Characterization of Plasmodium relictum, a cosmopolitan agent of avian malaria

To detect and isolate the Plasmodium parasite strain, the blood films from each captured bird were quickly examined microscopically in the field, as previously described [35]. One naturally infected Common chiffchaff was detected, with parasitaemia of 0.1%. Ten blood films were prepared for microscopic examination from this bird. Additionally, blood was also collected in heparinized microcapillaries and used to expose two uninfected domestic Common canaries Serinus canaria forma domestica by sub-inoculation into their pectoral muscle of about 250 μl of a freshly prepared mixture of infected blood, 3.7% sodium citrate (anticoagulant) and 0.9% saline (4:1:5) [36].

Parasitaemia developed in both exposed canaries, and blood of these birds was passaged as described above in three additional canaries. Two Zebra finches Taeniopygia guttate, one Budgerigar Melopsittacus undulatus and two European goldfinches Carduelis carduelis were also exposed, as described above. Six uninfected canaries were used as controls. Blood of all control and experimental birds was tested by microscopic examination and PCR-based methods (see description below) twice before the experiment to ensure that they were uninfected with malaria parasites.

Two canaries were observed for 57 and 94 days post exposure (dpe) and then euthanized for histology and chromogenic in situ hybridization research. Two European goldfinches were observed for 127 dpe. All remaining birds were observed for 131 dpe. Post-exposure blood samples were taken for microscopic examination and PCR-based testing as described above once every 4 days during the first post-exposure month, once every week during the second month and once every 1–2 weeks during the remaining experiment time. All experimental and control birds were kept indoors in a vector-free room under natural light–dark photoperiod. They were fed a standard diet for seed-eating bird species.

Control birds were maintained in the laboratory for further research. The donor Common chiffchaff and infected experimental birds were euthanized, and pieces of organs (brain, liver, lungs, heart, kidneys, spleen) and pieces of pectoral muscles were fixed in 10% neutral formalin, embedded in paraffin and processed using traditional histologic methods [7]. Histological sections of 4 μm were obtained, stained with haematoxylin–eosin, mounted in BioMount (BioGnost, Croatia) and examined microscopically. One smear of bone marrow was prepared from the tibia bone of each bird, air dried, fixed with absolute methanol and stained with Giemsa.

Laboratory-reared Culex quinquefasciatus mosquitoes were maintained and exposed to canaries infected with the isolated Plasmodium sp., as previously described [32]. Briefly, insects were kept in cages (45 × 45 × 45 cm) under 65–70% relative humidity, 16/8 h light/dark photoperiod and 26 ± 1 °C. One experimentally infected canary was used as the donor of parasites for infecting mosquitoes. Eleven female mosquitoes took blood meals on this canary. Parasitaemia was 0.02% with few visible mature gametocytes. Preparations of midgut contents (6 insects were dissected 24–48 h post exposure), one preparation of midgut wall (one insect on 12 dpe) and 4 preparations of salivary glands (on 15, 16 and 18 dpe) were prepared and examined, as previously described [32].

Detailed microscopic analysis was carried out with various Olympus light microscopes equipped with Olympus digital cameras and imaging software. Preparations of blood stages of the lineages pSGS1, pGRW11, pGRW4, and pLZFUS01 were from collections of voucher specimens which have been deposited at P. B. Šivickis Laboratory of Parasitology, Nature Research Centre Vilnius. These were blood films from canaries whose were exposed experimentally to the parasite lineages pSGS1 (parasitaemia varied between 0.6 and 1.8%, preparation accession number 48979–48981 NS), pGRW11 (1.1–6%, 48982–48984 NS), pGRW4 (0.2–2.1%, 48985–48987 NS), and the lineage pLZFUS01 (0.5%, 48694–48696 NS) from the blood of a naturally infected Red-backed shrike Lanius collurio (for exposure description see [30]). Additionally, preparations with the Hawaiian isolate of P. relictum (pGRW4) were used. These were (1) 12 blood films from two individual canaries that were exposed experimentally by inoculation of infected blood (parasitaemia varies between 0.6 and 10%, accession nos. 48988–48999 NS) (for infection details see [5]), (2) 11 blood films from one naturally infected Apapane Himatione sanguinea (parasitaemia 22%, accession nos. 49000–49010 NS).

