patterns of phylogenetic clustering by host species, suggesting cross-species transmission of these viruses between A. mellifera and A. cerana (Yañez et al., 2015;Yang et al., 2013). The lack of geographical separation of BQCV, DWV, IAPV, and SBV isolates from these six provinces implied that they freely spread across provinces. Previous studies have reported the cross-species transmission of A. cerana SBV from A. cerana to A. mellifera (Gong et al., 2016). In this study, we also noticed that four SBV isolates from A. mellifera were clustered with A. cerana SBV, which accounted for 14.28% (4/28) of the total sequenced A. mellifera F I G U R E 7 Phylogenetic tree of BQCV. Notes: Isolates identified in this study are flagged by asterisks. Isolates are marked with different colors according to their countries or continents of origin (right) and host (left