Material and methods

Sampling and culture establishment. Surface water samples from Acapulco Bay were obtained with a Van Dorn bottle (3L), (“Parque de la Reina”, 16° 50’ 58.8” N, 99° 54’ 1.6” W and “Naval”, 16° 50’ 18.33’’ N, 99° 51’ 27.16” W) in November 2021 and March 2024 (Figure 1). Living samples were placed in a dark cooler and transported to the laboratory. Each sample was first filtered through a 20 µm mesh and then passively (by gravity) concentrated using a 3 µm cellulose filter. Isolations of dinoflagellate cells were made with sterile capillary micropipettes using an inverted microscope (Invertoskop, ZEISS) and transferred to a 96 wells plate, previously filled with IMK and GSe culture medium (Daigo´s, Wako, Tokio, Japan). Strains were transferred to larger plates (24 wells), then scaled to 25 ml plastic flasks, to obtain a greater biomass for detailed studies. Culture conditions were temperature 24 °C, salinity 35 and a of 12:12 light/darkness photoperiod.

Figure 1 Map of the Area of study with sampling stations inside Acapulco Bay. The black dots represent sampling sites. 

Microscopy. Observations of living cells were carried out with a bright field light microscope (Olympus BX40F-3, Tokyo, Japan and Leica DMLB, Wetzlar, Alemania), at 40X and 100X. Micrographs were taken with a camera Canon G10 and Nikon Z6 (Tokyo, Japan). Additionally, an epifluorescence microscope (Olympus IX71, Tokyo, Japan) (with 470-490 nm excitation and 510-550 nm emission) was used to observe some details such as the thecal tabulation, number and position of chloroplasts, and size and position of nucleus, with the aid of Calcofluor and 4’, 6-diamidino-2 phenylindole dihydrochloride (DAPI) (Fisher Scientific, Madrid, Spain).

External body scales were observed by TEM (JEOL-JEM-1200) following the method described by ^{Iwataki et al. (2002)}, which briefly consists in placing 0.05 mL of glutaraldehyde fixed culture on a 100-mesh copper grid, which was rinsed three times with distilled water, and then stained with Uranyl acetate (2 %) for two minutes.

Ultrastructure. TEM subsamples were prepared according to Benico et al. (2019). Cells were prefixed in 2 % (v/v) glutaraldehyde and post-fixed in 1% (w/v) OsO₄ for 8 h, followed by dehydration through an ethanol series. This was followed by embedding with low - viscosity resin, sectioning and staining of the sample with uranyl acetate and lead citrate. Observations were carried out with a JEOL-JEM-1200 at an accelerating voltage of 100 kV.

DNA extraction, PCR and sequencing. DNA extraction was carried out with a kit (Zymo Research, USA), following the manufacturer’s instructions. For the amplification of the extracted DNA, a Master Mix (PROMEGA Go Taq^® G2 Hot Start) was used following the PCR conditions of ^{Escarcega-Bata et al. (2021)}. Single-cell PCR was performed directly, following the protocol described by ^{Hernández-Rosas et al. (2018)}, with minor modifications in PCR conditions (^{Escarcega-Bata et al. 2021}). Partial LSU rRNA gene sequences were amplified using primer pairs D1R-F and D2C-R (^{Scholin & Anderson 1994}). The D1-D2 region of LSU was selected due to the coverage of Heterocapsa species in GenBank database to compare our specimens. PCR amplificons were confirmed by electrophoresis with 1 % agarose gel for 20 min at 90 V; staining was performed with GelRed (Biotium). The visualization was carried out using a gel documentation system with UV light. PCR products were stored in -20 °C and sent to the Genomic Sequencing Laboratory, Instituto de Biología, Universidad Nacional Autónoma de México) (Mexico City) for purification and sequencing. LSU sequences were determined for the two chains, using the referenced primers.

