1. Ecology
  2. Microbiology and Infectious Disease
Download icon

Light-dependent single-cell heterogeneity in the chloroplast redox state regulates cell fate in a marine diatom

  1. Avia Mizrachi
  2. Shiri Graff van Creveld
  3. Orr H Shapiro
  4. Shilo Rosenwasser
  5. Assaf Vardi  Is a corresponding author
  1. Weizmann Institute of Science, Israel
  2. The Volcani Center, Israel
  3. The Hebrew University of Jerusalem, Israel
Research Article
  • Cited 0
  • Views 944
  • Annotations
Cite this article as: eLife 2019;8:e47732 doi: 10.7554/eLife.47732

Abstract

Diatoms are photosynthetic microorganisms of great ecological and biogeochemical importance, forming vast blooms in aquatic ecosystems. However, we are still lacking fundamental understanding of how individual cells sense and respond to diverse stress conditions, and what acclimation strategies are employed during bloom dynamics. We investigated cellular responses to environmental stress at the single-cell level using the redox sensor roGFP targeted to various organelles in the diatom Phaeodactylum tricornutum. We detected cell-to-cell variability using flow cytometry cell sorting and a microfluidics system for live imaging of oxidation dynamics. Chloroplast-targeted roGFP exhibited a light-dependent, bi-stable oxidation pattern in response to H2O2 and high light, revealing distinct subpopulations of sensitive oxidized cells and resilient reduced cells. Early oxidation in the chloroplast preceded commitment to cell death, and can be used for sensing stress cues and regulating cell fate. We propose that light-dependent metabolic heterogeneity regulates diatoms’ sensitivity to environmental stressors in the ocean.

https://doi.org/10.7554/eLife.47732.001

eLife digest

Microscopic algae, such as diatoms, are widely spread throughout the oceans, and are responsible for half of the oxygen we breathe. At certain times of the year these algae grow very rapidly to form large “blooms” that can be detected by satellites in space. These blooms are generally short-lived because the algae are either eaten by other marine organisms, run out of nutrients, or die as a result of being infected by viruses or bacteria. However, some diatom cells survive the end of the bloom and go on to generate new blooms in the future, but it is still not clear how.

As the bloom collapses, diatoms experience many stressful conditions which can cause active molecules known as reactive oxygen species, or ROS for short, to accumulate inside cells. Normally growing cells also produce low amounts of ROS, which regulate various processes that are important for maintaining a cell’s health. However, high amounts of ROS can cause damage, which may lead to a cell’s death. Now, Mizrachi et al. investigated why some algae survive while others die in response to stressful conditions, focusing on the amount of ROS that accumulates within the diatom Phaeodactylum tricornutum.

Laboratory experiments showed that individual cells of P. tricornutum respond differently to environmental stress, forming two distinct groups of either sensitive or resilient cells. Sensitive cells accumulated high levels of ROS within a cell compartment known as the chloroplast and eventually died. Whereas resilient cells were able to maintain low levels of ROS in the chloroplast and survived long after the other cells perished. Populations of genetically identical diatom cells also formed distinct groups of sensitive and resilient cells, demonstrating that these two opposing reactions to stress are not caused by genetic differences between cells.

Lastly, Mizrachi et al. showed that how diatoms acclimate to stress depends on the amount of light they are exposed to. When in the dark, all cells became sensitive to oxidative stress, without forming distinct groups. But, when exposed to strong light that mimics the ocean surface, cells formed distinct groups within the population. This suggests that light regulates how susceptible these microscopic algae are to environmental stress.

The different responses within a population may serve as a “bet-hedging” strategy, enabling at least some of the cells to survive unpredicted stressful conditions. The next challenge will be to find out whether algae growing in the oceans also use the same strategy and investigate what impact this has on diatom blooms.

https://doi.org/10.7554/eLife.47732.002

Introduction

Diatoms are considered amongst the most successful and diverse eukaryotic phytoplankton groups, and are estimated to contribute 20% of global net primary production (Armbrust, 2009; Nelson et al., 1995; Malviya et al., 2016). They form massive blooms and are thus central to the biogeochemical cycling of important elements such as carbon, nitrogen, phosphorus, iron and silica, in addition to their important role at the base of marine food webs (Armbrust, 2009; Nelson et al., 1995; Morel and Price, 2003; Strass and Nöthig, 1996; Tréguer et al., 2018). As other phytoplankton, diatoms need to constantly acclimate to physicochemical gradients in a fluctuating environment. They are exposed to stress from different biotic and abiotic origins such as grazing, viruses, bacteria, allelopathic interactions, light availability, and nutrient limitations (Bidle, 2015; Bidle, 2016; Ianora et al., 2006; Thamatrakoln et al., 2013; Thamatrakoln et al., 2012; Kimura and Tomaru, 2015; van Tol et al., 2017; Vardi et al., 2006). Importantly, induction of programmed cell death (PCD) in response to different stressors has been suggested as an important mechanism contributing to the fast turnover of phytoplankton and the rapid bloom demise (Bidle, 2015; Bidle, 2016; Vardi et al., 2007).

Recent studies suggested that diatoms can differentially respond to diverse environmental cues based on compartmentalized redox fluctuations that also mediate stress-induced PCD (Graff van Creveld et al., 2015; Rosenwasser et al., 2014; Volpert et al., 2018). Reactive oxygen species (ROS) are known to play an important role in sensing stress and additional signals across kingdoms, from bacteria to plants and animals (Vardi et al., 1999; Mittler et al., 2011; Suzuki et al., 2012; D'Autréaux and Toledano, 2007; Dietz et al., 2016). They are produced as byproducts of oxygen-based metabolism in respiration and photosynthesis, by ROS generating enzymes, and due to various stress conditions (Graff van Creveld et al., 2015; D'Autréaux and Toledano, 2007; Sheyn et al., 2016; Foyer and Noctor, 2016; Luo et al., 2014; Vardi et al., 2002; Waring et al., 2010). To maintain redox balance and avoid oxidative damage, cells harbor various ROS scavenging enzymes and small antioxidant molecules that regulate and buffer ROS levels, such as glutathione (GSH), ascorbate and NADPH. ROS can cause rapid post-translational modifications of pre-existing proteins through oxidation, affecting their activity faster than changes in gene expression (D'Autréaux and Toledano, 2007). The specificity of the ROS signal is derived from the specific chemical species, its concentration, sub-cellular localization, temporal dynamics, and available downstream ROS-sensitive targets (Graff van Creveld et al., 2015; D'Autréaux and Toledano, 2007; Sheyn et al., 2016; Noctor and Foyer, 2016; Owusu-Ansah and Banerjee, 2009; Tsukagoshi et al., 2010). Therefore, ROS production and redox metabolic networks can be used to sense and integrate information of both the metabolic state of the cell and its microenvironment.

H2O2 is a relatively mild and stable ROS that can accumulate in cells due to various stress conditions, thus often serves as a signaling molecule (Vardi et al., 1999; Mittler et al., 2011; Suzuki et al., 2012; D'Autréaux and Toledano, 2007; Dietz et al., 2016; Exposito-Rodriguez et al., 2017). It has a preferential activity towards cysteine residues, and can remodel the redox-sensitive proteome network (Rosenwasser et al., 2014; Dietz et al., 2016; Noctor and Foyer, 2016). In addition, H2O2 can diffuse across membranes (depending on membrane properties) and through aquaporins channels (Bienert and Chaumont, 2014). Combined properties as lower toxicity, diffusibility and selective reactivity make H2O2 suitable for studying signaling in various biological systems (Dietz et al., 2016; Noctor and Foyer, 2016). Since many environmental stressors induce ROS generation (Vardi et al., 1999; Sheyn et al., 2016; Waring et al., 2010; Exposito-Rodriguez et al., 2017; Asada, 2006; Diaz et al., 2018; Diaz et al., 2013), application of H2O2 can reproduce the downstream cellular response (Graff van Creveld et al., 2015; Rosenwasser et al., 2014). Importantly, H2O2 application in marine diatoms led to oxidation patterns similar to other environmental stressors such as nutrient limitations, toxic infochemicals and high light, demonstrating that it induces similar physiological responses within the cell (Figure 1) (Graff van Creveld et al., 2015; Rosenwasser et al., 2014). Although the H2O2 concentrations used in these studies were higher than measured in bulk seawater, which is typically in the nanomolar range (Zinser, 2018), local concentrations at the microenvironment of phytoplankton can be significantly higher due to patchiness in time and space (Stocker, 2012; Chrachri et al., 2018; Seymour et al., 2017). Local and temporal production of H2O2 by the cell itself or by its neighbors could lead to intracellular concentrations which are orders of magnitude higher than measured in bulk seawater in the field, especially during dense blooms or in aggregates and biofilms (Diaz et al., 2018; Diaz et al., 2013; Stocker, 2012; Chrachri et al., 2018; Seymour et al., 2017). H2O2 application in the model diatom P. tricornutum also led to the induction of cell death, in a dose-dependent manner, with characteristics of PCD that included externalization of phosphatidylserine, DNA laddering, and compromised cell membrane (Graff van Creveld et al., 2015). Moreover, early oxidation of the mitochondrial GSH pool preceded subsequent cell death at the population level following exposure to H2O2 or diatom-derived infochemicals (Graff van Creveld et al., 2015; Graff van Creveld et al., 2016).

Organelle-specific oxidation in response to environmental stressors in P. tricornutum.

Maximal roGFP degree of oxidation (OxD) measured in the chloroplast, nucleus and mitochondria in response to various environmental stress conditions based on bulk measurements using flow cytometry, data from previously published works. Cont – control untreated cells. N starvation - nitrate limitation 56 hr (See Rosenwasser et al., 2014). Fe starvation - iron limitation, day 3 (see Graff van Creveld et al., 2015; Graff van Creveld et al., 2016). High light, 700 µmol photons·m−2·sec−1, 38 min (see Graff van Creveld et al., 2015). BrCN - cyanogen bromide, 0.1 and 0.5 µM, 10 min (see Graff van Creveld et al., 2015). DD - 2E,4E/Z-decadienal (diatom-derived unsaturated aldehyde), 5, 25, 50 and 100 µM, 3 hr (see Graff van Creveld et al., 2015). H2O2 – 30, 80, and 150 µM, ~30 min (see Graff van Creveld et al., 2015). Bars represent average of biological triplicates, error bars are standard error of the mean.

https://doi.org/10.7554/eLife.47732.003

In the current work, we investigated phenotypic variability within diatom populations in response to oxidative stress. Classically, phenotypic variability has been studied primarily in established bacterial, yeast and mammalian model systems with little ecological relevance (Balaban, 2011; Raj and van Oudenaarden, 2008). However, examination of phenotypic heterogeneity in phytoplankton cells in response to stress is still underexplored, and research has been carried out primarily at the population level, masking heterogeneity. Cell-to-cell variability could result in different cellular strategies employed by the population to cope with environmental stress, and may provide important insights into cell survival during bloom succession. We established single-cell approaches to measure in vivo ROS dynamics in the model diatom P. tricornutum using flow cytometry and microfluidics live-imaging microscopy. We used P. tricornutum strains expressing redox-sensitive GFP (roGFP) targeted to different sub-cellular compartments (Rosenwasser et al., 2014) and exposed cells to oxidative stress and high light conditions. The oxidation of roGFP is reversible, and can be quantified using ratiometric fluorescence measurements (Meyer, 2008). The roGFP oxidation degree (OxD) reports the redox potential of the GSH pool (EGSH), which represents the balance between GSH and its oxidized form (Meyer, 2008). Therefore, roGFP OxD provides an important metabolic readout of the redox state of the cell, and represents the oxidation state of native proteins in the monitored organelles (Rosenwasser et al., 2014; Meyer, 2008). We uncovered a previously uncharacterized phenotypic heterogeneity in the response of a marine diatom to oxidative stress and high light. Furthermore, we revealed a specific link between oxidation patterns in the chloroplast and cell fate regulation.

