DAPI staining and DNA content estimation of uncultivable microbial eukaryotes’ nuclei (Arcellinida and Ciliates)

Abstract

Though acting as a major component of eukaryotic biodiversity, many microbial eukaryotes remain poorly studied, including the focus of the present work, testate amoebae of the order Arcellinida (Amoebozoa) and non-model lineages of ciliates (Alveolata). In particular, knowledge of their genome structures and changes in genome content over their often-complex life cycles remains enigmatic. However, the limited available knowledge suggests that microbial eukaryotes have the potential to challenge our textbook views on eukaryotic genomes and genome evolution. In this study, we developed protocols for DAPI (4’,6-diamidino-2-phenylindole) staining of Arcellinida nuclei and adapted protocols for ciliates. In addition, image analysis software was used to estimate the DNA content in the nuclei of Arcellinida and ciliates and to compare them to measurements of well-known model organisms. The results demonstrate that the methods we have developed for nuclear staining in these lineages are effective and can be easily applied to other microbial eukaryotic groups by adjusting certain stages in the protocols.

Introduction

The bulk of all biodiversity, and by extent eukaryotic diversity, is microbial. Microbial eukaryotes (i.e. protists) exhibit diverse and dynamic genome structures. Their genomes span a large range of sizes, from little over 2 Mb in some microsporidians to over 670,000 Mb in Amoeba dubia (reviewed in: McGrath and Katz, 2004). They also exhibit a number of unusual features, such as nuclear dualism (e.g. McGrath and Katz, 2004; Prescott, 1994), extensive genome fragmentation (e.g. Huang and Katz, 2014; Swart et al., 2013), and genome increase/reduction (Parfrey et al., 2008). Despite their unusual characteristics, studies of nuclear structures and genome sizes in microbial eukaryotes remain limited (e.g. Grattepanche et al., 2018). Given their diversity, characterizing the nuclear architectures and genome structures of microbial eukaryotes furthers our understanding of eukaryotic biodiversity and, more broadly, evolutionary principles.

Basic features such as nuclear number, structure, and estimated genome sizes are especially under-studied in testate amoebae (Arcellinida, Amoebozoa) and ciliates (Alveolata), the focal clades in this study. Testate amoebae are single-celled eukaryotes that build tests (shells) either from environmental materials or through biosynthesis (e.g. Mitchell et al., 2008; Nikolaev et al., 2005). These tests have been used traditionally as a feature of species identity. Arcellinida are mostly found in freshwater terrestrial habitats (e.g. Mitchell et al., 2008; Mitchell and Meisterfeld, 2005; Nikolaev et al., 2005) and since they are sensitive to environmental changes they are used as bioindicators for changing environmental conditions (Mieczan, 2009; Swindles et al., 2016). To date, we have only limited knowledge on Arcellinida genomes, mostly from transcriptome analyses (e.g. Lahr et al., 2019; Weiner et al., 2020) and no reference genome exists, to the best of our knowledge. Studies that have explored genomes in Amoebozoa have done so in pathogenic amoeba and slime molds (e.g. Bloomfield, 2016; Chávez-Munguía et al., 2006; Mukherjee et al., 2009), which are likely >500 million years divergent from Arcellinida. Notably, only few studies have attempted to explore Arcellinida life cycle stages due to the fact that they are uncultivable. However, even though life cycle stages of Arcellinida are not understood in depth, Cavallini (1926) provides evidence which suggest that testate forms of Arcellinida produce naked offsprings. In addition, Volkova and Smirnov (2016) showed that if Arcella are removed from their tests, they are capable of generating new tests although subsequent division was not observed.