Sequence alignment and phylogenetic analysis. The sequences obtained were aligned in BioEdit v. 7.0.5 (^{Hall 1999}) with those of a selected group of species of the genus Heterocapsa from the NCBI (GenBank). Prorocentrum cordatum (Ostenfeld) Dodge and Prorocentrum micans Ehrenberg were used as outgroup. Phylogenetic analyses were performed using Maximum Likelihood (ML) and Bayesian Inference (BI). ML analysis was performed using RAxML software (^{Stamatakis 2006}). Node support was obtained for each branch from 1,000 bootstrap replications. BI analysis was performed using TOPALi version 2 software (^{Milne et al. 2009}). The evolutionary model selected was GTR + G (general time reversible + gamma distribution) determined on the basis of the maximum likelihood ratio test implemented by TOPALi v. 2 software (^{Milne et al. 2009}). Four strings of the Markov Monte Carlo chain (MCMC) were used, starting with a random tree and sampling the data every 1,000 iterations for 5 × 10⁶ generations, discarding the first 25 % of the trees (heating the strings). Distance values per pair (p-distances) and distances corrected with Jukes Cantor model, were calculated using Mega v. 11 (^{Tamura et al. 2011}).

Results

Morphology. i) Heterocapsa borneoensis Teng, Hanifah, Leaw & Lim (Figures 2A-I, 3A-J). References: ^{Hanifah et al. 2022, p. 4}, Figures 1A-L, 2A-O.

Figure 2 A-I. Heterocapsa borneoensis, light and epifluorescence microscopy. A-F. Live cells showing the general outline of the species; pyrenoids (black arrows) and red bodies (white arrows) are shown. G-I. Cells stained with calcofluor and DAPI showing nucleus (N), chloroplast (Cl) and some thecal plates in ventral view. Scale bar: A-I = 5 µm 

Figure 3 A-J. Heterocapsa borneoensis, transmission electron microscopy. A, B, F. Longitudinal section with nucleus (N), Chloroplast (ch), Pyrenoids (Py), Starch grain (S) and mithocondria (m). C. Details of pyrenoid connection with chloroplast. D. Details of the chloroplast with thylakoids in stacks of three (white arrows). E. Trychocyst in longitudinal section. G-J. Body scale morphology. Scale bar: A, F = 3 µm; B = 2 µm; C = 1 µm; G-J = 200 nm. 

Cells solitary and motile with a very thin and transparent theca. Usually, the cells are ovoid to elongate ovoid in shape, with episome and hyposome of nearly equal length, although sometimes the episome may be slightly larger and wider. Measurements are: 10-15 μm in length and 6-10 μm in width (n = 50). The cingulum is wide and excavated, occupying about ¼ of the total length. The episome may be ovoid, showing a very slight or no protuberance in the apical end, at the apical pore complex level, whereas the hyposome is more rounded (Figure 2A-I). The nucleus is subspherical to elongate and is situated near the centre of the cell, more toward the hyposome (Figures 2C, G, H, 3A, F). Photosynthetic, with several (4-8) golden to brown elongate plate-like chloroplasts, and two to four large pyrenoids per cell (Figures 2A-F, 3A-D, F). Various red bodies were also found in the cells (Figure 2A-B, D-F). Few details of the thecal tabulation were seen in epifluorescence microscopy, showing some main plates in ventral view, such as the apical plates 1ʹ, 2ʹ and 5’, the precingular plates 1ʹʹ and 7ʹʹ, the cingular plates C₁, C₆, the postcingular plates 1ʹʹʹ, 5 ʹʹʹ, the antapical plates 1ʹʹʹʹ and 2 ʹʹʹʹ, and two sulcal plates Sa and Sp (Figure 2 I).

The general outline of the cells was also observed by TEM, with subequal epi- and hyposome, and wide and excavated cingulum (Figure 3A, F). Observations of the nucleus confirmed its position in the cell, slightly eccentric, toward the hyposome, and elongate shape (Figure 3A, F). Elongate plate-like chloroplasts were observed, distributed in the periphery of the cell, and showing some details of the thylakoids (Figure 3A-D, F). Observation of pyrenoids was especially interesting, as they were large and connected to the chloroplasts, always surrounded by almost continuous starch sheaths, whereas tubular invaginations were absent (Figure 3B-C, F); cells showed trichocysts (Figure 3E).

Body scales showed a pattern with a roundish to hexagonal basal plate finely perforated, ornamented by a central spine and six peripheral uprights, interconnected by radial and peripheral bars; the connections between peripheral uprights form three radial pentagons. Perforation of the basal plate was highly variable, with coarse to finer perforations were also present in the basal plate. The diameter of the scales ranged from 320 to 440 nm (Figure 3G-J).

ii) Heterocapsa iwatakii Teng, Hanifah, Leaw & Lim (Figure 4A-D). References: ^{Hanifah et al. 2022, p. 10}, Figures 5A-E, 6A-K.

