Leucocytozoon (Leucocytozoon) neotropicalis, Lotta & Valkiūnas & Pacheco & Escalante & Hernández & Matta, 2019

Lotta, Ingrid A., Valkiūnas, Gediminas, Pacheco, M. Andreína, Escalante, Ananías A., Hernández, Sandra Rocío & Matta, Nubia E., 2019, Disentangling Leucocytozoon parasite diversity in the neotropics: Descriptions of two new species and shortcomings of molecular diagnostics for leucocytozoids, International Journal for Parasitology: Parasites and Wildlife 9, pp. 159-173 : 165-169

3.2.4. Leucocytozoon (Leucocytozoon) neotropicalis sp. nov

Macrogametocytes ( Fig. 2A–E ) develop in fusiform host cells; the shape of gametocytes is oval-elongate (their length is greater than width, Table 3). However, the morphology of host-parasite complexes with fusiform processes is readily distinguishable from L. grallariae (compare Fig. 1 F, H and Fig. 2 A, F, I ). The host cell nuclei are pushed aside, deformed like a homogeneous band of variable width that extends close to half of the circumference of gametocyte ( Fig. 2 A, I ), but never extends into the cytoplasmic processes. This is similar in both L. grallariae and L. neotropicalis cells, so these parasites cannot be distinguished by this character (see the description of L. grallariae and compare Fig. 2 D, F, H and Fig. 2 A, G, I .). These species can be readily distinguished due to the length of the cytoplasmic processes. Mainly, the latter is significantly longer (Student's t-test, α = 0.05 P <0.0001) and narrower in their maximum width (P <0.0001) than in L. grallariae (see Table 2, compare Fig. 2. F, H and Fig. 2 A, F, I ).

Two long thin fusiform spindle-shaped processes of the host cells' cytoplasm reach up to 14 μm ( Table 3), and that never is observed in L. grallariae . The length of the cytoplasmic processes can be different in the same host-parasite complex ( Fig. 2J and K ), and that probably is a result of deformation during the preparation of blood films. Cytoplasmic processes are thin and often flattened appearing like a ribbon ( Fig. 3F–H ). It is important to note that mature gametocytes often have a flattened form on the side, located on the opposite side of the host cell nuclei ( Fig. 2 H , J-K). This character is not observed in L. grallariae ( Fig. 1I–K ).

In the type material, we observed tiny volutin granules and small vacuoles (up to 0.47 μm in diameter) in 47.9% of fully grown gametocytes ( Fig. 2A–E ). The parasite nucleus was roundish in 49.3% of 103 observed gametocytes ( Fig. 2A–C ) or elongated ( Fig. 2 E ); its position was mainly more or less central, but sometimes was off-centre. The nucleolus was variable both in shape and position, being visible in 53.2% of 103 observed parasites ( Fig. 2 B, D ).

Microgametocytes ( Fig. 2F–I ): General configuration and other features were similar to the macrogametocytes with the usual haemosporidian sexual dimorphic characters. The proportion of microgametocytes and macrogametocytes in the type material was approximately 1:2.

Taxonomic summary.

Type host: Green-and-black Fruiteater Pipreola riefferii ( Cotingidae , Passeriformes ).

Additional hosts: unknown.

Type locality: El Bosque, Los Nevados National Natural Park ( NNP) (4̊ 43 N; 75̊ 27 W, 3150 masl), Risaralda, Colombia .

Type specimens: Hapantotype (accession Nos. UNAL: GERPH: OT 1354 -II). The intensity of the infection of the lineage MK103894 is 0.33%, it was collected by Melisa Galarza (27 December 2015) and deposited in the biological collection GERPH ( Grupo de Estudio Relación Parásito Hospedero ) at the Universidad Nacional de Colombia, Bogotá, Colombia. Parahapantotypes (accession Nos. UNAL: GERPH: OT 1354 -I, UNAL: GERPH: OT 1354 -III, other data as for the hapantotype) are deposited in the same collection. Digital images of the blood stages of the parasite in the type preparations are available on request from GERPH .

Partial mitochondrial DNA genome (5811 bp) that includes cox1, cox3 and cytb genes (GenBank accession number MK103895) was obtained from the type host Pipreola riefferii.

Site of infection: Blood cells, the specific cell is unknown due to the marked deformation by developing gametocytes.

