Trypanosomatid parasites in Austrian mosquitoes

Trypanosomatid flagellates have not been studied in Austria in any detail. In this study, specific nested PCR, targeted on the ribosomal small subunit, was used to determine the occurrence and diversity of trypanosomatids in wild-caught mosquitoes sampled across Eastern Austria in the years 2014−2015. We collected a total of 29,975 mosquitoes of 19 species divided in 1680 pools. Of these, 298 (17.7%), representing 12 different mosquito species, were positive for trypanosomatid DNA. In total, seven trypanosomatid spp. were identified (three Trypanosoma, three Crithidia and one Herpetomonas species), with the highest parasite species diversity found in the mosquito host Coquillettidia richiardii. The most frequent parasite species belonged to the mammalian Trypanosoma theileri/cervi species complex (found in 105 pools; 6.3%). The avian species T. culicavium (found in 69 pools; 4.1%) was only detected in mosquitoes of the genus Culex, which corresponds to their preference for avian hosts. Monoxenous trypanosomatids of the genus Crithidia and Herpetomonas were found in 20 (1.3%) mosquito pools. One third (n = 98) of the trypanosomatid positive mosquito pools carried more than one parasite species. This is the first large scale study of trypanosomatid parasites in Austrian mosquitoes and our results are valuable in providing an overview of the diversity of these parasites in Austria.

There is a substantial support for the hypothesis that the dixenous life cycle emerged from the monoxenous one independently for representatives of the dixenous genera Trypanosoma, Leishmania, and Phytomonas [4,7,25,26]. Therefore, monoxenous trypanosomatids of mosquitoes and other bloodsucking insects can represent a crucial evolutionary link which is important for the elucidation of the emergence of a dixenous parasite life cycle.
Today, trypanosomatids are known primarily as important dixenous parasites of vertebrates, transmitted by various invertebrate vectors. Several species of the genus Trypanosoma cause serious and even life-threatening diseases in livestock [27,28] and two species, T. brucei s.l. [29,30] and T. cruzi [31,32], have a significant impact on human health. However, in the case of trypanosome infections, any serious impact on host health is rather an exception, and many trypanosome species occurring in wildlife and domestic animals may be considered as non-pathogenic parasites. Trypanosomes are common parasites of fish [33], birds, such as T. avium s.l. [16, [34][35][36][37][38], and of ungulates (especially domestic cattle), like the T. theileri complex [20, [39][40][41]. Both the T. avium and T. theileri species complexes are cosmopolitan and found worldwide [40,42,43].
Insect-borne trypanosomes found in Europe develop in the alimentary tract of bloodsucking insects [48] and are transmitted to vertebrates either by regurgitation of intestinal content [49], faecal matter deposited at the bite site [44] or by ingestion of the insect [47]. In the vertebrate hosts, these parasites can be found in blood [40,50], bone marrow [51] or inner organs [52]. In general, these trypanosomes are not regarded as pathogenic for the vertebrate hosts and T. theileri-like parasites and avian trypanosomes do not appear to overtly affect their hosts [40,50,52]. In the insect vectors, however, parasites can have a much larger impact, e.g. due to the blockage and destruction of the stomodeal valve facilitating the parasite transmission to the vertebrate host [49]. Host specificity of trypanosomes in vertebrates depends on the species and some have very broad host spectra like T. avium s.l., which has been found to infect a wide variety of bird orders and families [35,38,51].
Austrian mosquitoes have not been examined for trypanosomatid parasites before. We therefore screened female mosquitoes collected over two years in three Eastern Austrian provinces, namely Burgenland, Lower Austria and Vienna, to gain an overview which mosquitoborne trypanosomatids are present in the area, as well as to determine parasite diversity and prevalence in different mosquito species.