Blood films from each infected bird were examined and the observed blood stages were morphologically compared by skilled parasitologists of avian malaria parasites at the P. B. Šivickis Laboratory of Parasitology. At least 100 fields were studied at high magnification (1000×) in each preparation. Intensity of parasitaemia was estimated as a percentage by actual counting of the number of parasites per 10,000 erythrocytes. The morphometric features studied (Table 1) were those defined in [7]. The analyses were carried out using the ‘Statistica 7’ package as previously described [7].

Chromogenic in situ hybridization (ISH) was applied to increase detectability of tissue stages of the parasites. Organs (the same as for histological examination) from one naturally infected Common chiffchaff and two experimentally infected canaries (57 and 94 dpe, respectively) were tested using a previously described ISH protocol [37]. 3 μm paraffin-embedded tissue sections of all these organs were prepared. The sections were deparaffinized, subjected to proteolytic treatment with proteinase K (Roche, Basel, Switzerland) 6 μg/ml in Tris-buffered saline at 37 °C for 50 min. For hybridization, the slides with tissue sections were incubated overnight at 40 °C with hybridization mixture and a final probe concentration of 100 ng/ml. The used oligonucleotide probe (sequence: 5′-TTTAATAACTCGTTATATATATCAGTGTAGCAC-3′) was labelled with digoxigenin at the 3′ end (Eurofins MWG Operon, Ebersberg, Germany). This probe is specific to detect avian Plasmodium parasites [6, 37]. The digoxigenin-labelled hybrids were detected by incubating the slides with anti-digoxigenin-AP Fab fragments (Roche) (1:200) for 1 h at room temperature (RT). Visualization of the reaction was carried out using the colour substrates 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and 4-nitro blue tetrazolium chloride (NBT) (Roche). A positive control (sections of a lung of a Eurasian blackbird Turdus merula naturally infected with Plasmodium vaughani, which was proven to be positive by previous ISH) was used to assure that the protocol worked. Preparations were examined microscopically by skilled parasitologists and pathologists; at least 50 fields of each preparation were studied at low magnification (400×), and then each preparation was examined for 10–15 min at high magnification (1000×).

Total DNA was extracted from blood samples using an ammonium-acetate precipitation protocol [38]. Polymerase chain reaction using the primer set HaemNFI/NR3 and Haem/R2 was performed in order to amplify a 479 bp sequence of the parasite’s cytb gene [18, 39]. The total volume of the reaction mix for each sample was 25 µl, which included 12.5 µl of DreamTaq Master Mix (Thermo Fisher Scientific, Lithuania), 8.5 µl nuclease-free water, 1 µl of each primer and ~ 50 ng of a total genomic DNA template (2 µl). Negative controls (nuclease-free water) were used after each seven samples to detect possible false amplifications, and one positive control (extracted parasite DNA from a blood sample, which was confirmed positive during previous PCR testing) was used to evaluate the success of PCR if none of the samples would have been amplified.

Temperatures for the PCR were as described in the original protocols. The success of the performed PCR was evaluated by running electrophoresis on a 2% agarose gel. Successfully amplified DNA was precipitated using 11 µl of 8 M NH 4Ac, 37 µl of 96% and 150 µl of 70% ethanol. After centrifugation, the supernatant was discarded, the samples were air-dried overnight, and then 16 µl of nuclease-free water was added on the precipitated DNA. Big Dye Terminator V3.1 Cycle Sequencing Kit and ABI PRISM™ 3100 capillary sequencing robot (Applied Biosystems, Foster City, California) were used for sequencing. Sequences were edited and aligned using BioEdit software [40]. Absence of double-base calling in sequence electropherograms was used as an indication of single infections [41]. Nucleotide BLAST (megablast algorithm) (http://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi) was used to compare our amplified sequences with sequences deposited in the GenBank.

Molecular phylogenetic analysis was carried out using Bayesian and Maximum Likelihood algorithms. Sequences for the phylogenetic analysis were collected from GenBank and double-checked in MalAvi database [19]. Plasmodium falciparum was used as an outgroup. GenBank accession numbers and codes of the lineages are provided in the phylogenetic trees (Fig. 1). Bayesian phylogenetic tree (Fig. 1a) was constructed using MrBayes version 3.1 [42] software. The General Time Reversible Model (GTR) was used as suggested by the software MrModeltest 2.2 (https://​github.​com/​nylander/​MrModeltest2). Analysis was run for a total of 10 million generations with a sampling frequency of every 100 generations. Before the construction of the consensus tree, 25% of the initial trees were discarded as ‘burn in’ period. The tree was visualized using the software FigTree v1.4.3 (http://​tree.​bio.​ed.​ac.​uk/​software/​figtree/​). Maximum Likelihood tree (Fig. 1b) was constructed using the MEGA 7.0 [43] software; it was performed with 1,000 bootstrap replications using the GTR model and the same dataset as during the Bayesian analysis.

