Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
Sex allocation pattern of the diatom
Cyclotella meneghiniana
Y. Shirokawa and M. Shimada
rspb.royalsocietypublishing.org
Research
Cite this article: Shirokawa Y, Shimada M.
2013 Sex allocation pattern of the diatom
Cyclotella meneghiniana. Proc R Soc B 280:
20130503.
http://dx.doi.org/10.1098/rspb.2013.0503
Received: 25 February 2013
Accepted: 11 April 2013
Subject Areas:
ecology, cellular biology
Keywords:
sex allocation, split sex ratio, size advantage
model, fertility insurance, unicellular organism,
microfluidic system
Author for correspondence:
Y. Shirokawa
e-mail: [email protected]
Electronic supplementary material is available
at http://dx.doi.org/10.1098/rspb.2013.0503 or
via http://rspb.royalsocietypublishing.org.
Department of Systems Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro-ku, Tokyo 153-8902, Japan
Sex allocation is one of the most successful applications of evolutionary
game theory. This theory has usually been applied to multicellular organisms;
however, conditional sex allocation in unicellular organisms remains an unexplored field of research. Observations at the cellular level are indispensable for
an understanding of the phenotypic sex allocation strategy among individuals
within clonal unicellular organisms. The diatom Cyclotella meneghiniana, in
which the sexes are generated from vegetative cells, is suitable for investigating
effects of phenotypic plasticity factors on sex allocation while excluding genetic
differences. We designed a microfluidic system that allowed us to trace the fate of
individual cells. Sex allocation by individual mother cells was affected by cell
lineage, cell size and cell density. Sibling cell pairs tended to differentiate into
the same fates (split sex ratio). We found a significant negative correlation
between the cell area of the mother cell and sex ratio of the two sibling cells.
The male-biased sex ratio declined with higher local cell population density,
supporting the fertility insurance hypothesis. Our results characterize multiple
non-genetic factors that affect the phenotypic single cell-level sex allocation.
Sex allocation in diatoms may provide a model system for testing evolutionary
game theory in unicellular organisms.
1. Introduction
Sex allocation is one of the most successful areas in evolutionary game theory.
In theoretical analyses it can predict an offspring’s sex ratio in response to
environmental changes [1– 4]. Although the theory is usually applied to organisms with obligatory sexual systems, such as higher animals and plants, it has
not been applied to sex allocation in unicellular organisms (except in the case of
malaria [4]) with a facultative sexual cycle, which is assumed to be the ancestral
sex system [5,6]. Furthermore, conditional sex allocation [4] based on cell state
remains an unexplored field.
In this study, we investigate phenotypic sex allocation in a unicellular organism, the diatom Cyclotella meneghiniana. We address how each sex is generated
from vegetative clonal cells following induction of meiosis, using an on-chip
single-cell microfluidic system [7,8] that made it possible to trace the fate of
individual cells. Furthermore, we characterize the intrinsic and extrinsic factors
that affected sex allocation by individual mother cells to two sibling cells.
Diatoms undergo a cyclic process of size diminution and size restoration,
which is used to time the length of the life cycle and, therefore, the frequency of
sexual reproduction [9,10]. There is a vegetative phase, lasting months to years,
of size diminution [9], followed by a relatively short sexual reproduction phase,
during which the large cell size is restored. In centric diatoms, the cell diameter
diminishes during the vegetative phase of mitotic division because of the
unique silicified structure of the cell wall [10]. It is only when cells are below a
given size threshold (usually 40–50% of the maximum diameter) that sexual
reproduction can be induced, usually in response to a species-specific environmental cue [10]. Centric diatoms are monoecious; therefore, both sperm and
non-motile eggs can be formed within a single clone [10]. The C. meneghiniana
species complex [11] is sexualized by increased salinity, and each of its vegetative
cells differentiates into four sperm cells or one egg [12].
Studies of sex allocation have suggested that individuals within a population
should not adopt the same universal strategy when individual traits are
& 2013 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
(a) Culture conditions
The C. meneghiniana clone used in this study was maintained in a
freshwater medium. The clone was deposited as NIES-2364 at the
Natural Institute for Environmental Studies. The GenBank accession
number for the D1–D2 regions of the nuclear 28S rDNA is JN854149
and the slide number of the clone deposited in the National Museum
of Nature and Science (Tokyo) is TNS-AL- 57092. Sterile freshwater
medium was used under conditions of 208C, 16 L : 8 D illumination,
and a light intensity of approximately 30 mmol photons m22 s21 in a
1.5 ml centrifuge tube to prevent cell division as much as possible
(preservation stock). Freshwater medium was prepared by adding
10 mg l21 Na2SiO3 9H2O, 20 mg l21 vitamin B12, 1 mg l21 biotin
and 10 mg l21 thiamine to one-fifth diluted BBM [15]. For the
study, the preservation stock was reactivated by transferring to the
new media at 258C and under continuous illumination at a light
intensity of approximately 100 mmol photons m22 s21 until the
mid-to-late log phase, and then cultivating again under the same
conditions until the mid-log growth phase. Sexual reproduction
was induced by changing the medium from freshwater to dilute
seawater by dissolving 10.8 g l21 Daigo artificial seawater (Wako
Pure Chemical Industries, Osaka, Japan) in the freshwater
medium. Note that the seawater medium used in this study contained only 30 per cent of the amount of artificial seawater powder
prescribed in the manufacturer’s protocol.
