Integrity and quantity of salivary cell-free DNA as a potential molecular biomarker in oral cancer: A preliminary study

Vol.:(0123456789) 1 3 Microbial Ecology (2023) 86:2655–2665 https://doi.org/10.1007/s00248-023-02266-y RESEARCH Frequent Parasitism ofApis mellifera byTrypanosomatids inGeographically Isolated Areas withRestricted Beekeeping Movements DanielAguado‑López1· CarolinaBartolomé2· AnaRitaLopes3,4· DoraHenriques3,4· SaraKafafiSegura5· XulioMaside2· M.AlicePinto3,4· MarianoHiges1· RaquelMartín‑Hernández1,6 Received: 10 April 2023 / Accepted: 7 July 2023 / Published online: 22 July 2023 © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023 Abstract Trypanosomatids form a group of high prevalence protozoa that parasitise honey bees, with Lotmaria passim as the predominant species worldwide. However, the knowledge about the ecology of trypanosomatids in isolated areas is limited. The Portuguese archipelagos of Madeira and Azores provide an interesting setting to investigate these parasites because of their geographic isolation, and because they harbour honey bee populations devoid of two major enemies: Varroa destructor and Nosema ceranae. Hence, a total of 661 honey bee colonies from Madeira and the Azores were analysed using different molecular techniques, through which we found a high prevalence of trypanosomatids despite the isolation of these islands. L. passim was the predominant species and, in most colonies, was the only one found, even on islands free of V. destructor and/or N. ceranae with severe restrictions on colony movements to prevent the spread of them. However, islands with V. destructor had a significantly higher prevalence of L. passim and, conversely, islands with N. ceranae did not shown any significant correlation with the trypanosomatid. Crithidia bombi was detected in Madeira and on three islands of the Azores, almost always coincident with L. passim. By contrast, Crithidia mellificae was not detected in any sample. A high-throughput sequencing analysis distinguished two main haplotypes of L. passim, which accounted for 98% of the total sequence reads. This work suggests that L. passim and C. bombi are parasites that have been associated with honey bees predating the spread of V. destructor and N. ceranae. Keywords Honey bee· Trypanosomatids· Lotmaria passim· PCR· Sequencing· Varroa destructor Introduction Trypanosomatidae are a group of parasitic protozoa that can infect a wide range of organisms, from plants to insects and vertebrates [1, 2]. In recent years, particular interest has focused on their infection of honey bees because of their possible involvement in increased colony mortality. The first trypanosomatid described in honey bees was Crithidia mellificae [3] and, subsequently, Crithidia bombi was identified in bumble bees [4]. These parasites were not given special attention until the beginning of the twenty-first century [5–11] and until then, the trypanosomatids found in honey bees were thought to be limited to C. mellificae in honey bees and to C. bombi in bumble bees. However, the advances in molecular techniques led to the discovery of a new species: Lotmaria passim [8], whose existence had previously been suggested from genetic data [12]. The re-evaluation of all the molecular data obtained in previous works showed that many sequences were incorrectly assigned to C. mellificae and, in fact, it became clear that L. passim was the predominant species worldwide [13–15]. Thus, L. passim seems to have been present in honey bee colonies long before its description, although there is no scientific evidence as to whether it is a novel honey bee parasite that recently spread worldwide, similar to Varroa destructor and Nosema ceranae, or whether it is an old parasite. Due to the geographic isolation of the host populations, islands offer an ideal stage to assess these hypotheses and study the ecology and dynamics of invasive pathogen colonisation. Yet, there are very few trypanosomatid studies on islands, and these have been limited to the Pacific, including Japan [10, 16], New Zealand [15, 17, 18] and Hawaii [19]. Therefore, surveys from islands at other geographical * Raquel Martín-Hernández r[email protected] Extended author information available on the last page of the article 2656 D.Aguado-López et al. 1 3 latitudes would shed additional light as to whether these microorganisms are recent or ancient honey bee parasites. To this regard, the Portuguese archipelagos of the Azores and Madeira are located in the Atlantic Ocean, 1400km and 900km from the European mainland, Lisbon, respectively. The first introduction of honey bees on these islands dates back to the sixteenth century. These were taken by the Portuguese settlers, or by later inhabitants, when the islands were used as stopovers on the way to the Americas [20–22]. According to the latest report of the Portuguese Bee Health Programme [23], there are currently more than 10,000 honey bee colonies in Madeira