Figure 4 A-D. Heterocapsa iwatakii, light and epifluorescence microscopy. A, B, D. Live cells showing the general outline of the species; pyrenoids (black arrows), red bodies (white arrows) and chloroplast (Ch) are shown. B and C. Spherical nucleus (N) in episome. Scale bar: A-D = 5 µm. 

Solitary and motile cells, with an ellipsoid to elongated ellipsoid shape, the episome is larger than the hyposome, the episome is conical and the hyposome is hemispheric (Figures 4A-D). Measurements are: 8-12 μm in length and 6-8 μm in width (n = 50). Deep and excavated cingulum. The nucleus is spherical and located in the episome (Figure 4B-C). The species is photosynthetic, with several golden and parietal chloroplasts (Figure 4A-B, D). Cells have a single spherical pyrenoid, located in the central part of the cell or in the hyposome (Figure 4A-D). The presence of red bodies was recorded (Figure 4B), however, this was not found in all cells, only in those that were growing exponentially.

iii) Heterocapsa sp. (Figure 5A-F). Cells solitary and motile, ellipsoidal to ovoid in shape, with a slightly conical episome and hemispheric hyposome (Figure 5A-F). Measurements are: 15-18 μm in length and 9-12 μm in width (n = 50). The episome is slightly larger than the hyposome and the cingulum is excavated. Photosynthetic species with several golden and parietal chloroplasts (Figure 5A-B). One to three spherical pyrenoids located in the episome (Figure 5A-C). Nucleus ellipsoidal to ovoid centrally located, between the episome and the hyposome (Figure 5C, F). Occasional presence of red bodies, variable in shape and size (Figure 5 B).

Figure 5 A-F Heterocapsa sp. A-B. Light microscopy. General morphology of live cells with pyrenoids (black arrows) and red bodies (white arrows). C, F. Fixed cells in epifluorescence with elongated nucleus (N). D-E. Plate pattern in ventral view. Scale bar: A-F = 5 µm. 

The ventral view of the species observed with calcofluor showed some thecal plates, confirming the arrangement of them, the presence of five apical plates and seven precingular plates and the prominence of the anterior and posterior sulcal plates (Sa and Sp) (Figure 5D-E).

Phylogenetic analysis. Four partial LSU rDNA sequences were obtained for Heterocapsa. borneoensis and H. iwatakii. The final LSU alignment included 47 sequences, with a length of 900 bp. Phylogenetic reconstruction from the LSU dataset (with maximum phylogenetic support ML = 100 %, BI = 1), was made with sequences of species of the genus Heterocapsa. The resulting phylogeny showed three main clades (I-III, Figure 6). Clade I, with high phylogenetic support (ML = 78 %, BI = 0.84). Clade II, with low phylogenetic support (ML = 78 %, BI = 0.60), and Clade III, with low phylogenetic support (ML = 0.59 %, BI = 0.61). The sequences generated in this study were grouped in two subclades of Clade III (A and B). Subclade A, with high support (ML = 99 %, BI = 0.97) where sequences of H. iwatakii from the Acapulco Bay and those of the type species clustered together. Subclade B (ML = 81 %, BI = 0.80) clustered H. borneoensis together with several clades belonging to H. bohaiensis, H. limii, H. pygmaea and Heterocapsa sp. sequences.

Figure 6 Bayesian inference (BI) and Maximum likelihood (ML) tree based on LSU rADN sequences from species of Heterocapsa. BI (right) followed by ML bootstrap values (left) on branches. Asterisk indicated full support (ML = 100 %, BI= 1). Sequences generated in this study are highlighted in a green box. 

The genetic distances found between sequences of H. borneoensis and species of the genus Heterocapsa can be seen in Table 2, where uncorrected distances and corrected distances are shown. Distance values between the sequences of H. borneoensis, were not included because the variation between them was close to zero (0.001 %), the same as in H. iwatakii.

Discussion

Morphology. Four strains corresponding to the thecate dinoflagellate species Heterocapsa borneoensis (Strains: C9HBAB-C10HBAB), H. iwatakii (Strain: FPMBA01) and Heterocapsa sp. (Strain: B8HSPAB), were obtained from the Acapulco Bay, Mexico, which represents the first record of the species Heterocapsa borneoensis and H. iwatakii since their original description from the coast of Sarawak, Borneo, Malaysia (^{Hanifah et al. 2022}). The general and detailed morphology of these two species, studied by light microscopy, and in the case of H. borneoensis also by electron microscopy (e.g, the body scales), described in this study completely agree with the morphological characters described by ^{Hanifah et al. (2022)}.

