Parasitology Research

, Volume 102, Issue 6, pp 1185–1193

Polymerase chain reaction-based identification of Plasmodium (Huffia) elongatum, with remarks on species identity of haemosporidian lineages deposited in GenBank

  • Gediminas Valkiūnas
  • Pavel Zehtindjiev
  • Dimitar Dimitrov
  • Asta Križanauskienė
  • Tatjana A. Iezhova
  • Staffan Bensch
Original Paper

DOI: 10.1007/s00436-008-0892-9

Cite this article as:
Valkiūnas, G., Zehtindjiev, P., Dimitrov, D. et al. Parasitol Res (2008) 102: 1185. doi:10.1007/s00436-008-0892-9

Abstract

Numerous lineages of avian malaria parasites of the genus Plasmodium have been deposited in GenBank. However, only 11 morphospecies of Plasmodium have been linked to these lineages. Such linking is important because it provides opportunities to combine the existing knowledge of traditional parasitology with novel genetic information of these parasites obtained by molecular techniques. This study linked one mitochondrial cytochrome b (cyt b) gene lineage with morphospecies Plasmodium (Huffia) elongatum, a cosmopolitan avian malaria parasite which causes lethal disease in some birds. One species of Plasmodium (mitochondrial cyt b gene lineage P-GRW6) was isolated from naturally infected adult great reed warblers (Acrocephalus arundinaceus) and inoculated to one naive juvenile individual of the same host species. Heavy parasitaemia developed in the subinoculated bird, which enabled identification of the morphospecies and deposition of its voucher specimens. The parasite of this lineage belongs to P. elongatum. Illustrations of blood stages of this parasite are given. Other lineages closely related to P. elongatum were identified. The validity of the subgenus Huffia is supported by phylogenetic analysis. Mitochondrial cyt b gene lineages, with GenBank accession nos. AF069611 and AY733088, belong to Plasmodium cathemerium and P. elongatum, respectively; these lineages have been formerly attributed to P. elongatum and P. relictum, respectively. Some other incorrect species identifications of avian haematozoa in GenBank have been identified. We propose a strategy to minimise the number of such mistakes in GenBank in the future.

Introduction

Plasmodium elongatum Huff, 1930 is a cosmopolitan avian malaria parasite, which causes lethal disease in some naturally and experimentally infected birds. This parasite is the type species of the subgenus Huffia, which currently unites three species of Plasmodium (Garnham 1966; Valkiūnas 2005). Severe illness and even devastating outbreaks of P. elongatum-malaria amongst penguins have been recorded in zoos in North America and Eurasia (Cranfield et al. 1990; Valkiūnas 2005), so this infection is of practical significance. According to the older literature (Bennett et al. 1982), P. elongatum has been frequently recorded in all zoogeographical regions except the Australian and Antarctic. It has been especially frequently found in birds in the Holarctic (Garnham 1966). Surprisingly, there have been no records of this malaria parasite during the last 10 years, in spite of numerous studies of the distribution of avian blood parasites in wildlife. It is probable that P. elongatum has been frequently overlooked in blood films. Molecular diagnostic techniques are sensitive in detecting blood parasites, especially during low parasitaemias (Bensch et al. 2004; Ricklefs et al. 2004; Sehgal et al. 2006; Valkiūnas et al. 2006; Hellgren et al. 2007). However, only DNA sequences, which are clearly linked with their morphospecies, can be used for this purpose (Palinauskas et al. 2007). Unfortunately, sequences of haemosporidian parasites, which have been misidentified even at the level of genera, are present in GenBank (Valkiūnas et al. 2007b). Because GenBank data have been used increasingly, especially in evolutionary biology and veterinary medicine studies, incorrect linkages of sequences with morphospecies identity might be misleading, so it is important to compare DNA sequences and morphospecies of parasites precisely.