Prevalence: Only one individual of the host species was collected and found infected, so the sample size does not allow to estimate the prevalence. Parasite was detected by microscopy in 1 of out of 684 examined birds (0.12%). In the type locality, 1 of 686 birds captured at Los Nevados NNP (0.14%) was infected, as determined by microscopic examination.

Etymology: The species name ( neotropicalis ) was derived from the name of the zoogeographical region where this parasite was found.

3.2.5. Remarks

Leucocytozoon neotropicalis is one of the six Leucocytozoon species that parasitize passerine birds and possess gametocytes developing in fusiform host cells. The main differences between L. neotropicalis and L. grallariae are specified in the description of the former parasite. Both of these new leucocytozoids have gametocytes developing fusiform host cells, which can be distinguished from other leucocytozoids due to the unique shape of their host cell nuclei (see Remarks on L. grallariae ).

Due to the presence of long and narrow cytoplasmic processes in host cells, L. neotropicalis is similar to L. lovati ( Valkiūnas, 2005) and L. eurystomi ( Bennett et al., 1993; Valkiūnas, 2005) ( Table 3). However, the nuclei of host cells never reach the cytoplasmic processes in the last two parasites. Because of this character, these species can be readily distinguished.

Microscopic examination of blood smears from type series revealed the presence of co-infection of a parasite with gametocytes developing in roundish host cells. Overall, the configuration of the nuclei in roundish host cells resembles the same characters observed in the L. fringillinarum group ( Fig. 3 .). The reported gametocytes in roundish host cells are bigger than those observed in the sample with coinfection with L. grallariae (Student's t-test for parasite area: p <0.000 1, α = 0.05, Table S3). In contrast to the macrogametocytes of L. neotropicalis , the volutin granules are not pronounced or absent in roundish gametocytes of this parasite ( Fig. 3D–F ). It worth mentioning that, for both new species as well as for L. pterotenuis ( Lotta et al., 2015) , the gametocytes developing in fusiform host cells were the most common, and their parasitemia was on average ten to seventeen times higher than the species with gametocytes developing roundish host cells. Indeed, in the type material of L. neotropicalis parasitemia of gametocytes developing fusiform host cells was 0.21%, while it did not exceed 0.01% for gametocytes developing roundish host cells.

Two distantly related lineages with a genetic distance of 0.29 between them (L_PIRIEF_01, cytb gene GenBank No. MH909276 and

Morphometric parameters of gametocytes and host cells of Leucocytozoon neotropicalis sp. nov. Measurements of Leucocytozoon eurystomi and Leucocytozoon lovati are provided for comparison. Measurements are given in μm or μm 2 (for area). Minimum and maximum values as well as mean ± SD are provided.

Feature Leucocytozoon neotropicalis sp. nov Leucocytozoon eurystomi a,b Leucocytozoon lovati b

Macrogametocyte n = 6 Microgametocyte n = 1

Parasite

Length 14.4–16.0 (15.3 ± 0.6) 11.4–17.9 (13.6 ± 1.5) 22.6–29.6 (25.2 ± 1.3) 14.1–22.0 (17.5 ± 1.4) Width 7.0–8.2 (7.8 ± 1.0) 5.7–10.0 (7.8 ± 1.4) 7.0–12.2 (8.1 ± 1.0) 6.7–13.1 (11.0 ± 0.7) Area 86.1–101.2 (95.2 ± 4.9) 64.9–114.9 (83.7 ± 17.4) (182 ± 18.5)

Perimeter 34.1–40.9 (38.8 ± 1.6) 31.4–48.9 (37.4 ± 4.8) (63.0 ± 5.9)

Parasite nucleus

Length 2.0–3.1 (2.8 ± 0.5) 5.3–10.5 (8.0 ± 1.5) 2.8–6.4 (4.1 ± 0.4) 3.2–6.8 (4.2 ± 0.3) Width 3.3–4.1 (4.0 ± 0.4) 4.4–7.0 (5.6 ± 0.8) 1.4–5.7 (3.5 ± 0.3) 1.8–4.4 (3.6 ± 0.3)

Area 7.0–9.6 (8.6 ± 1.0) 32.5–48.4 (41.4 ± 4.7) (11.8 ± 2.6)