(b) On-chip single-cell cultivation system
An on-chip single-cell cultivation system [16] was constructed to
follow a particular single-cell sexual differentiation process. This
system contained a microchamber array plate, medium exchange
unit and microscope (see the electronic supplementary material,
figure S1). The microchamber array plate was a cover glass
of 0.1 mm thickness on which we constructed a 15 15 array of
micrometre-sized structures called microchambers using SU-8—a
(c) Morphological criteria for sex distinction
Morphological and nuclear patterns that occur during oogenesis
and spermatogenesis have been reported in many centric diatoms [10]. The process of meiotic division and sperm release
from the cell walls of C. meneghiniana [18,19], and the egg morphology and nucleus pattern [20] after sexual induction by
increasing salinity, have been described.
On the basis of the results of previous studies [10] and our preliminary observations, we focused on two notable differences
between egg and sperm production. The first is that the egg cytoplasm was continuously attached to the cell walls, whereas the
sperm cytoplasm became detached from the cell walls after a short
time. The second difference is that the exposed cell membrane of
the egg expanded more than the cell membrane of the released primary spermatogonia. The sperm that entered the first meiotic
cytokinesis (represented as a typical sperm by blue triangle in the
electronic supplementary material, figure S2) had qualitatively distinct cytoplasmic dynamics, where the cytoplasm formed into a
sphere and then split into two parts (figure 2a). However, the other
cells (black hollow circle in the electronic supplementary material,
figure S2a), which could be either eggs or aborted sperms before
the first meiotic cytokinesis, required a quantitative criterion for
sex distinction. The duration of attachment of the spherical
cytoplasm to the cell wall and the maximum amplification of
membrane expansion were measured to create the criterion. Sex distinction of cells was performed according to this criterion (electronic
supplementary material, figure S2).
(d) Measurement of sex ratio and cell size, cell density
and cell lineage
The field of view was calibrated using a Zeiss micrometre, and
areal projections of cells were determined using IMAGEJ software
[21]. Cell diameter and cell area were measured in the girdle (lateral) view (see the electronic supplementary material, figure S3).
The lateral view of the cell area changed during one cell cycle;
therefore, we focused on the cell area immediately after cell division (see the electronic supplementary material, figure S3). The
sex ratio of each sibling cell pair that included at least one differentiated cell was determined from all the sibling cell pairs that
appeared during a cell division event in the microchamber (corresponding to the sibling pairs ‘a’, ‘c’ and ‘d’ in the electronic
supplementary material, figure S4). The sex ratio of two sibling
cells was regarded as the number of cells differentiated into
sperm divided by the number of differentiated cells among the
two sibling cells. For a single pair of cells, the ratio could
2
Proc R Soc B 280: 20130503
2. Material and methods
thick, negative, photoresistive material (Microlithography Chemical, Newton, MA). Each microchamber was approximately 100 100 20 mm. After the sample cells were placed in the array, the
array was sealed with a semipermeable membrane (molecular
weight cut-off, 25 000; Spectrum Laboratories, Irving, TX) using
an avidin–biotin attachment to prevent the cells from escaping
[17]. Microchambers comprising various cell numbers were prepared using stochasticity for the process of cell application and
sealing with the membrane. A PDMS (polydimethylsiloxane)
plate of approximately 2 2 0.5 cm was mounted against the
microchamber array plate using an adhesive flame (Frame-Seal
incubation chambers, 25 mm capacity; MJ Research Inc., Waltham,
MA). Medium pumped from the medium tank at 2 ml h21 passed
through the space between the PDMS plate and the semipermeable
membrane, and flowed to a waste tank. The freshwater medium
flowed for approximately 24 h and was then changed to seawater
to induce sexual differentiation. Images were obtained using timelapse microscopy. For a detailed description of the microscopy
conditions, see the materials and methods in the electronic
supplementary material.