Material and methods
Trypanosomatid DNA for the study was obtained from mosquitoes sampled during a monitoring effort across the three provinces of Eastern Austria (Burgenland, Lower Austria, and Vienna) at 35 permanent and 25 non-permanent trapping sites. These sites were on public as well as private land, which was entered with the permission of the owners. Citizen Scientists in Lower Austria and Burgenland assisted with the sampling effort. At permanent sampling sites, mosquitoes were collected for a 24 hour time period on a regular basis every second week from April to October in 2014 and 2015, using BG-Sentinel traps (Biogents, Regensburg, Germany) equipped with bottled carbon dioxide (Air Liquide, Schwechat, Austria) as attractant. Non-permanent sampling sites were sampled at least once and up to six times over a 24 hour period during the summer months using CO 2 -baited BG-Sentinel traps as above or by hand aspirators. All mosquitoes were stored at −80˚C until further procedure.
Morphological identification of mosquito species was performed using the identification key of Becker et al. [53] and females were pooled by species, collection site and date, with a maximum number of 50 individuals. In 2014, three legs of each individual of Cx. pipiens s.l. / Cx. torrentium were taken and processed individually to identify the species/biotypes genetically in the frame of another project [54]. These mosquitoes were pooled after genetic identification which allowed us to determine the trypanosomatid parasite incidence in different biotypes of this species complex in more detail.
For amplifying trypanosomatid parasite DNA, each DNA sample was then subjected to nested PCR, described by [55] without modification. The used primers target a~2000 bp fragment of the ribosomal small subunit (SSU) gene. Obtained sequences were viewed and aligned using the software Geneious, version 10.0.6 [56]. Then the sequences were compared for similarity to sequences available on the GenBank 1 database. In the case of SSU rRNA gene sequence (Acc. No.: MG255960) of the most likely new Herpetomonas species (TR_SU106), a phylogenetic tree was constructed using all available sequences of Herpetomonas species retrieved from GenBank with Phytomonas spp. as an outgroup (Fig 1). Alignments for phylogenetic analysis were generated in Kalign [57]; the ambiguously aligned positions in the trimmed alignment were removed manually in BioEdit (Ibis Therapeutics, Carlsbad, US). The final dataset contained 46 taxa and 1,988 nucleotide positions. Analyses were done in MrBayes [58] and PhyML [59] with model optimization in ModelTest [60], version 3.06. A general time-reversible substitution model with a mixed model for amongsite rate variation (GTR + Γ + I) was chosen as the best fitting model of sequence evolution. Bootstrap analyses involved heuristic searches with 1,000 replicates (ML). Bayesian inference was accomplished in MrBayes 3.2.2 with analysis run for 5 million generations with covarion and sampling every 100 generations. Other parameters were left in their default states.

Minimum infection rate
To evaluate the infection rate of the collected mosquitoes, the minimum infection rate (MIR) of each mosquito species was calculated. If a mosquito pool was positive for trypanosomatid DNA, it was assumed that the pool contained at least one positive individual. Therefore, MIR (percentage) was calculated as follows: MIR ð%Þ ¼ n ðPCR positive poolsÞ =n ðtotal analysed mosquitoesÞ Â 100

Results
A total of 29,975 mosquitoes, belonging to 19 species and five genera, were collected in Vienna and Eastern Austria in the years 2014 and 2015 (S1 Fig). From these, 1680 pools were created using up to 50 mosquito females separated by species, time and site of sampling. A total of 298 (17.7%) mosquito pools were positive for trypanosomatid DNA (S1 Table). Of these, 243 pools (82.1%) belonged to 14 identified mosquito species and forma, whereas 53 pools (17.9%) were of unidentified individuals of the genera Aedes/Ochlerotatus, Culex, and Anopheles. Pools positive for trypanosomatids were of the following mosquito taxa: Ae