Another group of single cell eukaryotes, ciliates, has challenged the traditional views of eukaryotic nuclear structure. Ciliates are characterized by the presence of hair-like cilia and nuclear dimorphism, meaning presence of somatic macronuclei and germline micronuclei, within a single cell (e.g. Prescott, 1994). The somatic macronuclei are responsible for the majority of cellular activity, while germline micronuclei are quiescent for most of the life cycle. Within ciliates, there is a great diversity of nuclear structures. Our target organism Loxodes belongs to the class Karyorelictea, which is unique among ciliates in that their somatic macronuclei do not divide (Raikov, 1985). When the cell undergoes division, at least one somatic macronucleus is passed directly to the daughter cells, and germline micronuclei divide and differentiate to form new somatic macronuclei. Somatic macronuclei are believed to persist through several generations before degradation, and may experience changes in DNA content and morphology as they age (i.e. being kept for several generations; Raikov, 1985; Yan et al., 2017). Although the number and structure of nuclear groups in Loxodes species has been fairly well classified (Raikov, 1985; Ron and Urieli, 1977), these observations were conducted in an era before modern microscopy techniques.

Fluorescence microscopy is a powerful method that allows for detailed observations of nuclear structure in microbial eukaryotes as well as estimates of genome content and ploidy level (Bellec et al., 2014; Maurer-Alcalá and Katz, 2016; Parfrey and Katz, 2010a; Parfrey and Katz, 2010b; Wancura et al., 2018). The method of quantifying fluorescence to estimate genome content has been used in studies of plant species (Cousin et al., 2009; Loureiro et al., 2006; Suda and Trávníček, 2006), Foraminifera (Allogromia laticollaris, Parfrey and Katz, 2010b), ciliate species (Blepharisma americanum, Wancura et al., 2018) and myxomycetes (Therrien & Collins 1976; Ritch & Therrien 1988). However, previous studies that have used these approaches to analyze nuclear structures in testate amoeba have failed despite numerous attempts to modify protocols (e.g. Burdikova et al., 2010). Therefore, according to our knowledge, a reliable method for revealing nuclear architecture in testate amoebae has not yet been described.

In this study, we investigate the nuclear structure and estimate the genome content of two genera of microbial eukaryotes: the testate amoebae Hyalosphenia and the ciliate Loxodes. Here, we describe the staining protocols for testate amoebae and ciliates that were developed and adapted for this work, respectively, following methods from Parfrey et al., (2010b) and Wancura et al., (2018). We rely on DAPI (4’,6-diamidino-2-phenylindole), a standard nuclear stain but one with known limitations, notably that it preferentially binds to A-T rich chromosomal regions and thus may provide inaccurate estimates of genome content (Noirot et al., 2002). Furthermore, we use both protocols to stain organisms with known genome sizes and estimate the genome content of our organisms.

Material and Methods

Sample collection

Samples of freshwater and sediment were collected for Loxodes cells and Sphagnum moss for isolating Hyalosphenia species, both from Hawley bog (Hawley, MA) between June and November 2019. Loxodes cells were picked from the water samples using a hand pipette. The Sphagnum moss was washed in the lab using filtered (2 μm filter) bog water and poured over a 300 μm filter to isolate testate amoeba from larger plant material. The amoeba cells were then placed in a Petri dish from which they were individually picked using a hand pipette, washed again in filtered bog water and placed into a 3.0 ml tube.

Cell fixation and DAPI staining

Loxodes cells were stained following a protocol modified from Wancura et al. (2018). Cells were placed on Superfrost slides (Fisher, Waltham, MA) in 200 μL Volvic water. They were fixed directly on the slides with a mixture of 20% Paraformaldehyde (PFA), RNAlater, and Trizol for 30 minutes, and washed 3 times for 5 minutes with 1x Phosphate-buffered saline (PBS) buffer. Fixatives and buffers were added and removed from the slide using micropipettes. Cells were then incubated in 40 μL of 0.5 % Triton-X for 25 minutes and washed again 3 times for 5 minutes with 1x PBS buffer. The slides were then incubated in a pre-hybridization mix consisting of a 1:1 ratio of Formamide and 2x saline-sodium citrate buffer (SSC) for 30 minutes at room temperature, and hybridized in a solution of Nuclease-Free Water, Formamide, and 20x SSC in a 5:4:1 ratio for 1 hour at 37 °C. Slides were washed three times with 2x SSC for 5 minutes, and incubated in DAPI (5 mg/ml, 1:1000 or 1:100 dilution; Fisher) for 5 minutes. DAPI was washed off 3 times for 2 minutes each with 1x PBS. A drop of SlowFadeGold (Invitrogen, Carlsbad, CA) was then added. The slides were sealed with a coverslip and nail polish and kept in the dark at 4 °C before being examined under a microscope.