and around 8000 in the Azores. The severe restrictions on the importation of honey bees imposed since 2007 make the Azores archipelago epidemiologically unique, with six of the nine islands remaining free of V. destructor. Remarkably, in addition to being one of the few places in the world free of the mite [24], two of the Azorean islands (Flores and Santa Maria) have also escaped the worldwide spread of the microsporidium N. ceranae [25]. Therefore, the Azorean archipelago provides an incomparable setting for studying parasite ecology, as it is home to all kinds of parasite combinations. In light of this, we conducted a survey of trypanosomatids in the Azores as well as on Madeira, which was earlier colonised by V. destructor. The ultimate goal was to establish the prevailing trypanosomatid species and also to ascertain whether the queen marketing and colony movement that have been implicated in the spread of V. destructor have also influenced the distribution of trypanosomatids across the islands. To that end, we employed molecular methods on a cross-sectional honey bee sample that allow detection of an array of trypanosomatids, including C. mellificae and L. passim, which typically parasitize honey bees, and also species such as C. bombi, Crithidia expoeki and Crithidia acanthocephali, more rarely associated with honey bees [26]. Methods Survey andSample Collection This cross-sectional study is part of a larger survey conducted between 2014 and 2015 to describe the prevalence of pathogens in the Azores and Madeira archipelagos. The survey was carried out following the design described elsewhere [25, 27] and in accordance with the number of colonies on each island registered in 2013, with an expected pathogen prevalence of 15%, a precision rate of 10% and a confidence level of 95%. About 150 adult honey bee workers were collected from 159 georeferenced apiaries of the Azores in the Summer of 2014 and 2015, and from 23 georeferenced apiaries of Madeira in the Spring of 2014. Samples were collected from three random colonies, and in a few cases from two or four colonies per apiary, resulting in a total of 483 samples for the Azores and 89 for Madeira. In addition, the islands of Faial, São Jorge, Santa Maria and Terceira were re-sampled in the Summer of 2020 (89 colonies in 34 apiaries). In this later sampling, each sample comprised 20–30 workers. Samples from the Azores were shipped alive to the Centro de Investigação de Montanha (CIMO, Portugal) and then sent on dry ice to the Centro de Investigación Apícola y Agroambiental (CIAPA, Spain). Samples from Madeira were collected in ethanol and stored at − 20°C until shipping to CIAPA. All samples were kept at − 80°C for further analysis. DNA Extraction For each colony sampled in the Azores and Madeira in 2014–2015, a subsample of 120 workers was selected and processed as described previously [25]. In brief, the honey bees were macerated in 50% AL buffer (Qiagen®) and centrifugated at 3000rpm for 10min to obtain supernatant and sediment that were processed separately. The sediment was resuspended in 3mL of milli-Q water and the DNA was extracted as described previously [25]. Only for the 2014–2015 Azorean samples, 400 µL of each supernatant was transferred to a 96-well plate (Deepwell, Eppendorf), treated with 15 µL of protease (Qiagen®) and incubated at 70°C for 10min. Afterwards, the DNA was extracted as indicated before [25]. The samples collected in 2020 in the Azores were processed similarly, but in this case the 20–30 honey bees were macerated in 5mL of autoclaved milli-Q water in a Stomacher 80. In these samples, no phase separation was made so that 400 µL of each macerate was directly transferred to a 96-well Collection Microtube plate (Qiagen®) with glass beads (2mm diameter, Sigma) and shaken (3min, 30Hz) in a TissueLyser (Qiagen®). Then, 150 µL of each sample was dispensed into a plate with 150 µL of 50% AL buffer and 15 µL of protease. After incubation, the DNA was extracted as above mentioned [25]. Extracted negative controls were included in all the processing and DNA extraction steps (one every 20 samples) and processed in parallel. All samples were stored at − 80°C until further analysis. Sediment Versus Supernatant Samples from the Azores from 2014 to 2015 were first processed for the detection of other pathogens in a different study [25] and, due to their value, were reused in this study. Since sediment and supernatant were available and it was not known which phase was the best for detecting these parasites after honey bee maceration, DNA obtained from the two phases was analysed separately to determine 2657 Frequent Parasitism ofApis mellifera byTrypanosomatids inGeographically Isolated Areas… 1 3 which was the best substrate for the detection of Trypanosomatidae. For this purpose, DNA was extracted from the sediment and supernatant of each sample, as above, and analysed in parallel by PCR, in order to compare the number of positive samples