rspb.royalsocietypublishing.org
conditionally heterogeneous [3,4]. The effects of phenotypic
plasticity on sex allocation can be investigated by using a
clonal population because genetic differences are excluded
between individuals to the extent possible. In this study, we
investigated three factors that could affect sex allocation by individual mother cells—cell lineage, cell size and cell density—
using an on-chip single-cell microfluidic system [7,8]. First, we
investigated whether the sex allocation of individual mother
cells is distorted, whereby individuals specialize in producing
either male or female offspring (split sex ratio) [3,4]. Second,
we investigated the effects of the size of individual mother
cells on conditional sex allocation under the Trivers–Willard
hypothesis (i.e. if a physiological condition, such as body size,
differs among individuals, then, the fitness of the male and
female also becomes different; consequently, selection favours
conditional sex allocation), as shown across broad taxonomic
groups [3,4]. As the average cell size of a diatom population
decreases, the population becomes more male biased [10]. However, the individual diatom strategy has not been elucidated; for
instance, in what case does an individual diatom cell choose its
cell fate (egg, sperm, undifferentiated)? Third, we examined
the effect of local cell population density on the disproportionate
male ratio (the ‘fertility insurance hypothesis’ [13,14]).
The aims of this study were (i) to quantify how
C. meneghiniana forms male and female subtypes at the
single-cell level using the microfluidic system, and (ii) to investigate phenotypic sex allocation within a clonal population
from the perspective of evolutionary game theory.
Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
(a) 0 h
(b) 23.7 h
(c) 33.9 h
(d) 52.6 h
therefore be 1 (both sibling cells are male, or one sibling cell is
male and the other is undifferentiated), 0.5 (one sibling cell
is male and the other is female) or 0 (both sibling cells are
female, or one sibling cell is female and the other is undifferentiated). Data were collected from 344 mother cells in 49
chambers to analyse sex allocation by individual mother cells
according to cell size or cell density. Cells that could not be
assessed for fate or size because of overlap with other cells
were excluded. The cells were followed until they differentiated
or until the measurement time ended. All cell-tracking procedures were conducted manually to ensure accuracy. The cell
density within the microchamber was the number of vegetative
cells within the chamber at the sexual induction cue.
(e) Split sex ratio by the bootstrap test
Information from 722 cell pairs in 55 chambers was used to investigate the effects of cell lineage on sex ratio distortion. Cells whose
fate could not be assessed because of overlap with other cells were
excluded. The determined cell fate pairs were used from all sibling
cell pairs that appeared in the microchamber (corresponding to the
sibling pairs ‘a–d’ in the electronic supplementary material, figure
S4). The bootstrap method was used as described later. The frequency of each combination of cell fates was constructed by
random pair choice from the observed data and was duplicated
10 000 times. The observed number of each cell fate combination
of the two sibling cells was compared with the 99 per cent confidence limit of this virtual frequency distribution of the randomly
chosen pair. The calculations were performed using the original
code of R software v. 2.13.0 (R Foundation for Statistical
Computing; http://www.R-project.org).
(f ) Statistical analysis
Values are reported as the mean and standard error. A post hoc
power calculation was performed using G*POWER v. 3.1.5.1. Statistical tests including analysis of variance (Welch’s ANOVA) with
unequal error variances using oneway.test function, Tukey’s HSD
test and a generalized linear model (GLM) were performed
using R software. We examined Akaike’s information criterion
(AIC) [22] to determine the best-fitting model. A GLM with binomial errors and logit-link function was used to determine whether
cell area and cell density had an effect on sex ratios of the two sibling cells. The significance of independent variables was evaluated
by comparing the models using log-likelihood ratio tests, following
stepwise model selection using AIC [22] (see the electronic supplementary material, table S1). McFadden’s R2 was used to
estimate the degree to which cell area differences contributed to
differences in the fate of the two sibling cells.
3. Results
(a) Induction of sexual reproduction in a microfluidic
device
We designed an on-chip single-cell microfluidic system [16,17]
for tracking differentiation processes at the single-cell level.
Cells were randomly seeded on arrays of 100 100 mm microchambers fabricated on a glass slide (figure 1). Cells were
enclosed in each microchamber by a semipermeable membrane.
Multipoint time-lapse microscopy enabled tracking of multicell
lineages within microchambers simultaneously. The medium
was supplied to the microchambers through the semipermeable
membrane. A freshwater medium was used for the first 24 h,
after which seawater was used to induce sexual differentiation.