Host-cell parasite complex

Length 35.5–56.6 (44.7 ± 5.3) 29.8–55.8 (41.4 ± 8.6) (39.2 ± 6.3)

Width 9.5–10,8 (10.2 ± 0.5) 9.2–10.8 (9.9 ± 0.5) (9.1 ± 0.7)

Area 161.0–200.1 (181.8 ± 10.4) 127.4–181.7 (151.2 ± 13.0) (202.2 ± 21.1)

Host-cell nucleus

Length 14.8–19.7 (17.7 ± 1.8) 14.2–18.3 (16.3 ± 1.3) 18.3–26.0 (21.7 ± 2.3) 9.8–16.3 (13.5 ± 1.5) Width 2.4–3.7 (2.9 ± 3.0) 2.4–3.7 (3.0 ± 0.4)

Area 28.2–39.6 (35.7 ± 4.3) 23.6–42.0 (33.6 ± 5.3) (39.0 ± 7.2)

Perimeter of parasite covered 14.1–18.2 (16.5 ± 1.8) 14.2–18.3 (16.3 ± 1.3) (16.9 ± 2.3)

Cytoplasmic processes c

Length 11.3–22.0 (15.7 ± 2.8) 7.6–24.4 (14.8 ± 4.5)

Width 2.6–5.4 (3.8 ± 0.8) 1.9–5.3 (3.7 ± 0.9)

Area 18.2–29.0 (25.0 ± 5.3) 14.4–26.1 (21.2 ± 3.2)

a According to Bennet et al., (1993).

b According to Valkiūnas (2005).

c Only one of 2 cytoplasmic processes was measured for each parasite.

partial mtDNA genome that included cytb, cox1, cox3, GenBank No. MK103894), were amplified from the same sample, which makes it difficult to link the lineages with their morphotypes. Based on phylogenetic analysis, we suggest that the last one lineage (GenBank No.

MK103894) corresponds to L neotropicalis n. sp. (see discussion below).

3.3. Sequencing of the cytochrome b gene and the DNA mitochondrial genome

Two lineages were isolated from each of the samples containing L. grallariae or L. neotropicalis . The partial cytb fragments obtained using the primers suggested by Hellgren et al. (2004) were very distant from the cytb lineages obtained by the mtDNA genome amplification protocol ( Pacheco et al., 2011, 2018). In other words, different Leucocytozoon parasite sequences were obtained in the same sample using different protocols and that corresponded to the microscopic observation of possible co-infections in these samples.

Even though both protocols for the amplification of cytb fragments and the mtDNA genome were run at least two times independently, each protocol amplified different lineages (lineages isolated from G. squamigera : GenBank cytb accession No. MH909275 vs GenBank mtDNA accession No. MK103895; lineages isolated from P. riefferii : GenBank cytb accession No. MH909276 vs GenBank mtDNA accession No. MK103894). Similar results were obtained when new molecular studies were performed with the Grallaridae bird samples reported by Lotta et al. (2015), where parasites were described as Leucocytozoon pterotenuis . With the new analysis, we realized that the cytb fragment obtained along with the partial mitochondrial genomes (mtDNA) identified with GenBank accession No. KM272250 and the short cytb fragment amplified with Hellgren's primers identified with GenBank accession No. KY646032 were different. Thereby, we will be referring to the description of Leucocytozoon pteroteunuis ( Lotta et al., 2015) as a partial description of the parasite (according to the ( International Commission on Zoological Nomenclature, 1999) the description “in part”), because according to the molecular analyses performed, gametocytes in roundish host cells observed in the sample likely do not belong to L. pterotenuis , but to other Leucocytozoon species (see discussion). Thus, the species name L. pterotenuis is valid only in part, mainly for gametocytes developing in fusiform host cells, but not to gametocytes in roundish host cells.

