(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Regulation of sedimentation rate shapes the evolution of multicellularity in a close unicellular relative of animals

['Omaya Dudin', 'Institut De Biologia Evolutiva', 'Csic-Universitat Pompeu Fabra', 'Barcelona', 'Catalonia', 'Sébastien Wielgoss', 'Institute Of Integrative Biology', 'Department Of Environmental Systems Science', 'Eth Zürich', 'Zürich']

Date: 2022-04

Significant increases in sedimentation rate accompany the evolution of multicellularity. These increases should lead to rapid changes in ecological distribution, thereby affecting the costs and benefits of multicellularity and its likelihood to evolve. However, how genetic and cellular traits control this process, their likelihood of emergence over evolutionary timescales, and the variation in these traits as multicellularity evolves are still poorly understood. Here, using isolates of the ichthyosporean genus Sphaeroforma-close unicellular relatives of animals with brief transient multicellular life stages-we demonstrate that sedimentation rate is a highly variable and evolvable trait affected by at least 2 distinct physical mechanisms. First, we find extensive (>300×) variation in sedimentation rates for different Sphaeroforma species, mainly driven by size and density during the unicellular-to-multicellular life cycle transition. Second, using experimental evolution with sedimentation rate as a focal trait, we readily obtained, for the first time, fast settling and multicellular Sphaeroforma arctica isolates. Quantitative microscopy showed that increased sedimentation rates most often arose by incomplete cellular separation after cell division, leading to clonal “clumping” multicellular variants with increased size and density. Strikingly, density increases also arose by an acceleration of the nuclear doubling time relative to cell size. Similar size- and density-affecting phenotypes were observed in 4 additional species from the Sphaeroforma genus, suggesting that variation in these traits might be widespread in the marine habitat. By resequencing evolved isolates to high genomic coverage, we identified mutations in regulators of cytokinesis, plasma membrane remodeling, and chromatin condensation that may contribute to both clump formation and the increase in the nuclear number-to-volume ratio. Taken together, this study illustrates how extensive cellular control of density and size drive sedimentation rate variation, likely shaping the onset and further evolution of multicellularity.

Funding: This work was funded by European Research Council Consolidator Grant (ERC-2012-Co -616960) to IRT.OD was funded by a Swiss National Science Foundation Early PostDoc Mobility fellowship (P2LAP3_171815), a Marie-Sklodowska-Curie individual fellowship (MSCA-IF 746044), and an Ambizione fellowship from the Swiss National Science Foundation (PZ00P3_185859). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Beside the formation of new biological structures and increases in organismal size, the emergence of multicellularity is frequently accompanied with an increase in sedimentation rate. Indeed, S. cerevisiae snowflake yeast, multicellular C. reinhardtii, and S. rosetta colonies sediment faster when compared to their unicellular counterparts [ 15 , 16 ]. Such correlation has been described in several marine phytoplankton species where multicellular life stages show faster sedimentation than unicellular ones [ 38 – 41 ]. This phenotype has a large impact on where in the water column these microbes proliferate [ 38 , 42 – 44 ] and thus is presumed to be under strong genetic control and selective pressure. Despite its capacity to affect the depth at which marine species flourish, the role of sedimentation rate, the potential impact of its variation, and its connection to the emergence of multicellularity has not been systematically analyzed across unicellular marine organisms, including in the closest unicellular relatives of animals with transient multicellular life stages. Here, using EE, we characterize how regulation of sedimentation rate can influence the emergence of stable multicellular life forms in the ichthyosporean Sphaeroforma genus, a close unicellular relative of animals.