Arcellinida cells were fixed in 400 μL of 0.2 M, pH 7.2 PHEM (PIPES-Hepes-Ethylene glycol tetraacetic acid (EGTA)-MgCl₂) buffer (Electron Microscopy Sciences Hatfield, PA United States) in a microcentrifuge tube and were incubated for 2 hours at room temperature. After 2 hours, the fixed cells were gently spun on a mini centrifuge for 30 seconds to form a pellet at the bottom of the tube and the supernatant was removed. Fixed cells were then washed twice in 400 μL 1x PBS buffer. After the washing step, the cells were incubated for two hours in 400 μL of 10 % Triton-X for membrane permeabilization. After the incubation period, Triton-X was removed and the cells were washed twice in 400 μL 1x PBS. Fixed and permeabilized cells were incubated in 100 μL DAPI (5 mg/ml, 1:100 dilution; Fisher) for 2 hours in darkness. DAPI was then washed off twice using 400 μL 1x PBS. Stained cells were placed on a Superfrost slide (Fisher) with a drop of Slow Fade Gold (Invitrogen), covered with a cover slip and sealed with nail polish.

Both of these newly-developed fixation protocols (20% PFA for Loxodes vs PHEM buffer for Hyalosphenia) were also applied to Saccharomyces cerevisiae cells, Homo sapiens epithelial cells and Allium cepa root tip cells in order to understand the influence of fixation methods on staining intensity, the ratio of nuclear fluorescence to DNA content in Arcellinida and Loxodes, and to demonstrate that our protocols work on a variety of eukaryotic cells.

Fluorescent Imaging

Fluorescent images of all cells were collected on a Leica TCS SP5 laser-scanning confocal microscope (Leica, Mannheim, Germany) using a 63x oil immersion objective. A UV laser with an excitation wavelength of 405 nm, set to 20% intensity, was used to collect DAPI signals, and an argon laser with an excitation wavelength of 488 nm, set to 20% intensity, was used to collect differential interference contrast (DIC) images. Z-stacks of Loxodes, A. cepa, S. cerevisiae, and H. sapiens were collected at a resolution of 1024 × 1024 with an acquisition speed of 200 Hz, a line average of 4, and a step size of 0.13 μm. Z-stacks of Arcellinida were collected at a resolution of 1024 × 1024 with an acquisition speed of 200 Hz, a line average of 2, and a step size of 0.13 μm. The gain setting varied slightly across all images to adjust for variability in cell fixation and DAPI penetrance. We examined each cell’s morphology and image quality in the DIC images, and only considered those that were fixed and imaged well for our analyses of nuclear size, fluorescence, and DNA content. We counted the number of nuclei present in each cell, inspected them for the presence of nucleoli and measured the diameter of the nuclei using ImageJ software (Rasband, W.S. ImageJ. U.S National Institutes of Health, Bethesda, MD; Table 1).

Table 1.

Summary of observations and measurements conducted on the cells of each of the focal organisms in this study (Loxodes, Hyalosphenia papilio and Hyalosphenia elegans) as well as the standards (Homo sapiens, Allium cepa and Saccharomyces cerevisiae). Atypical cells are those with more than two macronuclei for Loxodes, and more than one nucleus for H. papilio. Average nuclear diameter is calculated from cell volume that measured in NIS-Elements Advanced Research software. MAC = somatic macronucleus, MIC = germline micronucleus.