obtained in each phase. Trypanosomatidae Detection Polymerase chain reaction (PCR) was used to detect the Trypanosomatidae in all the samples (sediment from Madeira samples, supernatant and sediment from the 2014–2015 Azorean samples, and macerate from the 2020 Azorean samples) with the Tryp RPB1 primers described elsewhere [26] and shown in Table1. These primers target the DNA-dependent RNA polymerase I (rpb1) gene and they amplify a fragment of 283bp from all Trypanosomatidae species that have been detected in honey bees to date. PCRs were performed in a 25-µL volume containing 13.25 µL of H2O, 5 µL of 5X Phusion™ HF buffer, 0.5 µL of a 10mM dNTP mix, 2.5 µL of each primer (5µM), 0.25 µL of Phusion™ DNA Polymerase (Thermo Fisher) and 1 µL of the DNA template. The PCR temperature profile was set according to the manufacturer’s instructions, and consisted of an initial denaturation at 98°C for 30s, followed by 45 cycles at 98°C for 10s, 62.2°C for 30s and 72°C for 10s, and a final extension of 8min at 72°C. PCRs were carried out in a Mastercycler® ep gradient S (Eppendorf) and the resulting amplicons were analysed in a QIAxcel Advanced System (Qiagen®), storing the remaining PCR product at − 20°C for further processing. Extraction and PCR negative controls and a positive control were included in all the analyses and run in parallel. In the case of 2014–2015 Azorean samples, as both sediment and supernatant were analysed, a sample was considered positive if there was amplification in at least one of the phases. High‑Throughput Sequencing A subset of 91 positive samples that were positive for Trypanosomatidae were sequenced on the Illumina MiSeq platform together with negative controls. The PCR product of the positive samples was quantified in a NanoDrop™ 2000 Spectrophotometer (ThermoScientific™) to determine the DNA concentration and sent frozen to Universidade de Santiago de Compostela (CIMUS, Spain) for library preparation and sequencing. Amplicons from each sample were used as input to prepare the library using the KAPA HyperPrep kit (Roche Sequencing Solutions Inc.), following the manufacturer’s protocol directly from the end-repair and A-tailing step. Library pools were normalised to a concentration of 4nM and loaded at a concentration of 12pM on an Illumina MiSeq instrument for 1 × 300bp single-end sequencing (Flowcell Nano V2, 2 × 150bp). The reads generated in the sequencing run were de-mul- tiplexed according to the barcodes assigned to each sample, and then processed with fastp [29] for adaptor removal and quality filtering. The reads were then organised into individual files that contained the number and sequence haplotypes detected for each amplicon in a sample, as well as the number of reads in each direction (forward and reverse). Sequences with less than 5 reads were discarded. Subsequently, the files were converted to fasta format for sequence alignment with MACSE v2.05 (Multiple Alignment of Coding SEquences), a program that aligns protein-coding gene datasets without disrupting the underlying codon structure [29]. The reference sequences of the rpb1 were obtained from GenBank for L. passim (MT558272.1 and LT976801. 1), C. mellificae (MT558227.1 and MT558204.1), C. bombi (MT558162.1 and MT558134.1), C. acanthocephali (MW28878781.1) and a new Trypanosomatidae species (Trypanosomatidae sp. MN038411.1), recently described [26]. Sequences were visualised with BioEdit [30] and those Table 1 Primers and probes used for the molecular analysis of the honey bee samples References Target Primers Sequence 5ʹ–3ʹ [26] Trypanosomatidae Tryp RPB1-F1 GTG GCT GGA YCT GTG GGA GC Tryp RPB1-R1 GCC RTT GAT GAA CTT CGC CAC [28]C. mellificae qCmell_Cytb_F TTT TGC CAT GCA CTA TGA TGTCT qCmell_Cytb_R AAC CTA TTA CAG GCA CAG TTG CTA AA qCmell_Cytb_P 6FAM-ATT GAG GAT TAA CAG TGT TTAGT-BHQ1 L. passim qLpass_TOPII_F GGC CAT GGA AAT ACT CGA GTCT qLpass_TOPII_R ACC TTG CCT TCC TTC TTG AGATT qLpass_TOPII_P 6FAM-CCT CGA CACGC + T + TA + GT-BHQ1 C. bombi qCbom_RPB1_F TGG TGG GTG CGA TTA CGA A qCbom_RPB1_R TCA TTG AAG ATG ACG TGG ATA AGC qCbom_RPB1_P 6FAM-CGT TGT CGG CGC CG-BHQ1 2658 D.Aguado-López et al. 1 3 containing indels or stop codons were removed from the final dataset. Species‑Specific Detection All Trypanosomatidae-positive samples that were not sequenced by high-throughput sequencing (HTS) were analysed by real-time qPCR (qPCR) for the species-specific detection of L. passim, C. mellificae and C. bombi using the primers described elsewhere [28] and shown in Table1. The qPCR was performed on DNA obtained from Madeira samples, from both sediment and supernatant of the 2014–2015 Azorean samples, and from samples without phase separation from the 2020 Azorean samples. Negative and positive controls were also tested in parallel in all PCRs. As for Trypanosomatidae detection in the 2014–2015 Azorean samples, a sample was considered positive if there was amplification from either of the two phases. Amplicons from samples