(A) Cladogram representing the position of ichthyosporeans including Sphaeroforma species within the eukaryotic tree. (B) Schematic representation of the coenocytic life cycle of S. arctica. (C) Data derived from Fig 1 [ 37 ] (gray points) were used to scale velocity measurements determined in our sedimentation assay to physical (μm/s) units (red points) (Methods). Error bars represent the 95% CI for each unique genotype, time point, and temperature measurement presented in our study (N = 3 for each of 1 or 2 independent replications). Values were log-transformed prior to calculation of error. This figure is to illustrate where our data fit in the scheme of known plankton sedimentation rates. For our best estimations of cellular density and velocity in meters per second, a subset of this data from Smayda’s Appendix Table 1 was used (Methods). (D) Relying on the Smayda dataset as reference, we used measurements of sedimentation rate from our assay along with cellular perimeter measurements to calculate maximum likelihood estimates of excess cellular density. These estimates are plotted on a landscape illustrating the relationship between density and size on sedimentation rate (gray contour lines). Gray contour lines represent the predicted settling velocity of each genotype in pure water (excess density = 1,000 kg/m 3 ) in units μm/s. (E) Sedimentation rates of Sphaeroforma during the life cycle at 17°C. Every trace represents an independent experiment. (F) Distributions of nuclear content of S. arctica cells during the life cycle at 17°C, measured by microscopy (n > 500 cells per time point). (G) Quantification of mean DNA content per time point (measured by geometric mean) for cells grown in marine broth (MB) at 17°C (n > 500 cells per time point). (H) Average coenocyte volume per time point at 17°C (n = 100 coenocytes per time point). (I) Sedimentation rates of S. arctica coenocytes during the life cycle at 17°C. The data underlying this figure may be found at https://doi.org/10.6084/m9.figshare.18319232.v1 .

Alternatively, non-model organisms with key evolutionary positions can be used to better understand the emergence of multicellularity. In particular, the study of unicellular holozoans ( Fig 1A ), the closest unicellular relatives of animals, revealed that these organisms contain a rich repertoire of genes required for cell adhesion, cell signaling, and transcriptional regulation and that each unicellular holozoan lineage uses a distinct developmental mode that includes transient multicellular forms [ 2 , 23 – 27 ]. For instance, the choanoflagellate Salpingoeca rosetta can form clonal multicellular colonies through serial cell division in response to a sulphonolipid of bacterial origin [ 28 – 30 ], whereas the filasterean Capsaspora owczarzaki can form multicellular structures by aggregation [ 31 ]. Ichthyosporeans display a coenocytic life cycle unique among unicellular holozoan lineages and pass through a short and transient clonal multicellular life stage prior to the release of newborn cells [ 32 – 36 ]. Despite their pivotal phylogenetic position, their rich “animal” genetic toolkit and the capacity to undergo transient multicellularity, to date, unicellular holozoans lineages and EE, have never been combined.

For example, EE under controlled conditions allows selection for diverse phenotypes [ 12 , 13 ], including multicellularity [ 14 – 18 ]. Using EE, Ratcliff and colleagues repeatedly observed the evolution of a simple form of multicellularity in Saccharomyces cerevisiae and Chlamydomonas reinhardtii in response to gravitational selection [ 14 – 19 ]. Similarly, multicellularity emerged in yeast as a mechanism to improve the use of public goods [ 17 ]. In all these cases, cells form clumps by incomplete separation of daughters from mother cells, instead of by postmitotic aggregation [ 20 – 22 ].

The emergence of multicellularity from single-celled life represents a major transition, which has occurred many times independently across the tree of life [ 1 – 8 ]. Multicellularity can arise either by aggregation of single cells that come together or from single cells that are maintained together clonally after division [ 9 – 11 ]. The unicellular and intermediate multicellular ancestors, which led to present-day multicellular organisms, have long been extinct [ 3 ], obscuring direct investigation of how multicellular life has emerged. However, several strategies have been used to study the emergence of multicellularity, including the use of experimental evolution (EE) approaches and the investigation of novel non-model organisms at pivotal positions in the tree of life.