Trypanosomatid parasite diversity
The most common trypanosomatid species found in the tested mosquito pools were trypanosomes belonging to the Trypanosoma theileri/cervi complex (No. of positive pools = 105, which represents 35.5% of all positive pools and 6.3% of all tested pools) and T. culicavium (n = 69; 23.2% / 4.1%), followed by T. avium s.l. (n = 3; 1.0% / 0.2%) and monoxenous species belonging to the genus Crithidia and Herpetomonas (n = 20; 7.0% / 1.3%). A total of three dixenous and four monoxenous trypanosomatid species were identified by the analysis of their SSU (S1 Table and Fig 1). One third of the examined mosquito pools positive for trypanosomatid DNA (n = 98; 33.1%) carried more than one parasite species, as could be seen on the electropherogram where different peaks superimposed on each other. The mosquito species with the highest diversity of different trypanosomatid parasites was Cq. richiardii, in which we found the T. theileri complex and all four detected species of monoxenous trypanosomatids (S1 Table). The trypanosome species T. culicavium was only detected in mosquitoes of the genus Culex and the species T. avium s.l. was found only in 2014. Since the whole bodies of mosquitoes were used in pools, and no dissection and microscopy was performed, it was not possible to assert competent vector status of trypanosomes on the sampled mosquito species.

Trypanosomatid prevalence
In 2014, a total of 10,575 individual mosquitoes, consisting of 830 pools, were collected. Of these, 110 pools (13.3%) of six identified mosquito species were positive for trypanosomatids (S2 Table). The most commonly collected mosquito in 2014 was Ae. vexans with 4420 individuals (41.2%), this mosquito also yielded the highest number of positive pools (n = 33) and the second highest prevalence in the identified mosquito species (21.7%; S2 Table). The highest prevalence of trypanosomatids (23.8%) was found in Oc. sticticus.

Trypanosomatid prevalence expressed by minimum infection rate (MIR)
The minimum infection rate varied between the mosquito species and between the years (S2 Table and Table 1, Figs 2 and 3). The average total MIR (both years and all sampling events) was 0.99%, with the highest prevalence for Oc. geniculatus (3.9%), An. maculipennis (1.8%), and An. plumbeus (1.7%) (S1 Table). The overall highest MIR was found in Cx. martinii (3.0%) in 2014 and Oc. geniculatus (16.7%) in 2015 (S2 Table); however the calculated prevalence could be overestimated due to generally low number of tested pools. In 2014, when morphologically undistinguishable mosquitoes of the Cx. pipiens complex and Cx. torrentium were identified genetically, it was possible to determine the MIR in the different biotypes comprising this complex (Table 1). Here, Cx. torrentium presented with the highest MIR (5.4%), followed by Cx. pipiens f. pipiens/molestus hybrids (1.3%) and Cx. pipiens f. pipiens (1.1%). No parasite DNA was found in Cx. pipiens f. molestus.

Monthly changes and differences between 2014 and 2015
The total number and proportion of positive pools in 2014 and 2015 (as well as MIR) was highest in early and mid-summer in both years and became gradually less towards autumn (Fig 2). This is also evident for T. theileri/cervi, where the numbers (2014) and pool positivity (2015) was highest in June and was gradually tapering off towards October. The trypanosome species T. avium was only found in 2014, and that year, only in late spring and early summer (May/ June). Compared to mammalian trypanosome T. theileri/cervi, avian T. culicavium showed the opposite trend; total numbers, pool positivity and MIR increase towards late summer and were highest in August in both years (Figs 2 and 3).

Discussion
This is the first large scale study of trypanosomatid flagellates in mosquitoes with emphasis on Austria. Our results are of special interest, because the used genetic identification of the Cx. pipiens complex and Cx. torrentium mosquito species enabled the determination of trypanosomatid parasites in the morphologically undistinguishable species, biotypes and their hybrids of this complex for the first time. Since we used pools of whole body insects and did not perform microdissections, microscopy or experimental infections, we cannot assert any vector competence and/or host specificity in any of the examined mosquitoes for the detected trypanosome species. However, our results are non-the-less valuable in providing an overview of the dixenous as well as monoxenous trypanosomatid species present in Central Europe.