3.4. Analysis of primer affinities

The success of amplification is highly dependent on the primer's affinity for the target sequence and the parasitemia that determines the amount of parasite present DNA ( Perez-Tris and Bensch, 2005; Pacheco et al., 2018a). After verification of affinities of the primers proposed by Hellgren et al. (2004) with the all complete cytb gene sequences of Leucocytozoon parasites available for passerine birds, it was noticed that oligo-sequences matched the cytb sequences of L. fringillinarum (Genbank accession No. KY653765) and L. dubreuili (Genbank accession No. KY653795), which both have gametocytes developing roundish host cells ( Fig. 4 A, B ). In contrast, we noticed that the primer HaemR2L did not completely match with the mtDNA sequences obtained using the mtDNA genome amplification protocol used for L. pterotenuis (in part) (Genbank No.KM610046). Indeed, the base pairs at the 3′ end of the primer, as well as the two last base pair of the 5’, did not match with the cytb gene sequences obtained from the mtDNA genome of L. grallariae (Genbank No. MK103895) nor L. neotropicalis (Genbank No. MK103894) ( Fig. 4 ).

3.5. Phylogenetic analysis

In the phylogenetic reconstructions based on partial mitochondrial genomes and 476 cytb fragments ( Fig. 5 and S 1 ), two main clades that resemble the classification of parasites according to morphological features were observed. Thus, parasite lineages of leucocytozoids with gametocytes developing round host cells were part of a separate clade ( Fig. 5 and S 1 clade I). An exception is L. danilewskyi , in which the gametocytes develop both roundish and fusiform host cells. Meanwhile, parasites that produce gametocytes in fusiform cells are part of a separate monophyletic group (identified as clade II). Within this, lineages Genbank No. MK 103894 of L. neotroplcalis and Genbank No. MK103895 of L. grallariae samples form a clade that is the sister lineage to L. pterotenuis (in part) (KM272250) and Leucocytozoon sp. (KM272251). These parasites are closely related to L. sabrazesi (a morphological synonym of Leucocytozoon macleani, AB 299369), a parasite infecting Galliformes birds, in whose gametocytes develop in fusiform host cells ( Fig. 5A , Table 4). Thus, the phylogenetic analyses suggested a link between the parasite morphotypes and their sequences for both samples with co-infection.

It is worth noting that, since parasite mitochondrial genomes (mtDNA) corresponding to the partial cytb fragments of the MH909275 and MH909276 sequences could not be amplified, they were not included in the phylogenetic hypothesis constructed with mtDNA ( Fig. 5A ).

Interestingly, in clade I, the partial cytb lineages MH 909275, MH909276 and KY646032 obtained from samples of L. neotropicalis , L. grallariae and L. pterotenuis (in part) respectively using the primers proposed by Hellgren et al. (2004) share a recent common ancestor with L. fringillinarum and L. quynzae ( Fig. 5B and S 1 ). This suggests that these lineages likely correspond to the roundish host cell morphotype coexisting with the fusiform host cell morphospecies present in the samples infected with L. neotropicalis , L. grallariae and L. pterotenuis (in part). On the other hand, cytb fragments obtained from mtDNA genome lineages KM 272250, MK103894 and MK103895 form a monophyletic group ( Fig. 5B clade II) that is a sister clade of parasites developing roundish host cells plus L. sabrazesi (synonym of Leucocytozoon macleani ) (AB299369) and L. danilewskyi (KY653781) ( Fig. 5B clade II). Both Leucocytozoon sabrazesi and L. danilewskyi are parasites with gametocytes that develop both roundish and fusiform host cells.

Phylogenetic relationships of parasite lineages with sequences isolated from South American birds are depicted in Supplementary Figure S1 (see also Supplementary Table S1). It is noteworthy that the cytb lineage L_GRSQU 02 (Genbank No. MH909275) obtained by PCR from the sample infected with L. grallariae was placed in a well-supported clade along with the lineage KY646032 isolated from the type material of L. pterotenuis (in part) ( Fig. S1 , clade E). Within the clade E, a partial cytb sequence of L. quynzae and the lineage KF874769 obtained from a Peruvian Grallaria erythroleuca specimen were included. Genetic distances between the isolate L_GRSQU_02 (Genbank No. MH909275) and the lineages of parasites previously reported in other species of Grallaridae (KY646032, KY646033, and KF874764) were 0.05, and 0.06 respectively; while it was 0.05 for L. quynzae ( Fig.S1 , Table 5). Furthermore, the linage L_PIRIE 02 (Genbank No. MH909276) fell into a clade composed by parasites infecting other species of Pipreola ( Fig. S1 , Clade IB). Genetic distances between this lineage and the Peruvian lineages isolated from Pipreola (P.) intermedia (Genbank accession No. KF874740, KF874814) and Pipreola arcuata (Genbank accession Nos. KF874701, KF874796) ranged between 0,01 and 0.03 (clade B).