Organism	# Cells measured	# Nuclei measured	# Atypical cells	Avg. Nuclear Diameter (μm)
Loxodes sp. MAC	29	59	1	6.6
Loxodes sp. MIC	29	67	5	3.5
Hyalosphenia papilio	25	35	5	18.3 (uninucleate) 11.4 (multinucleate)
Hyalosphenia elegans	3	3	0	10.7
Saccharomyces cerevisiae	63	63	0	1.8
Homo sapiens	43	43	0	7.8
Allium cepa	19	19	0	11.1

DNA Content Estimation

In addition to nuclear number and diameter, we also measured the fluorescence intensities and nuclear volumes for Loxodes macronuclei, Loxodes micronuclei and the nuclei of Hyalosphenia papilio, Hyalosphenia elegans, S. cerevisiae, H. sapiens, and A. cepa (Table 1; Table 2). Z-stacks of nuclei were analyzed using the General Analysis 3 feature of NIS-Elements Advanced Research software (Nikon, Tokyo, Japan). The threshold setting was manually determined for each z-stack analyzed to ensure that nuclear volumes were defined accurately before measurement. For each nucleus, the volume, total fluorescence intensity (measured in K), and mean fluorescence intensity were measured (Table 2). S. cerevisiae, H. sapiens, and A. cepa nuclei were used as standards for comparison of Loxodes and Arcellinida measurements to assess the variability in fluorescent intensities produced by DAPI staining, as the genome content of these cells is well known.

Table 2:

Average fluorescence intensity (in thousands of fluorescence units), average genome size estimates with standard deviation, and range of estimated DNA content observed from standards (Saccharomyces cerevisiae, Homo sapiens and Allium cepa) and the focal microbes used in this study. DNA contents of Loxodes sp. and Hyalosphenia papilio were estimated based on the average ratio of fluorescence to DNA content across the three standards. MAC = somatic macronucleus, MIC = germline micronucleus.

Organism	Avg. fluor. (K)	Average DNA content (Mb) ± SD	Min. DNA content (Mb) – Max. DNA content (Mb)
Loxodes sp. MAC	2,120 ± 1,037	3,500 ± 1,732	705 – 7,325
Loxodes sp. MIC	841 ± 555	1,400 ± 927	182 – 3,782
Uninucleate Hyalosphenia. papilio	15,032 ± 10,634	20,900 ± 17,759	4,407 – 61,021
Multinucleate Hyalosphenia. papilio	12,538 ± 4,915	11,200 ± 8,209	1,872 – 32,727
Saccharomyces cerevisiae	98 ± 39	13 ± 5	6 – 25
Homo sapiens	3,139 ± 1,031	3,300 ± 1,084	953 – 5,216
Allium cepa	4,145 ± 1,249	15,876 ± 4,783	8,376 – 24,067

The ratio of fluorescence to DNA content was calibrated following methods from Wancura et al. (2018). For each standard, the ratio of fluorescence to DNA content was calculated using the average measurement of nuclear fluorescence in that organism and its published genome size. The average of these three calculations was used as the final ratio by which we estimated the DNA content in Loxodes and Hyalosphenia nuclei.

Results

Protocols for fixation and DAPI staining of uncultivable microbial eukaryotes

We developed protocols to successfully stain the nuclei of two lineages of uncultivable microbial eukaryotes: the ciliate genus Loxodes (Ciliophora: Karyorelictea) and the testate amoeba genus Hyalosphenia (Tubulinea: Arcellinida). Steps for these protocols involve isolation of cells from nature, and fixation in buffers that vary between the two lineages (Fig.1). Protocol development required many trials as each species requires specific fixation methods. For example, Loxodes cells burst when spun in tubes or immersed in ‘standard’ fixatives such as ethanol or methanol, but we demonstrate that their nuclei can be stained following fixation in 20% PFA and membrane permeabilization in a low concentration of Triton-X. In contrast, Arcellinida cell membranes remain impermeable in many common fixative chemicals including PFA, ethanol, and methanol; instead, we found that PHEM buffer followed by membrane permeabilization using Triton X allows staining of Arcellinida nuclei. The robust methods we developed worked for visualizing nuclear number and structures in our study organisms and, in addition, we applied them to estimate nuclear volume and DNA content of Arcellinida and Loxodes.

Figure 1.