Results

Bi-stable chloroplast roGFP oxidation in response to oxidative stress reveals distinct subpopulations

Previous works demonstrated that P. tricornutum exhibits organelle-specific oxidation patterns in responses to diverse environmental stress conditions (Figure 1) (Graff van Creveld et al., 2015; Rosenwasser et al., 2014; Graff van Creveld et al., 2016). However, these findings were based on bulk measurements averaging the phenotypes within the population. In order to investigate heterogeneity within the population, we measured the response to oxidative stress in P. tricornutum strains expressing roGFP targeted to the chloroplast, nucleus and mitochondria at single-cell resolution using flow cytometry. At steady-state conditions without perturbations, roGFP OxD distribution in the population had a single distinct peak, representing a reduced state at all examined compartments (Figure 2A,E,I and Figure 2—figure supplements 12). Next, we examined the response of cells to oxidative stress by treatment with H2O2 at concentrations that led to oxidation patterns similar to other environmental stressors (Figure 1), and that also led to death of part of the population (Graff van Creveld et al., 2015).

Figure 2 with 6 supplements see all
Organelle-specific oxidation of roGFP in response to H2O2 reveals heterogeneity at the single-cell level.

The distribution of roGFP OxD in the population over time was measured by flow cytometry in P. tricornutum cells expressing roGFP targeted to the chloroplast (chl-roGFP, (A–D, M)), nucleus (nuc-roGFP, (E–H, N)) and mitochondria (mit-roGFP, (I–L, O)). (A–L) Oxidation of roGFP in response to 0 µM (A,E,I), 50 µM (B,F,J), 80 µM (C,G,K), and 100 µM (D,H,L) H2O2. Maximum reduction (blue) and oxidation (red) of roGFP following additions of 2 mM Dithiothreitol (DTT) or 200 µM H2O2 respectively are shown as a reference. The ‘oxidized’ and ‘reduced’ subpopulations are marked by red and blue dashed boxes respectively (C, G, K). The experiment was done in triplicates, for visualization the first replica is shown except for the first 4 min in which all replicates are shown for higher temporal resolution. Each histogram consists of >8000 (A–D), >5900 (E–H) and >1400 (I–L) roGFP-positive cells. Measurements of >100% OxD may result from increased auto-fluorescence leakage after long exposure to stress (Figure 2—figure supplements 56). (M–O) The fraction of the ‘oxidized’ subpopulation over time upon exposure to 0–100 µM H2O2. Mean ± SEM, n = 3 biological repeats. SEM lower than 0.018 are not shown.

https://doi.org/10.7554/eLife.47732.005

Chloroplast-targeted roGFP (chl-roGFP) exhibited a distinct bimodal distribution following treatments of 50–100 µM H2O2, revealing two distinct subpopulations of ‘oxidized’ and ‘reduced’ cells (Figure 2B–D and Figure 2—figure supplement 3A–C). These subpopulations emerged within the first few minutes after H2O2 addition (Figure 2B–D). The existence of these subpopulation is masked in bulk analysis (Figure 2—figure supplement 1), demonstrating the importance of single-cell measurements. In the ‘oxidized’ subpopulation, roGFP completely oxidized (~100%) in response to H2O2, reaching a similar distribution of the fully oxidized positive control (200 µM H2O2) (Figure 2B–D and Figure 2—figure supplement 3G). However, in the ‘reduced’ subpopulation roGFP reached lower values of 30–43% OxD within 2–24 min post treatments, and then gradually recovered (Figure 2B–D and Figure 2—figure supplement 3G). Only a minor fraction of the cells displayed intermediate oxidation, suggesting that these subpopulations represent discrete redox states. Interestingly, a larger fraction of cells was within the ‘oxidized’ subpopulation at 20–25 min post treatment compared to later time points, indicating that some cells were able to recover during this time (Figure 2M). The proportion between these subpopulations stabilized after 46–51 min post treatment, and was H2O2-dose dependent, as more cells were within the ‘oxidized’ subpopulation at higher H2O2 concentrations (Figure 2M). The quick emergence of stable co-existing ‘oxidized’ and ‘reduced’ subpopulations exposed underlying heterogeneity within the diatom population, resulting in a differential response to oxidative stress.

This clear bi-stable pattern was unique to the chloroplast-targeted roGFP. Nuclear-targeted roGFP displayed a continuous distribution in response to H2O2 treatments, and no distinct subpopulations could be observed (Figure 2F–H). Within minutes post treatment, nuclear roGFP exhibited fast oxidation even in response to a low H2O2 concentration of 50 µM, which had only a mild effect on the chloroplast (Figure 2B,F). At that concentration, nuclear roGFP oxidation was followed by a gradual and much slower recovery, which lasted >5 hr post treatment (Figure 2F). At higher concentrations, the entire population was oxidized within 3 min post treatment, and most cells remained stably oxidized >5 hr post treatment (Figure 2G–H). The mitochondria-targeted roGFP exhibited a heterogeneous redox response, as seen in the 80 µM and 100 µM H2O2 treatments starting at ~24 min post treatment (Figure 2K–L). However, distinct subpopulations were not clearly separated until later stages, and were not detected consistently in different experiments (Figure 2K–L and Figure 2—figure supplement 4). Therefore, we chose to focus on the chl-roGFP strain, which revealed two discrete subpopulations.

Oxidation of chl-roGFP precedes the induction of cell death

Next, we examined the possible link between early chloroplast EGSH oxidation and subsequent H2O2-dependent cell death. We quantified cell death 24 hr post H2O2 treatment using flow cytometry measurements of Sytox green staining, which selectively stains nuclei of dead and dying cells. The fraction of ‘oxidized’ cells 1–2 hr post treatment was correlated with the fraction of dead cells at 24 hr (Figure 3A, R2 = 0.89, p=2.2·10−16), suggesting that early oxidation in the chloroplast in distinct subpopulations may predict cell fate at much later stages.

Figure 3 with 2 supplements see all
Early oxidation of chl-roGFP in a subpopulation precedes the induction of cell death.

(A) The fraction of dead cells measured by Sytox positive staining 24 hr post 50–200 H2O2 treatment as a function of the fraction of chl-roGFP ‘oxidized’ cells 1–2 hr post treatment (see Figure 2C). N ≥ 3 biological repeats per treatment, individual samples are shown, data of 5 independent experiments. Linear model is shown with 95% confidence interval in gray, formula: y = 1.09x-0.16, R2 = 0.89. (B) Colony forming units (CFU) survival of individual cells that were sorted into fresh media for regrowth based on their chl-roGFP oxidation at different times following treatment with 80 µM H2O2. Control – untreated chl-roGFP positive cells that were sorted regardless of degree of oxidation. CFU % survival was measured by the number of CFU divided by the number of sorted cells. N ≥ 6 biological repeats, individual repeats are shown, each of 24–48 separately sorted single cells. Oxidized – red circles; reduced – blue triangles; control – gray triangles. *** – p<0.001; ns – non-significant.

https://doi.org/10.7554/eLife.47732.013

To investigate directly the link between early chl-roGFP oxidation and subsequent cell death, we used fluorescence-activated cell sorting (FACS) to sort cells based on chl-roGFP oxidation and measured their survival. Single cells of the ‘oxidized’ and ‘reduced’ subpopulations were sorted into fresh media at different time-points following the addition of 80 µM H2O2, and colony forming units (CFU) were counted to assess survival 3.5–9 weeks later (Figure 3B and Figure 3—figure supplement 1). The CFU assay provides a direct link between chl-roGFP oxidation and the ability of individual cells to proliferate, and in addition enables to generate clonal populations which were also used for downstream analyses (Figure 3—figure supplement 1). When sorted 30 min post treatment, the ‘oxidized’ subpopulation exhibited a high survival rate (92.3 ± 1.4%) that was similar to the ‘reduced’ subpopulation (94.1 ± 1.1%, p=0.24, paired t-test) and to sorted untreated control (96 ± 0.9%, p=0.29, Dunnett test, Figure 3B). However, at later time-points the survival of the ‘oxidized’ subpopulation gradually diminished, and was significantly lower than both the ‘reduced’ and control sorted cells (p<0.001 for all comparisons, paired t-test for comparisons with ‘reduced’ cells, Dunnett test for comparisons with control). When sorted 60 min post treatment almost half of the ‘oxidized’ cells recovered (45.1 ± 2%), but when sorted 100 min following treatment only 12.7 ± 2.1% survived (p<0.001, Tukey test, Figure 3B). These results suggest that after a distinct exposure time, cell death is induced in an irreversible manner in the ‘oxidized’ subpopulation. In contrast, the ‘reduced’ subpopulation from the same culture and treatment exhibited a high survival rate similar to the control at all time-points examined, demonstrating its resilience to the stress (Figure 3B, P≥0.86 for all comparisons, Dunnett test). Corroborating these findings, cell death measurements using Sytox staining of sorted subpopulations also showed higher mortality in the ‘oxidized’ cells compared to the ‘reduced’ and control cells, which remained viable (Figure 3—figure supplement 2). Taken together, these results demonstrate that the ‘oxidized’ subpopulation was sensitive to the oxidative stress which led to induction of cell death in those cells, while the ‘reduced’ subpopulation was able to survive. In addition, we detected a distinct phase of ‘pre-commitment’ to cell death, ranging approximately 30–60 min in most cells, during which despite the strong oxidation in the chloroplast the fate of the ‘oxidized’ subpopulation is still reversible upon removal of the stress, and they are still able to survive. After this initial phase, the ‘oxidized’ cells were not able to survive even when the stress was removed from the system by sorting into fresh media.

Early oxidation of chloroplast EGSH is linked to cell fate determination at the single-cell level

In order to track oxidation dynamics and subsequent cell fate of individual cells, we established a microfluidics platform for in vivo long-term epifluorescence imaging, under controlled flow, light and temperature conditions customized specifically for diatom cells (Figure 4—figure supplement 1 and Videos 12). We introduced cells expressing chl-roGFP into a custom-made microfluidics device, let the cells settle, and introduced treatments of either 80 µM H2O2 or fresh media (control) continuously for 2.5–3 hr, after which the treatment was washed by fresh media (see Materials and methods, Figure 4—figure supplement 1A). In addition, the use of microfluidics enabled imaging of the basal OxD state of single cells prior to treatment, as well as the introduction of Sytox green at the end of the experiment to visualize cell death. We detected the distinct ‘oxidized’ and ‘reduced’ subpopulations following 80 µM H2O2 treatment, similar to the flow cytometry experiments (Figure 4C,F, and Video 1). However, no clear differences were observed in their OxD prior to treatment (Figure 4—figure supplement 2 and Video 1). The separation between the subpopulations emerged within 20 min of exposure to 80 µM H2O2, and remained stable over the course of the experiment with the ‘oxidized’ subpopulation maintaining a high OxD above 80% (Figure 4F, Figure 4—figure supplement 2B and Video 1). The ‘reduced’ subpopulation exhibited an immediate response to H2O2 comparable with flow cytometry measurements, from 25–45% OxD before treatment to 30–65% OxD during the first 20 min post 80 µM H2O2 treatment (Figure 4F, Figure 4—figure supplement 2B, and Video 1). Following this initial oxidation, the ‘reduced’ cells recovered gradually over the next hours, reducing to 5–25% OxD 8 hr post treatment, even below the initial basal state (Figure 4F, Video 1). A gradual slow reduction was also observed in control cells over the course of the experiment (Figure 4E, Video 2), which may represent acclimation to the experimental setup or a diurnal redox alteration. Control cells did not oxidize in response to addition of fresh media (Figure 4E and Figure 4—figure supplement 2A), excluding the possibility that the oxidation observed in 80 µM H2O2 treated cells was due to shear stress during treatment.

Figure 4 with 4 supplements see all
Tracking redox dynamics of individual cells using long term in vivo imaging in a microfluidics setup.

Oxidation of chl-roGFP was imaged over time using a customized microfluidic setup and epifluorescence microscopy. Cells were imaged following treatment with either fresh media (control; A, B, E) or 80 µM H2O2 (C, D, F). To quantify cell death, cells were stained with Sytox ~23 hr post treatment. (A, C) Representative frames depicted in pseudo-color of calculated roGFP OxD at different times post treatment (hh:mm). Two subpopulations of ‘oxidized’ (red arrows) and ‘reduced’ (blue arrows) cells were detected in treated cells. (B, D) Overlay of Sytox staining (green, dead cells) and chlorophyll auto-fluorescence (red) at 23:10 hr post treatment. (E, F) Quantification of chl-roGFP OxD per cell over time post treatment. Color is based on cell fate as measured ~23 hr post treatment: blue – alive (Sytox negative), red – dead (Sytox positive). OxD values > 110% likely result from auto-fluorescence leakage, and are not shown. (G) Box-plot of chl-roGFP OxD 40 min post treatment (arrow in E, F) of cells that were grouped based on their cell fate as measured ~23 hr post treatment. Gray – control alive; blue – 80 µM H2O2 alive; red – 80 µM H2O2 dead. Circles - single cells; cross – mean; colored box – 1st to 3rd quartiles; horizontal line within the box – median. Individual cells are shown. (E–G) N ≥ 250 cells per treatment from ≥3 different fields of view. Horizontal dashed line represents the chl-roGFP oxidation threshold used for discriminating cells that subsequently died or survived.

https://doi.org/10.7554/eLife.47732.018
Video 1
Microfluidics in vivo imaging of chl-roGFP oxidation over time following H2O2 treatment.