Two mitochondrial cytochrome b (cyt b) gene lineages have been attributed to P. elongatum, with GenBank accession nos. AF069611 and DQ659588. Surprisingly, these two lineages belong to clearly different clades on the phylogenetic tree (Beadell et al. 2006); they cluster well with lineages of Plasmodium cathemerium/P. relictum and P. elongatum, respectively. According to information from GenBank, the genetic distance between these mitochondrial lineages of P. elongatum (AF069611 and DQ659588, 5.8%) is much greater, for instance, than the genetic distances between P. elongatum (DQ659588) and P. relictum (AY733088; 0.3%) and between P. elongatum (AF069611) and P. cathemerium (AY377128; 1.7%). It is important to note that P. (Huffia) elongatum, on one hand and P. (Haemamoeba) relictum and P. (Haemamoeba) cathemerium, on the other hand, are clearly different in numerous morphological features and life history traits; they belong to different subgenera of avian malaria parasites (Garnham 1966; Valkiūnas 2005). If the information about the relationships of lineages and morphospecies currently present in GenBank is correct, the use of morphological and life history characters in the taxonomy of haemosporidian parasites should be discouraged. However, some recent studies have shown good concordance between traditional taxonomy of haemosporidians and molecular clustering of their lineages (Hellgren et al. 2007; Martinsen et al. 2007; Perkins et al. 2007; Palinauskas et al. 2007; Valkiūnas et al. 2007a, b).

Because identification of P. elongatum is of practical and theoretical significance, the main aim of this study was to develop a method for identification of this malaria parasite based on partial sequences of its mitochondrial cyt b gene and using the polymerase chain reaction (PCR), which has been successfully used in evolutionary biology studies of haemosporidian blood parasites (Escalante et al. 1998; Fallon et al. 2003; Bensch et al. 2004; Pérez-Tris and Bensch 2005; Križanauskienė et al. 2006; Hellgren et al. 2007; Palinauskas et al. 2007; Perkins et al. 2007). To achieve this aim, we used experimental infections with high parasitaemia in naive birds, as recommended by Palinauskas et al. (2007) and Valkiūnas et al. (2007b). This permitted the clear morphological identification of P. elongatum, the deposition of voucher specimens of its blood stages and the determination of mitochondrial cyt b lineages, which can be used for molecular diagnosis of P. elongatum infection. We have also detected some incorrect linkages between morphospecies of avian blood parasites and their DNA sequences which are present in GenBank, and we have suggested a possible strategy to minimise the number of such incorrect identifications of organisms in GenBank in the future.

Materials and methods

Experimental birds and parasites

Eighty-two adult great reed warblers (Acrocephalus arundinaceus) were caught at the Kalimok Biological station in north-eastern Bulgaria (44°01′ N, 26°26′ E) in July 2006. In August 2006, three juvenile great reed warblers were captured near the station and kept in captivity for later use in the infection experiment. The birds were trapped under license from the Bulgarian Ministry of Environment and Waters.

All birds were examined for Plasmodium spp. by both PCR (see below) and microscopy. Blood was taken by puncturing the brachial vein. Two blood films were prepared from each bird, air dried, fixed in methanol and stained with Giemsa’s stain, as described by Valkiūnas (2005). The slides were examined for 10–15 min at low magnification (×400), and then at least 100 fields were studied at high magnification (×1,000). Intensity of infection was estimated as a percentage by counting the number of parasites per 1,000 erythrocytes in moderate or heavy infections and per 10,000 erythrocytes at low parasitaemia, i.e., <0.01%, as recommended by Godfrey et al. (1987). Malaria parasites were identified according to Valkiūnas (2005).

Uninfected great reed warblers were released. One naturally infected adult great reed warbler with light parasitaemia (<0.001%) of Plasmodium sp. lineage P-GRW6 was used as a donor of the infection. The experimental and control birds were kept indoors in cages supplied with suitable vegetation from typical habitats of this species. The food given included mealworms, fly larvae and a commercial mix based on hens’ eggs enriched with vitamins and minerals. The cages were protected from vectors by mosquito nets.

One juvenile bird was infected by inoculation into the pectoral muscle of 300 μl of a freshly prepared mixture of infected blood (four parts) and 3.7% sodium citrate solution (one part) as described by Iezhova et al. (2005). Two juvenile birds were used as negative controls. Blood films were prepared every 2–5 days between day 15 after inoculation (PI) and the end of the experiment (126 days PI).

DNA extraction, PCR amplification and sequencing

DNA was extracted using either a chloroform/isoamylalcohol method (Sambrook et al. 2002) or the standard ammonium acetate procedure. For the sequence-based analysis, we amplified part of the mitochondrial cyt b gene (479 bp) using a nested PCR protocol, as described by Hellgren et al. (2004). In the first PCR, a segment of the parasite cyt b gene (750 bp) was amplified using the initial primers HaemFNI and HaemNR3 which are general for species of Haemoproteus, Plasmodium and Leucocytozoon (Hellgren et al. 2004). For the second PCR, we used the primers HaemF and HaemR2, which are specific to Haemoproteus and Plasmodium spp. (479 bp; Bensch et al. 2000). One negative control was used after processing every eighth sample to test for false amplifications; the false positive results were not recorded.