Consistently with phylogenetic reconstructions performed with mtDNA genome, lineages MK 103895 and MK103894 fell into a clade that included L. sabrazesi and L. pterotenuis (in part) ( Fig. 5A , clade II). The genetic distance between the mtDNA genome lineage MK103894 of L. neotropicalis and its sister taxa (KM272250) was 0.25, while the genetic distance of lineage MK103895 L. sabrazesi , L. pterotenuis (in part) and MK103894 were 0.21, 0.17, and 0.25 respectively. The large genetic distances estimated using both the partial mtDNA and the cytb sequences obtained from the new species as well as L. pterotenuis (in part) using the methodologies above mentioned are reported in Table 5.

Figures

Fig. 2. Leucocytozoon neotropicalis sp. nov. from the peripheral blood of its type vertebrate host Greenand-black Fruiteater (Pipreola riefferii) captured at Los Nevados NNP, Colombia. Macrogametocytes (A–E) and microgametocytes (F–I). Black arrows () indicate the deformed host cell nuclei. Parasite nuclei are indicated by white arrow () and nucleoli are shown by double white arrowtips (). Volutin granules are indicated by double black arrowtips () and vacuoles – by white arrowtips (). Uneven cytoplasmic processes may acquire a ribbon-like appearance (asterisk *). Giemsa-stained thin blood films. Scale bar = 10 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 1. Leucocytozoon grallariae sp. nov. from Undulated Antpitta (Grallaria squamigera) captured at Palacio forest in the Chingaza National Natural Park (NNP), Colombia. Immature gametocytes (A–C), macrogametocytes (D–F) and microgametocytes (G–I) in fusiform host cells. Nucleus of host cells (black arrows) possessing mature gametocytes assumes a slender waning moon shape (D–I). Parasite nuclei are indicated by white arrows () and parasite nucleolus – by double white arrow tips (). Host cell cytoplasm is distorted by developing parasites forming a thin rim that develops into the cytoplasmic processes (asterisk *). In mature gametocytes, vacuoles are indicated by white arrow tips (). Volutin granules are indicated by double black arrow tips () and azurophilic granule by black arrow tips (). Giemsa-stained thin blood films. Scale bar = 10 μm.

Fig. 3. Gametocytes developing roundish host cells observed in type material of Leucocytozoon grallariae (A–C) and Leucocytozoon neotropicalis (D–F). Host cell nucleus (black arrows) deformed as a cap re- sembles the parasites of the Leucocytozoon fringilli- narum group. The white arrow indicates the parasite nuclei (), double white arrow tips shows the nu- cleolus (). Vacuoles (white arrow tips) are pre- sent, while volutin granules (double black arrow tips ) are few or absent in some gametocytes. Giemsa- stained thin blood films. Although gametocytes in roundish host cells were observed in both samples, those co-existing with Leucocytozoon grallariae (A–C), are significative smaller than Leucocytozoon sp. ga- metocytes found in type host of L. neotropicalis (D–F) (see Table S3). Scale bar = 10 μm.

Fig. 4. Primer affinity analyses of the primers suggested by Hellgren et al. (2004). Parasites with gametocytes developing roundish host cells are: (a) L. fringillinarum, and (b) L. dubreuili; and parasites with fusiform host cell are (c) L. pterotenuis (in part), (d) L. neotropicalis sp. nov, and (e) L. grallariae sp. nov. An asterisk over the base pair highlights mismatches between the sequences and the primers. Note that primers HaemNR3 and HaemR2L are presented in 3′-5′ sense to fit the parasite sequences.

Fig. 5. (A) A Bayesian phylogenetic hypothesis of Leucocytozoon species constructed only with partial mitochondrial genomes (5485 bp excluding gaps) and (B) partial cytb gene sequences of leucocytozoids. Branch colors indicate the parasite morphology, with green branches representing parasites in fusiform host cells, and blue branches correspond to a species that develops in roundish host cells. Notice that, since parasite mitochondrial genomes (mtDNA) corresponding to the partial cytb fragments of the MH909275 and MH909276 sequences could not be amplified, they were not included in the phylogenetic hypothesis constructed with mtDNA (Fig. 5A). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)