Flowchart showing the main methodological steps used in DAPI staining of both, Hyalosphenia and Loxodes. The drawings in the center illustrate the sampling sites for the uncultivable focal taxa Hyalosphenia and Loxodes, respectively. The two protocols for Arcellinida and ciliates mainly differ in the fixation step (Step 2, PHEM buffer for Arcellinida vs. 20% PFA for Loxodes).

As control organisms for estimates of genome size, we stained the nuclei of Saccharomyces cerevisiae, Allium cepa and Homo sapiens cheek cells according to both protocols: the PFA protocol developed specifically for Loxodes and the PHEM protocol developed specifically for Hyalosphenia. We observed similar fixation quality, cell morphology, and fluorescent intensity among high-quality slides of standard cells imaged according to both protocols (Table S1, Fig. S2). On cells from slides with spurious issues unrelated to the specific staining protocol, we measured artificially low nuclear fluorescent intensity (Table S1, Fig. S2). As such, we selected the cells from high-quality slides with the best morphology and image quality to use in our analyses. Therefore, for the final estimates of DNA content, we used S. cerevisiae cells stained according to the Loxodes protocol, while the chosen A. cepa and H. sapiens cells were stained according to the PHEM protocol for Hyalosphenia. In our analyses, we omitted S. cerevisiae cells generated from the PHEM protocol for Hyalosphenia, because the cells were so densely arranged on the slide—as seen through DIC images—that their DAPI signal was compromised and they posed an interference to the measurements. We omitted A. cepa cells stained according to the PFA protocol for Loxodes because cells on some slides were obscured by debris, which affected DAPI penetration. H. sapiens cells from both protocols all came from high-quality slides and yielded similar fixation quality, cell morphology, and fluorescent signal across both protocols (Table S1, Figure S2). We chose to use H. sapiens cells stained according to the PHEM protocol for Hyalosphenia in our analyses because the positions of these cells lent themselves to more effective detection of nuclear volume with our image analysis software than some cells stained according to the PFA protocol for Loxodes.

Nuclear number and structures in Loxodes and Hyalosphenia

In total, we imaged and analyzed the nuclei of 29 Loxodes cells, 25 Hyalosphenia papilio cells, 3 Hyalosphenia elegans cells, 63 S. cerevisiae cells, 43 H. sapiens cells, and 19 A. cepa cells (Table 1). Our results show that Loxodes cells generally have two nuclear groups, each consisting of a spherical somatic nucleus and a germline nucleus (Fig. 2). Our observation of the nuclear architecture in Loxodes is consistent with previous studies (eg.Raikov, 1982; Raikov, 1985). The germline nucleus is smaller in size (3.5 μm in diameter on average, Table 1), appears to be evenly stained and has a stronger fluorescent signal, while the somatic nuclei have a greater diameter (6.6 μm on average, Table 1) and show a much fainter and uneven DAPI signal (Fig. 2). A large unstained round area (about 3.6 μm in diameter on average) is located in the center of most somatic nuclei, making them appear ring-shaped in the fluorescent images (Fig. 2). We suggest that this unstained area represents the nucleolus, in accordance with Raikov (1985), who noted that this area stains intensely for RNA and protein. We never observed more than one nucleolus in a single nucleus, and in five of 29 cells, a smaller nucleolus or even no obvious nucleolus was detected. However, in one of 29 cells we observed more than four somatic nuclei and in five of 29 cells we observed more than two, and a maximum of five, germline nuclei (Table 1).

Figure 2:

Exemplar cells of Loxodes sp. (A, B), Hyalosphenia elegans (C, D) and H. papilio (E, F) successfully stained with DAPI and the corresponding DIC images. Red arrows show the location of nuclei in the DIC images. MIC: germline micronuclei; MAC: somatic macronuclei. Scales bars: 50 μm.