Oxidation of chl-roGFP cells was imaged over time using a customized microfluidic setup and epifluorescence microscopy with controlled flow, light and temperature conditions (see Materials and methods). Movie of chl-roGFP OxD in pseudo-color of cells treated with 80 µM H2O2 at time 0, the first frame is before treatment. Time stamp represents time post treatment (hh:mm). Color bar as in Figure 4.

https://doi.org/10.7554/eLife.47732.024
Video 2
Microfluidics in vivo imaging of basal chl-roGFP oxidation over time.

Oxidation of chl-roGFP cells was imaged over time using a customized microfluidic setup and epifluorescence microscopy with controlled flow, light and temperature conditions (see Materials and methods). Movie of chl-roGFP OxD in pseudo-color of cells treated with fresh media (control) at time 0, the first frame is before treatment. Time stamp represents time post treatment (hh:mm). Color bar as in Figure 4.

https://doi.org/10.7554/eLife.47732.025

We detected a clear correlation between initial oxidation in the chloroplast in response to oxidative stress and subsequent cell fate (Figure 4A–G). Cells that exhibited high chl-roGFP oxidation within the first 40 min also died at a much later stage, while cells that maintained a lower OxD were able to recover (Figure 4G). In addition, cells of the ‘reduced’ subpopulation and of control treatment were able to proliferate, further demonstrating their viability under these conditions (Videos 12). Logistic regression modeling of cell death as a function of chl-roGFP OxD at this time-point revealed a threshold of ~74% OxD, which discriminated with high accuracy (98.8%) between cells that subsequently died and cells that survived the treatment (0.8% false positive, 1.7% false negative; Figure 4G and Figure 4—figure supplement 3). These results corroborate the flow-cytometry analysis, and demonstrate that under these conditions early chloroplast EGSH response is linked to subsequent cell fate determination at the single-cell level.

The distinct subpopulations derive from phenotypic variability and not from variable genetic backgrounds

The differential chloroplast oxidation of the observed subpopulations could be due to genetic variability or due to phenotypic plasticity within the population. To differentiate between the two scenarios, we sorted chl-roGFP individual cells of the ‘oxidized’ and ‘reduced’ subpopulations 30 and 100 min post 80 µM H2O2 treatment as well as untreated control cells, and regrew them to generate clonal populations derived from cells exhibiting specific phenotypes. The clonal progeny cultures were subsequently exposed to 80 µM H2O2 and their chl-roGFP oxidation was measured (Figure 3—figure supplement 1). The two distinct subpopulations were detected in all the clones measured, and the fraction of the ‘oxidized’ subpopulation was again correlated with cell death (Figure 5 and Figure 5—figure supplement 1). Therefore, the different subpopulations observed did not originate from genetic differences, but rather represent phenotypic variability within isogenic populations.

Figure 5 with 1 supplement see all
Clonal populations derived from sorted cells maintain the bi-stable phenotype of chl-roGFP response to H2O2.

(A–C) The distribution of chl-roGFP OxD (%) 40–45 min post 80 µM H2O2 treatment in clonal populations derived from sorted single cells of different origins, 3 weeks post sorting. The ‘reduced’ (B) and ‘oxidized’ (C) subpopulations were sorted 30 min post 80 µM H2O2 treatment based on chl-roGFP oxidation (Figure 3—figure supplement 1); ‘control’ (A) – clones of untreated cells that were sorted based on positive roGFP fluorescence. Each histogram is of a single clone,≥9900 cells per histogram, six representative clones per group are shown. (D) The fraction of dead cells 24 hr post H2O2 treatment of the different clones shown in (A–C) as measured by positive Sytox staining. Data is shown as mean ± SEM, n = 6 clones per group per treatment.

https://doi.org/10.7554/eLife.47732.026

The ‘oxidized’ subpopulation is enriched with cells at G1 phase

One possible source for phenotypic variability in genetically homogenous populations can be explained by differences in the cell cycle phase, as the cell cycle is linked to metabolic changes including redox oscilations (Papagiannakis et al., 2017; Burhans and Heintz, 2009; Mathis and Ackermann, 2016; Diaz-Vivancos et al., 2015). Therefore, we sorted the ‘oxidized’ and ‘reduced’ subpopulations 30 min following 80 µM H2O2 treatment into a fixation solution, and stained the fixed cells with 4′,6-diamidino-2-phenylindole (DAPI) to quantify DNA content for cell cycle analysis. The sorted ‘oxidized’ subpopulation had a higher fraction of cells at G1 (86.9 ± 1.8%) compared to control untreated cells (76.8 ± 0.7%, p=0.0024, Tukey test, Figure 6). The ‘reduced’ subpopulation on the other hand had a smaller fraction of G1 cells (68.7 ± 2.2%) compared to both control (p=0.011, Tukey test) and ‘oxidized’ cells (p=0.0001, paired t-test), and exhibited a larger fraction of G2/M cells (Figure 6). These results demonstrate that although cell cycle phase alone cannot explain the differences between the subpopulations, it is linked to the chloroplast EGSH response to oxidative stress and may represent an important factor that affects H2O2 sensitivity in the population.

Cell cycle analysis of sorted ‘reduced’ and ‘oxidized’ subpopulations.

(A–C) Cell cycle analysis of FACS sorted control untreated cells (A) and of ‘reduced’ (B) and ‘oxidized’ (C) subpopulations that were sorted 30 min post 80 µM H2O2 treatment, based on chl-roGFP OxD. DAPI staining was used for DNA content quantification, gates used to discriminate G1 (two genome copies) and G2/M (four copies) are marked. (D) Fraction of cells within the G1 gate (marked in A-C) in sorted subpopulations. Data points of n = 6 biological repeats of 2 independent experiments (exp 1 and 2 marked with circles and triangles respectively) and the mean (orange plus sign) are shown, 1200–2000 cells analyzed per sample. P-values: *=0.011, **=0.0024, ***=0.0001. Tukey test was used for comparisons with control, paired t-test was used for comparing ‘reduced’ and ‘oxidized’ subpopulations.

https://doi.org/10.7554/eLife.47732.029

The bimodal chloroplast redox response is light dependent

Photosynthesis is a major source for reductive power as well as ROS in algal cells, and exposure to dark was shown to increase sensitivity to oxidative stress in another marine diatom (Volpert et al., 2018). Therefore, we hypothesized that light regime will affect the bimodal pattern of chl-roGFP following oxidative stress, and investigated the effects of short exposure to darkness during daytime. Cells were treated with 0–100 µM H2O2 and were immediately moved to the dark for 90 min, after which they were moved back to the light (dark treated, Figure 7A–C and Figure 7—figure supplement 1). These cells were compared to cells that were kept in the light during this time (light treated). The transition to the dark caused an immediate oxidation of the basal chl-roGFP OxD (without H2O2 treatment), reaching a peak within 15 min (Figure 7A,D). Then, while still under dark, chl-roGFP gradually reduced while maintaining a continuous distribution (Figure 7A,D). Upon shifting back to the light, chl-roGFP reduced within 2 min back to its basal state prior to dark exposure (Figure 7A,D). The dark mediated oxidation was specific to the chloroplast and was not detected in the nucleus (Figure 7—figure supplement 2), demonstrating organelle specificity of these redox fluctuations.

Figure 7 with 2 supplements see all
The bimodal chl-roGFP oxidation in response to H2O2 is light-dependent.

The effects of a short exposure to darkness during daytime on chl-roGFP oxidation patterns were examined. (A–C) Flow cytometry measurements of chl-roGFP OxD distribution in the population over time. Cells were treated with 0 µM (control, A), 50 µM (B), and 80 µM H2O2 (C), and were then transitioned to the dark at time 0 (within 5 min post H2O2 treatment). Cells were kept in the dark for 90 min (green) and were then transferred back to the light (cyan). The same H2O2 treatment without transition to the dark (light ctrl, black) and maximum oxidation (200 µM H2O2, red) and reduction (2 mM DTT, blue) are shown for reference. The experiment was done in triplicates that were highly similar, for visualization the first replica is shown. Each histogram consists of >8000 cells. (D) Mean ± SEM basal (control) chl-roGFP OxD over time of cells transitioned to the dark for 90 min (gray box) at time 0 (‘dark’) and cells kept in light conditions (‘light’), n = 3 biological repeats. Pre-treatment baseline under light conditions is shown. SEM lower than 0.5% are not shown. (E) The fraction of dead cells 24 hr post H2O2 treatment, with or without transition to the dark (‘dark’ and ‘light’ respectively), as measured by positive Sytox staining. Data is shown as mean ± SEM, n = 3 biological repeats. P values: *=0.0064, **=0.0026, ***=2·10−5, t-test.

https://doi.org/10.7554/eLife.47732.031

The transition to the dark eliminated the bimodal pattern of chl-roGFP oxidation in response to H2O2 (Figure 7C). No distinct subpopulations were observed while cells were under darkness even in cells treated with low H2O2 doses (Figure 7B and Figure 7—figure supplement 1B–C). The transition to the dark increased H2O2 sensitivity in the entire population, and following treatment of 80 µM H2O2 and transition to the dark chl-roGFP fully oxidized in the entire population and remained stably oxidized even after transition back to the light (Figure 7C). The bimodal pattern was regained only upon transition back to the light, and only at lower doses of 30 µM and 50 µM H2O2, in which some or most cells were able to recover following this transition (Figure 7B and Figure 7—figure supplement 1C). In accordance with the higher chl-roGFP oxidation, ‘dark’ treated cultures also exhibited higher mortality compared to ‘light’ treated cells (p≤0.0064 for all pairs in 50–100 µM H2O2 treatments, t-test, Figure 7E). Therefore, we conclude that the mechanism generating the bimodal response in the chloroplast is light dependent and plays an important role in cell fate regulation in diatoms.

High light induces variability in chloroplast ROS accumulation and subsequent cell survival

Next, we investigated the effects of high light (HL), a key environmental stressor in the marine ecosystem, which was shown to induce ROS generation in the chloroplast (Waring et al., 2010; Exposito-Rodriguez et al., 2017; Asada, 2006). We hypothesized that HL may cause H2O2 accumulation in the chloroplast, leading to a similar phenotypic variability as detected in response to H2O2 treatment. Cells were exposed to HL treatment of 2,000 µmol photons m−2 s−1 in order to evaluate the effect of HL exposure equivalent to full sunlight in nature (Long et al., 1994), and as was used in previous studies in P. tricornutum (Lepetit et al., 2013). Indeed, 1.5 hr after exposure to HL an ‘oxidized’ subpopulation started to emerge (Figure 8A and Figure 8—figure supplement 1B). The fraction of ‘oxidized’ cells gradually increased over time of HL exposure (Figure 8A–B and Figure 8—figure supplement 1A–F). These subpopulations were not detected in control cells that were kept under low light (LL, Figure 8A and Figure 8—figure supplement 1G–H), nor in cells expressing nuclear targeted roGFP that were exposed to HL (Figure 8—figure supplement 2A–B), demonstrating the specificity of the redox signal to the chloroplast.

Figure 8 with 2 supplements see all
High light induces the emergence of an ‘oxidized’ subpopulation with decreased survival.