The new sequence of lineage pPHCOL01 was deposited in GenBank (accession MG724747). Genetic differences between different lineages of P. relictum were calculated using the Jukes–Cantor model of substitution, as implemented in the programme MEGA 7.0 [43].

Single infections of P. relictum (cytb lineage pPHCOL01) was identified in the donor Common chiffchaff both by microscopic examination of blood films and PCR-based amplification and sequencing. All exposed canaries were susceptible and developed a single infection with the same malaria parasite, as determined both by microscopic examination of blood films and PCR-based testing. Parasitaemia developed in one exposed European goldfinch. Two Zebra finches, one Budgerigar and one European goldfinch were resistant. All control canaries remained non-infected during this study.

The reported lineage of P. relictum (pPHCOL01) was new. It clustered with other morphologically characterized lineages of P. relictum (pSGS1, pGRW4, pGRW11, pLZFUS01) in both phylogenetic analyses, supporting the close phylogenetic relationships among them (Fig. 1a, b). Genetic differences among five lineages of P. relictum varied between 0.2% (minimum, the lineages pSGS1 and pGRW11) and 3% (maximum, the lineages pSGS1 and pLZFUS01).

See (Fig. 2, Table 1).

Examination of all blood films with the P. relictum lineages pSGS1, pGRW4, pGRW11, pLZFUS01 (Fig. 3) revealed the morphological identity of trophozoites, meronts and macro- and microgametocytes of these parasites in all infections that were examined. Number of merozoites in mature erythrocytic meronts of all parasite lineages and different isolates of the same lineage is markedly variable during development in the same and different species of avian hosts; it varied between 10 and 32 merozoites, but most often reported to be between 12 and 24 merozoites in all examined infections. These lineages of P. relictum cannot be distinguished based on this character. Additionally, the main morphological forms of blood stages reported in parasites of the new lineage pPHCOL01 (Fig. 2) were seen in blood films with single infection of all other lineage of P. relictum in the same and different species of avian hosts (Fig. 3). Variation in shape of each blood stage of P. relictum occurs, but all observed morphological forms of blood stages (Figs. 2a–y, 3a–x) were seen in parasites belonging to each examined parasite lineage. In other words, the morphological forms of all blood stages (trophozoites, growing and mature meronts, growing and mature gametocytes) in all examined P. relictum lineages were indistinguishable.

Interestingly, two different isolates of the lineage GRW4 (the Hawaiian strain and the strain isolated in Europe) produced indistinguishable trophozoites, meronts and gametocytes during development in canaries. Additionally, extensive microscopic examination showed that morphological and morphometric characters of blood stages of the widespread lineages pGRW4 and pSGS1 were variable during development in same and different avian hosts, and they markedly overlapped among these lineages. In other words, blood stages of the lineages pSGS1 and pGRW4 were indistinguishable from each other during their development in canary and other avian hosts (see Additional file 1: Figure S1, Additional file 2: Figure S2, Additional file 3: Figure S3, Additional file 4: Figure S4).

Exo-erythrocytic stages were readily visible in a positive control, assuring that the ISH protocol worked. Microscopic examination of the histological sections stained with H&E and the same organ sections treated for ISH did not reveal tissue stages of the parasite lineage pPHCOL01.

