Short communication. Presence, quantification and phylogeny of Israeli acute paralysis virus of honeybees in Andalusia (Spain)

This study aimed to assess the possible relationship between the presence of Israeli acute paralysis virus (IAPV) of honeybees and disease symptoms development at the colony level, to describe the IAPV load in field colonies and to illustrate phylogenetic relationships between IAPV isolates in Andalusia (Spain). Presence and load of IAPV was studied in 96 colonies from all provinces in Andalusia. Epidemiological surveys were performed in all the colonies to assess their sanitary status. IAPV was found in 13.5% of the sampled colonies, and no association was observed between the presence of IAPV and disease symptoms at the colony level. An average IAPV load was established in 4.9·10 genome equivalent copies per bee. Phylogenetic analysis revealed that Andalusian isolates belong to a different lineage to a previously described isolate found in Valencia (2010). The results of this study will help us understand the epidemiology and effect of IAPV on Spanish colonies. Additional key words: Apis mellifera; viral load; phylogenetic analysis; epidemiological survey. * Corresponding author: marina@sanidadanimal.info Received: 17-09-12. Accepted: 08-07-13. Abbreviations used: ABPV (Acute bee paralysis virus); CCD (colony collapse disorder); CERA (the Beekeeping Reference Centre of Andalusia); COAG (Coordination of Agricultural and Livestock Organisations); GEC (genome equivalent copies); IAPV (Israeli acute paralysis virus), IGR (intergenic region); KBV (Kashmir bee virus); ORF (open reading frame); PBS (phosphate-buffered saline); RT-PCR (reverse transcription polymerase chain reaction); RT-qPCR (real-time reverse transcription polymerase chain reaction). Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA) Spanish Journal of Agricultural Research 2013 11(3): 708-713 Available online at www.inia.es/sjar ISSN: 1695-971-X http://dx.doi.org/10.5424/sjar/2013113-3531 eISSN: 2171-9292

Israeli acute paralysis virus (IAPV) of honeybees is a Dicistrovirus that was first described in 2004 in Israeli colonies that had suffered from heavy losses (Maori et al., 2007). This virus gained prominence after a report in 2007 associated it with the colony collapse disorder (CCD) in the USA, where IAPV presence was considered a statistically significant marker of CCD (Cox-Foster et al., 2007). IAPV has since been described in numerous countries such as China (Xun et al., 2009), Canada and Australia (Palacios et al., 2008), or Argentina (Reynaldi et al., 2011). In Europe, IAPV has been found in France (Blanchard et al., 2008) and more recently in Spain (Kukielka & Sánchez-Vizcaíno, 2010) and Poland (Pohorecka et al., 2011). However, the high frequencies found in some countries (up to 41% in Argentina) and lack of obvious disease symptoms in most sampled colonies suggest that IAPV is a widespread virus that usually appears in covert infections, like most Dicistroviruses (De Miranda et al., 2010). Criteria identification to differentiate covert from overt infections in honeybee colonies is essential to identify risk factors of honeybee diseases, including CCD, where viruses may play an important role. Viral load may influence the development of disease symptoms, thus studies that differentiate the virus load causing covert and overt infections at the colony level are required. Some studies have quantified IAPV loads in experimental infections (Maori et al., 2007(Maori et al., , 2009) and have observed high mortality upon inoculation and feeding with purified viral particles. However, none of these studies has assessed IAPV loads in field colonies (commercial hives) until recently (Martin et al., 2012). Another explanation for the covert infection of IAPV is the virulence of the isolate. Most published isolates are low-virulent, as IAPV normally persists in the colony in covert infections, showing no obvious symptoms at the individual or colony level (De Miranda et al., 2010). In Spain, only one sequence has been phylogenetically analysed from one healthy colony in Valencia (Kukielka & Sánchez-Vizcaíno, 2010

Abstract
This study aimed to assess the possible relationship between the presence of Israeli acute paralysis virus (IAPV) of honeybees and disease symptoms development at the colony level, to describe the IAPV load in field colonies and to illustrate phylogenetic relationships between IAPV isolates in Andalusia (Spain). Presence and load of IAPV was studied in 96 colonies from all provinces in Andalusia. Epidemiological surveys were performed in all the colonies to assess their sanitary status. IAPV was found in 13.5% of the sampled colonies, and no association was observed between the presence of IAPV and disease symptoms at the colony level. An average IAPV load was established in 4.9·10 5 genome equivalent copies per bee. Phylogenetic analysis revealed that Andalusian isolates belong to a different lineage to a previously described isolate found in Valencia (2010). The results of this study will help us understand the epidemiology and effect of IAPV on Spanish colonies.