Cells expressing chl-roGFP were exposed to either high light (HL, 2000 µmol photons m−2 s−1) or low light (LL, 14 µmol photons m−2 s−1), and were FACS sorted based on their chl-roGFP oxidation at different times of light exposure to measure single-cell survival using the colony forming units (CFU) assay. (A) Flow cytometry measurements of chl-roGFP OxD distribution in the population over time of HL exposure. Cells were kept in HL for 6.3 hr (green) and were then transferred back to LL for 1 hr (cyan). LL control (black) and maximum oxidation (200 µM H2O2, red) and reduction (2 mM DTT, blue) are shown for reference. The experiment was done in triplicates, for visualization the first replica is shown. Each histogram consists of >9000 cells. (B) The fraction of oxidized chl-roGFP cells over time of exposure to HL. Mean ± SEM, N = 3. (C) Images of chlorophyll auto-fluorescence of colonies generated from single cells that were sorted based on chl-roGFP oxidation following exposure to HL or LL for the CFU assay. Panels from top to bottom: oxidized (red box) and reduced (blue box) subpopulations sorted after 5.3 hr of HL, reduced subpopulation sorted under LL control conditions (gray box). CFU survival (%) was measured by the number of colonies divided by number of sorted cells, as shown at the right of each panel. Gates used for sorting are shown in Figure 8—figure supplement 1. (D) CFU survival (%) of single cells that were sorted based on chl-roGFP oxidation at different times of light exposure onto agar plates. Data of 2 independent experiments each with three biological repeats, 64–96 sorted cells per repeat, individual repeats are shown. (A–D) Data points of 6.3 hr HL exposure were exposed to HL for 6.3 hr and then moved back to LL for 1 hr.

https://doi.org/10.7554/eLife.47732.036

To measure the survival of the subpopulations that emerged under HL conditions, individual cells from the ‘oxidized’ and ‘reduced’ subpopulations were FACS sorted after different HL exposure times into agar plates for the single-cell CFU survival assay (see Materials and methods, Figure 8C–D). CFU survival of the ‘oxidized’ subpopulation gradually decreased over time of HL exposure and was significantly lower than both control and ‘reduced’ cells (p<0.001 for all comparisons. Tukey test was used for comparison with control, ANCOVA was used for comparison with ‘reduced’ and for the interaction with exposure time, Figure 8D). While most ‘oxidized’ cells survived ≤3 hr HL exposure (77.3 ± 4.2% CFU survival), longer exposure of >6 hr to HL led to only 3.1 ± 0.9% CFU survival (Figure 8D). In contrast, the ‘reduced’ subpopulation in the same HL treatment exhibited high CFU survival of 92.7 ± 3.6%, and maintained high CFU survival at all time-points examined similar to the control (p=0.93, Tukey test, Figure 8D). Interestingly, when cells were exposed to 6.3 hr HL and then moved to 1 hr LL the separation between the subpopulations became clearer, with almost no cells with intermediate oxidation states (Figure 8A and Figure 8—figure supplement 1E–F), resembling the response to H2O2 treatment (Figure 2—figure supplement 3B). To conclude, these findings demonstrate that HL can generate heterogeneity in ROS accumulation in the chloroplast within diatom populations, leading to differences in survival probability and likely affecting sensitivity to additional stressors in the marine environment.

Discussion

Our current understanding of the mechanisms that mediate acclimation to environmental stressors in marine microorganisms, including diatoms, has been derived primarily from observations at the population level, neglecting any heterogeneity at the single-cell level. Co-existence of distinct subpopulations that employ diverse cellular strategies can be significant for the survival of this globally important phytoplankton group, as was shown in other microorganisms (Balaban, 2011; Schreiber et al., 2016; Sengupta et al., 2017). Here, we established a novel system for studying phenotypic variability in the marine diatom P. tricornutum using flow cytometry and a microfluidics system for live imaging microscopy. Based on a metabolic readout of chloroplast EGSH oxidation, we uniquely identified two distinct subpopulations that emerged as an early response to H2O2 and high light, demonstrating the importance of phenotypic variability in cell fate regulation in diatoms.

We propose that in diatoms, chloroplast EGSH is involved in sensing specific environmental stress cues and in regulating cell fate (Figure 9). The chloroplast is a major source for generation of both ROS and reductive power to generate and recycle NADPH, thioredoxin and GSH (Dietz et al., 2016). In plants, chloroplast-generated ROS were demonstrated to be involved in retrograde signaling from the chloroplast and in hypersensitive response cell death (Dietz et al., 2016; Exposito-Rodriguez et al., 2017; Van Aken and Van Breusegem, 2015; Liu et al., 2007). In diatoms, specific stress cues can lead to ROS accumulation and EGSH oxidation in the chloroplast, as was shown in response to nitrogen limitation, the diatom-derived toxic infochemicals cyanogen bromide, and HL (Figure 8A and Figure 1) (Graff van Creveld et al., 2015; Rosenwasser et al., 2014). Specifically, HL can lead to ROS accumulation in the chloroplast by generation of singlet oxygen (1O2) in PSII, and by photoreduction of O2 to superoxide (O2-) in PSI, which in turn is rapidly converted to H2O2 by superoxide dismutase (SOD) (Waring et al., 2010; Exposito-Rodriguez et al., 2017; Asada, 2006). Redox fluctuations in the chloroplast can serve as a rapid mechanism to perceive specific environmental cues, by regulating key metabolic pathways on the post-translational level. Analysis of the redox-sensitive proteome in P. tricornutum revealed over-representation of chloroplast-targeted proteins, that were also oxidized to a greater degree under H2O2 treatment as compared to other subcellular compartments, further supporting the existence of a redox-based signaling network in the chloroplast (Rosenwasser et al., 2014; Woehle et al., 2017).

A conceptual model: phenotypic variability within diatom populations affects cell fate determination in response to stress conditions.

At steady state conditions, cells within the population have diverse metabolic states due to various factors such as local ROS levels, antioxidant capacity, metabolic activity, growth phase and cell cycle position. We propose that this variability could lead to a differential response to environmental stressors. Exposure to specific stress conditions, such as high light (HL) or toxic infochemicals, can lead to ROS accumulation at different subcellular compartments, including the chloroplast, which is used to sense the stress cue and regulate cell fate. Cells at a more susceptible metabolic state may accumulate high ROS levels and will subsequently die, as observed in the ‘oxidized’ subpopulation. More resilient cells may exhibit milder oxidation and may be able to acclimate, as observed in the ‘reduced’ subpopulation. Chloroplast EGSH oxidation is an early stage in this stress response and precedes the commitment to cell fate.

https://doi.org/10.7554/eLife.47732.041

The role of chloroplast EGSH perturbations in sensing specific stress cues gains further support by the early chl-roGFP oxidation, which preceded the ‘point of no return’, after which cell death was irreversibly activated in the ‘oxidized’ subpopulation (Figure 3B). This ‘pre-commitment’ phase provides an opportunity for cells to recover if conditions change during a narrow time frame of ~30–60 min following oxidative stress in most cells (Figure 3B), before the cell has accumulated damage beyond repair or a PCD cascade was fully activated. This ‘pre-commitment’ phase was shown previously in diatoms, as exogenous application of the antioxidant GSH rescued cells from otherwise lethal treatments of infochemicals or H2O2 only during the first hour (Graff van Creveld et al., 2015; Volpert et al., 2018). These findings shed light on the timeline of events in PCD progression in diatoms.

To date, the role of the chloroplast in mediating PCD has remained elusive, although mitochondria-generated ROS are known to play a key role in PCD in plants and animals (Van Aken and Van Breusegem, 2015; Lam et al., 2001). This knowledge gap is even greater in unicellular marine algae, for which the molecular basis for the PCD machinery is largely unknown (Bidle, 2016). In P. tricornutum, early mitochondrial oxidation was shown to precede subsequent cell-death in response to various stress conditions at the population level, but the link with chloroplast EGSH was less clear and depended on the specific stress cue (Graff van Creveld et al., 2015). In another diatom, chloroplast EGSH was shown to mediate changes in oxidative stress sensitivity upon light-dark transitions (Volpert et al., 2018). A recent model in plants suggested possible mitochondria-chloroplast cooperative interactions in the execution of ROS-mediated PCD (Van Aken and Van Breusegem, 2015). In addition, there are evidence for energetic coupling of chloroplasts and mitochondria in diatoms, and they use extensive energetic exchanges between these organelles to regulate the ATP/NADPH ratio (Bailleul et al., 2015). Taken together with the results presented here, we suggest that redox dynamics of both the mitochondria and the chloroplast are involved in cell fate regulation in diatoms.

We propose that cells that accumulate ROS above a certain threshold are likely to induce cell death with PCD-like hallmarks (Figure 9), as was shown in response to H2O2 (Graff van Creveld et al., 2015) and as observed in the death of the ‘oxidized’ subpopulation (Figures 3B and 4F-G, and Figure 8D). Cells that do not cross this threshold are able to recover and acclimate, as in the ‘reduced’ subpopulation (Figures 3B and 4F-G, and Figure 8D). Based on the data from the microfluidics setup, which allowed cell tracking throughout the entire dynamics during exposure to H2O2, we propose that such a ‘death threshold’ could be detected by early chl-roGFP oxidation (Figure 4G). This ‘death threshold’ may be dependent on additional factors, such as ecological context or the specific stressor. The balance between the cellular metabolic state, antioxidant capacity, and the magnitude of the applied stress determines whether a cell will cross the ‘death threshold’, leading to a differential response within the population. Harsher stress conditions will have a stronger effect on the population, leading to more cells crossing the threshold and exhibiting early oxidation and subsequent cell death, as shown with increasing H2O2 doses (Figure 2M) or prolonged exposure to HL (Figure 8B).

The source for the cell-to-cell variability observed in our system is yet to be further explored, but the results provide insights into factors that may drive it. Since clonal populations originating from single-cell isolates maintained the bi-stable chloroplast response, the variability does not result from genetic differences but rather from phenotypic plasticity (Figure 5). The combination of factors such as life history (Graff van Creveld et al., 2016; Murik et al., 2014), cell cycle phase (Papagiannakis et al., 2017; Mathis and Ackermann, 2016), cell age (Levy et al., 2012; Radzinski et al., 2018), metabolic activity (Şimşek and Kim, 2018; Campbell et al., 2018), heterogeneous microenvironment (Stocker, 2012) and biological noise (Balaban, 2011; Raj and van Oudenaarden, 2008) results in a distribution of different metabolic states within the population (Ackermann, 2015; Takhaveev and Heinemann, 2018), which could lead to differential sensitivity to oxidative stress. In yeast for example, redox-based heterogeneity was linked to proliferation and aging (Radzinski et al., 2018). It remains to be investigated whether the emergence of the subpopulations represents heterogeneity that occurs following exposure to stress, or rather a pre-existing variability within the population. Nevertheless, differences in cell-cycle phase distribution between the subpopulations likely represent pre-existing disparities, supporting the latter. However, differences in cell cycle do not completely explain the variability, and are likely to be a contributing factor rather than the source, for example by antioxidants oscillations (Papagiannakis et al., 2017; Burhans and Heintz, 2009; Mathis and Ackermann, 2016; Diaz-Vivancos et al., 2015). In addition, the microfluidics experiments showed no clear differences in chl-roGFP OxD between the subpopulations prior to the treatment (Figure 4—figure supplement 2B), demonstrating that the possible pre-existing heterogeneity is not reflected in the chloroplast EGSH basal level, but rather is based on a different parameter that is yet to be identified.

Unlike in previous studies conducted on heterogeneity, the mechanism that generates variability in our system is light-dependent, as the bi-stable chl-roGFP pattern was abolished when the cells were under darkness and the entire population became more sensitive to oxidative stress (Figure 7A–C). The antioxidant capacity of a diatom cell depends on photosynthesis-generated NADPH, which is also used for GSH recycling. The transition to the dark may have compromised the biosynthesis and recycling of GSH, therefore enhancing sensitivity to oxidative stress, as was shown in another diatom (Volpert et al., 2018). Taken together, we propose that the source for heterogeneity could be variability in the flux of photosynthesis-derived reductive power, regulating the recycling rates of antioxidants.

The light-dependent emergence of distinct subpopulations and the increased sensitivity under dark suggest important implications for environmental scenarios. Fluctuating light conditions are frequent in natural environments due to mixing, shading, the diel cycle, and tide (in coastal and intertidal regions), and may greatly affect diatoms’ susceptibility to diverse abiotic stresses and biotic interactions with pathogens. Phenotypic variability can provide an important strategy to cope with fluctuating environments in microbial populations (Ackermann, 2015). Co-existence of subpopulations with different susceptibilities to specific stressors can be viewed as a ‘bet-hedging’ strategy of the population, enabling at least a portion of the population to survive unpredicted stress events and subsequently leads to a growth benefit at the population level (Schreiber et al., 2016; Sengupta et al., 2017; Levy et al., 2012; Ackermann, 2015). In diatoms, phenotypic variability in cell size, shape and susceptibility to stress conditions was suggested (Armbrust, 2009; Graff van Creveld et al., 2015; Volpert et al., 2018; De Martino et al., 2011), but until now the experimental setups were not designed to study individuality in stress response. Future studies are required to investigate the possible tradeoff involved in maintaining high antioxidant capacity. For example, resilience to oxidative stress may come with a cost in the ability to sense environmental cues with high precision, as high ROS buffering capacity may mask milder ROS cues (Woehle et al., 2017). Redox-based phenotypic variability may provide a rapid and adjustable strategy to cope with unpredicted stress conditions as compared to relying only on genetic diversity.