that gave a negative result by the specific qPCR but a positive result by the standard PCR amplification with the Tryp RPB1 primers were purified with QIAquick (Qiagen®) and Sanger sequenced in both directions (Genomic Unit, Universidad Complutense, Spain) on a 3730 Genetic Analyzer (Applied Biosystems). The sequences were checked manually with BioEdit [30] and compared with the sequences downloaded from GenBank using BLAST. The geographical distribution of the colonies was compiled and plotted for each island using the ArcGIS desktop software [31]. Statistical Analysis Chi-square tests were performed to compare the sensitivity of Trypanosomatidae detection in the sediment and supernatant phases. Generalised linear models with a binomial family and a logit link function were used to determine whether the presence of V. destructor or N. ceranae influenced the prevalence of L. passim at the island scale, including island as covariable. In the latter case, only data from samples from the Azores collected in 2014–2015 were used, as this sampling was conducted and analysed homogeneously. Results Comparison Between Sediment andSupernatant To determine the best phase for Trypanosomatidae detection, sediment and supernatant analysis was performed on 477 Azorean samples collected in 2014–2015 (Table2; six samples out of 483 had insufficient supernatant for this assay). The supernatant produced the highest number of positive samples (55.34%) as compared to the sediment (50.73%), although no statistically significant differences were found between both phases (X2=2.04, gl=1; p=0.15). No trypanosomatid DNA was detected in any negative controls. Prevalence ofTrypanosomatidae The prevalence of Trypanosomatidae is shown in Table3. As the Madeira, Azores 2014–2015 and Azores 2020 samples were processed differently, no comparisons were made between those 3 groups. In Madeira, the prevalence was 66.3%, whereas in the Azores was 72% in 2014–2015, with the highest number of infected colonies detected on Pico (92%) and the lowest on Graciosa (22.7%). In the later sampling period (2020), 31.5% of the colonies were positive, Table 2 Number (and percentage) of Trypanosomatidae-positive ( +) and Trypanosomatidae-negative ( −) samples detected on the supernatant and sediment phases obtained from the honey bee macerates Supernatant ( +) Supernatant ( −) Total Sediment ( +) 161 (33.75%) 81 (16.98%) 242 (50.73%) Sediment ( −) 103 (21.59%) 132 (27.67%) 235 (49.27%) Total 264 (55.34%) 213 (44.65%) 477 (100%) Table 3 Number of colonies and apiaries sampled in each island, number of colonies positive for Trypanosomatidae and corresponding prevalence (%). 1Varroa-free islands; 2Nosema-free islands Apiaries (N) Colonies (N) Positive colonies (N) Prevalence (%) 2014–2015 Madeira 23 89 59 66.3 Azores Santa Maria1, 2 19 57 48 84.2 São Miguel130 105 84 80.0 São Jorge117 37 19 51.4 Faial 20 60 43 71.7 Pico 25 75 69 92.0 Graciosa17 22 5 22.7 Terceira126 80 43 53.8 Flores215 47 37 78.7 Total 159 483 348 72.0 2020 Santa Maria1, 2 12 28 12 42.9 São Jorge110 30 5 16.7 Faial 2 8 2 25.0 Terceira110 23 9 39.1 Total 34 89 28 31.5 Total 216 661 435 2659 Frequent Parasitism ofApis mellifera byTrypanosomatids inGeographically Isolated Areas… 1 3 with the largest percentage detected on Santa Maria (42.9%) and the lowest on São Jorge (16.7%). Species Identification High‑Throughput Sequencing A total of 91 samples identified as positive to Trypanosomatidae by the PCR assay were sequenced on the Illumina MiSeq platform. The sequencing run produced 1,385,021 raw reads, which yielded 1,208,665 Trypanosomatidae sequences after filtering, representing an average of 13,282 sequences per sample. Strikingly, identification of all sequence reads revealed that L. passim was the only species present in the dataset. Two main haplotypes were found in all samples and on all islands and accounted for about 98% of the sequences found (Table4) and these matched the reference sequences MT558272.1 and LT976801.1 (hereafter referred to as Lp1 and Lp2, respectively) downloaded from GenBank. These two haplotypes were found distributed within each island without any defined geographical pattern and they differed by a single nucleotide in the amplicon at position 29, with Lp1 having a T and Lp2, a C (Supp. Figure1). On average, 72.9% of the reads from all the islands corresponded to the Lp2 haplotype, whereas 25.3% of the sequences corresponded to the Lp1 haplotype. This difference was most notable in samples collected in Madeira and in samples collected in 2014–2015 from all islands of central group, where more than the 70% of sequences corresponded to the Lp2 haplotype (Table4; Fig.1). The remaining 2% of sequences were found at a very low frequency and appeared to be variants from the majority haplotypes with additional single nucleotide changes and did not match any sequence available in GenBank. The sequences of the haplotypes found (both Lp1 and Lp2 and those at low