This study demonstrate that the new rare lineage pPHCOL01 can be linked to P. relictum on both morphological and molecular grounds and provide new data about specificity and development of this infection in experimentally infected avian hosts. This is the first study to compare morphology of blood stages of different lineages of P. relictum using the same methodology. Parasites of all examined lineages are typical representatives of sub-genus Haemamoeba, whose inclusive species produce large erythrocytic meronts and gametocytes, both of which markedly influence host cell nuclear position (Figs. 2, 3). Morphological forms of blood stages of the parasite lineage pPHCOL01 found in Common chiffchaff, canary and European goldfinch were indistinguishable morphologically. Extensive comparison of blood stages of other P. relictum lineages gave the same results (Figs. 2, 3; Additional file 1: Figure S1, Additional file 2: Figure S2, Additional file 3: Figure S3, Additional file 4: Figure S4). Morphological characters, which might be used for distinguishing different lineages of P. relictum, were not found because of marked variability of these features during each single infection in all parasite lineages. These data are in accordance with former morphological observations on blood stages of the lineages pSGS1, pGRW4 and pLZFUS01 accessed during experimental exposure of different avian hosts [28‐30]. The lineages pSGS1, pGRW4, pGRW11, pLZFUS01, and pPHCOLL01 of P. relictum belong to the same P. relictum morphotype. Interestingly, the same is true for sporogonic stages of the lineages pSGS1, pGRW11 and pGRW4, which complete development in Culex pipiens forma molestus mosquitoes synchronously and produce morphologically indistinguishable ookinetes, oocysts and sporozoites at same conditions [32, 33, 46].

None of the bird species that were experimentally infected with lineage pPHCOL01 was good host for investigating dynamics of parasitaemia, and they cannot be recommended for experimental research aimed at studying blood stage infections. Canaries were susceptible, but parasitaemia was transient and light. Zebra finches and one Budgerigar were resistant. Interestingly, the available field observations indicate that the latter two avian species are likely resistant to other P. relictum lineages as well. Zebra finches and Budgerigars have never been reported as host of Plasmodium parasites by microscopic examination of blood films (this method provides opportunities to visualize blood stages), and probably might resist or tolerate many species of malaria parasites [7, 16, 17]. It is worth noting that Baron et al. [47] reported the lineage pGRW4 in New Zealand budgerigars, indicating that these birds were exposed naturally, but provided no information about whether this parasite completed its life cycle and produced erythrocytic infection in this host species. Development of this parasite might be abortive in Budgerigars, as is the case in many haemosporidian infections [48]. Thus, both Zebra finches and Budgerigars might be excellent model hosts for better understanding mechanisms of innate resistance during avian malaria.

Molecular techniques that amplify parasite cytb genes provide new opportunities to readily distinguish genetically different isolates of P. relictum and to identify infections caused by these parasites in avian hosts. This was impossible during the pre-molecular era of malaria research. Numerous molecular studies reported P. relictum in naturally infected birds [19, 23, 24, 34, 49], resulting in a solid body of information about the occurrence of these parasite lineages in various avian hosts and ecosystems all over the world (Table 2). However, comparative research on development and virulence of P. relictum lineages in different avian hosts and vectors has lagged behind and remains uncommon. This missing information is an obstacle to developing a better understanding of the biological properties of infections caused by different P. relictum lineages, limits the ability to predict disease outbreaks, and makes it more difficult to develop adequate steps for improving bird health and conservation.

Experimental research is essential for better understanding the biology of malaria parasites [5, 8, 11, 33, 50‐55]. Controlled experimental studies with P. relictum are relatively easy to design due to availability of laboratory-friendly experimental vertebrate hosts (canaries and some species of other common birds), laboratory-colonized susceptible mosquitoes (species of the Culex pipiens complex) and worldwide high prevalence in many wild bird species (donors of natural infections). This makes P. relictum a convenient and even unique model organism to approach numerous questions about mechanisms of host-parasite interactions, including the immunological aspects during malaria infections [56‐58], the ecology and evolution of host-parasite associations [25, 59‐63], the host adaptations to tolerate malaria infections [10, 31, 47, 64, 65], patterns of mosquito transmission [32, 46, 53, 66‐68] and many other questions.

Unfortunately, experimental information about different lineages of P. relictum is still limited, but available data indicate that different lineages and even different isolates of the same lineage might differ remarkably in their ability to develop in different avian hosts and in other biological properties [11]. Brief review what is known about this biological variation is given in the following sections.

The pathology of known lineages of P. relictum is highly variable in host species or incompletely known. For example, the same P. relictum lineage might cause severe disease in one species of avian host, but other bird species might be tolerant or even resistant [5, 8, 50, 69]. Experimental observations show that the same isolate of pSGS1 behave markedly differently in different species of birds, with the susceptibility ranging from complete resistance to light subclinical (< 0.1%) and high (> 10%) parasitaemia [8, 69]. The variation in parasitaemia dynamics and maximum intensity are often great in different individuals of the same bird species infected with pSGS1 parasite [55]. Similarly, the susceptibility of same bird species to different isolates of the same P. relictum lineage also might be markedly different. For example, Hawaiian isolates of pGRW4 readily infect canaries, with maximum parasitaemia ranging from light (about 0.1%) to high (up to 30% and greater) reported in birds exposed by inoculation of infected blood ([70], CTA, pers comm.). However, this bird species was either resistant or had mainly light (< 0.1%) and transient parasitaemia, which rapidly turned to chronic or even latent stages of infection after exposure to European isolates of the same parasite lineage by the same mode of infection ([11], GV, unpublished observation).