In both species of Hyalosphenia: H. papilio and H. elegans, we observed a single nucleus. H. papilio nuclei appear spherical and range from 8.0–30.8 μm (18.3 μm on average) in diameter within uninucleate cells (Table 1). H. elegans nuclei also appear spherical, however they are generally smaller than H. papilio nuclei, ranging from 7.3–14.3 μm (10.7 μm on average) in diameter (Table 1). The location of the nucleus varied in different cells: in some cells the nucleus was in the center of the cell while in other cells it was at the edge, close to the shell (Fig. 2), though location may be driven by fixation and cell orientation on the microscope slide. In the DIC images, nuclei are discernable in only some H. papilio and are easier to see in H. elegans (e.g. Fig. 2). In five of 25 H. papilio cells, we saw multiple nuclei ranging in size from 8.5 −17.2 μm in diameter (11.4 μm on average), with up to nine in one case (Fig. S1, Table 1).

Estimates of nuclear size and DNA content in target species

In the present study, we report DNA content estimates for Hyalosphenia papilio, both the somatic macronuclei and germline micronuclei of Loxodes, and the three “standards” onion, yeast and human cheek cells. The fluorescence of our three standard organisms’ nuclei was consistent with their relative genome sizes, i.e., A. cepa has larger genomes compared to H. sapiens and then S. cerevisiae (e.g. Palazzo and Gregory, 2014). We exclude the measurements of H. elegans cells for this analysis, because of its small sample size. Interpolating from the standards with known genome size, we estimate that DNA content in the Loxodes somatic macronuclei (3,500 ± 1,732 Mb) is approximately 2.5 times higher than the DNA content in the germline micronuclei (1,400 ± 927 Mb; Fig. 4, Table 2). Additionally, in five of our 29 Loxodes cells, the measurement of one macronucleus is more than twice the estimated DNA content of the other, which may indicate the ‘age’ difference between the two macronuclei. For H. papilio, we estimate a genome size of 20,900 ± 17,759 Mb for uninucleate cells and 11,200 ± 8,209 Mb for multinucleated cells, which is a very large size and surpasses even the onion genome (15,876 ± 4,783 Mb; Fig. 4, Table 2). This result is consistent with our observations of nuclear morphology, in which the H. papilio nuclei were by far the largest in terms of diameter and volume (Table 2, Table S2). Multinucleated and uninucleate H. papilio cells not only differed in terms of nuclear number, but also in size and DNA content (Fig. 4, Table 2). Uninucleate H. papilio nuclei measured ~ 830 μm³ on average and nuclei of multinucleated cells reached only half the size with ~ 474 μm³ on average (Table S2). DNA content for uninucleate cells was ~21 Gb while multinucleated cells had ~11.2 Gb (Table 2).

Figure 4:

Estimates of DNA content from study organisms, Loxodes sp. and Hyalosphenia papilio, plus control organisms (Homo sapiens, Allium cepa, Saccharomyces cerevisiae). (A) Scatter plot showing a linear relationship between total nuclear fluorescence intensity (F) and nuclear volume (in μm³). (B) boxplot showing estimates of DNA content (bp) for all organisms (see Table 2).

Discussion

Methods development for nuclear staining in uncultivable microbial eukaryotes

We have successfully developed protocols for nuclear fluorescent staining in members of two distinct clades of uncultivable eukaryotic microbes, Loxodes (Ciliophora) and Hyalosphenia (Arcellinida) with the emphasize on effective fixation. PHEM buffer, which acts as fixative in Hyalosphenia species, has been shown to be an excellent fixative agent for marine invertebrates (Montanaro et al., 2016) and for foraminifera (Parfrey and Katz, 2010b; Weber and Pawlowski, 2013) because of its ability to permeabilize tough membranes while preserving cell morphological structures. It is also noticeable that though Loxodes cells can be fixed successfully in multiple solutions used for morphological studies (e.g. silver staining), such as osmium tetroxide (Finlay and Berninger, 1984), Nissenbaum’s sublimate mixture (Bobyleva et al., 1980), Lugol’s iodide, and mercuric chloride (Sime-Ngando et al., 1990), DAPI staining protocols that can successfully reveal nuclear structures in other ciliate groups (e.g. Bellec et al., 2014; Sun et al., 2009) cannot be directly applied to Loxodes. The cells burst when incubated in common fixatives used for fluorescent microscopy work, such as ethanol and methanol (unpublished data), while Arcellinida cells are difficult to penetrate by these fixatives. This indicates that cell membrane properties vary considerably. Therefore, potential adjustment might be required when using the present respective protocols to other related organisms.