frequencies) were deposited in GenBank (accession numbers OR117383-OR117469). Real‑time qPCR Species Confirmation The rest of the Tryp-RPB1 positive samples (n=344) were analysed by qPCR to detect the presence of L. passim, C. mellificae and/or C. bombi. At least one of these species was identified in 341 samples, whereas three samples (from Madeira, Santa Maria and São Miguel) were negative (Table5). No amplification was observed in any of the negative controls. The qPCR analysis confirmed that L. passim was the predominant but not the only species on Madeira and in the Azores, as C. bombi was also found on some islands (Table5). In the first sampling period, infection by L. passim was identified alone in 64.2% of the positive samples on Madeira, while C. bombi was found in 34% of the positive samples, always together with L. passim (Table5). On São Jorge, Pico, Graciosa, Santa Maria and Terceira, L. passim was the only species detected, whereas C. bombi was found as a mono-infection in one colony on Flores (3.7%). Yet, on São Miguel and Faial, this species was always associated with L. passim infection. In the samples collected in 2020, L. passim was the only Trypanosomatidae detected (Table5). Finally, C. mellificae was not found in any of the samples analysed throughout the study. Sanger sequencing of the Tryp RPB1 primer amplicons of the three negative samples after qPCR showed 100% homology with the MT558162.1 sequence of C. bombi in the Madeira sample, in the Santa Maria sample the sequence was identical to Lp2, while in the São Miguel sample, two peaks for C/T at position 29 of the amplicon, corresponding to the Lp1 and Lp2 haplotypes, were easily distinguishable. Total Occurrence ofTrypanosomatidae Species onMadeira andtheAzores The occurrence of each Trypanosomatidae species calculated by compiling all the results obtained using the different techniques (HTS, qPCR and Sanger sequencing) is shown in Table6 and Fig.1. On Madeira, the proportion of colonies in which Trypanosomatidae were detected was 66.3%, with L. passim being the only species detected in 44.9%, C. bombi in 1.1% and both species concomitantly in 20.2% of the samples analysed. In the 2014–2015 samples from the Azores, 348 colonies were positive for Trypanosomatidae Table 4 Average number of total and valid reads (sequences with < 5 reads were removed), generated by the MiSeq platform for 91 Trypanosomatidae-positive samples, and proportion of the two main L. passim haplotypes (Lp1 and Lp2) detected on each island. 1Varroa-free islands; 2Nosema-free islands NTotal reads Valid reads % Lp1 % Lp2 2014–2015 Madeira 6 12,678 11,509 34.2 64.1 Azores Santa Maria1, 2 10 15,809 14,547 39.2 59.0 São Miguel110 12,940 11,812 39.5 58.8 São Jorge17 12,982 11,700 6.5 91.0 Faial 10 12,886 11,883 14.5 83.6 Pico 10 12,642 11,652 14.9 83.2 Graciosa14 12,827 11,725 27.9 70.9 Terceira19 12,204 11,152 9.0 89.3 Flores210 13,256 12,137 47.6 50.8 2020 Santa Maria1, 2 5 14,321 13,146 37.7 60.5 São Jorge15 13,459 12,490 19.8 78.4 Terceira15 13,261 12,116 12.8 84.7 Total 91 13,282 12,168 25.3 72.9 2660 D.Aguado-López et al. 1 3 (72.0%). L. passim was found on all the islands, and it was the only species detected in 71.2% of samples. C. bombi was rarely found in the Azores, being the only species detected in one colony on Flores (0.2%) and occurred concurrently with L. passim on São Miguel and Faial (0.6%). In the 2020 samples, all the colonies that were positive for Trypanosomatidae (31.5%) were also positive for L. passim, and this was the only species found in this sampling period (Table6). Finally, as previously mentioned, C. mellificae was not detected in any of the samples tested. As L. passim was the main species found, the influence of V. destructor and N. ceranae was only studied for this trypanosomatid. Thus, the prevalence of L. passim was significantly higher on the Azores islands (2014–2015 samples) with V. destructor (Table7, regression coefficient = 0.77; z value = 3.41; p < 0.01) than on those without the mite. In Fig. 1 Geographical distribution of C. bombi and L. passim (and its haplotypes) in the Azores archipelago (a) and Madeira (b). indicates the islands where V. destructor is present. indicates the islands where N. ceranae is present. The red colour in sector charts indicates the percentage of sequences that matched with Lp1 haplotype while the pink colour indicates the percentage of sequences that matched with Lp2 haplotype. C. bombi + L. passim;C. bombi;L. passim;Negative sample 2661 Frequent Parasitism ofApis mellifera byTrypanosomatids inGeographically Isolated Areas… 1 3 contrast, on islands where N. ceranae was established, the prevalence of L. passim was lower (Table7; regression coefficient = − 0.56; z value = − 2.03; p < 0.01). However, when island was included as covariable, the effect of N. ceranae on the presence of L. passim became statistically no significant (regression coefficient = − 0.29; z value = − 0.66; p = 0.51) while the significance was kept for V. destructor (regression coefficient = 1.06; z value = 2.15; p = 0.03). Discussion The aim of this study was to assess the status of Trypanosomatidae species in honey bee colonies in Madeira and the Azores, and to determine whether the presence of V. destructor and N. ceranae on some islands influenced their distribution. Our results show that islands with the mite had a significantly Table 5 Percentage of each Trypanosomatidae species detected in positive samples. ND, not detected. 