Morphologically, H. borneoensis is similar to H. bohaiensis, overlapping in size, (10-15/ 10-17 µm long and 6-10/ 6-12 µm wide, respectively), and the main difference between the two species is the shape of the hyposome (Figure 1A-B; ^{Xiao et al. 2018}). However, in the strain established by ^{Hanifah et al. (2022)}, the ellipsoidal shape of H. bohaniensis strongly resembled H. borneoensis. In this study, the shape and size match the original description of H. borneoensis. The presence of multiple pyrenoids is another shared character, although in the description of ^{Xiao et al. (2018)}, H. bohaiensis showed 1 to 3 pyrenoids, whereas ^{Hanifah et al. (2022)} reported 1 to 4 pyrenoids, the same number of pyrenoids reported for H. borneoensis (2-4 pyrenoids).

Other species show multiple pyrenoids, as H. pygmaea (2 pyrenoids; ^{Iwataki et al. 2004}, ²⁰⁰⁸, ^{Hernández-Becerril et al. 2010}), which differs from H. borneoensis in the displacement of the cingulum and the shape and location of the nucleus, H. limii (2 pyrenoids; ^{Hanifah et al. 2022}), which has a close relationship with H. borneoensis. However, the differences are the shape of the nucleus, the invariable presence of 2 pyrenoids and the absence of red bodies in H. limii. Finally, H. huensis Iwataki & Matsuok has 1-3 pyrenoids, but differs in the general form of the cell (^{Iwataki et al. 2009}, ^{Hanifah et al. 2022}).

^{Iwataki (2008)} provided a general identification guide, in which the first character to consider is the proportions of the episome and hyposome, as well as the size and shape of the cell. However, the main character that allows differentiation between species of the genus Heterocapsa is the morphology of external body scales (^{Iwataki 2008}). Considering this fact, we found the difference between H. borneoensis and H. bohaiensis to be in the morphology of the scales, which differ in size (320-460 nm and 300-350 nm), outline (triangular and circular), crests (six and three), spines and peripheral arches respectively. Additionally, the identification of H. borneoensis is also supported by LSU rDNA phylogenetic inference.

The observations of the ultrastructure of H. borneoensis made in this study coincided with the ultrastructure of other species of the genus: peripheral chloroplasts with pyrenoid connection, pyrenoid surrounded by starch grain, mitochondria, nucleus with condensed chromosomes and the presence of trichocysts (^{Horiguchi 1997}, ^{Iwataki et al. 2003}, ^{Rintala et al. 2010}, ^{Tillmann et al. 2017}, ^{Xiao et al. 2018}, ^{Sunesen et al. 2020}, ^{Benico et al. 2021}). In terms of ultrastructure, within the genus Heterocapsa, some species have tubular invaginations in the pyrenoid matrix, which have an unclear function (^{Iwataki et al. 2003}, ^{Benico et al. 2021}). In H. borneoensis, a relatively small species, tubular invaginations are absent, as these structures are thought to occur only in large species (^{Iwataki et al. 2003}).

Heterocapsa iwatakii is a small species of the genus, ^{Hanifah et al. (2022)} described in detail the characteristics that distinguish it from those morphologically similar species, H. horiguchii, H. busaensis and H. ovata, in which, despite sharing characters, H. iwatakii can be distinguished due to its size (7-12 um length and 5-8 um width) as it does not overlap with any of the above-mentioned species. The strain from Acapulco Bay coincides with the original description of ^{Hanifah et al. (2022)}: episome larger than hyposome, round nucleus located in the episome, a pyrenoid located in the hyposome and the presence of red bodies. Unfortunately, it was not possible to observe the morphology of the scales, but its identification could be corroborated by the LSU rDNA phylogenetic inference.