Additional key words: Apis mellifera; viral load; phylogenetic analysis; epidemiological survey. a report of 18% IAPV frequency in 2006 (Garrido-Bailón et al., 2010) and 13% and 25.7% in 200613% and 25.7% in and 200713% and 25.7% in , respectively (Antúnez et al., 2012. This Valencian IAPV isolate was surprisingly clustered with isolates from collapsed colonies from geographically far countries. Thus, new phylogenetic analyses of IAPV are required in Spain to confirm its phylogeny in relation to virulence as compared with other isolates from different countries. This study aimed to: assess the possible relationship between IAPV presence of honeybees and disease symptoms at the colony level; describe IAPV load in commercial hives; and illustrate the phylogenetic relationships between IAPV isolates by focusing on Andalusia, an important Spanish beekeeping region. This work studied Andalusia (south-east Spain), which ranks first in hive censuses and is the second most important Spanish honey-producing region. Technicians from the Coordination of Agricultural and Livestock Organisations (COAG) and the Beekeeping Reference Centre of Andalusia (CERA, University of Córdoba) sampled 96 honeybee colonies from all provinces of Andalusia in winter 2010/2011. The study was designed to describe frequencies of honey bee viruses in this autonomous community. Assuming the worse prevalence scenario (50%), sample size was calculated by considering hive censuses with an expected error of 10% and a 95% of confidence interval with WinEpiscope v2.0 (De Blas et al., 2000). One hundred different aged adult bees (nurse bees, guard bees, foragers) were collected per colony and frozen immediately afterwards. Epidemiological data were also collected during sampling by surveys to assess sanitary conditions. No clear CCD symptoms, collapse nor severe winter losses were found in the sampled colonies. Thus the present study focused on the potential association of IAPV with colony weakening represented by colonies with worse health status. Worse health status indicators included presence of clear IAPV disease symptoms, such as paralysis, loss of hair or inability to fly, depopulation, kleptoparasitism of stronger neighbouring colonies, diagnosed disease potentially associated with immune depletion such as chalkbrood (Glinski & Buczek, 2003;Aronstein & Murray, 2010); and problems to control Varroa destructor -mite infestation might lead to immunosuppression (Yang & Cox-Foster, 2005), which may facilitate IAPV replication, as Varroa destructor has been described as an IAPV vector (Di Prisco et al., 2011). Sanitary status data were collected from the epidemiological surveys con-ducted by technicians upon sampling. Colonies were classified as "healthy" or "weak" depending on the above-described symptoms appearing or not. Each analytical sample consisted of 50 bees, which were homogenized with 6 mL of sterile phosphate-buffered saline (PBS) using a mortar and pestle. RNA extraction was carried out using the column-based Nucleospin II Virus ® kit (Macherey Nagel) following the manufacturer's instructions.
Samples were analysed for IAPV presence by amplifying virus-specific nucleic acid. One-step real-time reverse transcription polymerase chain reaction (RT-qPCR) was carried out using SYBR-Green dye. Primers targeting a 220-bp region of the gene ORF-2 were used (Palacios et al., 2008), and amplification was done following the thermocycler protocol described by Palacios et al. (2008).
Absolute quantification of IAPV-positive samples was performed to determine their viral load. The ORF-2 fragment of IAPV was cloned into a PGemT ® TA cloning vector (Promega) following the manufacturer's instructions. The standard curve was constructed with triplicates of serial dilutions of known amounts of plasmid DNA. The viral load results were expressed in genome equivalent copies (GEC) per bee (Gauthier et al., 2007). One sample from Valencia, previously analysed for IAPV presence in 2010 (Kukielka & Sánchez-Vizcaíno, 2010), was included in the samples quantification with identical procedures because its viral load was not quantified in the previous study. This sample was also included in the phylogenetic analysis as it was the first IAPV isolate found in Spain.
After conf irming IAPV presence, RNA samples were first re-amplified by conventional RT-PCR targeting of a 705-bp sequence, which contains the 3' end of ORF1 (including part of the RdRp polymerase), the intergenic region (IGR) (183-bp) and the 5' end of the ORF2 region -including part of viral protein 2 (VP2)of the IAPV genome (Palacios et al., 2008), and were then sequenced. This fragment contains a variable region considered adequate to perform the IAPV phylogenetic analyses, as the highly conserved nature of the RdRp region used in other bee virus phylogenetic analyses can produce cross-reactivity between close KBV and IAPV viruses (Palacios et al., 2008). The resulting 705-bp sequences were utilised to construct a phylogeny of the identified isolates. Complete sequences of the amplified region were aligned with other IAPV sequences from different countries using MEGA 4 (Tamura et al., 2007) (Table 1). A phylogenetic tree was created by using the neighbour-joining test (Fig. 1).