1Varroa-free islands; 2Nosema-free islands NOnly L. passim Only C. mellificae Only C. bombi Co-occurrence L. passim/C. bombi 2014–2015 Madeira 53 64.2 ND ND 34 Azores Santa Maria1, 2 38 97.4 ND ND ND São Miguel174 95.9 ND ND 2.7 São Jorge112 100 ND ND ND Faial 33 97 ND ND 3 Pico 59 100 ND ND ND Graciosa11 100 ND ND ND Terceira134 100 ND ND ND Flores227 96.3 ND 3.7 ND Total 278 97.8 ND 0.4 1.1 2020 Santa Maria1,2 7 100 ND ND ND Faial 2 100 ND ND ND Terceira14 100 ND ND ND Total 13 100 ND ND ND Table 6 Prevalence of Trypanosomatidae (%) detected either alone or simultaneously on each island grouping the data from all the tools used (conventional PCR, HTS, qPCR and Sanger sequencing). Number of samples analysed as shown in Table3. ND, not detected. 1Varroa-free islands; 2Nosema-free islands Trypanosomatidae Only L. passim Only C. mellificae Only C. bombi Co-occurrence L. passim/C. bombi 2014–2015 Madeira 66.3 44.9 ND 1.1 20.2 Azores Santa Maria1, 2 84.2 84.2 ND ND ND São Miguel180.0 78.1 ND ND 1.9 São Jorge151.4 51.4 ND ND ND Faial 71.7 70.0 ND ND 1.7 Pico 92.0 92.0 ND ND ND Graciosa122.7 22.7 ND ND ND Terceira153.8 53.8 ND ND ND Flores278.7 76.6 ND 2.1 ND Total 72.0 71.2 ND 0.2 0.6 2020 Santa Maria1, 2 42.9 42.9 ND ND ND São Jorge116.7 16.7 ND ND ND Faial 25.0 25.0 ND ND ND Terceira139.1 39.1 ND ND ND Total 31.5 31.5 ND ND ND 2662 D.Aguado-López et al. 1 3 higher prevalence of L. passim. Other studies had reported an association between them [32] even detecting DNA of this trypanosomatid in the acari, which is still of uncertain biological significance [33]. In contrast, when the island was included as covariable, no effect of N. ceranae on L. passim prevalence was observed as it was previously reported by other authors [14, 34]. However, a positive correlation between infection levels in colonies with both parasites has been also described [35]. Despite these differences related to the presence of V. destructor and N. ceranae, it is clear that trypanosomatids are well established in the Azores and Madeira, so that the strong restrictions on the introduction of honey bees on the Varroa-free islands of the Azores have not prevented the presence of trypanosomatids. Furthermore, L. passim is the main species found and the only identified in most of the colonies analysed. Interestingly, C. mellificae, the species associated with honey bees until the discovery of L. passim [8, 12], is not present in the territories analysed or is below our detection limit. Neither C. acanthocephali nor C. expoeki were detected in this study, although they could only have been confirmed by HTS as they were not analysed by qPCR. This scenario where L. passim is the main trypanosomatid species in honey bee colonies seems to be repeated worldwide [13, 14, 28, 32, 35–39]. Given the high prevalence found herein and elsewhere, it is likely that L. passim, rather than C. mellificae, was the primary species infecting honey bees in studies conducted prior to its discovery [9–12]. Indeed, the frequencies observed and the absence of C. mellificae in many areas of the world corroborate the hypothesis that L. passim is a common parasite of Apis mellifera and that it is currently the dominant trypanosomatid species in adult worker bees [8, 13, 28, 39, 40]. On the other hand, C. bombi was identified in three islands and almost always coexisted with L. passim although it also appeared alone in some colonies. These results further confirm previous findings [41, 42], in which C. bombi was detected in honey bees. How trypanosomatids entered the Azores is still unclear since the islands are located far from the mainland; the natural arrival of bees is not possible, so a human-made introduction was very probably. Although it could be that Pico was the point of entry, as it hosts the highest number of trypanosomatid-positive colonies and was also the point of entry of V. destructor into the Azores, the data from Santa Maria, where neither V. destructor nor Nosema spp. are present [25] but where L. passim was detected in a large proportion of the colonies, do not support this hypothesis. The widespread distribution of L. passim on the Azorean islands that remain free of V. destructor strongly suggests that this species has been associated with A. mellifera since before the worldwide spread of the mite. A similar scenario was found in Hawaii archipelago, in which still has some islands where the mite is not present and where L. passim has been detected [19]. In New Zealand (Varroa-free until 2000), significant colony losses began to be observed in 2014 and an analysis of the surviving colonies revealed that most were infected with L. passim [15, 17, 18]. On the other hand, the presence of bumblebees in Madeira [43] and in the Azores [44] could explain the presence of C. bombi in honey bees due to feeding on common floral resources [45–47]. Our study supports that both sediment and supernatant phases are suitable for trypanosomatid detection. Also, the use of different technologies was key to identify all the species present in this area of Macaronesia. Thus, on the one hand, HTS technology allowed us to detect the haplotypes present but not all the species, as the cost of this tool limited the number of samples analysed (n = 91). On the other hand, the use of qPCR served to complement the analysis by allowing us to analyse the remaining trypanosomatid-positive samples (n = 344), which served to identify C. bombi present in lower prevalence. There is little data available on the population genetics of the rpb1 in the trypanosomatids that infect honey bees or bumble bees. Recently, this gene was analysed and genetic polymorphisms were observed between C. mellificae, C. bombi and L. passim [26, 48, 49]. Our results are consistent with those obtained for the RPB1 gene in these three studies, indicating that there were two core haplotypes differing in only one nucleotide. It cannot be ruled out that the sequences found at low frequencies correspond to true rare haplotypes, although further experimental confirmation would be needed. Other genes like 18S rRNA, 28S rRNA and its-2 have been studied to evaluate intraspecific variation, and their sequences were almost identical between the different strains of L. passim, with the exception of the virulence factor gp63, which seems to have a variable stretch between strains [40]. Therefore, it is possible that the rpb1 does not have sufficient variation to distinguish strains, but that this may occur in other genes. In conclusion, despite geographical isolation, trypanosomatids frequently infect colonies in Madeira and the Azores, even when there are restrictions on honey bee movements to mite-free islands. L. passim is the predominant Trypanosomatidae species, occurring as the only species in many colonies and showing a higher prevalence on islands with V. destructor and lower prevalence on islands with N. ceranae. Additionally, C. bombi also occurs, often co-occurring with L. passim. The detection of L. passim and C. bombi in areas Table 7 Number (and percentage) of L. passim–positive ( +) and L. passim–negative ( −) samples collected in 2014–2015 in the Azores distributed according to whether they were on islands where V. destructor or N. ceranae were present ( +) or not ( −) L. passim ( +) L. passim ( −) Total V. destructor ( +) 148 (81.32%) 34 (18.68%) 182 (100%) V. destructor ( −) 199 (66.11%) 102 (33.89%) 301 (100%) N. ceranae ( +) 263 (69.39%) 116 (30.61%) 379 (100%) N. ceranae ( −) 84 (80.77%) 20 (19.23%) 104 (100%) 2663 Frequent Parasitism ofApis mellifera byTrypanosomatids inGeographically Isolated Areas… 1 3 where V. destructor and N. ceranae are not present suggests that they have been associated with A. mellifera for much longer than originally thought. Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s00248- 023- 02266-y. Acknowledgements The authors wish to thank V. Albendea, T. Corrales, M. Gajero, C. Uceta and M. Buendía of the Honey Bee Pathology Laboratory (CIAPA) and D. da Silva (Santiago de Compostela University). We are grateful to all veterinarians and technicians of the “Direção Regional da Agricultura e Desenvolvimento Rural dos Açores” and “Divisão de Experimentação e Melhoria Agrıcola, Direcão Regional de Agricultura da Madeira” who collected the samples across the Azores and Madeira, namely Frank Aguiar, Nuno Salvador, Janyne Sousa, Ivan castro, Célia Mesquita, Ana Jorge, José Dias, Paulo Rico, Pedro Leal, Vagner Paulos, Luis Xavier, Luís Silva, Martins Silva, Carlos Gouveia, Ana Carina Coimbra, João Ramos, João Arruda, Edgardo Melo, João Luís, Moniz da Ponte and Berta Correia. Special thanks are due to Paula Vieira, Frank Aguiar and Berta Correia for coordinating the sampling in the Azores and Madeira. Author Contribution Conceptualization: R.M.-H., M.A.P. and M.H.; methodology: all authors; formal analysis: all authors; investigation: D.A-L., A.R.L. and S.K.S.; data curation: D.A.L.; writing—original draft preparation: D.A-L. and R.M.-H.; writing—review and editing: D.A-L., R.M.-H., C.B, X.M, M.A.P. and M.H.; supervision: R.M.-H.; project administration and funding acquisition: R.M.