Trypanosomatid diversity and prevalence
The trypanosomatid parasites we found in mosquitoes belonged to three trypanosome species (T. theileri complex, T. culicavium, and less frequently T. avium s.l.) and four monoxenous insect species, three of the genus Crithidia (C. fasciculata and C. brevis were the most frequent, whereas C. pragensis was found in one pool only) and one of the genus Herpetomonas (found in one mosquito pool only). In the previous study performed in neighboring Czechia [16], different bloodsucking dipterans (Culex spp., Simuliidae, and Hippoboscidae) were examined for trypanosomatids. T. culicavium, T. avium, and Crithidia brevicula were detected in both Cx. pipiens and Cx. modestus, with overall trypanosomatid prevalence 8.2% and 5.1% in Cx. pipiens and Cx. modestus, respectively. In Czechia, the prevalence of Trypanosoma culicavium in Cx. pipiens s.l. and Cx. modestus varies between 0.3% and 5.4% and between 0.05% and 1.4%, respectively [16,47]. Similar prevalence of T. culicavium was detected in Culex mosquitoes during our study, ranging from 1 to 3%. Despite the fact that we examined a wider range of mosquito species and a larger amount of individuals than these previous studies in Czechia [16,47], we only found a small number of T. avium s.l. positives (only in May and June 2014). Trypanosoma avium s.l. is a common parasite in various avian orders worldwide and the prevalence in birds in Europe ranges between 1 to 87.2% [43,38,[61][62][63]. Although these parasites have not been studied in Austria in any detail before, a similar range of prevalence would be expected for local birds and the relatively low prevalence we observed in the mosquitoes is surprising. The habitat where the sampling takes place has a great impact on mosquito diversity [64] and the parasites they carry, which is influenced by the available vertebrate host species [16], it is therefore possible that our sampling locations had only low numbers of bird species carrying T. avium sensu lato.
The parasites T. theileri, T. cervi, T. cf. cervi belong to a complex of species which cannot be resolved using the SSU gene and it is therefore not possible to determine the exact taxonomy of the parasites belonging to this complex found in this study. Further research on the collected material will take more genes into account to resolve this ambiguity.
Aedes vexans is a mammalophilic mosquito, and previous studies have shown that wild game animals like red (Cervus elaphus) and roe deer (Capreolus capreolus) are commonly bitten [65,66]. In a study in Switzerland, Ae. vexans blood meals taken from wild game animals were the second most common after blood meals from cattle, with 18.25% of all examined blood meals from red deer, and 5.1% from roe deer [66]. Börstler et al. [65] reported a very similar result for Ae. vexans in their study on host preferences of different mosquito species in Mixes consisted of ( §) T. culicavium dominant, with unidentified smaller peaks on electropherogram (n = 6); C. brevicula/fasciculata dominant with unidentified smaller peak on electropherogram (n = 1); unidentified mix (n = 1) (+) mix Crithidia sp. possibly C. pragensis (n = 1) (x) unidentified mix, unable to BLAST (n = 1). Germany. The information concerning the occurrence of Trypanosoma theileri/cervi in wild animals, especially cervids, in Central Europe is very limited. The presence of these parasites in Germany and Poland [67,68] is supported by two trypanosome sequences available in Gen-Bank and obtained from a red deer (Cervus elaphus) and fallow deer (Dama dama) sampled in Poland. Our frequent findings of T. theileri/cervi in Austrian mosquitoes (preferably in the genera Aedes/Ochlerotatus and Coquillettidia) is no evidence for the involvement of these mosquitoes in the transmission cycle of the parasite; on the other hand, it proves the abundance of the trypanosomes in the vertebrate hosts (probably game ungulates) in the studied areas. Monoxenous trypanosomatids infect a broad range of insects, including those of the order Diptera. Due to their limited impact on human and animal health, monoxenous trypanosomatids have received only little attention. Based on PCR screening, three species of the genus Crithidia, common parasites of the insect alimentary canal, were detected in mosquito pools. Whereas C. fasciculata, a well-known laboratory model, infects many mosquito species [69], Crithidia brevicula is known mainly from true heteropteran bugs [66,70]. In Czechia, however, the parasite was found in Culex mosquitoes [16]. While the two previous species have been found in mosquitoes repeatedly, the third Crithidia species, C. pragensis, was found only in one pool of Cq. richiardii. The parasite species was recently described in neighboring Czechia [71] from a brachyceran fly Cordilura albipes (Scatophagidae) and our finding therefore extends the possible host spectrum and area of the distribution.
We did not find any Paratrypanosoma parasites, repeatedly reported from mosquitoes in the neighboring Czechia [16,25], but also in the USA [72]. However, in one pool of Cq. richiardii we have found an unknown species of Herpetomonas. This parasite genus is predominately found in dipterans, mainly in brachyceran flies [65], however several studies demonstrated the occurrence of Herpetomonas parasites in blood sucking nematoceran insects, specifically in biting midges [17,18].
One third of all positive mosquito pools examined in this study carried a mix of either two or more different trypanosomatid species. Detecting mixes in this study is a byproduct of examining pools of mosquitoes instead of looking at individuals, although the presence of several species, visible as double peaks on the chromatogram, has been found in other studies examining other haematozoa (avian malaria parasites) even in single mosquitoes [73].