An example of the time course of C. meneghiniana sexual
differentiation is shown in figure 1a–d and in the electronic
supplementary material, movie S1. After changing the
medium to seawater, spermatogenesis occurred more rapidly
than oogenesis (figure 1c,d; electronic supplementary material,
figure S5). Figure 2 presents a depiction of identifying each cell
type and the time course of differentiation. During spermatogenesis (around 15 h after induction; figure 2a; electronic
supplementary material, figure S5), we observed that the cells
released spherical spermatogonia from their cell walls and
then underwent the first meiotic cytokinesis (figure 2a). Meiotic
cytokinesis occurred within the cell walls before extrusion in
approximately 40 per cent of sperm (figure 2a). Individual
clone cells produced four sperms. Oogenesis occurred relatively later than spermatogenesis (approx. 30 h after induction;
figure 2b; electronic supplementary material, figure S5). We
observed that an egg cell opened its cell wall and expanded the
spherical cytoplasm attached to the cell wall (figure 2b). The
nuclei dynamics by meiosis and insemination were investigated
by DAPI (40 ,6-diamidino-2-phenylindole) staining (see electronic
supplementary material, figures S6 and S7). Expansion of the
spherical cytoplasm of egg cells occurred even without insemination (see electronic supplementary material, figure S7b–e). In
addition, we confirmed insemination within the clone (see
electronic supplementary material, movie S2 and figure S8),
and the clone was able to generate vegetative cells of the next
generation (see electronic supplementary material, figure
S7g,h); however, the relationship between insemination and formation of the next vegetative cell was not determined.
Cells that did not open their cell walls until termination of
the recording (60.8 h after sex induction) were identified as
‘undifferentiated cells’ (figure 2c). The majority of these
Proc R Soc B 280: 20130503
Figure 1. Sexual differentiation in Cyclotella meneghiniana using a microfluidic system. (a,b) Cells within a microchamber under freshwater media. (c) The cell
started spermatogenesis (arrowhead) under seawater. (d) Cells differentiated into male gametes (arrowheads) or female gametes (arrows). Scale bar, 10 mm.
rspb.royalsocietypublishing.org
seawater medium
freshwater medium
3
Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
sex induction cue
4
experimental time
cell wall
short-time
cell wall attachment
*
15 h
after induction cue
hatch
identified as sperm
*
meiotic cytokinesis
Proc R Soc B 280: 20130503
(b)
oogenesis
rspb.royalsocietypublishing.org
spherical
cytoplasm
(a)
spermatogenesis
identified as sperm
long-time
cell wall attachment
cytoplasm expansion
30 h
after induction cue
(c)
undifferentiated
identified as egg
fertilization
identified as
undifferentiated cell
undifferentiated cells tended to terminate cell division 20–30 h
after sex induction.
On the basis of our observations (figure 2; electronic
supplementary material, figure S2) and a previous report
[10], the processes of diatom oogenesis and spermatogenesis
appear to have morphologically distinct patterns. The sex of
each cell can be distinguished by a differentiation pattern.
We focused on the following three notable differences
between egg and sperm:
cell fate combination number
Figure 2. Cell-type identification and differentiation process observed by movies. (a) Typical spermatogenesis process. Spermatogenesis started approximately 15 h
after sexual induction. The cytoplasm extruded for a short time after cell wall attachment. Two male gametes (asterisks, *) after the first meiotic cytokinesis and four
sperm (arrowhead) resulted from the second cytokinesis. (b) Typical oogenesis process. Oogenesis started approximately 30 h after sexual induction. Cytoplasm
expansion occurred with the cell wall attached. (c) Cells that did not expose the cytoplasm out of the cell wall were identified as undifferentiated cells.
400
300
200
100
0
(i) after sex induction, spermatogenesis started faster
than oogenesis by approximately 15 h;
(ii) the egg cytoplasm was continuously attached to the
cell wall, but the sperm cytoplasm was detached
from the cell wall after a short time (figure 2a,b); and
(iii) the exposed cell membrane of the egg was more
expanded than the primary spermatogonia that were
released (figure 2a,b).
We recorded the start time of a cell’s sexual reproduction,
and determined sex discrimination by the duration of attachment of the cytoplasm to the cell wall and by the maximum
amplification of cytoplasm expansion (see the electronic
supplementary material, figure S2). The proportional cell
fate of the population was as follows: egg, 6.5 per cent (n ¼
103); sperm, 61.45 per cent (n ¼ 974); and undifferentiated
cells, 32.05 per cent (n ¼ 508) (total n ¼ 1585).
(b) Distorted cell fate allocation based on cell lineage
To examine the dependence of sex ratio distortion on cell lineage, we investigated whether two sibling cells were likely to
E/E
E/S
E/U
S/S
S/U
U/U
cell fate combination
Figure 3. Cell fate similarity according to cell lineage (n ¼ 722 cell pairs).