The viral diagnosis results showed that IAPV was present in 13 of the 96 samples (13.5%) with 4.9 · 10 5 GEC/bee on average (minimum-maximum GEC/bee: 8.6 · 10 3 -1.2 · 10 7 ). The IAPV load of the Valencian sample was 1.43·10 5 GEC/bee. The Andalusian IAPV frequency results were similar to those previously obtained (Garrido-Bailón et al., 2010) with IAPV present in 18 out of 100 samples from 33 Spanish provinces; and more recently, to those of a retrospective study (Antúnez et al., 2012) describing IAPV frequencies of 13% and 25.7% in 2006 and 2007, respectively. Other than IAPV frequency in Andalusia, viral quantification showed that IAPV was present in substantial loads (up to 10 7 GEC/bee). Another study on adult honeybees at the colony level found viral loads of six different honeybee viruses in the order of 10 8 and 10 9 GEC/bee (Gauthier et al., 2007). More recently, Martin et al. (2012) reported similar results on the Hawaiian islands with IAPV loads in the order of 10 7 GEC/bee in covert infections not associated with colony collapse, but with lower prevalences (only 3 colonies of 293 were infected with IAPV). Based on the recent description of the V. destructor mite as a transmitter of IAPV (Di Prisco et al., 2011), future studies should also determine IAPV loads in pupae and V. destructor to better understand the biology of the virus. In this study, only the effect of IAPV presence and load was considered on weakening colonies in Andalusia. Future weakening colonies research should also contemplate other pathogens, the effect of co-infections and the host-pathogen-environment interaction. Given the IAPV frequency and load in Andalusia, these results suggest that IAPV presence is not an anecdotic finding in Andalusia.
According to the epidemiological surveys, 32 colonies were classified as "weak" and showed one or more above-described disease symptoms. No colony reported CCD or collapse. Eight of these weak colonies were positive to IAPV (25%), but IAPV presence was not   associated with weakening (χ 2 test, p > 0.05). No specific symptom contemplated in the epidemiological surveys was statistically associated with IAPV presence. Besides, the highest viral loads were present in both healthy and weak colonies, indicating that IAPV loads in the order of 10 7 GEC/bee do not produce obvious disease symptoms at the colony level themselves. However, it would be interesting to perform molecular and histopathological analyses of individual bees to describe the proportion of infected bees per colony as pools of bees were used, thus we cannot ensure that these viral loads do not produce clinical symptoms in individual bees. Future studies should consider this issue to establish the IAPV load baseline in healthy commercial hives. The phylogenetic analysis revealed two main lineages of IAPV (Fig. 1). The first one grouped all the IAPV sequences from Andalusia in the present study with IAPV isolates mostly from colonies suffering severe winter losses in France. The Valencian isolate (Kukielka & Sánchez-Vizcaíno, 2010) was, however, grouped with the second lineage, which included isolates from collapsed colonies in USA, China, Australia, Israel and Canada. These results confirm the great genetic diversity of IAPV sequences identif ied in Spain and suggest at least two differentiated evolutionary IAPV lineages in Spain. Andalusian samples may relate more closely to French samples given geographical proximity since natural bee migration between Spain and France can occur in lower-altitude areas of the Pyrenees, near the Atlantic and Mediterranean coasts. Hence, colonies separated by approximately 1,000 kilometres, like those in Andalusia and the northern Spain, can come into contact when beehives are moved to pollinate crops during transhumance, a common practice in Spain where 30.2% (7,323 of 24,251) of colonies are transhumant (MAGRAMA, 2012). The Valencian isolate is phylogenetically divergent from the IAPV isolates described in Andalusia and relates instead to isolates from geographically remote regions. Trade is the most likely explanation for the spread of IAPV between countries; imports of IAPV-positive honeybees from Australia to the USA have been reported (Palacios et al., 2008). The fact that IAPV isolates from healthy and symptomatic hives cluster together suggests that the symptoms development may be linked with viral replication activation and presence of high viral loads. This fact has been suggested , but is still to be verified in IAPV. Since IAPV has been found in many Spanish regions, further phylogenetic analyses are recommended in future studies on IAPV in honeybees. To confirm the results described herein, these future analyses should include the whole genome of the studied IAPV isolates.
Here the results suggest that IAPV is no sporadic finding in Andalusia because its frequency is 13.5% and IAPV presence does not significantly associate with disease symptoms presence at the colony level in this region. Therefore, we report IAPV loads in the order of up to 10 7 GEC/bee in healthy commercial hives, which indicates that presence of these viral loads is not indicative of overt infection at the colony level. Therefore, the development of obvious disease symptoms at this level must be produced by higher viral loads, probably in conjunction with other factors and especially in immunosuppressed colonies. Moreover, Andalusian IAPV isolates are phylogenetically distant from the Valencian isolate, suggesting that IAPV has at least two different evolutionary lineages in Spain. Spanish isolates from healthy colonies relate to isolates from colonies suffering depopulation or collapse in other countries, thus further investigation into isolates virulence in field samples is required. Indeed, phylogenetic analysis may help investigate relevant honeybee virus epidemiology aspects, such as their spread via trade routes, which must be considered when implementing control measures.
In conclusion, this study reports the IAPV frequency in an important Spanish beekeeping region, Andalusia, and lack of association between IAPV presence and disease symptoms at the colony level. It is also the first quantification of IAPV load in Spain and it has established phylogenetic similarities between Andalusian IAPV isolates and those mainly from France by identifying at least two different evolutionary IAPV lineages in Spain.