The novel approaches developed here provide new insights into individuality in marine phytoplankton, and enable studying dynamic processes at the single-cell level in diatoms and possibly other ecologically relevant microorganisms. The mechanisms that underlie differential sensitivity to oxidative stress are yet to be explored. The findings presented here show promising ecological implications for light-dependent heterogeneity, and future studies will unravel its ecological significance in the marine environment.

Materials and methods

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional
information
Strain, strain background (Phaeodactylum tricornutum)Pt1 8.6 (WT)National Center of Marine Algae and Microbiota (NCMA)CCMP2561
Strain, strain background (Phaeodactylum tricornutum)chl-roGFP(Rosenwasser et al., 2014)
Strain, strain background (Phaeodactylum tricornutum)nuc-roGFP(Rosenwasser et al., 2014)
Strain, strain background (Phaeodactylum tricornutum)mit-roGFP(Rosenwasser et al., 2014)
Commercial assay or kitSytox GreenInvitrogen
Software, algorithmImage analysis MATLAB scriptThis paper and on GitHub (Mizrachi, 2018): https://github.com/aviamiz/ITRIA

Culture growth

Request a detailed protocol

P. tricornutum accession Pt1 8.6 (CCMP2561 in the Provasoli-Guillard National Center for Culture of Marine Phytoplankton) was purchased from the National Center of Marine Algae and Microbiota (NCMA, formerly known as CCMP). Cultures were grown in filtered seawater (FSW) supplemented with F/2 media (Guillard and Ryther, 1962) at 18°C with 16:8 hr light:dark cycle and 80 μmol photons m−2 sec−1 light intensity supplied by cool-white LED lights (Edison, New Taipei, Taiwan). Strains expressing roGFP were obtained as described previously (Graff van Creveld et al., 2015; Rosenwasser et al., 2014). Cultures were kept in exponential phase (<2·106 cells·ml−1) for at least 1 week and were sequentially diluted at least three times prior to experiments, experiments were performed in ~0.5–1·106 cells·ml−1. All cultures were counted and diluted a day before the experiment to ensure the same cell concentrations between samples. Cell concentration was measured using Multisizer 4 COULTER COUNTER (Beckman Coulter).

roGFP measurements

Request a detailed protocol

roGFP oxidation was measured over time following the addition of H2O2 or in untreated control using the ratio between two fluorescence channels, i405 and i488, by fluorescence microscopy (described below) and by flow cytometry using BD LSRII analyzer, BD FACSAria II and BD FACSAria III. The roGFP ratio (i405/i488) increases upon oxidation of the probe (Schwarzländer et al., 2008). The oxidation degree of roGFP (OxD) was calculated according to Schwarzländer et al. (Schwarzländer et al., 2008):

OxDroGFP=R-Rredi488oxi488redRox-R+R-Rred

Where R is the roGFP ratio of i405/i488, Rred is the ratio of fully reduced form (15–50 min post treatment with 2 mM Dithiothreitol, DTT), Rox is the ratio of the fully oxidized form (7–30 min post treatment with 200 μM H2O2), and i488ox and i488red are the i488 of the maximum oxidized and maximum reduced forms respectively. For sorting purposes, roGFP ratio was used, as exact OxD cannot be calculated prior to sorting, both parameters give a similar partition between the subpopulations (data not shown). In flow cytometry measurements, i405 was measured using excitation (ex) 407 nm, emission (em) 530/30 nm or 525/25 nm, and i488 was measured using ex 488 nm, em 530/30 nm. Only roGFP positive cells (roGFP+) which had a clear roGFP fluorescence signal separated from WT auto-fluorescence (AF) were included in the analysis. roGFP+ gate was determined either by roGFP relative expression level, which was measured by multiplication of i405 and i488 (Figure 2—figure supplement 2D), or based on i405 and i488 intensity in sorting experiments, as relative roGFP expression could be calculated only post acquisition (see Figure 8—figure supplement 1 for example). Dynamic range of roGFP was calculated by the ratio of Rox/Rred (Table 1). For H2O2 treatments, H2O2 was added at time 0 from a freshly prepared 20 mM stock to P. tricornutum cultures to a final concentration of 5–200 µM. Flow cytometry measurements were done under ambient light (5.5–14 μmol photons m−2 sec−1) and temperature (20–22°C) conditions, unless stated otherwise. For dark treatment samples were covered with aluminum foil. High light of 1,700–2,200 μmol photons m−2 sec−1 was applied using a LED lamp (deviations are due to uneven illumination depending on the distance from the lamp center). To avoid heating by the HL lamp, samples were kept at 16°C using a chilled stage (LCI, Korea).

Table 1
Dynamic range of roGFP targeted to different organelles.

The dynamic range (Rox/Rred) was calculated by dividing the mean roGFP ratio under maximum oxidation conditions (200 µM H2O2) by that under maximum reduction conditions (2 mM DTT) using data from flow cytometry and microfluidics imaging experiments.

https://doi.org/10.7554/eLife.47732.042
P. tricornutum strainFlow cytometry dynamic rangeMicrofluidics imaging
dynamic range
chl-roGFP5.575.42
nuc-roGFP5.94-
mit-roGFP3.50-

Cell death analysis

Request a detailed protocol

For cell death analysis, samples were stained with a final concentration of 1 μM Sytox Green (Invitrogen), incubated in the dark for 30–60 min at RT and analyzed using an Eclipse flow cytometer (ex 488 nm, em 525/50 nm). Unstained samples were used as control to discriminate background signal. In microfluidics experiments, Sytox was dissolved in fresh media (FSW + F/2) to a final concentration of 1–2 µM, and inserted into the system at 21.5–23 hr post treatment, without changing the flow rate (1 µl/min). Fresh stain was continuously flowing through the system for at least 1.5 hr during which cells were imaged for Sytox staining (ex 470/40 nm, em 525/50 nm) as described below. For Sytox staining analysis, images of 30–106 min incubation time were used based on highest staining and best focus.

Sorting of subpopulations and generation of clonal populations

Request a detailed protocol

To measure survival and generate clonal populations originating from different subpopulations, cells expressing chl-roGFP of the ‘oxidized’ and ‘reduced’ subpopulations were sorted according to their roGFP ratio at different times post 80 µM H2O2 treatment or HL using BD FACS AriaII and BD FACS AriaIII. The ‘oxidized’ and ‘reduced’ subpopulations gates were based on visible separation and avoiding cells near intermediate values. Untreated roGFP+ control cells were sorted based on clear separation of roGFP fluorescence from WT AF as described above (the gate upstream of the subpopulations’ gates), regardless of roGFP oxidation. To avoid oxidation due to darkness within the FACS machine, sorting times were minimized and the ‘oxidized’ subpopulation was sorted first followed immediately by sorting of the ‘reduced’ subpopulation. However, in longer sorting sessions as for the Sytox and cell cycle analyses, some cells within the sorted ‘oxidized’ subpopulation may have been oxidized due to the combined effect of H2O2 and exposure to the dark within the FACS. For Sytox analysis of cell death post sorting, 10,000 cells/well were sorted into fresh media (FSW + F/2) in triplicate biological repeats. For single cell survival 1 cell/well was sorted into 96-well plates containing either ‘agar’ (1.5% agarose +FSW/2 + F/2 + antibiotics) or ‘liquid’ (FSW + F/20) fresh media. For the HL CFU assay, 384 cells were sorted onto one-well agar plates. Cells grown in liquid were further diluted and then spotted on agar plates. 3.5–9 weeks post sorting colonies were counted manually (H2O2 experiments) or scanned using Amersham Typhoon 5 Biomolecular Imager and quantified using ImageQuant (HL experiments), assuming each colony originates from a single surviving cell. CFU survival was calculated as CFU number divided by the number of sorted cells. Since survival was highly similar in liquid and in agar the results of these two methods were combined together. Each method was done in biological triplicates per medium type per experiment. In Figure 3B data is shown for two independent experiments for time points 30 and 100 min and one experiment for the 60 min time point. For generation of clonal populations, single cells sorted into liquid medium were used. Clones were exposed to 80 µM and 100 µM H2O2 ~ 3–6 weeks post sorting, and their chl-roGFP oxidation was measured using flow cytometry. A total of 18 ‘control’, 29 ‘oxidized’ and 32 ‘reduced’ clones were examined in two independent experiments.

Microfluidics chip preparation

Request a detailed protocol

Microfluidics chip design was based on Shapiro et al. (Shapiro et al., 2016) and was modified for P. tricornutum cells. Each chip contained 4 channels of 2 cm length X 0.2 cm width X 150 µm height with one circular widening of 0.4 cm diameter, with a total volume of ~12.7 µl per channel (Figure 4—figure supplement 1B–C). The microfluidics chip was etched into a silicone elastomer (SYLGARD 184, Dow Corning) using soft lithography. Silicone elastomers were prepared by mixing the two components in a 10:1 ratio and were poured onto the dust-free wafer, de-aired in a desiccator to eliminate air bubbles, and incubated overnight at 60°C for curing to generate the Polydimethylsiloxane (PDMS) microfluidics chips. Inlet and outlet holes were punched at both ends of each channel using a 1 mm biopsy punch (AcuDerm, FL, USA). The PDMS chip was placed on the clean surface of a new glass microscope 60 × 24 mm cover slip using plasma bonding with a BD-20AC corona treater (Electro-Technic Products) followed by heating of 100°C for >15 min to ensure covalent bonding of the PDMS and the glass.

Microfluidics live imaging

Request a detailed protocol

Light and epifluorescence microscopy imaging was performed using a fully motorized Olympus IX81 microscope (Olympus) equipped with ZDC component for focus drift compensation, 20X air objective (numerical aperture 0.5), and Lumen 200PRO illumination system (Prior Scientific). Images were captured using a Coolsnap HQ2 CCD camera (Photometrics, Tuscon, AZ, USA). The microfluidics chip was mounted on a motorized XY stage (Prior Scientific, MA, USA) with a temperature-controlled inset (LCI, Korea) set to 18°C (Figure 4—figure supplement 1D). The outlet tubes were connected to syringe pumps (New Era Pump Systems, NY, USA) set to withdrawal mode, using negative pressure for flow generation. The inlet tubes were connected to Eppendorf reservoirs, containing the fluid to be inserted into the system. Experiment layout is shown in Figure 4—figure supplement 1A. Chambers were washed with at least 500 µl of pure ethanol, then double-distilled water and then fresh media prior to the introduction of cells. Cells were introduced into the system and settled on the glass bottom.

Flow rate was kept at 1 µl/min for the duration of the experiment, except during cell introduction (100 µl/min), cell settlement (up to 20 µl/min with occasional stops), and treatment introduction (10 µl/min for the initial 10 min for rapid replacement of media). Following settlement and at least 1 hr after cells were introduced to the system, cells were imaged for roGFP measurements (roGFP i405: ex 405/20 nm, em 525/50 nm; roGFP i488: ex 470/40 nm, em 525/50 nm), chlorophyll auto-fluorescence (ex 470/40 nm; em 590 lp) and bright field (BF, without a condenser). Each chip contained four chambers that were imaged sequentially: chl-roGFP control, chl-roGFP 80 µM H2O2 treated, WT 80 µM H2O2 treated and WT control (WT strain can be used to monitor auto-fluorescence changes and leakage during experiments). In each chamber, 5–6 different fields were imaged every 20 min over the course of >24 hr to avoid photo-toxicity. Ambient light was provided during light period using the microscope’s BF illumination without a condenser, light intensity ranging between 34 (at the very edge, outside the imaging region) to 80 (center) μmol photons m−2 sec−1. No images were obtained during the night to avoid disturbance to the diurnal cycle. After imaging the basal state of the cells, treatments of either 80 µM H2O2 dissolved in fresh media (FSW + F/2) or fresh media control were introduced to the system continuously for ~2.5–3 hr, after which they were gradually washed away by fresh media. To quantify cell death, Sytox green was introduced into the system at 21.5–23 hr post treatment (see above) and was imaged using the roGFP i488 channel with a shorter exposure time. The Sytox signal was separated from the roGFP i488 based on its stronger fluorescence and localization to the nucleus. Only a small fraction of cells within the control treatment were Sytox positive (0.0054%), and in addition some control and ‘reduced’ cells proliferated during the experiments (Videos 12), indicating that cells remained viable in this experimental setup.