The observed number of each cell fate combination of the two sibling cells
(red circles) and mean virtual frequency distribution of a randomly chosen
pair (black diamonds) were compared. Error bars indicate 99% CI of virtual
random-pairing distribution. E, egg; S, sperm; U, undifferentiated cell. The
pale red bar shows that the observed frequency of the cell fate combination
was significantly higher than the 99% CI of the virtual distribution. The pale
green bar shows that the observed frequency was significantly lower.
experience the same fate (egg, sperm or undifferentiated). We
determined the cell fates of terminal sibling pairs of a cell lineage (see electronic supplementary material, figure S4). Figure 3
illustrates that sibling cell pairs tended to experience the same
fate (i.e. the frequencies of egg–egg, sperm–sperm and undifferentiated–undifferentiated pairs were significantly higher,
and the frequency of egg–undifferentiated, sperm–undifferentiated pairs were significantly lower, than the 99% confidence
limit of a virtual random-pairing distribution estimated by
the bootstrap method). Therefore, the two sibling cells
demonstrated the split sex ratio.
Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
1.0
6
0.8
sex ratio
8
4
egg
sperm
undifferentiated
(b)
0
low
mediumlow
mediumhigh
high
cell density
60
Figure 5. Effect of local cell population density on sex ratio (n ¼ 344
mother cells). Cell density within microchambers at sexual induction and
its effect on the sex ratio of the two sibling cells are shown. Cell density
was divided into quartiles, as shown in figure 4. Error bar ¼ s.e.m.
40
20
0
egg
sperm
undifferentiated
determined cell fate
1.0
0.8
0.6
0.4
0.2
0
small
medium- mediumlarge
small
large
mother cell area in vegetative state
Figure 4. Effects of cell size on sex ratio. (a) Effects of individual cell diameter on subsequent cell fate (egg, n ¼ 91; sperm, n ¼ 519; undifferentiated,
n ¼ 176). (b) Effects of individual cell area on subsequent cell fate.
(c) Effects of mother cell area on sex ratio of two sibling cells (n ¼ 344
mother cells). Cell area was divided into quartiles: small (57.2 – 85.9 mm2;
first quartile), medium-small (86.1– 99.2 mm2; second quartile), mediumlarge (99.3 – 117.1 mm2; third quartile) and large (117.3 – 311.2 mm2;
fourth quartile). Error bar ¼ s.e.m.
(c) Sex allocation based on cell size
We initially compared two cell size indices: the cell diameter,
which is a proxy for cell line ‘age’ in diatom species, and the
lateral view of the cell area, which can change with the differential ability of cells to elongate. The diameter range of the
clone was 5.97 + 0.03 mm and the cell fate (sperm, egg or
undifferentiated) did not depend significantly on cell diameter in this clone (ANOVA, F ¼ 1.73, d.f. ¼ 2, p ¼ 0.18;
figure 4a), because of the use of a single clone with cells of
homogeneous diameter. The sample size had 34 per cent
power to detect the effect of individual diameter on cell
fate, at a p-value of 0.05.
The lateral view of the cell area changed during one cell
cycle in the vegetative stage; therefore, to unify the timing of
cell area measurement, we focused on the cell area immediately
after cell division (see electronic supplementary material,
figure S3). We measured the cell area of cells whose fate was
determined (see electronic supplementary material, figure S4).
Our results revealed that individual cell area in the lateral
view was significantly different among the groups categorized
by subsequent cell fate (ANOVA, F ¼ 20.41, d.f. ¼ 2, p , 0.0001;
Tukey’s test, for all combinations of cell fates: egg .
undifferentiated . sperm, p , 0.05; figure 4b). The sample
size had 100 per cent power to detect the effect of individual
cell area on cell fate, at a p-value of 0.05. In addition, the lateral
view of the cell area of the mother cell affected the combination
of two sibling cell fates (see electronic supplementary material,
figure S9). Note that the mother cell area is the sum of the areas
A1 and A2 described in the electronic supplementary material,
figure S3.
Sex allocation is usually evaluated as the ratio of individuals that become males among all offspring [3,4]. In the
present study, the sex ratio of the two sibling cells was
regarded as the number of cells differentiating into sperm
divided by the number of differentiated cells among the two
sibling cells. The results showed that the sex ratio of the
two sibling cells decreased as cell area increased (average sex
ratio, 0.83; n ¼ 344 mother cells; GLM with binomial error:
deviance ¼ 68.63, p , 0.001; figure 4c; electronic supplementary material, table S1a). Therefore, if the mother cell had
elongated to a greater extent before dividing, there was a
greater tendency for daughter cells to differentiate into eggs.
The sample size had 100 per cent power to detect effects of
the mother cell area on sibling sex ratio, at a p-value of 0.05.
We found a significant negative correlation between the
similarity of cell fate (same ¼ P, different ¼ 1 2 P; logit) and
the difference in cell area between the two sibling cells
(GLM with binomial error: deviance ¼ 17.73, p , 0.001;
pseudo R 2 ¼ 0.05). Therefore, sibling cells with the same
fate are likely to divide equally.