Results

Sphaeroforma species exhibit large variation in sedimentation rates Similar to other ichthyosporeans, Sphaeroforma species proliferate through continuous rounds of nuclear divisions without cytokinesis to form a multinucleated coenocyte [34,35,45,46]. Sphaeroforma coenocytes then undergo a coordinated cellularization process leading to the formation of a transient multicellular life stage resembling an epithelium [32]. This layer of cells then detaches and cell-walled newborn cells are released to the environment (Fig 1B) [32]. The entire life cycle prior to cellularization occurs in highly spherical (multinucleated) cells. Extensive literature has documented a positive correlation between cell size and sedimentation rate, including during the life cycle of marine phytoplankton (Fig 1C) [37–41].This is consistent with Stoke’s law, which shows that the relationship between a spherical particle’s terminal sedimentation rate v in a fluid and its radius R should be determined by where c represents the scaled ratio of gravitational to viscosity constants, and p p is the difference between the particle’s and the fluid’s densities (see Methods). Therefore, even small shifts in a particle’s radius will lead to pronounced (i.e., quadratic) changes in sedimentation rate. Similarly, for particles sitting near the buoyancy threshold, small changes in density can lead to proportionally large changes in settling rate (Fig 1D). Due to the nature of the coenocytic life cycle (Fig 1B), which is associated with an increase in the number of nuclei and coenocyte volume, we expected to observe an increase in cellular sedimentation rates over time [44,47,48].To better understand this relationship, throughout this study, we conducted overlapping experiments characterizing cell volume and sedimentation rates of Sphaeroforma species for cultures growing at 17°C for 72 hours. For certain replicates of this core dataset, we included measurements of various genetic variants, the temperature dependence of phenotypes, fitness, as well as the speed of nuclear duplication. Overall, the measurements we report are highly reproducible with >95% variance explained for replicate measurements across phenotypes (S1A Fig and S1–S4 Data) (see Methods). This reflected a high heritability of the different phenotypes in a given environment. To begin, we measured the sedimentation rate of 5 different Sphaeroforma species that have been isolated from different habitats either free-living or host-derived, namely S. arctica, S. sirkka, S. napiecek, S. gastrica, and S. nootakensis, using a quantitative sedimentation rate assay based on changes in optical density over time [32,45,49–52]. According to these estimates, cell sedimentation rates varied greatly, from between 0.4 to 125 μm per second (i.e., up to 0.45 meters per day) (Fig 1E). This broad variability over the life cycle and among Sphaeroforma species suggested that appreciable changes in size and/or cellular density should accompany different stages of the life cycle.

The transient multicellular life stage of S. arctica is associated with an increase in sedimentation rate To better understand the cellular basis of sedimentation rate variation, we focused on S. arctica, the most studied Sphaeroforma species to date [32,34,50]. Using fixed cell imaging, we observed that synchronized cultures of S. arctica undergo a complete life cycle in about 48 hours and can reach up to 256 nuclei per coenocyte before undergoing cellularization and releasing newborn cells (Fig 1F). Prior to this release, all cellular division occurs in highly spherical mother coenocytes. Consistent with previous results [34], nuclear division cycles were periodic during coenocytic growth and occurred on average every 9 to 10 hours, as measured indirectly from changes in DNA content (Fig 1G). Average cell volume increased throughout the coenocytic cycle to reach its maximum value at 36 hours prior to the release of newborn cells (Fig 1H). Similarly, the sedimentation rate increased up to >300-fold the initial value after 36 hours (Fig 1I). Altogether, we observe that every cycle of nuclear division is associated with a significant increase in nuclear content, volume, and sedimentation rate with a distinct peak just prior to cell release. As the transient multicellular life stage of S. arctica occurs during the latest stage of cellularization and ahead of cell release [32], our results suggest that it is tightly associated with an increased sedimentation rate.