Nuclear features and estimates of genome size in Hyalosphenia and Loxodes

Our analyses demonstrate that the majority of Hyalosphenia cells have one spherical nucleus, consistent with observations of other Arcellinida genera including Phryganella acropoda (Dumack et al., 2020) and Difflugia sp. (Griffin, 1972; Mazei and Warren, 2014; Volkova and Smirnov, 2016). A few H. papilio cells with more nuclei than expected were also observed (Table 1; Fig. 3). We hypothesize that the multinucleated cells may be undergoing cell division and/or may represent specific life-history stages consistent with Mignot and Raikov (1992) who suggested that meiosis occurs in cysts of Arcella vulgaris. Most strikingly, the multinucleated cells show smaller nuclei (avg. 11.4 μm) compared to uninucleate cells (18.3 μm; Table 1) indicating DNA reduction in relation with nuclear size.

Figure 3:

Examples of DAPI and corresponding DIC images of cells from model organisms used for establishing standard curves. (A, B) Homo sapiens; (C, D) Allium cepa; (E, F) Saccharomyces cerevisiae. DAPI staining carried out following the Hyalosphenia protocol. Scale bars = 25 μm (A, B, E, F), 200 μm (C, D).

Our results also suggest that H. papilio bears a huge genome, twice the genome size of humans (Fig. 4, Table 2), which is consistent with previous estimates of Amoebozoa genomes; for example, Amoeba dubia has the largest eukaryotic genome size known to date with an estimate of ~ 670 Gb (Friz, 1968). Also, it must be acknowledged that genome dynamics in Amoebozoa are very complex. For example, Goodkov et al., (2020) report DNA extrusion during the life cycle of Amoeba proteus. Similarly, the amoebozoan parasite Entamoeba varies in DNA content at different life stages, perhaps because of its poor control in DNA segregation (Mukherjee et al., 2009). Myxomycetes are characterized by a wide range of genome sizes and differences in ploidy even within individual strains, as extensively reviewed in Clark & Haskins (2013). We therefore hypothesize that H. papilio has variable ploidy levels consistent with findings of aneuploidy in other amoebozoan lineages (Byers, 1986; Friz, 1968). Additionally, we speculate that the change in nuclear volume and DNA content between multinucleate and uninucleate H. papilio suggests a possible reduction in DNA during meiosis, but additional data and a more robust sample size will be required to test this possibility.

The observations on Loxodes nuclear architecture are congruent with previous work on this genus (e.g. Raikov, 1982; Raikov, 1985) as we consistently observe a minimum of two somatic macronuclei and germline micronuclei per cell. The somatic macronucleus contains only one nucleolus located in the center of the nucleus, and the micronuclear architecture is similar to other ciliate species in that DNA appears densely and uniformly distributed (Prescott, 1994). The estimate of the germline micronuclear genome in Loxodes sp. is ~1.4 Gb, which is larger than that of previously characterized ciliates species, e.g. 82.9 Mb for Paramecium tetraurelia (Oligohymenophorea: Arnaiz et al., 2012),157 Mb for Tetrahymena thermophila (Oligohymenophorea: Hamilton et al., 2016), and ~500 Mb for Oxytricha trifallax (Spirotrichea: Chen et al., 2014). It is possible that Loxodes indeed has a large and complex genome, though the AT binding preference of DAPI staining might have contributed to an overestimation of the genome size. The Loxodes genome likely has a higher AT content than the genomes of the three standard organisms used to calibrate the ratio of fluorescence units to DNA content for our staining protocol (Piovesan et al., 2019; Ricroch and Brown, 1997; Wang and Gao, 2019). Also, DAPI is known to overestimate DNA content in AT-rich genomes relative to GC-rich genomes (Button and Robertson, 2001; Wheeler et al., 2012). A previous study showed that karyorelictids, the group Loxodes belongs to, tend to have less transcripts in conserved gene families referring to smaller gene families (Yan et al., 2019). It could be assumed that the germline micronuclear genome in Loxodes is enriched with germline specific information that is not protein coding.