(d) Dependence of the sex ratio on local cell population
density
In this system, cells should have experienced different cell densities in the microchambers because they were randomly
trapped there. We focused on this isolated, local cell population
density, which was thought to be important for mating success.
A significant negative correlation was observed between cell
density at the time of the sexual induction cue and the sex
ratio of the two sibling cells (n ¼ 344 mother cells; GLM
with binomial error: deviance ¼ 15.16, p , 0.0001; figure 5;
electronic supplementary material, table S1a). The sample
size had 100 per cent power to detect effects of cell density
on sibling sex ratio, at a p-value of 0.05.
Proc R Soc B 280: 20130503
individual vegetative
cell area (µm2)
0.4
0.2
determined cell fate
sex ratio
0.6
2
0
(c)
5
rspb.royalsocietypublishing.org
diameter of individual
vegetative cell (µm2)
(a)
Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
4. Discussion
(b) Intrinsic and extrinsic factors in sex allocation by
mother cells
We demonstrated that larger mother cells tend to differentiate
into eggs at a higher frequency (figure 4c). Sex allocation of a
maternal intrinsic factor (body size) has been investigated in
many organisms, and the size-advantage model of evolutionary game theory predicts a biased sex ratio when one sex gains
a greater fitness benefit from increased resources [2,3]. The size
of the cell that differentiates into an egg has a positive correlation with the size of the fully expanded auxospore (and
hence with the initial vegetative cell size) in several centric diatoms [27]. By contrast, the number of sperm produced by one
cell is not affected by cell size in C. meneghiniana [12]. These
observations suggest that allocation of larger cells to females
more effectively restores cell size in the next vegetative cycle.
Furthermore, the difference in cell fate was reflected by cell
area, rather than diameter, in the lateral view (figure 4). Our
use of a single clone with homogeneous diameter, in contrast
to previous studies that used plural clones [28,29], might have
resulted in the lack of effect of cell diameter on the sex ratio
reported here. Decrease in diameter by vegetative division is
remarkably slowed in cells with diameters less than 5 mm in several Cyclotella species, including C. meneghiniana [30]. Within
(c) Sex allocation in the unicellular organism
We have characterized, for the first time, how non-genetic intrinsic (cell linage and cell area in the lateral view) and extrinsic (cell
density) factors affect sex allocation strategy by individual
mother cells to two sibling cells, using an on-chip single-cell
microfluidic system. Evolutionary game theory may help explain
the sex allocation of diatoms, as well as that of other unicellular
organisms. Phenotypic plasticity plays a largely underappreciated role in driving phenotypic diversification and speciation
[36,37]. Diatom populations may provide a model system for
investigating evolution through phenotypic sex allocation.
Theoretical modelling and integrating measurements of the molecular processes at the single-cell level will further reveal the sex
allocation mechanism within unicellular organisms.
We thank Y. Wakamoto, D. H. Jewson, M. Bonsall and M. Montresor
for their critical reading of the manuscript. We also thank A. Kuwata,
A. Mizuno, K. Sasakawa and laboratory members for discussions and
technical advice. This study was supported, in part, by a Research
Fellowship for Young Scientists to Y.S. (no. 22-9326) and by a
Grant-in-Aid for challenging Exploratory Research (no. 24657014)
to M.S. from JSPS.
References
1.
Hamilton WD. 1967 Extraordinary sex ratios. A sexratio theory for sex linkage and inbreeding has new
implications in cytogenetics and entomology.
Science 156, 477–488. (doi:10.1126/science.156.
3774.477)
2.
3.
Trivers RL, Willard DE. 1973 Natural selection of parental
ability to vary the sex ratio of offspring. Science 179,
90–92. (doi:10.1126/science.179.4068.90)
Charnov EL. 1982 The theory of sex allocation.
Princeton, NJ: Princeton University Press.
4.
5.
West S. 2009 Sex allocation. Princeton, NJ: Princeton
University Press.
Dacks J, Roger AJ. 1999 The first sexual lineage and
the relevance of facultative sex. J. Mol. Evol. 48,
779–783. (doi:10.1007/PL00013156)
Proc R Soc B 280: 20130503
We have demonstrated that the sex allocation of individual
mother cells was distorted (figure 3). Creating outbreeding
opportunity may be one of the factors that favours the evolution of split sex ratios among mother cells. Although
many centric diatoms can reproduce homothallically, selffertilization deleteriously affects the viability and fertility of
offspring in several species [10,23]. Thus, a strategy that distributes the same sex cells as sibling cells that come close to
each other may decrease the inbreeding rate depending on
population structure. The present sex allocation among individual mother cells suggests that common epigenetic states,
such as the number of transcription factors that regulate cell
fate, are passed to daughter cells. Cellular states can be
passed from one generation to the next without alterations
in DNA sequence information by various cis and trans signals
[24], and many reports provide examples of this in prokaryotes and single-celled eukaryotes [25]. We found that
sibling cells with the same fate are likely to divide equally;
therefore, the common epigenetic states may establish a linkage between the symmetry of cell division and cell fate
similarity in the two sibling cells. The molecular mechanisms
of such epigenetic inheritance are unknown; however,
applying genetic engineering tools that have been recently
developed for diatoms [26] will help in identifying the
mechanisms underlying the observed distorted sex allocation.