The first PCR, with the primers HaemFNI–HaemNR3, was carried out in a 25 μl volume and included 50 ng of total genomic DNA, 1.5 mM MgCl2, 1× PCR buffer, 1.25 mM of each deoxynucleoside triphosphate, 0.6 mM of each primer and 0.5 U Taq DNA polymerase. Cycling conditions were initial denaturation for 3 min at 94°C, 30 s at 94°C, 30 s at 50°C, 45 s at 72°C for 20 cycles, followed by final extension at 72°C for 10 min. For the second PCR, we used 2 μl of the first PCR product as template in a 25 μl volume with the primers HaemF–HaemR2, including the same reagents and under the same thermal conditions as the first reaction, except that 35 cycles were used instead of 20 cycles. Positive or negative amplifications were evaluated as the presence or absence of bands on 2% agarose gels. Samples that showed positive amplification were sequenced using sequencing robot ABI PRISM™ 3100 and “Big Dye” sequencing kit.

Morphological analysis

An Olympus BX61 light microscope equipped with an Olympus DP70 digital camera and AnalySIS FIVE imaging software was used to prepare illustrations (Fig. 1) and to take measurements (Table 1). The morphometric features studied were those defined by Valkiūnas (2005). The morphology of the parasites was compared with neohapantotype material of P. elongatum from its type vertebrate host, the canary Serinus canaria, deposited in the Garnham Collection at the Natural History Museum, London (blood slides accession nos. 216 and 217).
Fig. 1

Plasmodium elongatum (lineage P-GRW6) from the blood of the great reed warbler Acrocephalus arundinaceus: a, b—trophozoites; cj—erythrocytic meronts; kn—macrogametocytes; o, p—microgametocytes. Scale bar = 10 μm

Table 1

Morphometry of mature erythrocytic meronts and gametocytes of Plasmodium elongatum (n = 21)

Feature

Measurements (μm)a

Meront

 

Length

3.8–6.1 (4.7 ± 0.5)

Width

2.8–4.7 (3.7 ± 0.5)

Area

8.8–19.5 (13.5 ± 2.6)

Area of pigment granules

0.1–0.3 (0.2 ± 0.03)

No. of merozoites

4–8 (5.5 ± 1.3)

Macrogametocyte

 

Length

11.9–16.9 (13.9 ± 1.1)

Width

1.4–2.0 (1.7 ± 0.1)

Area

18.6–25.3 (21.8 ± 1.8)

Microgametocyte

 

Length

11.9–16.6 (14.8 ± 1.2)

Width

1.1–2.0 (1.6 ± 0.2)

Area

18.8–28.1 (23.4 ± 2.7)

aMinimum and maximum values are provided, followed in parentheses by the arithmetic mean and standard deviation.

Phylogenetic analysis

Sixteen mitochondrial cyt b sequences of avian Plasmodium spp. from our survey and GenBank were used during this study. The GenBank accession numbers of these sequences are given in Fig. 2. The sequences were aligned by eye using BioEdit 6.0.6 (Hall 1999). All grouped sequences matched in 365 bp.
Fig. 2

Bayesian (BY) and neighbour-joining (NJ) phylogeny of 16 mitochondrial cytochrome b gene lineages of species of avian Plasmodium and five lineages of Haemoproteus spp. as outgroups. Numbers above branches indicate bootstrap support based on 1,000 replicates. Numbers before species names correspond to the lineage numbers in Table 2. Names of the lineages (when available) and GenBank accession numbers of the sequences are given after the species names of parasites. Lineages of malaria parasites which are closely related (within a genetic distance of ≤4%) to Plasmodium (Novyella) ashfordi, P. (Huffia) elongatum, P. (Giovannolaia) circumflexum and P. (Haemamoeba) relictum are marked by bars a, b, c and d, respectively; they represent malaria parasites of the different subgenera