Seasonal changes
During both years, the total trypanosomatid pool prevalence (%) and MIR was highest in the mid of summer and decreased towards autumn. Differences between the years can be explained by climatic differences, since the period April to September in the year 2014 was on average cooler with more precipitation than the same period in 2015. This had an impact on which mosquito species and the numbers of individuals were caught, which was reported in a previous paper [64]. The differences in seasonality between the two dominant trypanosome species are more remarkable. Compared to mammalian T. theileri/cervi, the total numbers, prevalence (%) and MIR of avian T. culicavium appeared to increase towards autumn. These noticeable differences can be explained by the different host and vector preferences of both mentioned trypanosomes. While avian T. culicavium develops in mosquitoes of the genus Culex, mammalian T. theileri/cervi is found mainly in mosquitoes of the genera Aedes/Ochlerotatus and Coquillettidia. Unlike monoxenous trypanosomatids, dixenous trypanosomes infect mosquitoes when sucking blood, and the different behavior and seasonality of various mosquito species/genera may also result in different seasonality and occurrence of transmitted parasites.
Trypanosoma avium was only detected in two months (May and June) in 2014, and total numbers, prevalence (%) and MIR were higher in May. It is known that temperature has an impact on the development of trypanosomes in invertebrate hosts. Experiments performed on T. avium in Ae. aegypti mosquitoes showed that higher temperatures were detrimental for parasite development and the optimal temperature was around 20˚C. This might be the reason we only observed these parasites during late spring/early summer in 2014. The temperature requirements might be similar for the development of T. theileri/cervi, although no studies have been performed on this parasite and the only report of seasonal changes in prevalence of T. theileri in the Northern hemisphere noted an increase in the infection rate of domestic cattle in the state of New York from May to September [74]. In contrast to our detected vectorborne trypanosomes, the monoxenous Crithidia spp. appeared evenly distributed over the year, probably due to the horizontal transmission between mosquito hosts via contamination of sugar food sources by parasites.

Trypanosomatid parasites in mosquitoes of the Cx. pipiens s.l. / Cx. torrentium
The second most common mosquitoes caught in this study were species belonging to the morphologically indistinguishable Cx. pipiens s.l. and Cx. torrentium. During the previous study, these taxa, sampled in 2014, were identified genetically [54] and this provided us with an opportunity to determine trypanosomatid diversity and prevalence in these Culex mosquitoes. The most common trypanosome species we detected in this species complex was T. culicavium. On the other hand, Trypanosoma avium s.l., Crithidia brevicula and C. fasciculata were found only in Cx. pipiens f. pipiens. During our sampling, the most common mosquito of this species group caught was Cx. pipiens f. pipiens and subsequently, the largest total number of trypanosomatids as well as the largest proportion of positive pools was found in this biotype. However, when comparing the MIR of the different biotypes and the hybrids in the species complex, differences are evident. Culex torrentium showed the highest MIR, followed by the Cx. pipiens f. pipiens/molestus hybrids, while the MIR for Cx. pipiens was lowest and no trypanosomatids were detected in Cx. pipiens f. molestus. It is unclear if these differences could be explained by the much lower sample size of Cx. pipiens f. molestus, Cx. torrentium and hybrids or if these mosquitoes in general bite birds infected with T. culicavium more frequently and therefore have a higher chance of acquiring these parasites.