Image analysis

Request a detailed protocol

Image analysis was performed using a designated MATLAB based script (see overview in Figure 4—figure supplement 4) that is available on GitHub: https://github.com/aviamiz/ITRIA (Mizrachi, 2018; copy archived at https://github.com/elifesciences-publications/ITRIA). Images were imported using bio-formats (Linkert et al., 2010). Then, image registration for XY drift correction was done using the Image Stabilizer plugin (Li, 2008) for FIJI (Fiji Is Just ImageJ) and using MIJI (Prodanov and Tinevez, 2012) to access FIJI from MATLAB. Then images were normalized by bit-depth. Background subtraction was done based on mean value of a user-defined region of interest (ROI) that did not include cells. All fluorescence channels (i405, i488 and chlorophyll) were thresholded by a user-defined value to generate masks of positive expression. The roGFP relative expression level was calculated pixel-by-pixel by multiplication of the i405 and i488, only at pixels that were co-localized in the i405 and i488 masks. Then, roGFP relative expression (i405 * i488) was thresholded in order to include only pixels with high enough signal, based on a user-defined threshold. The roGFP ratio and OxD were calculated pixel-by-pixel as described above, pixels that were not included in the roGFP expression mask were excluded and set to NaN (not a number). For values of maximum oxidation and reduction of roGFP, cells were imaged in the same microfluidics imaging setup following treatments of 200 µM H2O2 and 2 mM DTT respectively (see ‘roGFP calculations’). Cell segmentation was based on i405 (chl-roGFP strain) or chlorophyll (WT strain) masks and fluorescence intensity using watershed transformation. Cells were filtered based on area, major and minor axis length, and eccentricity in order to exclude clumps of cells and doublets. Cell tracking was adapted and modified from a MATLAB code kindly provided by Vicente I. Fernandez and Roman Stocker (Smriga et al., 2016; Shapiro et al., 2014). In short, particles were tracked based on minimizing the distance between particle centroids in adjacent frames within a distance limit. Sytox analysis was based on a user defined threshold and co-localization of the Sytox with the extended cell region within the cell segmentation mask. Images from the same experiment were analyzed using the same values for all thresholds and parameters, except for Sytox analysis in which the threshold was adjusted manually to validate correct assignment of cell-fate and to avoid effects of focus differences. Cells that were not detected in the frame used for Sytox analysis or were not tracked for at least six consecutive frames were excluded from further analysis. The 74% OxD threshold used for early discrimination between cells that subsequently died or survived (Figure 4E–G and Figure 4—figure supplement 2) was based on logistic regression modeling of cell fate at the end of the experiments as a function of chl-roGFP OxD 40 min post 80 µM H2O2 treatment (Figure 4—figure supplement 3). The observed roGFP OxD of more than 100% oxidation in some cells could result from increased auto-fluorescence leakage to the i405 channel at later times post treatment (see Figure 2—figure supplements 56).

Cell cycle analysis

Request a detailed protocol

Cell cycle analysis was based on Huysman et al. (2010) and modified for sorted cells. 30,000 cells of ‘oxidized’ and ‘reduced’ sub-populations were sorted 30 min post 80 µM H2O2 treatment into 260 µl 80% ethanol kept at 5 ˚C, reaching a final concentration of 70% ethanol. Control untreated cells and synchronized cells for reference (20 hr dark, as previously described; Huysman et al., 2010) were sorted based on positive roGFP fluorescence. Cells were then gently mixed and kept at 4 ˚C until further processing. Then, 500 µl of 0.1% bovine serum albumin in phosphate-buffered saline (PBS) was added to improve pellet yield, and samples were centrifuged at 4000 rcf at 4 ˚C for 10 min to discard supernatant. Cells were then washed with PBS, re-suspended and stained with 4',6-diamidino-2-phenylindole (DAPI, Sigma) at a final concentration of 10 ng/ml. Samples were analyzed using BD LSRII analyzer, with ex 355 nm and em 450/50 nm. Synchronized cells were used as a reference to validate the gates for G1 and G2/M phases (data not shown). S phase was not clearly detected in this analysis.

Statistics

Request a detailed protocol

All statistical analyses were done in R. ANOVA or ANCOVA were used for multiple comparisons, and then Dunnett test or Tukey test were performed were applicable. For comparisons of two samples, t-test was used, and paired t-test was used where applicable. Values are represented as mean ± SEM unless specified otherwise. Box-plot was generated using the web tool BoxPlotR http://shiny.chemgrid.org/boxplotr/ (Spitzer et al., 2014) using Tukey whiskers, which extend to data points that are less than 1.5 x Interquartile range away from 1 st/3rd quartile.

Data availability

Request a detailed protocol

All relevant data supporting the findings of the study are available in this article and its Supplementary Information, or from the corresponding author upon request.All data generated or analysed during this study are included in the manuscript and supporting files. Source data files have been provided for Figures 1– 8. MATLAB script used for image analysis is available at GitHub, as referenced in the methods section: https://github.com/aviamiz/ITRIA.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
    The image stabilizer plugin for ImageJ
    1. K Li
    (2008)
    Accessed March 10, 2017.
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
    MIJ: making interoperability between ImageJ and matlab possible
    1. D Prodanov
    2. D Tinevez
    (2012)
    ImageJ User & Developer Conference (IUDC'12).
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76

Decision letter

  1. Dianne K Newman
    Reviewing Editor; California Institute of Technology, United States
  2. Ian T Baldwin
    Senior Editor; Max Planck Institute for Chemical Ecology, Germany

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "Single-cell heterogeneity in the chloroplast redox state mediates acclimation to stress in a marine diatom" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and a Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: E. Virginia Armbrust (Reviewer #1).

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife at this time.

While there is much to like about the technical approach used in this study and the rigor of its execution, whether the conditions used in these studies are representative of natural conditions, and thereby meaningful outside of the laboratory, is unclear. This mitigates enthusiasm for publishing the story in eLife in its current form. Which environmental stressors generate H2O2 intracellularly at the levels used in these studies? How do these stressors stimulate intracellular H2O2 production? The topic of how marine diatoms respond to environmental stress is interesting, but the authors have more work to do to establish that the conditions used in this work are environmentally and physiologically representative. If this can be done, we encourage a revised version of the manuscript to be resubmitted to eLife. We recognize that providing satisfying answers to these questions is non-trivial, this we are rejecting this version of the story to provide the authors sufficient time to address them. If the authors elect to perform more work and resubmit, we will send the manuscript back out to the same reviewers in the future.

Reviewer #1:

The manuscript by Mizrachi et al. is a carefully crafted documentation of phenotypic heterogeneity within the model diatom Phaeodactylum tricornutum. Their target phenotype is responsiveness to exogenously added H2O2, which serves as a surrogate for oxidative stress. The authors rely on their previously developed roGFP sensors for plastid, mitochondrial, and nuclear oxidation state to show that two phenotypic variants in plastid oxidation are found within exposed populations of cells. Those that display greater oxidative stress are strongly correlated with subsequent cell death. This manuscript combines multiple state of the art methods to study individual cell phenotypes including the use of the roGFP tags, flow cytometry sorting, and microfluidics. Phenotypic variability between individual cells is emerging as an important feature that influences how populations of cells respond to environmental change. I have relatively few suggestions for further improvement of the manuscript.

1) Overall, the manuscript is well-written, although it was a little repetitive in sections. This is a manuscript packed with data and figures. Removing any repetitiveness would help with the flow. One example is where the authors state:

"To further explore the physiological factors that may affect the bi-stable distribution, we investigated the effects of different growth phases on oxidation of chl-roGFP in response to H2O2. Since aging has been associated with changes in redox states in several model systems40-42, we hypothesized that culture growth phase may affect the susceptibility of the population to oxidative stress."

This could be simplified to:

"To further explore the physiological factors that may affect the bi-stable distribution, we investigated the effects of different growth phases on oxidation of chl-roGFP in response to H2O2 as aging has been associated with changes in redox states in several model systems40-42."

2) Introduction, second paragraph – typo: "response" should be replaced with "respond".

3) Materials and methods – growth state is clearly an important parameter for phenotypic variability. The authors should provide a bit more information than "Cultures were kept in exponential phase." How were cultures maintained in exponential phase? This is particularly important for interpreting Figure 6A as the initial slopes appears to differ between the different curves.

4) Materials and methods – There are a number of places within the Materials and methods that include results (see for example, in the subsection “roGFP measurements”) and references to figures that are described in the main text. This can be confusing for the reader and I suggest that the authors remove references to figures and results from the Materials and methods section.

5) Figure 1A, control. Could the authors comment on the slight shift to a more reduced state at 50 minutes and then the return to the initial "control" state at 282 minutes? Could this reflect cell cycle transitions?

6) Figure 1. The authors could consider adding a control distribution to panels 1B-D in a manner similar to what was done with the distributions for fully oxidized and reduced. This would help illustrate any shifts away from/return to the control state over the course of the experiment.

7) Figure 2B. The schematic panel is not required as part of main text.

8) Figure 2C legend. I realize this was discussed in the Materials and methods, but in the legend, it is not clear what is meant by "sorted based on positive roGFP fluorescence" given that the sort window is indicated in the 2C panel.

9) The use of colony forming units to detect survival was initially confusing to me. I realized after reading further that this was done to allow the authors to generate clonal cell lines for subsequent analysis. This could be clarified when first referenced in the main text.

10) Figure 3H (the diagram) should be moved to the supplement rather than included as part of the main figure. In fact, this panel is not currently referenced in the main text.

11) The third sentence of the subsection “The “oxidized” subpopulation is enriched with cells at G1 phase”, seems out of place in the Results.

12) In the Introduction, the authors reference earlier work to state "In addition, early oxidation of the mitochondrial GSH pool preceded subsequent cell death at the population level following exposure to H2O2 and diatom-derived infochemicals in the model diatom Phaeodactylum tricornutum." A key conclusion of their study is that oxidation of the plastid, not the mitochondrium, is a good predictor of cell death. The authors should expand on this in the Discussion.

Reviewer #2:

This is an excellent study on the versatility of the chloroplast oxidation state and subsequent cell fate in a model marine diatom exposed to a range of external oxidative stress. The authors have leveraged a number of powerful and elegant techniques to reveal the heterogeneity of the metabolic stress response at the single-cell level. The study is technically impressive, the graphics are clear and well made, and the manuscript is organized and well written overall. I do have some concerns with the paper, particularly in terms of the discussion of ROS and the relevance of the experimental conditions to the natural environment, which are detailed below.

Hydrogen peroxide is typically in the nanomolar range in seawater (Zinser, 2018. The levels of hydrogen peroxide used in experiments (10's of μM) are therefore much higher than would be expected in most natural marine environments. Can the authors explain in more detail how the chosen hydrogen peroxide concentrations are representative of stress conditions that may be encountered in nature?

The discussion is focused on the paradigm of stress = ROS accumulation = PCD. While this is not incorrect, I do feel that it is oversimplified, and that a broader discussion of the results is more appropriate. For example, ROS were not directly measured, so it cannot be said that changes in EGSH were driven conclusively by ROS dynamics. Accordingly, a discussion of other potential factors besides ROS that could affect EGSH in the chloroplast would be helpful. Indeed, the results showed that chl-roGFP oxidation in the presence of H2O2 occurred more readily and more permanently in the dark than in the light (Figure 7, Figure 7—figure supplement 2). This is opposite to what would be expected if ROS were involved, because ROS production in the chloroplast should not increase in the dark (it should increase in the light). So it seems that other mechanisms of chl-roGFP oxidation and subsequent mortality are likely at play. Similarly, there is a focus on PCD being the mechanism of mortality. Is PCD the only way for cells to die in monoculture? Or could another mechanism of mortality be involved in association with chloroplast oxidation state? I believe a fuller discussion than currently given would be helpful.