6
rspb.royalsocietypublishing.org
(a) Cell fate similarity in a sibling pair
homogeneous populations of cells with smaller sexualizable
diameter, the differential ability of cells to elongate longitudinally
would be important for sex allocation. Furthermore, we
measured the cell area immediately after cell division; therefore,
our results indicate that cell elongation ability in the mother cell
generation affects the cell fate decision-making process without
waiting for elongation after cell division (figure 4b,c; electronic
supplementary material, figure S3).
In this study, the sex ratio became less male-biased as the cell
population density increased (figure 5). This extrinsic sex allocation may agree with the fertility insurance hypothesis, which
has been previously discussed for sexual reproduction of a
malarial parasite [31,32]. The C. meneghiniana complex is
known to form aggregations in culture condition [33], and exhibits local blooms and dominates in estuaries [34,35]. Therefore, it
may be important for these diatoms to adjust their sex ratio by
changing local cell density. On the basis of our results and a previous study [12] of C. meneghiniana, a vegetative cell that is
triggered to become a sperm can differentiate into no more
than four sperms. According to the fertility insurance hypothesis,
the optimal sex ratio could be affected by the number of gametes
(four) released from one male gametocyte [13,14]. Therefore, in
the case of this diatom, an insufficient number of gametes
released from a male gametocyte may accelerate male bias. As
an overall trend, the sex ratio became less male-biased as the
cell population density increased. However, this trend was
reversed at the highest cell density (figure 5). Further investigation is required to understand the exact mechanism and
reason for this response.
Downloaded from http://rspb.royalsocietypublishing.org/ on April 2, 2017
6.
8.
9.
11.
12.
13.
14.
15.
16.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28. Rao VNR. 1996 Size dependent variable
reproductive behaviour in Cyclotella meneghiniana
(Bacillariophyta). Nova Hedwigia Beiheft 112, 235–
238.
29. Koester JA, Brawley SH, Karp-Boss L, Mann DG.
2007 Sexual reproduction in the marine centric
diatom Ditylum brightwellii (Bacillariophyta).
Eur. J. Phycol. 42, 351– 366. (doi:10.1080/
09670260701562100)
30. Desikachary TV, Rao VNR. 1973 Studies on Cyclotella
meneghiniana Kütz. III. The frustule. Proc. Ind. Acad.
Sci. B 77, 78 –91.
31. Reece SE, Drew DR, Gardner A. 2008 Sex ratio
adjustment and kin discrimination in malaria
parasites. Nature 453, 609–614. (doi:10.1038/
nature06954)
32. Mitri C, Thiery I, Bourgouin C, Paul REL. 2009
Density-dependent impact of the human malaria
parasite Plasmodium falciparum gametocyte sex
ratio on mosquito infection rates. Proc. R. Soc. B
276, 3721– 3726. (doi:10.1098/rspb.2009.0962)
33. Herth W, Zugenmaier P. 1977 Ultrastructure of the
chitin fibrils of the centric diatom Cyclotella cryptica.
J. Ultra. Res. 61, 230 –239. (doi:10.1016/S00225320(77)80090-7)
34. De Se‘ve MA. 1993 Diatom bloom in the tidal
freshwater zone of a turbid and shallow estuary,
Rupert Bay (James Bay, Canada). Hydrobiologia 269,
225–233. (doi:10.1007/BF00028021)
35. Muylaert K, Sabbe K. 1999 Spring phytoplankton
assemblages in and around the maximum turbidity
zone of the estuaries of the Elbe (Germany), the
Schelde (Belgium/The Netherlands) and the
Gironede (France). J. Mar. Syst. 22, 133–149.
(doi:10.1016/S0924-7963(99)00037-8)
36. West-Eberhard MJ. 2003 Developmental
plasticity and evolution. Oxford, NY: Oxford
University Press.
37. Kirshner MW, Gerhart JC. 2005 The plausibility of
life: resolving Darwin‘s dilemma. New Haven, CT:
Yale University Press.
7
Proc R Soc B 280: 20130503
10.