Increased nuclear number-to-volume ratio leads to faster sedimentation Above, we observed that S4 and S9 mutants can sediment as fast as S1 despite their smaller coenocyte volumes, suggesting an alternative regulation mechanism of sedimentation rate. Across all experiments reported in this study, we found that approximately 31% of the variance in observations of sedimentation rate could not be explained by cell size only, suggesting that cellular density might also contribute to this variation [37,40,41]. From Stoke’s law, we calculated that excess cellular density, i.e., cellular density minus that of distilled water (1,000 kg/m3), might vary between 40 and 300 kg/m3 for S. arctica wild-type and evolved clones across their life cycle—the upper limits between the densities of pure protein and pure cellulose. Values reached approximately 650 kg/m3 for wild S. nootakensis (Snoo) soon after cellularization, approaching the excess density of pure nucleic acid (Fig 1D). During cell cycle stages prior to cellularization, when cells were most spherical (<36 hours), excess density varied from 40 to 200 kg/m3 across Sphaeroforma isolates. To better characterize the relationship between sedimentation rate, cell cycle, and size, we performed higher resolution measurements of sedimentation rates over the complete life cycle of the ancestor and all 3 fast-settling mutants at 12°C and 17°C. Consistent with their capacity to form clumps, we observed that the sedimentation rate of all fast-settling mutants increases during growth but, unlike the ancestor, does not recover after cell release to their original levels (Figs 4A and S4A). Interestingly, we noticed that individual S4 and S9 coenocytes sediment faster (approximately 2.5× and approximately 1.6×, respectively) than S1 or AN even before clump formation (24 to 36 hours time points) (Figs 4A and S4A). Such increase in sedimentation rate was not due to a rise in cell size or change in cell shape as both S4 and S9 exhibit smaller cell perimeters throughout the cell cycle (Figs 4B, 4C, S4B, and S4C). Rather, excess cellular density estimations show that both S4 and S9, even prior to cell release and clump formation, tend to be on average 3× denser when compared to the ancestor (Figs 4D and S4D). Altogether, these results show that both cell size and cell density contribute to sedimentation rate variation in S. arctica. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Sedimentation rates variation in fast-settling mutants is associated with variation in cell size and cellular density. (A) Sedimentation rates of S. arctica AN and evolved mutants during the life cycle at 17°C. Every trace represents an independent experiment. (B) Average perimeter measured from fixed cells every 12 hours over a complete life cycle of 72 hours at 17°C shows that fast-settling mutants increase their size upon cell release. Every trace represents an independent experiment (n > 180 measurements per time point for each independent experiment). (C) Average perimeter of fast-settling cells and clumps at 24 and 60 hours, respectively, show that S4 and S9 cells and clumps have a smaller size when compared to S1. Every square represents an independent experiment, and the white circle represents the median (n > 180 coenocytes per time point for each independent experiment). (D) Excess cellular density of fast-settling individual coenocytes (before cellularization) and clumps (after cell release) at 24 hours and 60 hours, respectively, show that S4 and S9 single cells are denser when compared to S1 and AN. Every square represents an independent experiment, and the white circle represents the median. (E) Quantification of mean DNA content per time point for fast-settling mutants grown in MB at 17°C. Every trace represents an independent experiment (n > 400 coenocytes per time point for each independent experiment). (F) Nuclear doubling time, calculated by linear regression of mean nuclear content at time points from 0 hour to 24 hours. Every square represents an independent experiment, and the white circle represents the median (n > 400 coenocytes per time point for each independent experiment). (G) Boxplots of cell volume measurements of DAPI-stained fixed cells. For 1-, 4-, 16-, and 64-nuclei cells. Cells with 1 nucleus represent newborn cells at the end of the experiment (n > 80 coenocytes per DNA content). (H) Boxplots of nuclear number-to-volume ratio of DAPI-stained cells show significant increase for S4 and S9 fast-settling mutants. Every square represents an independent experiment, and the white circle represents the median (n > 600 coenocytes per strain). (I-K) Temporal transcript abundance of genes mutated in fast-settling phenotype across the native life cycle of S. arctica. The data underlying these graphs may be found in S5 Table. The data underlying this figure may be found at https://doi.org/10.6084/m9.figshare.18319232.v1. https://doi.org/10.1371/journal.pbio.3001551.g004 As cell size and nuclear division cycles are decoupled in S. arctica [34], we reasoned that increased cell density in S4 and S9 could be caused by an acceleration of nuclear divisions leading to a rise in the number of nuclei per volume. Using DAPI staining to label nuclear DNA, we observed that S4 and S9 undergo nuclear duplication faster (approximately 2 hours) than both AN and S1 (Figs 4E, 4F, and S4E–S4H). By carefully examining the volumes of coenocytes containing the same number of nuclei at the single-cell level, we show that for the same nuclear content, S4 and S9 tend to be 30% to 45% smaller in volume when compared to the ancestor (Figs 4G and S4I). Consequently, both S4 and S9 exhibited the highest number of nuclei per volume (nuclear number-to-volume ratio) (Figs 4H and S4J). Taken together, these results argue that cell density can contribute an appreciable amount to cellular sedimentation rates (up to approximately 50 μm/s) and that mechanistically this could arise by faster nuclear doubling times relative to cell size.