There is growing recognition that ROS play a variety of beneficial signaling and regulatory roles across a range of cell types. The authors allude to this lightly (e.g., "milder ROS cues", "ROS accumulation.… is used to sense the stress and regulate cell fate" (Figure 9 legend). However, the predominant focus of the discussion is on the deleterious potential of biological ROS production. Since endogenous ROS were not measured and linked to any deleterious impacts on cell vitality in this study, I think a more balanced presentation of the potential functions of biological ROS production, including a more explicit acknowledgement of the possible beneficial roles, is appropriate. For instance, endogenous ROS can act as signaling molecules or oxidative stress inducers based on their concentration (Buetler et al., 2004, News Physiol Sci 19:120-123; Saran 2003, Free Radical Research 37(10): 1045). Although the authors did use exogenous H2O2 concentrations high enough to induce more deleterious effects and the focus of this paper is on oxidative stress, there was not enough mention that lower endogenous ROS levels could be beneficial. This is highly pertinent information to provide in the context of ROS in cell physiology and cellular redox health.

Please describe specifically what kinds of stressors would be expected to induce the chloroplast redox and cell death responses reported in this study. In various instances, the authors refer to "specific" stress conditions or cues (Figure 9 legend), but aside from a general list provided in the Introduction (first paragraph), the exact types of stressors that may be relevant in the context of the findings of this study are not clear.

Reviewer #3:

This manuscript examines how individual cells and populations of a model diatom species respond to stress from hydrogen peroxide exposure. The authors found that a particular dose of stress leads to two subpopulation of cells, which are ultimately indicators of cell fate (survival/death). The authors go on to show that this subpopulation effect is not due to genetic variability in the population, that it is correlated with the cell cycle and physiological state of the population, and that it is dependent on light.

The main contribution of this paper is that it demonstrates phenotypic heterogeneity exists in this particular species response to this level of stress. While the authors do a lot of work to show that these subpopulations are predictive of cell fate, I don't quite get why this is a breakthrough. The physiological response is not really new or surprising to me (i.e. that more heavily damaged cells are more likely to die) and the fact that this only happens at a very particular range of stress exposure (at one concentration, not ~25% lower or higher, and for no more than 100 minutes) implies to me that this subpopulation physiological response occurs rarely in nature.

While I find this paper to be a rigorous examination of this phenomenon (with impressive single cell data!), I do wonder how influential the observation of this phenomenon is and how this paper drives its field forward. I can evaluate this paper based on my own expertise in the field of phenotypic heterogeneity, but have little ability to comment on diatom physiology beyond my knowledge of bacterial physiology. Simply showing that within population heterogeneity exists after a perturbation is no longer a significant contribution from my perspective (see Ackermann, 2015). I would be more inclined to see it as exciting if the authors were able to demonstrate the mechanisms underlying the cause of the heterogeneity or that this heterogeneity has important consequences for the organism or its ecosystem. The authors address this in the Discussion and state that they don't have much evidence for the causes of this heterogeneity or its ecological consequences (Discussion, fifth and last paragraphs). I may be missing something regarding this study’s findings and its relevance to what is known about programmed cell death, diatom physiology, or general marine microbiology. If this is the case I am open to learning more about this, but the paper doesn't frame this point as its central contribution to the field.

Outside of this big picture issue of influence/significance to a general audience, I do find the study to be a strong demonstration of phenotypic heterogeneity which excludes other possible sources of cell-to-cell variability. The methods used and analyses performed are strong. The data is persuasive that phenotypic heterogeneity is happening in response to this particular perturbation.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Light-dependent single-cell heterogeneity in the chloroplast redox state regulates cell fate in a marine diatom" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Ian Baldwin as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The authors have done an excellent job addressing many of the previous concerns, and the manuscript is significantly improved and will be publishable in eLife pending one more round of revision. Upon return, the manuscript will not be sent for re-review; an editorial decision will be made expeditiously.

Essential revisions:

There are a few important yet straightforward editorial changes that should be made prior to publication. The most important of these concern the presentation of the "death threshold" data as well as referencing assertions about environmental H2O2 concentrations and light intensities.

Reviewer #1:

The authors have substantially improved their manuscript with the changes that have been made. The review process has convinced me that this paper does contribute a more broadly interesting story to readers beyond the scope of the phenotypic heterogeneity story. They have now linked the heterogeneity in chloroplast oxidation state from H2O2 experiments to a similar phenomenon in a separate environmentally relevant stressor (high light) with new experiments. The clarity and focus of the manuscript is also much improved, particularly in the Introduction/Discussion sections. The discussion of potential mechanisms underlying the light-dependent heterogeneity is also an important improvement. The extra data in Figure 2A also made it much more convincing.

The authors have addressed my previous comments well. I have one final major issue with the way the authors present the OxD% "death threshold" findings in Figure 3 and in the text (subsection “Early oxidation of chloroplast EGSH predicts cell fate at the single-cell level”, last paragraph; Discussion, fifth paragraph). The authors appear to use the same dataset to define a OxD% threshold for predicting live-dead cells, and then test the accuracy of using that threshold for predicting cell fate. This is obviously circular, and a true test of the accuracy needs to use a different dataset/experiment. If the authors did an independent experiment they need to state it. If not, the authors need to modify these statements from the text because they will mislead readers to think that independent validation experiments have been done. Conceptually, I think they have evidence for a death threshold, but they don't have enough data to estimate the true value of that threshold or if the threshold changes with different conditions/stressors. While the exact single cell experiments weren't done to test this for high light stress, it appears that OxD% thresholds leading to death aren't equivalent across at least high-light and H2O2 (Figures 1, 2, 4, and 8). I would invite the authors to discuss the concept of the "death threshold", but withhold from declaring they have measured a threshold that can accurately predict cell fate (which implies independent validation). I would also suggest they discuss the possibility of the threshold being dependent on the stressor or ecological context.

Reviewer #2:

The authors have significantly revised their manuscript in response to the reviewers' comments. The authors made several changes in response to my major concern regarding the environmental relevance of the study – these include new figures and new data demonstrating that the chloroplast redox response to H2O2 application is similar in magnitude to other environmentally relevant stimuli, such as high light. I appreciate the additional experiments, but I'm still not sure about the choice of the high light level (~2000 μmol/photons/m^2), which was said to be similar to values measured in surface seawater (subsection “High light induces variability in chloroplast ROS accumulation and subsequent cell survival”, first paragraph), yet this is not general knowledge and I couldn't find any support for this statement in the manuscript. I also accept the argument that H2O2 concentrations could be higher in microenvironments compared to bulk seawater, but without additional references, I am skeptical that these microenvironments could reach tens of μm H2O2. Overall, I am partially satisfied that these changes resolve my original concerns. For example, it seems pertinent to discuss what the typical H2O2 concentrations are within cells under "normal" and oxidative stress conditions, and how these concentrations compared to the levels of H2O2 used in the experiments (the authors allude to this in their response, but I couldn't find anything on this subject in the manuscript). I accept that not all of the environmental relevance question needs to be addressed in this single manuscript. For example, I agree with the authors that the ecological significance in the marine environment should be addressed in future studies (last line of the Discussion), but I still think there is a bit more that could be easily done to address the environmental relevance of the present work. I would support acceptance of this version of the manuscript, provided that the authors are encouraged to consider the above comments during the proofs stage.

https://doi.org/10.7554/eLife.47732.045

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #1:

[…] 1) Overall, the manuscript is well-written, although it was a little repetitive in sections. This is a manuscript packed with data and figures. Removing any repetitiveness would help with the flow. One example is where the authors state:

"To further explore the physiological factors that may affect the bi-stable distribution, we investigated the effects of different growth phases on oxidation of chl-roGFP in response to H2O2. Since aging has been associated with changes in redox states in several model systems40-42, we hypothesized that culture growth phase may affect the susceptibility of the population to oxidative stress."

This could be simplified to:

"To further explore the physiological factors that may affect the bi-stable distribution, we investigated the effects of different growth phases on oxidation of chl-roGFP in response to H2O2 as aging has been associated with changes in redox states in several model systems40-42."

We thank the reviewer for acknowledging the importance of the work and for providing helpful suggestions. After adding the high light section in the Results, we decided to remove the section about growth phase for better flow and readability, and it will be included instead in future research. Therefore, these repetitive parts were removed. In addition, we modified the Discussion and additional parts in the paper to further improve the flow.

2) Introduction, second paragraph – typo: "response" should be replaced with "respond".

Changed accordingly.

3) Materials and methods – growth state is clearly an important parameter for phenotypic variability. The authors should provide a bit more information than "Cultures were kept in exponential phase." How were cultures maintained in exponential phase? This is particularly important for interpreting Figure 6A as the initial slopes appears to differ between the different curves.

Additional information regarding culture growth and maintenance is now provided in the Materials and methods, see subsection “Culture growth”.

4) Materials and methods – There are a number of places within the Materials and methods that include results (see for example, in the subsection “roGFP measurements”) and references to figures that are described in the main text. This can be confusing for the reader and I suggest that the authors remove references to figures and results from the Materials and methods section.

We thank the reviewer, we have removed the results from the revised version accordingly. However, we kept in the Materials and methods section some references to specific figures that are essential for the understanding of the experimental setup.

5) Figure 1A, control. Could the authors comment on the slight shift to a more reduced state at 50 minutes and then the return to the initial "control" state at 282 minutes? Could this reflect cell cycle transitions?

We thank the reviewer for pointing out this interesting observation, however it is beyond the scope of the current work. This observation requires further investigation and validation, and it may be related to cell cycle transitions, diel cycle, or other technical or biological sources. As the manuscript is already packed with data, we feel that adding additional information regarding this observation will divert the reader from the main findings and will not benefit the paper.

6) Figure 1. The authors could consider adding a control distribution to panels 1b-d in a manner similar to what was done with the distributions for fully oxidized and reduced. This would help illustrate any shifts away from/return to the control state over the course of the experiment.

We respectfully disagree, as the control distribution over the entire time course of the experiment is shown in Figure 1A, E and I, and therefore we believe that adding it within the other figures will be redundant. Since the x axis is kept this comparison should be visually clear. In addition, the mean oxidation over time of each subpopulation as well as control cells is shown in Figure 2—figure supplement 3G, further demonstrating shifts in oxidation state.

7) Figure 2B. The schematic panel is not required as part of main text.

We thank the reviewer for the comment, and we therefore rearranged the figure layout: what was previously Figure 2B was moved to the supplementary information and an illustration for the sorting procedure was added (new Figure 2—figure supplement 3), and Figure 2C was removed.

8) Figure 2C legend. I realize this was discussed in the Materials and methods, but in the legend, it is not clear what is meant by "sorted based on positive roGFP fluorescence" given that the sort window is indicated in the 2C panel.

The meaning in “positive roGFP fluorescence” is a clear fluorescence signal of roGFP in both i405 and i488 channels, which is well separated from wild type cells, which possess only auto-fluorescence signal (i.e. cells that don’t express roGFP). This gate is not affected by the degree of oxidation of roGFP. To clarify it, we changed the description in Figure 3B legend as follows: “Control – untreated chl-roGFP positive cells that were sorted regardless of degree of oxidation”. Figure 3C was removed, and instead an illustration for the sorting procedure was added as part of the schematic representation (Figure 3—figure supplement 1). In addition, we further clarified the sorting strategy in the Materials and methods section (see subsections “roGFP measurements” and “Sorting of subpopulations and generation of clonal populations”). Furthermore, in the newly added Figure 8—figure supplement 1 of the high light section the gate for positive roGFP expression is shown.

9) The use of colony forming units to detect survival was initially confusing to me. I realized after reading further that this was done to allow the authors to generate clonal cell lines for subsequent analysis. This could be clarified when first referenced in the main text.

In order to clarify the use of the CFU assay, we added the following sentence: “The CFU assay provides a direct link between chl-roGFP oxidation and the ability of individual cells to proliferate, and in addition enables to generate clonal populations which were also used for downstream analyses (Figure 3—figure supplement 1).”

The use of CFU to measure survival was conducted in order to measure directly the survival of the same cells that their oxidation was measured. In other methods, such as the Sytox staining shown in Figure 3A, there is a correlation between the fraction of cells that were oxidized and the fraction of cells that were dead but these are not the same cells. In addition, Sytox staining will only detect cells that are dead or dying at the time of measurement, and not cells that were already completely disintegrated or cells that will die at a later stage. Furthermore, the Sytox measurement of sorted subpopulations (shown in Figure 3—figure supplement 2) requires a larger quantity of sorted cells for the analysis and prolongs sorting times. However, the CFU assay measures survival in a different manner, by the ability of the cell to proliferate, which may be even more biologically relevant. In that assay, the “time of death” is not relevant, and also cells with permanent growth arrest will appear as non-viable, providing a direct method to investigate the subsequent fate of individual cells. In addition, the same assay was also used for generating clonal populations.