A Bayesian phylogenetic tree was constructed using mrBayes version 3.1.1 (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003) after finding an appropriate model of sequence evolution using the mrModeltest (Nylander 2004) and MEGA 3.0 software (Kumar et al. 2004). A General Time Reversible model including invariable sites (GTR+I) was used for the Bayesian approach. The Bayesian phylogeny was obtained with the use of four heated and one cold mcmc chain, sampled every 200 generations over 20 million generations; 100,000 trees were generated. After visualising the parameters and controlling for a burn-in period using the softwares AWTY (Wilgenbusch et al. 2004) and TRACER (Rambaut and Drummond 2003, available at http://evolve.zoo.ox.ac.uk), 25% of the trees were discarded as burn-in material. The remaining 75,000 trees were used to construct a major-consensus tree. Support for internal branches of the maximum likelihood tree was estimated by bootstrap analyses with 1,000 replicates.

To construct the neighbour-joining (NJ) tree, we used the program MEGA, version 3.1 (Kumar et. al. 2004) and the neighbour-joining method with a Kimura 2-parameter distance matrix.

Five mitochondrial cyt b lineages of well-identified avian haemoproteids (Haemoproteus lanii, H. belopolskyi, H. parabelopolskyi, H. majoris and H. minutus; Hellgren et al. 2007) were used as outgroups (Fig. 2).

The sequence divergence between and within the different lineages (Table 2) was calculated using a Jukes–Cantor model of substitution, with all substitutions weighted equally, implemented in the program MEGA 3.1 (Kumar et al. 2004).
Table 2

The sequence divergence (in percentage) between 16 mitochondrial cytochrome b gene lineages of avian Plasmodium spp.

 

Lineage

1a

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

1

0b

2

3.4

3

9.1

7.6

4

9.1

7.6

0.0

5

9.4

7.6

0.3

0.3

6

9.4

7.6

0.3

0.3

0.0

7

11.7

10.1

6.7

6.7

6.4

6.4

8

8.8

8.2

5.8

5.8

5.5

5.5

5.5

9

9.1

8.2

5.8

5.8

5.5

5.5

5.5

0.3

10

11.7

10.1

10.1

10.1

9.7

9.7

8.8

7.6

7.6

11

10.7

8.8

6.7

6.7

6.7

6.7

6.7

4.6

4.6

8.5

12

10.4

8.5

6.4

6.4

6.4

6.4

6.4

4.9

4.9

8.2

0.3

13

10.4

8.5

5.8

5.8

5.8

5.8

6.4

4.9

4.9

8.2

2.5

2.2

14

10.1

8.2

5.5

5.5

5.5

5.5

6.1

4.6

4.6

7.9

2.2

2.0

0.3

15

9.1

8.2

5.8

5.8

6.1

6.1

5.5

4.0

3.7

7.0

3.4

3.1

3.7

3.4

16

9.7

8.5

5.5

5.5

5.8

5.8

5.2

4.9

4.9

8.2

3.7

3.4

4.0

3.7

1.7

0

aNumbers correspond to the numbers of mitochondrial lineages in Fig. 2, in which the morphospecies of the parasites and the GenBank accession numbers of the lineages are given.

bThe sequence divergence was calculated with the use of a Jukes–Cantor model of substitution.

Results

Experimental infection

According to the PCR analysis and microscopical examination of blood films, infection of the lineage P-GRW6 was present in one experimentally infected bird at 15 days PI (the first day of observation after the experimental infection). Maximum parasitaemia reached 2.0% at 15 days of patency and then dropped to less than 1%; it remained <0.1% to the end of the experiment. The infected bird died at 126 days PI; no post-mortem investigation was performed. Blood films with high parasitaemia were used for description of the morphology of the parasite and deposition of its voucher specimens. Malaria parasites were not recorded in the control birds, all of which survived to the end of the study.

Description of the parasite

Plasmodium (Huffia) elongatum Huff, 1930 (Fig. 1, Table 1).

DNA sequences

Mitochondrial cyt b lineage P-GRW6 (479 bp, GenBank accession no. DQ368381).

Avian hosts

The lineage P-GRW6 of P. elongatum has been recorded in the great reed warbler (this study), house sparrow Passer domesticus, wheatear Oenanthe oenanthe, icterine warbler Hippolais icterina and oriole finch Linurgus olivaceus (S. Bensch, unpublished observations).