17.
correlations. Analyst 130, 311 –317. (doi:10.1039/
b409860a)
Inoue I, Wakamoto Y, Moriguchi H, Okano K, Yasuda
K. 2001 On-chip culture system for observation of
isolated individual cells. Lab. Chip. 1, 50 –55.
(doi:10.1039/b103931h)
Schultz ME, Trainor FR. 1970 Production of male
gametes and auxospores in a polymorphic clone of
the centric diatom Cyclotella. Can. J. Bot. 48,
947 –951. (doi:10.1139/b70-133)
Rao VNR. 1970 Studies on Cyclotella meneghiniana
Kütz. Proc. Indian. Acad. Sci. B 73, 285–287.
Hoops JH, Floyd GL. 1979 Ultrastructure of the
centric diatom, Cyclotella meneghiniana: vegetative
cell and auxospore development. Phycologia 18,
424 –435. (doi:10.2216/i0031-8884-18-4-424.1)
Abràmoff M, Magalhães P. 2004 Image processing
with ImageJ. Biophotonics Int. 11, 36 –42.
Akaike H. 1973 Maximum likelihood identification of
Gaussian autoregressive moving average models.
Biometrika 60, 255–265. (doi:10.1093/biomet/60.
2.255)
Hiltz M, Bates SS, Kaczmarska I. 2000 Effect of light:
dark cycles and cell apical length on the sexual
reproduction of the pennate diatom Pseudonitzschia multiseries (Bacillariophyceae) in culture.
Phycologia 39, 59– 66. (doi:10.2216/i0031-888439-1-59.1)
Bonasio R, Tu S, Reinberg D. 2010 Molecular signals
of epigenetic states. Science 330, 612 –616. (doi:10.
1126/science.1191078)
Veening JW, Smits WK, Kuipers OP. 2008 Bistability,
epigenetics, and bet-hedging in bacteria. Annu. Rev.
Microbiol. 62, 193–210. (doi:10.1146/annurev.
micro.62.081307.163002)
Poulsen N, Kröger N. 2005 A new molecular tool for
transgenic diatoms. FEBS. J. 272, 3413–3423.
(doi:10.1111/j.1742-4658.2005.04760.x)
Davidovich N. 1994 Factors controlling the size of
initial cells in diatoms. Russ. J. Plant. physiol. 41,
220 –224.
rspb.royalsocietypublishing.org
7.
Ramesh MA, Malik SB, Logsdon JM. 2005 A
phylogenomic inventory of meiotic genes: evidence
for sex in Giardia and an early eukaryotic origin of
meiosis. Curr. Biol. 15, 185 –191. (doi:10.1016/j.
cub.2005.01.003)
Weibel DB, DiLuzio WR, Whitesides GM. 2007
Microfabrication meets microbiology. Nat. Rev.
Microbiol. 5, 209 –218. (doi:10.1038/nrmicro1616)
El-Ali J, Sorger PK, Jensen KF. 2006 Cells on chips.
Nature 442, 403 –411. (doi:10.1038/nature05063)
Mann DG. 2011 Size and sex. In The diatom world
(eds J Seckbach, JP Kociolek), pp. 145–166.
Dordrecht, The Netherlands: Springer.
Chepurnov VA, Mann DG, Sabbe K, Vyverman W.
2004 Experimental studies on sexual reproduction in
diatoms. Int. Rev. Cytol. 237, 91– 154. (doi:10.
1016/S0074-7696(04)37003-8)
Beszteri B, John U, Medlin L. 2007 An assessment
of cryptic genetic diversity within the Cyclotella
meneghiniana species complex (Bacillariophyta)
based on nuclear and plastid genes, and amplified
fragment length polymorphisms. Eur. J. Phycol. 42,
47 –60. (doi:10.1080/09670260601044068)
Schultz M, Trainor FR. 1968 Production of male
gametes and auxospores in the centric diatoms
Cyclotella meneghiniana and C. cryptica. J. Phycol. 4,
85 –88. (doi:10.1111/j.1529-8817.1968.tb04681.x)
West SA, Smith TG, Nee S, Read AF. 2002 Fertility
insurance and the sex ratios of malaria and related
hemospororin blood parasites. J. Parasitol. 88, 1–6.
(doi:10.1645/0022-3395(2002)088[0258:FIATSR]2.0.
CO;2)
Gardner A, Reece S, West S. 2003 Even more
extreme fertility insurance and the sex ratios of
protozoan blood parasites. J. Theor. Biol. 223,
515–521. (doi:10.1016/S0022-5193(03)00142-5)
Bischoff H. 1963 Phycological studies. IV. Some soil algae
from Enchanted Rock and related algal species, vol. 6318,
pp. 1–95. Austin, TX: University of Texas Publications.
Wakamoto Y, Ramsden J, Yasuda K. 2005 Single-cell
growth and division dynamics showing epigenetic