It remains unclear why different geographical isolates of the same lineage of P. relictum (pGRW4) behave so differently in the same species of birds. The differences between different geographic isolates of P. relictum lineages might be due to different clonal intra-lineage genetic diversity, which is great in Hawaiian strains of the lineage pGRW4, but remains insufficiently documented in European isolates of the same lineage [21, 31]. Marked variation in the susceptibility of same experimental bird species to different parasite lineages provide opportunities to use this host-parasite model system for comparative research aimed at a better understanding of the genetic mechanisms of tolerance and virulence during parasitic infections.

Without question, the lineages pSGS1 and pGRW4 are virulent in birds and can cause marked blood pathology and even mortality in susceptible hosts [5, 8, 11, 29, 50, 69]. The negative effects of P. relictum (pSGS1) on bird physiological parameters and behaviour are documented due to delicate experimental studies [54, 55]. Observations of infected, naive birds in zoos and rehabilitation centres provided evidence of the severity of disease caused by these and related parasite lineages in wild birds [71‐74]. These studies are the basis of understanding the predictions and conclusions of field observations about negative influence of P. relictum on population decline or even extinction, particularly on oceanic islands [63, 75‐78]. However, to evaluate the true virulence of a malaria parasite lineage in certain avian host species, experimental and field observations are needed, ideally in each targeting host-parasite system separately.

Even though there are numerous reports of exo-erythrocytic stages of P. relictum from the pre-molecular research era [1, 7, 13, 84], information about these stages and associated tissue pathology in avian hosts is still absent for parasites of all lineages of P. relictum. This is an obstacle to understanding of the mechanisms of persistence in birds, as well as, the association between tissue merogony and pathogenicity caused by different parasite lineages in different avian hosts. This study shows that exo-erythrocytic stages of P. relictum can be difficult to find during chronic infections even in experimentally infected birds with visible parasitaemia. This indicates that large multinuclear tissue stages, which are easy to see under light microscopy [6, 13], might persist for a short time and their development might be markedly dependent on the stage of infection. Application of in situ hybridization methods is promising in the investigation of tissue merogony of haemosporidians [6, 37, 78], but may not be sensitive enough to detect uninuclear hypnozoite-like intracellular stages should they occur in P. relictum, as is the case in human Plasmodium vivax infection. This suggests application of more sensitive immunofluorescent diagnostic techniques in parallel with traditional histology and in situ hybridization methods in research of exo-erythrocytic development of different lineage parasites [1, 6, 35, 37, 78].

Observation of parasites in blood films and determination of morphological characters of their blood stages remain important not only in identification of haemosporidian species [11, 27, 79], but also for distinguishing competent and abortive haemosporidian infections, which might have different consequences for the bird health. During abortive infections, the parasites might circulate within avian hosts as sporozoites or even undergo partial development within non-erythroid tissues, providing templates for PCR amplification, but the parasite would not be able to complete its life cycle due to an inability to enter red blood cells. This would result in absence of gametocytes and other blood stages in the circulation, but severe disease might occur due to damage of internal organs [48]. In the latter case, a positive PCR signal might be obtained, but parasitaemia would be absent or barely detectable due to difficulties in microscopic detection of remnants of tissue stages in the circulation [80‐82]. This highlights the relevance of microscopic detection of blood stages and knowledge about morphological features of haemosporidians in pathology and epidemiological studies when used in parallel with molecular diagnostic tools.

Longevity of the prepatent period cannot be used for distinguishing infections caused by different P. relictum lineages. Duration of the prepatent period following sporozoite-induced infection of different lineages of P. relictum remains largely undetermined. Prepatent periods have been observed in the Hawaiian parasite lineages pGRW4 where it was within 4 dpe in Iiwi Drepanis coccinea and 8 dpe in Hawaii Amakihi Chlorodrepanis virens [5, 50]. The prepatent period was about 5 dpe after sporozoite-induced infection of unknown lineage of P. relictum in canaries [83, 84].