The strain (B8HSPBA) was studied only by LM, using live and fixed material. This strain is morphologically close to H. borneoensis, H. bohaiensis and H. limii, all these species overlap in size and overall cell shape. Heterocapsa sp. has a higher resemblance to H. bohaiensis (Figure 5A-B; ^{Hanifah et al. 2022}) since both species have a larger episome. The next character to consider is the number and position of the pyrenoids, in which the strain showed 2 pyrenoids located in the episome and only one cell was observed with 3 (Figure 5B). Following the identification guide provided by ^{Hanifah et al. 2022} (Figure 12A-B), only five species have multiple pyrenoids: H. pygmaea, which has a spherical nucleus and has no red bodies (^{Loeblich et al. 1981}, Morril & Loeblich 1981, ^{Iwataki et al. 2004}, ^{Hernández-Becerril et al. 2010}), H. limii, which differs from Heterocapsa sp. by the presence of a third pyrenoid and the presence of red bodies (^{Hanifah et al. 2022}), H. huensis, which is larger and differs in cell shape (^{Loeblich et al. 1981}, ^{Hanifah et al. 2022}), whereas H. borneoensis and H. bohaiensis have 1 to 4 pyrenoids. Heterocapsa sp. differs from H. borneoensis in the proportion of episome to hyposome and in the position and shape of the nucleus, which is found in the episome in H. borneoensis and is centrally located in Heterocapsa sp. (Figure 5F), although in some cells it can be seen covering more area of the hyposome (Figure 5C). The plate pattern of Heterocapsa sp. coincides with H. borneoensis (^{Hanifah et al. 2022}). The morphological resemblance and overlap of characters from Heterocapsa sp. with other species of the genus have made identification difficult and unfortunately it has not been possible to obtain sequences or observe scale morphology. This overlap of characters can be explained by the deterioration or modification of the morphology of Heterocapsa sp. in fixed samples.

Molecular phylogeny. Following both the morphological observations and the molecular phylogeny obtained (LSU rDNA sequences), the occurrence of the species H. borneoensis and H. iwatakii in the eastern Pacific coast was confirmed. The phylogeny of the analyzed data set showed that the four sequences obtained in this study (OQ946954-OQ946955 and PP911072-PP911073), were grouped into separate clades with sequences from H. borneoensis and H. iwatakii respectively, collected from coasts of Malaysia (^{Hanifah et al. 2022}). However, the topology of the LSU tree in ^{Hanifah et al. (2022)}H. borneoensis, showed to be paraphyletic. In our tree, the clade of H. borneoensis (subclade B, Figure 6) is well differentiated with maximum support and in the same way as in ^{Hanifah et al. (2022)}, the sequence GBHetBol (LC621347, Genebank access number) of ^{Benico et al. (2021)} clustered together with the sequences of H. bohaiensis. The divergence of the sequences of H. borneoensis from this study compared to those provided by ^{Hanifah et al. (2022)} is zero, implying the same species.

In H. iwatakii, our sequences are grouped together with those provided by ^{Hanifah et al. (2022)} (subclade A, Figure 6), and the divergence of the sequences compared to those of the type species is zero. Therefore we interpreted the correspondence of one and the same species. The position of H. iwatakii in the LSU tree is provisional as shown by ^{Hanifah et al. (2022)}, given the lack of sequences of closely related species, this information is sufficient to determine the identity of the species using the LSU marker.

Distribution. The genus Heterocapsa has a wide global distribution (^{Hanifah et al. 2022}, ^{Tillmann et al. 2017}, ^{Xiao et al. 2018}, ^{Sunesen et al. 2020}), some potentially toxic species such as H. rotundata and H. triquetra, usually have a greater range of distribution (^{Lee et al. 2019}). In the case of the western Pacific there is a greater number of studies, given the ecological importance of this genus due to the potential of some species to form harmful algal blooms (HAB’s) (e.g, H. circularisquama, H. bohaiensis; ^{Matsuyama et al. 1995}, ^{Iwataki 2008}, ^{Choi & Kim 2021}). Studies of the genus in Mexican coasts are scarce, and it may be justified that there are no records of algal blooms that have been harmful or generated economic losses. For H. borneoensis and H. iwatakii, current information is limited, so it is not possible to determine its global distribution, because only two records are known yet: Borneo, Malaysia and Acapulco, Mexico (^{Hanifah et al. 2022} and this study).

The confirmation of the species Heterocapsa borneoensis and H. iwatakii in the eastern Pacific Ocean by morphological and molecular identification suggests a wide range of distribution of this potentially harmful genus, which has been little studied along this coast. The information provided in this study contributes to the current knowledge of this genus, and it is necessary to maintain monitoring and research on these species, so that future analyses can provide a more accurate estimate of the diversity in Mexican waters.