Distribution

We isolated the lineage P-GRW6 at Kalimok biological station, NE Bulgaria; it is common in the great reed warbler and house sparrow. This lineage has been also recorded in great reed warblers in Sweden where the prevalence is around 2% (Bensch et al. 2007).

Site of infection

Trophozoites and erythrocytic meronts inhabit all types of red blood cells but are more frequently present in young erythrocytes, including erythroblasts. Gametocytes develop in mature red blood cells; no other information is available.

Representative blood films

Voucher specimens were deposited in the Institute of Ecology, Vilnius University, Vilnius, Lithuania (accession nos. 45604 NS, 45607 NS, 45609–45611 NS); in the United States National Parasite Collection, Beltsville, USA (USNPC 100359.00); in the Queensland Museum, Queensland, Australia (G464984); and in the Institute of Zoology, Bulgarian Academy of Sciences, Bulgaria (2007–001MS). All preparations were derived from the blood of the great reed warbler; they were prepared by P. Zehtindjiev.

The morphology of trophozoites (Fig. 1a,b), erythrocytic meronts (Fig. 1c–j), macrogametocytes (Fig. 1k–n) and microgametocytes (Fig. 1o,p) and their influence on the host cells are similar to those of the same blood stages of P. elongatum in its type vertebrate host, the canary, as described by Valkiūnas (2005), with one exception. A large clear vacuole is frequently present in growing trophozoites (Fig. 1b). This character can be used in morphological identification of this lineage of P. elongatum.

Phylogenetic relationships of parasites

There were no differences between the topology of phylogenetic trees constructed using either the Bayesian or neighbour-joining approaches (Fig. 2).

Lineage P-GRW6 of P. elongatum belongs to a clearly separated clade (Fig. 2, clade b), which is in accord with the morphology of this parasite (Fig. 1) and traditional taxonomy. This lineage is identical to the lineage of P. elongatum with the accession no. DQ659588. Both these lineages of P. elongatum are closely related to two other lineages of Plasmodium (Huffia) of which the species identity is uncertain (Fig. 2, clade b; Table 2).

The lineage of Plasmodium sp. (AY733088, P. relictum according to GenBank) is identical to the lineages of P. (Huffia) sp. (EF011175), so one of these lineages must be incorrectly identified (Fig. 2, clade b). Because the genetic difference amongst four lineages of the well-supported clade b (Fig. 2) is ≤0.3% (Table 2), it is probable that all these lineages belong to P. elongatum.

The following two lineages deposited in GenBank were incorrectly identified. The lineage Plasmodium sp. (AF069611) was attributed to P. (Huffia) elongatum in GenBank (Fig. 2, clade d), and the lineage Plasmodium sp. (AY733088) was attributed to P. (Haemamoeba) relictum (Fig. 2, clade b). These two lineages cluster well with lineages belonging to parasites of the subgenera Haemamoeba and Huffia, respectively (Fig. 2; Table 2).

Well-identified species of different subgenera of avian malaria parasites belong to clearly different clades in the phylogenetic tree (Fig. 2, clades a,b,c,d), in accordance with morphological and biological characteristics of these parasites.

Discussion

Numerous lineages of avian blood parasites have been deposited in GenBank. However, only 11 morphospecies of avian Plasmodium have been linked to these lineages (Palinauskas et al. 2007; Valkiūnas et al. 2007b). Two identical sequences (DQ368381 and DQ659588) can be used for the molecular diagnosis of P. elongatum infection. Because genetic differences between lineages of the clade b are small (Fig. 2; Table 2), it is possible that all these lineages represent genetic variations of the morphospecies P. elongatum, as is probably the case with numerous lineages of P. relictum (Palinauskas et al. 2007) and several Haemoproteus spp. (Hellgren et al. 2007). It is worth noting that the genetic difference between the mitochondrial cyt b gene lineages of all well-identified morphospecies of avian Haemoproteus and Plasmodium is <5%, which probably reflects the level of intraspecific genetic variation of morphospecies of these haemosporidian parasites (Hellgren et al. 2007; Palinauskas et al. 2007; Valkiūnas et al. 2007a). This is important in the future development of the taxonomy of avian haemosporidians. In relation to P. elongatum, the following incorrectly identified sequences have been deposited in GenBank.