This study demonstrated that prepatent period of infection is markedly variable in different bird species and individuals of the same species during blood-induced infection of the lineage pPHCOL01. The prepatent period is often about 1 week after the blood-induced infections of pSGS1, but varies markedly in different species of avian hosts and even individuals of the same species even after the same mode and dose of infection, and it might be as long as several weeks after infected blood-induced exposure, indicating the possibility of parasite persistence in internal organs [7, 8, 13, 69, this study].

In all investigated lineages of P. relictum, parasitaemia was asynchronous, with trophozoites, growing and mature meronts as well as gametocytes present in the same blood films at the same time in all species of exposed birds at any stage of parasitaemia [8, 29, 30, 33, 70, this study]. This provides opportunities to design vector research with all lineages at any stage of parasitaemia using susceptible avian hosts as donors of infections to expose mosquitoes, but all work carried out to date with different lineages has failed to demonstrate significant differences.

However, infectivity and patterns of development of different lineages and even different isolates of the same lineage might be different, sometimes significantly in canary [11, 70]. A moderate to high (> 0.1% and greater) long-lasting (several months) parasitaemia usually develops during infections with lineages pSGS1 and pGRW11 in canaries exposed by inoculation of infected blood [22, 32, 46]. The same is true for the parasite lineage pGRW4 during development in canaries, but not for all its isolates. For example, the Hawaiian and European isolates of the lineage pGRW4 develop differently in canaries. Hawaiian pGRW4 isolates develop naturally in canaries when caged birds are exposed in habitats with active natural transmission and can develop high (up to 30% and higher) long-lasting parasitaemia after sub-inoculation of infected blood, although significant individual variation is present ([70], CTA, unpublished data). Attempts to induce a long-lasting parasitaemia (several weeks or longer) and gametocytaemia exceeding 0.01% with European isolates of lineage pGRW4 were either completely unsuccessful (compete resistance was recognized in nine exposed birds) or only partially successful with extremely light transient parasitaemia (few gametocytes reported after examination of 100 microscopic fields at high magnification in four birds) ([11], GV, unpublished data). In other words, the canary is not a good host for experimental studies of erythrocytic infections with the European isolates of the lineage pGRW4, but can be used in experiments with the Hawaiian isolate. Experimental studies with other geographical isolates of P. relictum (pGRW4) infection have not been performed. Due to relative resistance of canaries to European isolates of lineage pGRW4, Eurasian siskin Carduelis spinus has been used in experiments with this parasite lineage, and this species is an excellent experimental host [33].

The lineages pSGS1, pGRW4, pGRW11, pLZFUS01, and pPHCOL01 of P. relictum are closely related based on similarities in cytb sequence (Fig. 1) and cannot be distinguished by morphology (Figs. 2, 3, Additional file 1: Figure S1, Additional file 2: Figure S2, Additional file 3: Figure S3, Additional file 4: Figure S4). Do these lineages represent distinct species of the P. relictum group or are they different genetic variants of the same morpho-species? Do parasites of these lineages maintain the ability to mate? Does the available information provide opportunities to approach answering these questions? This study and available experimental observations [28‐30, 32, 33, 46] show that morphological data both of blood and vector stages cannot help in distinguishing parasites of the lineages pSGS1, pGRW4, pGRW11, pLZFUS01, pPHCOL01, indicating that they might belong to the same P. relictum morphotype, but some of them also might represent cryptic species of the P. relictum group.

Between-lineage hybridization experiments provide opportunities to obtain direct information about the possibility that different lineages of haemosporidian parasites can mate and exchange genetic information. Sexual processes and between-lineage hybridization of Haemoproteus parasites (sister genus to Plasmodium) can be readily induced in vitro [7]. These experiments indicate probable development of between-lineage Haemoproteus parasite hybrids in vitro, which can be readily distinguished morphologically on ookinete stage, but genetic information is lacking, primarily due to obstacles in accessing nuclear genetic information from single cells [85]. A recent molecular study [86] revealed that cytb lineages belonging to Haemoproteus majoris have unique alleles in 4 investigated nuclear genes and may represent cryptic species. These lineages of Haemoproteus majoris are closely related and differ by only 1–6 substitutions over the 479 bp of sequenced cytb gene (0.2–1.3% difference). By contrast, an experimental observation in vivo [22] has demonstrated that parasites of the closely related lineages pSGS1 and pGRW11 can mate in mosquitoes Culex pipiens forma molestus and produce hybrid oocysts. Genetic differences between these lineages in the cytb gene are small (0.2%). According to hybridization experiments [22], the parasites of the lineages pSGS1 and pGRW11 are different variants of the same species, but information about hybridization of other lineages of P. relictum and other avian haemosporidian parasites is absent.