Evolved genetic variation correlating with fast sedimentation Up to now, our results suggest that S. arctica mutants evolved faster sedimentation using 2 strategies: (i) clump formation; and (ii) increased nuclear number-to-volume ratio. We found that sedimentation rate variation was highly heritable, persisting for 780 generations of passaging for all 10 isolates without selection for sedimentation phenotype, suggesting that the phenotypes have a genetic basis. To test this, we resequenced the whole genomes of both the ancestral clone (AN) and 1 evolved clone per lineage (S1 to S10) obtained at the conclusion of the evolution experiment (Week 8) with high (>30-fold) coverage of the very large genome size of S. arctica, at 143 Mbp. Following very careful variant filtering (S2 and S3 Tables), we identified a total of only 26 independently evolved variants with an average of 2.6 mutations per clone (range 1 to 5 per clone) (S4 Table). Of the 26 variants, 24 (approximately 92.3%) were SNPs (11 coding, and 13 intergenic or intronic), and 2 were insertions (1 coding, 1 intergenic) (S4 Table). Beside the fact that 2 of these variants are identical SNPs at the same position (Sarc4_g3900) and have independently evolved in 2 different backgrounds (S4 Table), no other form of genetic parallelism was detected. While it is very likely that some of these mutations hitchhiked in the background of beneficial driver mutations, we found that the coding SNP variants were clearly skewed toward nonsynonymous changes (9:2), with a cumulative dN:dS ratio of 1.32. This indicates the general presence of positive selection and, hence, adaptative evolution driving this molecular pattern. Based on COG assignments (S6 Table), 4 of the changes are orthologous to genes implicated in signal transduction, and 2 are related to genes with functions in DNA binding or chromosome condensation. In the absence of molecular genetic tools and functional knowledge of many of the mutated gene targets, we set out to better understand how the distinct genetic variants could have influenced sedimentation rates and clump formation by examining the predicted expression dynamics of mutated genes across the cell cycle. The data were derived from a recently published time-resolved transcriptomics dataset of the S. arctica life cycle [32]. Of the mutated genes, 9 showed no expression during the native life cycle, whereas 12 displayed dynamical expression during cellularization, and the remaining 5 genes were more or less stably expressed (Figs 4I–4K and S4K, S5 Table). We also annotated all mutation-associated genes based on a recent comprehensive orthology search [53] (S6 Table). The fastest-settling and clumpiest mutant isolated from the population S1 bore a synonymous mutation of a homolog of increased sodium tolerance 1 (Ist1) superfamily and as such likely impacts gene expression rather than gene function. This gene shows a dynamic expression during cellularization and codes for a conserved protein involved in multivesicular body (MVB) protein sorting (Fig 4I, S6 Table) [54,55]. In humans, hIST1, also known as KIAA0174, is a regulator of the endosomal sorting complex required for transport (ESCRT) pathway and has been shown to be essential for cytokinesis in mammalian cells [56]. Similarly, Ist1 orthologs in both budding and fission yeasts play a role in MVB sorting pathway and, when deleted, exhibit a multiseptated phenotype consistent with a role in cytokinesis and cell separation [57,58]. In the clone derived from population S4, we observed 5 distinct mutations, 2 nonsynonymous, 1 intergenic, and 2 intronic SNPs. Among the 2 nonsynonymous SNPs, one causes a E90G change in a homolog of human Kanadaptin (SLC4A1AP), which may play a role in signal transduction [59,60]. Among the noncoding SNPs, 1 mutation is found in an intron of the 7-dehydrocholesterol reductase (DHCR7), expressed during cellularization and known to be key in the cholesterol biosynthesis pathway [61,62], and the second intronic mutation codes for a STE20-like kinase (SLK), which plays numerous roles in cell cycle signaling and actin cytoskeleton regulation (Fig 4J) [63–66]. Finally, among the 5 mutations discovered in the clone derived from population S9, 2 mutations are in transcription factors that are continually expressed during the cell cycle: an intronic SNP in a basic helix–loop–helix (bHLH) transcription factor, and the sole nonsynonymous SNP leading to A923V change in a gene predicted to encode a nucleotide binding C2H2 Zn finger domain [67]. A third mutation was found in an intron of the highly and dynamically expressed homolog of the regulator of chromosome condensation 1 (RCC1; Fig 4K) [68–70]. RCC1 is a chromatin-associated protein implicated in several processes including nuclear formation, mRNA splicing, and DNA replication [69,71–74]. Thus, it may contribute to the accelerated nuclear duplication cycle observed in S9 by impacting cell cycle progression. Altogether, the mutations identified in both S4 and S9 may affect both cellularization and cell separation. Among the variants detected in evolved clones with intermediate-settling phenotype, we highlight an intergenic mutation 118 bp downstream of Dynamin-1 known to be essential for cytokinesis across different taxa [75–77], and a nonsynonymous mutation in a protein similar to Fibrillin-2 (Sarc4_g7365T), which is an extracellular matrix (ECM) glycoprotein essential for the formation of elastic fibers in animals (S4K Fig) [78–80]. Altogether, our results across all isolates suggest that a large mutational target affects cellular sedimentation and multicellularity.

[END]

[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001551

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.