First, the sequence with accession no. AF069611 (Fig. 1, clade d), which was isolated from the house sparrow P. domesticus and attributed to P. elongatum by Escalante et al. (1998). Because the genetic distance between this sequence and two other lineages, which certainly belong to P. elongatum (DQ659588 and DQ368381; Fig. 1, clade b) is great (Table 2), it is probable that this sequence was not derived from P. (Huffia) elongatum but from a species of Plasmodium of the subgenus Haemamoeba, with lineages of which it is well clustered (Fig. 2, clade d). It is important to note that several mitochondrial lineages of P. relictum have been well identified morphologically; they cluster together on phylogenetic trees and are easy to distinguish from P. elongatum both by morphology of their blood stages and mitochondrial cyt b sequences (Fig. 2; Palinauskas et al. 2007; Valkiūnas et al. 2007b).

Second, the sequence with GenBank accession no. AY733088, which was isolated from the African penguin Spheniscus demersus and attributed to P. relictum (Beadell et al. 2006). The genetic distance between this sequence and two clearly identified P. elongatum lineages (DQ368381 and DQ659588) is small (Fig. 2, clade b; Table 2), so this lineage probably belongs to P. elongatum. Importantly, the sequence AY733088 is identical to the sequence EF011175, which was attributed to P. (Huffia) sp. by Martinsen et al. (2007).

The incorrect assignment between genetic lineages and morphospecies identity of malaria parasites might be explained by (a) the difficulties of identification of parasites during light parasitaemias and (b) the shortcomings of nested PCR protocols in determination of simultaneous infections of more than one species of haemosporidian parasites.

First, the identification of malaria parasites and other haemosporidians in blood films from naturally infected birds is difficult even for experts because parasitaemia frequently is light and certain blood stages, which are necessary for identification of parasites, are often absent (Valkiūnas 2005). One should therefore be careful when comparing genetic lineages with morphospecies which have been identified using blood films with light parasitaemias. PCR-based tools are promising for the diagnosis of haematozoan infections, especially in wildlife. However, such techniques should be developed carefully by precise comparison of both microscopy and genetic data (Perkins 2000; Martinsen et al. 2006; Palinauskas et al. 2007).

Second, molecular tools have greatly increased the sensitivity of diagnosis of parasites. However, the increased sensitivity of detection methods often creates specificity problems (Pérez-Tris and Bensch 2005). In the case of natural infections of blood parasites, one should always be aware that the PCR is not a direct detection method. The amplification of DNA fragments by PCR includes not only target DNA but often also DNA from simultaneous infections, so the validation of molecular techniques is important (Pérez-Tris and Bensch 2005; Valkiūnas et al. 2006). Simultaneous infections with different haematozoa are common in wildlife, but the nested PCR techniques often do not detect this (Valkiūnas et al. 2006; Palinauskas et al. 2007). One should therefore be careful when interpreting sequence data obtained from naturally infected birds. During natural infections, it is difficult to rule out completely the possibility that low-level mixed infections may be present in a sample but may not be detected even by long-duration microscopy. In this case, a sequence obtained and assumed to be that of the parasite seen under the microscope might in fact be derived from a low-level infection of another species and thus linked with the wrong morphospecies. This must be taken into consideration in molecular systematic and phylogenetic studies and probably accounts for the aforementioned incorrectly identified sequences of P. elongatum from the house sparrow and the African penguin. Simultaneous infections of P. elongatum and P. relictum are common in these bird species and have been well documented (Beier and Stoskopf 1980; Cranfield et al. 1990; Valkiūnas 2005).

Unfortunately, the number of incorrectly identified species of avian blood parasites in GenBank is increasing. For example, it is clear that the sequences of Haemoproteus (Haemoproteus) columbae (AF069613), Haemoproteus (Parahaemoproteus) silvae (AY099040), Plasmodium (Novyella) nucleophilum (AF254962), P. (Huffia) elongatum (AF069611) and P. (Haemamoeba) relictum (AY33088) were originally identified incorrectly; they belong to P. relictum (see Bensch et al. 2000; Valkiūnas et al. 2007b), H. (P) payevskyi (see Valkiūnas et al. 2007b), Plasmodium (Novyella) ashfordi (see Valkiūnas et al. 2007b), P. relictum (this study) and P. elongatum (this study), respectively. Phylogenetic analyses based on such incorrect identifications may be unreliable. Unfortunately, it is difficult to determine how frequently such mistakes have occurred in the GenBank collection, because details of voucher specimens of parasites and their place of deposition are rarely specified in current molecular studies.