It is worth noting that partial sequences of merozoite surface protein 1 (msp1) gene were determined in 3 P. relictum lineages (pSGS1, pGRW11, pGRW4) in samples collected from different geographic sites using nuclear markers [21]. All three lineages were from markedly randomly sampled birds, with unclear geographical origin of infection. Four different alleles were reported in the lineage pSGS1, and three of them were shared with the lineage pGRW11, indicating possible hybridization. This is in accordance with the available experimental observations [22]. However, five different alleles were revealed in the lineage pGRW4 [21], suggesting the lack of gene flow between parasites of this lineage and the lineages pSGS1 and pGRW11. However, due to the markedly random sampling (many lineage isolates came from different species of African migrants with unclear geographical origin of infection), it is difficult to rule out that the reported genetic difference might reflect strain varieties, but not species differences. Additionally, due to common co-infections of malaria parasites in naturally infected hosts and possible selective amplification of different lineages using general primers [87], it is possible that some samples contained co-infections of different lineages. Because of this, the possibility to create between-lineage nuclear gene artefacts cannot be ruled out as well. In other words, the quality of the haemosporidian sequences should be carefully considered if samples from wildlife are used [88].

Plasmodium relictum is a unique among malaria parasites in regard to the enormous range of its avian hosts and mosquito species involved in its transmission. Therefore, direct in vivo experimental hybridization of different P. relictum lineages [22] would be most useful if they involved lineage isolates which are transmitted at the same site by the same mosquito species as this would make experimental studies closer to real epidemiological situations that are observed in wildlife.

Data about vertebrate host and geographical distribution of different P. relictum lineages are summarized in Table 2. The lineages pLZFUS01, pPHCOL01 of P. relictum have been reported occasionally, mainly in birds wintering or resident in tropical countries where transmission occurs [30, this study). The parasite lineage pGRW4 has both broad host and worldwide geographical distribution, but is rare in Europe [2, 11, 21, 33]. The lineage pSGS1 and pGRW11 are also broadly distributed, but neither has been reported in several extensive studies in the mainland Americas [2, 21, 89‐91]. However, Marzal et al. [3] found P. relictum (pSGS1) in 8 native bird species belonging to two orders in Peru, and Quillfeldt et al. [92] reported this parasite in seabirds on Falkland Islands, indicating presence of transmission, at least in South America.

The reported differences in geographical distribution of the lineages pSGS1 and pGRW11 on the one hand, and GRW4 on the other hand are difficult to explain bearing in mind the enormously broad range of their susceptible avian hosts (Table 2) and mosquito vectors, such as the globally distributed Culex pipiens, Culex quinquefasciatus and other mosquito species of the Culex pipiens complex, which are of global distribution [93‐95]. It is worth noting that recent experimental studies have demonstrated complete sporogony of the pGRW4 parasites from European birds, in cosmopolitan Culex pipiens forma molestus mosquitoes at relatively low temperatures. This indicates that there are no obstacles preventing transmission of this infection in Europe during the warm period of the year [33]. The following explanations of the observed phylogeographic data are worth discussion.

First, the existence of still unclear mechanisms of geographically related limitations in transmission of the parasite lineages pSGS1 and pGRW4 cannot be ruled out. However, the observed results in the phylogeography of these parasites might also originate, at least in part, from bias in DNA amplification of different lineages during co-infections while using general primers [87]. Failure in detection of mixed infections of Plasmodium parasites have often been reported [41, 87, 96‐98], but have not been investigated among P. relictum lineages. In other words, a sensitive issue is that the majority of available studies on P. relictum used only general primers for haemosporidian parasite DNA amplification. Such primers are selective and often do not indicate the presence of co-infection of parasites of different lineages [87]. Parasite lineage-specific primers have not been applied in phylogeographic studies of P. relictum lineages pSGS1 and pGRW4 and others so far. It remains unclear whether some P. relictum lineages are preferably amplified over others, particularly in cases of co-infections of different lineages. Relatively simple experimental studies using the protocol by Bernotienė et al. [87] might be helpful in answering this question. Co-infections of malaria parasites are common and even predominate in some bird populations [87, 96, 97]. This information is essential for better understanding of true distribution of P. relictum lineages both by hosts and geographically. Application of specific primers might contribute to better understanding patterns of geographical distribution of these invasive bird infections.