-H and M.A.P; All authors have read and agreed to the published version of the manuscript. Funding This work was funded by the Consejería de Educación, Cultura y Deportes, of the Junta de Castilla – La Mancha (European Regional Development Fund) Project No. SBPLY/19/180501/000334, and through the program COMPETE 2020—POCI (Programa Operacional para a Competividade e Internacionalização) and FCT (Fundação para a Ciência e a Tecnologia) in the framework of the project BeeHappy (POCI-01–0145-FEDER-029871). INCRECYT program was funded by ESF/EC (European Social Funds). D.A-L contract was funded by the Ministerio de Asuntos Económicos y Transformación Digital (Grant No. PRE2018-084878, RTA2017- 00004-C02-01). A.R.L was funded by Fundação para a Ciência e Tecnologia (FCT) through the individual research grant SFRH/ BD/143627/2019. FCT provided financial support by national funds (FCT/MCTES) to CIMO (UIDB/00690/2020 and UIDP/00690/2020) and SusTEC (LA/P/0007/2021). Data Availability Lotmaria passim sequences obtained by HTS are available at Genbank (NCBI) data base. Total and valid readsgenerated by the MiSeq platform are available at Supplementary Table1. Other datasets generated during and/oranalyzed during the current study are available from the corresponding author upon reasonable request. Declarations Competing Interests The authors declare no competing interests. References 1. Lopes HA (2010) Trypanosomatids: odd organisms, devastating diseases. Open Parasitol J 4:30–59. https:// doi. org/ 10. 2174/ 18744 21401 00401 0030 2. Kaufer A, Ellis J, Stark D, Barratt J (2017) The evolution of trypanosomatid taxonomy. Parasit Vectors 10:287. https:// doi. org/ 10. 1186/ s13071- 017- 2204-7 3. Langridge DF, McGhee RB (1967) Crithidia mellificae n. sp. an acidophilic trypanosomatid of the honey bee Apis mellifera. J Protozool 14:485–487. https:// doi. org/ 10. 1111/j. 1550- 7408. 1967. tb020 33.x 4. Lipa JJ, Triggiani O (1988) Crithidia bombi sp. n. a flagellated parasite of a bumble-bee Bombus terrestris L. (Hymenoptera, Apidae) 27:287–290 5. Meeus I, De Graaf DC, Jans K, Smagghe G (2010) Multiplex PCR detection of slowly-evolving trypanosomatids and neogregarines in bumblebees using broad-range primers. J Appl Microbiol 109:107–115. https:// doi. org/ 10. 1111/j. 1365- 2672. 2009. 04635.x 6. Schmid-Hempel R, Tognazzo M (2010) Molecular divergence defines two distinct lineages of Crithidia bombi (Trypanosomatidae), parasites of bumblebees. J Eukaryot Microbiol 57:337–345. https:// doi. org/ 10. 1111/j. 1550- 7408. 2010. 00480.x 7. Runckel C, Flenniken ML, Engel JC etal (2011) Temporal analysis of the honey bee microbiome reveals four novel viruses and seasonal prevalence of known viruses, Nosema, and Crithidia. PLoS One 6:e20656. https:// doi. org/ 10. 1371/ journ al. pone. 00206 56 8. Schwarz RS, Bauchan GR, Murphy CA etal (2015) Characterization of two species of Trypanosomatidae from the honey bee Apis mellifera: Crithidia mellificae Langridge and McGhee, and Lotmaria passim n. gen., n. sp. J Eukaryot Microbiol 62:567–583. https:// doi. org/ 10. 1111/ jeu. 12209 9. Ravoet J, Maharramov J, Meeus I etal (2013) Comprehensive bee pathogen screening in Belgium reveals Crithidia mellificae as a new contributory factor to winter mortality. PLoS One 8:e72443. https:// doi. org/ 10. 1371/ journ al. pone. 00724 43 10. Morimoto T, Kojima Y, Yoshiyama M etal (2013) Molecular detection of protozoan parasites infecting Apis mellifera colonies in Japan. Environ Microbiol Rep 5:74–77. https:// doi. org/ 10. 1111/j. 1758- 2229. 2012. 00385.x 11. Yang B, Peng G, Li T, Kadowaki T (2013) Molecular and phylogenetic characterization of honey bee viruses, Nosema microsporidia, protozoan parasites, and parasitic mites in China. Ecol Evol 3:298–311. https:// doi. org/ 10. 1002/ ece3. 464 12. Cepero A, Ravoet J, Gómez-Moracho T etal (2014) Holistic screening of collapsing honey bee colonies in Spain: a case study. BMC Res Notes 7:649. https:// doi. org/ 10. 1186/ 1756- 0500-7- 649 13. Arismendi N, Bruna A, Zapata N, Vargas M (2016) PCR-specific detection of recently described Lotmaria passim (Trypanosomatidae) in Chilean apiaries. J Invertebr Pathol 134:1–5. https:// doi. org/ 10. 1016/j. jip. 2015. 12. 008 14. Stevanovic J, Schwarz RS, Vejnovic B etal (2016) Species-spe- cific diagnostics of Apis mellifera trypanosomatids: a nine-year survey (2007–2015) for trypanosomatids and microsporidians in Serbian honey bees. J Invertebr Pathol 139:6–11. https:// doi. org/ 10. 1016/j. jip. 2016. 07. 001 15. Hall RJ, Pragert H, Phiri BJ etal (2021) Apicultural practice and disease prevalence in Apis mellifera, New Zealand: a longitudinal study. J Apic Res 60:644–658. https:// doi. org/ 10. 1080/ 00218 839. 2021. 19364 22 16. Yamamoto S, Nakamura S, Nakayama K etal (2023) Molecular detection of Lotmaria passim in honeybees in Japan. Parasitol Int 93:102711. https:// doi. org/ 10. 1016/j. parint. 2022. 102711 17. McFadden AMJ, Mackay J, Borowick O, Goodwin RM (2016) Investigation of bee mortality in a beekeeper operation in the Coromandel district. Surveill 43:9–11 18. Waters TL (2018) The distribution and population dynamics of the honey bee pathogens Crithidia mellificae and Lotmaria passim in New Zealand. Dissertation, Victoria University of Wellington