If the work on comparison of morphospecies of organisms and their DNA lineages is to be continued without the expertise of taxonomists, there is a risk that GenBank data will include increasing numbers of incorrect identifications of species and even higher taxa. That would be misleading and might lead to a limitation of the use of GenBank in the future. One way to avoid this situation might be for journals to require authors to deposit voucher specimens of blood parasites (mainly blood films with blood stages) and other organisms in well-recognised museums, so that they would be available for examination by experts. Ideally, GenBank databases of accepted sequences should be supplemented with information about the place of deposition and accession numbers of voucher specimens of the parasites and other organisms from which the sequences were derived.

This study has contributed to the discussion about the validity of the subgenera Huffia, Haemamoeba, Giovannolaia and Novyella (Fig. 2, clades a,b,c,d). A molecular study by Martinsen et al. (2007) reached conclusions about validity of the subgenera Haemamoeba and Bennettinia, but not of Giovannolaia and Novyella. Sequences of more well-identified Plasmodium morphospecies from each subgenus are required to answer this question and to develop the traditional taxonomy. DNA lineages of the majority of species of different subgenera of Plasmodium have not been determined as yet; that is an urgent necessity.

It is important to note that Plasmodium (Novyella) rouxi is markedly different genetically from other well-identified species of avian malaria parasites (Fig. 2; Table 2), including the recently described species of the same subgenus, P. (Novyella) ashfordi (Valkiūnas et al. 2007b). P. rouxi was originally described from the Spanish sparrow Passer hispaniolensis in Algeria; it is common in the Mediterranean region (Garnham 1966; Valkiūnas 2005). Because the lineage AY178904 was isolated from a non-type vertebrate host far distant from the type locality (China), its morphospecies identity is questionable and should be further investigated.

Importantly, lineage P-GRW6 of P. elongatum is identical to the lineage of P. elongatum (DQ659588) which was isolated from the great blue heron Ardea herodias in the USA (Beadell et al. 2006). That is the first molecular evidence for intercontinental distribution of P. elongatum, which certainly parasitises birds of different orders and thus is important in conservation biology. This PCR-based information coincides with the results of former microscopical studies of the cosmopolitan distribution of this pathogenic avian malaria parasite. Based on microscopical data, P. elongatum has been recorded in over 60 species of birds all over the world, except Australia; it is particularly common in passerines (Garnham 1966; Valkiūnas 2005).

Our study showed complete identity in the topology of the phylogenetic trees constructed using the Bayesian and NJ approaches. The MEGA software for the NJ method (Tamura et al. 2007) is freely available (http://www.megasoftware.net/). The NJ approach has given good results in phylogenetic analysis of short sequences (<500 bp) of the mitochondrial cyt b gene of haemosporidian parasites (Fig. 2), so it can be recommended for future use in phylogenetic studies of this particular group of haematozoa.

In conclusion, this study highlights the increasing importance of PCR-based methods in the diagnosis of parasitic infections and illustrates the enhancement in diagnostic ability possible with properly validated molecular tools. Such validation is an urgent requirement for both practical and theoretical reasons relating to parasitology and evolutionary biology; it will require the combined efforts of traditional, evolutionary and molecular biologists.

Acknowledgements

This article benefited from comments made by John R. Baker. The authors are grateful to Alan Warren, Natural History Museum, London for providing the type material of P. elongatum, and Vaidas Palinauskas for assistance during the preparation of Fig. 1. The study was supported in part by the Swedish Research Council, Carl Tryggers Foundation, the Lithuanian State Science and Studies Foundation, and SYNTHESYS. The experiments described herein comply with the current laws of Sweden, Bulgaria and Lithuania.

Copyright information

© Springer-Verlag 2008

Authors and Affiliations

  • Gediminas Valkiūnas
    • 1
  • Pavel Zehtindjiev
    • 2
  • Dimitar Dimitrov
    • 2
  • Asta Križanauskienė
    • 1
  • Tatjana A. Iezhova
    • 1
  • Staffan Bensch
    • 3
  1. 1.Institute of EcologyVilnius UniversityVilnius LTLithuania
  2. 2.Institute of ZoologyBulgarian Academy of SciencesSofiaBulgaria
  3. 3.Department of Animal Ecology, Ecology BuildingLund UniversityLundSweden

Personalised recommendations