The macronuclear DNA content is estimated to be greater than micronuclear DNA content, consistent with previous fluorescent studies of ciliate genomes (e.g. Wancura et al., 2018). The estimate of the Loxodes somatic macronuclear genome size, approximately 3.5 Gb, is large compared to other ciliates. For example, the model lineages O. trifallax, P. tetraurelia, and T. thermophila, have macronuclear genomes ranging from 50 to 103 Mb (Aury et al., 2006; Eisen et al., 2006; Swart et al., 2013). However, our estimate for Loxodes is not unfeasible for a ciliate, as the estimated DNA content of the Blepharisma americanum MAC is 42.6 Gb (Wancura et al., 2018). During the somatic macronuclear development from germline micronuclei, differential chromosome/gene replication and whole genome scale amplification may both occur, which usually results in high ploidy levels and larger amount of DNA in the somatic macronuclei (Prescott, 1994; Raikov, 1982). For instance, the ratio between the DNA content of macronuclei to micronuclei in Bursaria truncatella and Spirostomum ambiguum is approximately 5,240 and 13,150, respectively (Ovchinnikova et al., 1965; Ruthmann, 1964).

Both H. papilio and Loxodes display variation in nuclear number and genome content estimates (Fig 4; Table 2). This variation could result from the staining and imaging process, as cells do not always fix in the same orientation on the microscope slide, as well as sometimes substantial differences in nuclear volume among cells. Also, the variability could be influenced by the cell’s life cycle stage, at the time when it was captured for the experiment. Loxodes micronuclei are generally diploid (e.g. Raikov, 1985; Parfrey et al, 2008), and thus we expect that cell cycle differences among samples will yield DNA content variability on the order of 2n to 4n. Macronuclear DNA content may also vary substantially throughout Loxodes, and all ciliate, life cycles. Ploidy in adult Loxodes MACs has been observed to range from 4.5n-10n in some cases (Raikov, 1985). In some cases, encysted ciliates may decrease their DNA content through macronuclear extrusion (Akematsu and Matsuoka, 2008; Gutiérrez et al., 1998). Cyclical endopolyploidzation and other examples of genome dynamics are widespread among eukaryotes (Parfrey and Katz, 2010b), and may contribute to the broad range of DNA content estimates in both Loxodes and H. papilio.

We also observe differences in DNA content between the macronuclei within an individual Loxodes cell, which concurs with Raikov’s observations that young macronuclei tend to have less DNA than mature macronuclei, evident by less intense staining and smaller nuclear size (Raikov, 1985). Since the macronuclei in Loxodes are not capable of dividing, in every vegetative division, the daughter cell receives half of all macronuclei while the other half are generated anew from the germline micronuclei. Therefore, in Loxodes, the two macronuclei have gone through different numbers of vegetative divisions, which, in other words, shows the varied ‘age’ in the two somatic macronuclei (Raikov, 1985). As suggested in Bobyleva et al. (1980), mature macronuclei might undergo partial DNA amplification resulting in an increase in DNA content. The measured differences in Loxodes cells are supported by fluorescent images, in which we observe that one macronucleus in each cell is substantially smaller, dimmer, or has a less-developed nucleolus than the other. These data are a step towards validating the hypothesis that nuclear age differences cause differences in DNA content among macronuclei in Karyorelictea.

Synthesis

The newly developed and adapted protocols for uncultivable testate amoebae and ciliates, specifically, for Hyalosphenia spp. and Loxodes sp., successfully revealed the nuclear structure of the target lineages. Furthermore, we provide approximations of genome sizes using DAPI with three other model organisms. The protocols present in this work can be used for staining and estimating DNA content in uncultivable protists with modifications, which provides a useful way to advance our knowledge in nuclear properties of diverse microbial eukaryotes, especially for those with few genomic/molecular data.