Second, parasite prevalence data depend on both force of infection and the longevity of infection. If local transmission is occurring, the low prevalence of GRW4 infection in European bird populations might be a result of (1) mortality of some European birds due to this infection, as is the case with some endemic Hawaiian birds [1]; (2) resistance and ability of some bird species to tolerate the pGRW4 malaria infection [11]; or, (3) a combination of these two factors. Naive Hawaiian and New Zealand endemic birds suffer mortality from infection with P. relictum pGRW4 [5, 50, 75, 77, 99, 100], but introduced bird species are less susceptible and might tolerate this disease [5, 50, 70]. Little is known about the virulence of the pGRW4 infection in resident European birds and other birds worldwide [33]. Preliminary observations indicate that several European bird species (Fringilla coelebs, Sylvia atricapilla, Passer domesticus) can resist pGRW4 strains, which were isolated from African migrating Great read warblers Acrocephalus arundinaceus [11]. Further experimental studies and application of lineage specific primers might provide more certain information about distribution of these parasite lineages, their co-existence in the same avian hosts and study sites, and better understanding infections in bird health.

The list of mosquito species, which are susceptible to P. relictum includes over 20 species [7, 13], however, information about vectors at parasite lineage levels is insufficient [101]. Widespread Culex pipiens, Culex quinquefasciatus and Culex tarsalis mosquitoes are excellent vectors for pSGS1, pGRW4, pGRW11 [22, 32, 33, 46, 52, 102‐105], but data about vectors of the pLZFUS01 and pPHCOLL01 parasites are absent. It is interesting to note that mosquitoes belonging to three genera, Aedes albopictus, Wyeomyia mitchellii and Culex quinquefasciatus, are susceptible to the pGRW4 parasite, and the sporogony was completed in all these mosquito species, but prevalence varied significantly between species. The latter mosquito is the main vector, but other mosquito species might be involved in transmission as well [106]. However, it worth mentioning that, while sporogony was completed in a small fraction of Wyeomyia mitchellii, the authors [106] did express doubt in the viability of aberrant sporozoites in this mosquito species.

Culex quinquefasciatus is absent in Lithuania. This insect was used in experiments because the new P. relictum lineage (pPHCOL01) was isolated from a bird species wintering in Africa where Culex quinquefasciatus is widespread [93, 94]. Sporogony of the parasite lineage pPHCOL01 was not initiated in Culex quinquefasciatus probably because the donor bird has light gametocytaemia (single gametocytes were seen in donor canaries during mosquito exposure), and that might have been the main obstacle.

Numerous mosquitoes were incriminated as possible P. relictum vectors using microscopic methods, but mainly only oocysts were reported in the majority of the studied insects, and the development of sporozoites were accessed in a few species [101]. This questions the conclusions about true possibility and involvement of mosquitoes belonging to different genera to act as effective vectors of P. relictum in wildlife. More delicate studies, including the observation of sporozoites in the salivary gland are needed to reach conclusions about ability of certain mosquito species to act as vectors. It is important to note that even presence of sporozoites of Plasmodium parasites in salivary glands does not always guarantee that the insects can transmit infection by bite. For example, sporozoites of Plasmodium hermani were reported in mosquito Wyeomyia vanduzeei, and these sporozoites were used successfully to induce infection in turkeys by syringe inoculation, but this mosquito was unable to transmit infection by bite [107]. This example calls for more delicate vector studies for better understanding transmission of avian haemosporidians. Determination of vectors is time consuming in wildlife studies where diversity of blood-sucking dipteran insects is high. The PCR-based reports of P. relictum lineages in wild-caught dipteran insects markedly speed search for possible vectors by indicating significant links between insects, avian hosts and parasites [103, 104, 108‐118], but cannot prove that sporozoites develop and can be transmitted by the PCR-positive insects. The observation of Plasmodium spp. sporozoites in salivary glands and the studies of transmission by mosquito bites remain the gold standards for determining vector competence. Combination of molecular diagnostic, experimental procedures and microscopic tools remain essential in haemosporidian vector research [33, 46, 101, 106, 119‐121].