10) Figure 3H (the diagram) should be moved to the supplement rather than included as part of the main figure. In fact, this panel is not currently referenced in the main text.

As suggested by the reviewer the figure was moved to the supplementary accordingly (new Figure 4—figure supplement 1B).

11) The third sentence of the subsection “The “oxidized” subpopulation is enriched with cells at G1 phase”, seems out of place in the Results.

This sentence was removed, and the subject is addressed in the Discussion (sixth paragraph).

12) In the Introduction, the authors reference earlier work to state "In addition, early oxidation of the mitochondrial GSH pool preceded subsequent cell death at the population level following exposure to H2O2 and diatom-derived infochemicals in the model diatom Phaeodactylum tricornutum." A key conclusion of their study is that oxidation of the plastid, not the mitochondrium, is a good predictor of cell death. The authors should expand on this in the Discussion.

We now elaborate on the subject in the Discussion (fourth paragraph).

Reviewer #2:

[…] I do have some concerns with the paper, particularly in terms of the discussion of ROS and the relevance of the experimental conditions to the natural environment, which are detailed below.

Hydrogen peroxide is typically in the nanomolar range in seawater (Zinser, 2018). The levels of hydrogen peroxide used in experiments (10's of μM) are therefore much higher than would be expected in most natural marine environments. Can the authors explain in more detail how the chosen hydrogen peroxide concentrations are representative of stress conditions that may be encountered in nature?

We thank the reviewer for raising this important point. We added an explanation regarding the H2O2 concentrations measured in the environment accordingly (Introduction, third paragraph). To clarify the reason for choosing the specific H2O2 concentrations used in the study we added Figure 1 and the following explanation: “Next, we examined the response of cells to oxidative stress by treatment with H2O2 at concentrations that led to oxidation patterns similar to other environmental stressors (Figure 1) and that also led to death of part of the population16”. Importantly, the H2O2 concentrations applied in our study (10’s µM) are three orders of magnitude lower (at least) compared to concentrations that are commonly used in numerous papers in plants, green algae, yeast and other protists (10’s mM).

We agree with the reviewer that H2O2 concentrations measured in averaged bulk seawater are much lower than the concentrations used in this study. However, these measurements are not derived necessary from high dense algal blooms and do not accurately depict the actual conditions that a single-celled phytoplankton will actually experience. The marine environment is not homogenous, but rather patchy in time and space8–10. The cell densities experienced by microorganisms in their microenvironment can be significantly higher due to local high densities in marine snow, biofilm, aggregates, and at the peak of coastal blooms, when cell densities can reach up to 105 cells per mL of seawater8,9. Intracellular ROS concretions produced by the cells themselves or by bystanders cells could lead to higher intracellular concentrations than measured in bulk seawater in the field, and can generate a gradient of a few orders of magnitude across the phycosphere8–10. As was shown in previous studies, diverse environmental stressors lead to similar oxidation patterns (assessed in vivo by roGFP) as observed in response to H2O2 (see the added new Figure 1)1,2, or may lead to H2O2 production3–7. More specifically, high light, nitrogen limitation, and cyanogen bromide (a toxic infochemical produced by some diatoms, applied in environmentally relevant concentrations) were shown to cause chl-roGFP oxidation in P. tricornutum, similarly to the response to H2O2 in concentrations used in this study. To further address the issue of environmental relevance we examined the effects of high light, as shown in the newly added section in the Results (subsection “High light induces variability in chloroplast ROS accumulation and subsequent cell survival”). High light can lead to H2O2 accumulation in the chloroplast through photoreduction of O2 to superoxide, which in turn is rapidly converted to H2O2 by superoxide dismutase (SOD)4–6. Indeed, following exposure to high light a distinct oxidized subpopulation emerged (Figure 7A-B, Figure 8—figure supplement 1A-F), similarly to the response to external addition of H2O2. The survival of the oxidized subpopulation was initially high, but then gradually decreased over time of exposure to high light (Figure 7D). These findings demonstrate a similar pre-commitment phase to that observed in response to H2O2 addition, during which cells were still able to recover upon change of conditions (i.e. by sorting into fresh media, Figure 7D). Taken together with the increased sensitivity to oxidized stress observed under dark conditions, these results demonstrate that light regime plays an important role in determining cell fate and sensitivity to additional stressors in marine diatoms. Light fluctuations are frequent in marine environments due to mixing, shading, diel cycle and tide (in coastal and intertidal regions), and therefore these results are highly relevant to environmental scenarios.

The discussion is focused on the paradigm of stress = ROS accumulation = PCD. While this is not incorrect, I do feel that it is oversimplified, and that a broader discussion of the results is more appropriate. For example, ROS were not directly measured, so it cannot be said that changes in EGSH were driven conclusively by ROS dynamics. Accordingly, a discussion of other potential factors besides ROS that could affect EGSH in the chloroplast would be helpful. Indeed, the results showed that chl-roGFP oxidation in the presence of H2O2 occurred more readily and more permanently in the dark than in the light (Figure 7, Figure 7—figure supplement 2). This is opposite to what would be expected if ROS were involved, because ROS production in the chloroplast should not increase in the dark (it should increase in the light). So it seems that other mechanisms of chl-roGFP oxidation and subsequent mortality are likely at play.

Indeed, the oxidation observed in chl-roGFP in the dark is surprising, and the sources for ROS production in the dark are yet to be further investigated. One possible explanation could be that the recycling of antioxidants is impaired in the dark, as light-generated NADPH is used for the recycling and biosynthesis of GSH and other antioxidants. The abrupt transition to the dark during daytime may have caused a perturbation in the redox balance through the inability to recycle GSSG to GSH, leading to oxidation of chl-roGFP. This could also explain the increase in sensitivity when cells are under dark, which was also observed in another diatom species11 (see Discussion, seventh paragraph).

As detailed in previous works, the redox sensor roGFP is coupled to the redox state of GSH through the activity of glutaredoxins (GRX), for which roGFP serves as an artificial target that is mimicking native proteins with similar redox potential (for a comprehensive review on roGFP sensors see

Meyer and Dick, 2010)12,13. Perturbations in the balance between GSHGSSG are due to changes in the redox state in their subcellular microenvironment, which are mainly driven by ROS dynamics. Therefore, roGFP measures the biochemical response of the cell to ROS through its coupling to the EGSH balance. Changes in the size of the GSH and GSSG pools without changing the balance between them (i.e. without changes in EGSH) should not affect roGFP oxidation. Therefore, we are not sure to which additional sources for roGFP oxidation the reviewer is referring. The great advantage of using the roGFP as an in vivo sensor over any ROS chemical fluorescent stains is that it can be reversibly oxidized and reduced, and thus is sensitive to the actual fluctuations of the redox state within a given microenvironment. An additional big difference is that the roGFP is a genetically encoded probe, and thus can be targeted to measure ROS in a specific organelle, in contrast to a chemical stain that accumulates by diffusion in a non-specific manner, and may also accumulate in different cells to a different degree due to permeability differences. The roGFP is a ratiometric probe, and its oxidation measurements are therefore less affected by bleaching or variability in expression levels.

Regarding the role of ROS in simply mediating PCD, we respectfully disagree with the reviewer. The whole ideology of our body of works in the past 5 years is around how specificity in redox singling can regulate not only PCD, but also intricate signaling cascades that lead to cell survival under diverse stress conditions1–3,14,15. Over the years, we revealed how redox changes due to environmental stresses can generate highly specific redox patterns in specific organelles and to activate a redox-sensitive proteome network that regulates metabolic fluxes and cell fate decisions. In our current manuscript, we take a step further to show that under the same given stress of high light or H2O2, individual cells within the same isogenic population can display distinct cell fates, of PCD and also survival. Furthermore, cells of the “reduced” subpopulation are shown to exhibit mild chl-roGFP oxidation in response to H2O2 treatment (Figures1C, 3F and Figure 2—figure supplement 3G), and are also shown to divide and survive under exposure to oxidative stress (Figures 2B, 3F, and Figure 4—video 1). Please see additional information below.

Similarly, there is a focus on PCD being the mechanism of mortality. Is PCD the only way for cells to die in monoculture? Or could another mechanism of mortality be involved in association with chloroplast oxidation state? I believe a fuller discussion than currently given would be helpful.

To the best of our knowledge, the main mode of death in response to H2O2 in concentrations that were used in this manuscript has clear PCD markers. Our previously published results (as mentioned in the third paragraph of the Introduction and in the fifth paragraph of the Discussion) demonstrated that H2O2 led to the induction of cell death with PCD characteristics in P. tricornutum, consisting of a cascade of events of DNA fragmentation (starting in at least part of the population within 4 h of treatment, measured by TUNEL assay), followed by phosphatidylserine externalization (starting 8 h post treatment, assayed by Annexin V), and eventually irreversible membrane permeabilization (positive Sytox staining 24 h post treatment)1. Lastly, we have recently characterized the first PCD molecular component from P. tricornutum, metacaspase 5, which is regulated by H2O2 and infochemicals produced by diatoms16. Nevertheless, we agree with the reviewer that there are other modes of death in monoculture, such as direct cell lysis, necrosis and autophagic cell death etc. We clarified the use of the PCD term for describing the mode of death in our system in the Introduction and the Discussion (see the aforementioned paragraphs), however we feel that further mechanistic discussion of alternative modes of death in response to other stressors is out of the scope of this paper.

There is growing recognition that ROS play a variety of beneficial signaling and regulatory roles across a range of cell types. The authors allude to this lightly (e.g., "milder ROS cues", "ROS accumulation.… is used to sense the stress and regulate cell fate" (Figure 8 legend). However, the predominant focus of the discussion is on the deleterious potential of biological ROS production. Since endogenous ROS were not measured and linked to any deleterious impacts on cell vitality in this study, I think a more balanced presentation of the potential functions of biological ROS production, including a more explicit acknowledgement of the possible beneficial roles, is appropriate. For instance, endogenous ROS can act as signaling molecules or oxidative stress inducers based on their concentration (Buetler et al. 2004, News Physiol Sci 19:120-123; Saran 2003, Free Radical Research 37(10): 1045). Although the authors did use exogenous H2O2 concentrations high enough to induce more deleterious effects and the focus of this paper is on oxidative stress, there was not enough mention that lower endogenous ROS levels could be beneficial. This is highly pertinent information to provide in the context of ROS in cell physiology and cellular redox health.

In this work and in our previous publications1–3,14,15 we strongly emphasize the role of ROS as signaling molecules used in sensing stress cues and mediating the response to these cues. In our work, we propose that redoxbased responses to stress serve as a mechanism for a rapid and reversible response to variable conditions, and act as a powerful surveillance system to accurately assess the chemo-physical gradients in the cell’s microenvironment, enabling discrimination between multiple environmental cues (for example see the Introduction, Discussion, second and third paragraphs, and the added Figure 1). In addition, in the Introduction we specifically emphasize the role of H2O2 in signaling (third paragraph). The results presented in this paper further support a signaling role for ROS, which mediate cell fate determination to either survival or death. For example, we show that also in the “reduced” subpopulation chl-roGFP underwent rapid oxidation in response to H2O2 treatment, however to much lower values than the “oxidized” subpopulation (30-65% OxD compared to ~100% OxD), and afterwards recovered from the stress (Figures 2B-D, 4F, Figure 2—figure supplement 3G, and Figure 4—video 1). Moreover, cells of the “reduced” subpopulation were also able to proliferate despite the stress conditions (Figure 4—video 1). In addition, the EGSH oxidation in the chloroplast is proposed as a sensing mechanism, as chl-roGFP oxidation preceded the cell fate determination in the “oxidized” subpopulation (Discussion, third paragraph). Furthermore, we demonstrate that upon transition of P. tricornutum cells to the dark, chl-roGFP oxidized immediately and then gradually reduced over time, and this response was not accompanied with increased death, and was specific to the chloroplast as it was not observed in the nucleus (Figure 7A and E, Figure 7—figure supplement 2, subsection “The bimodal chloroplast redox response is light dependent”, first paragraph). To further clarify the subject we modified the Discussion section.

Please describe specifically what kinds of stressors would be expected to induce the chloroplast redox and cell death responses reported in this study. In various instances, the authors refer to "specific" stress conditions or cues (Figure 8 legend), but aside from a general list provided in the Introduction (first paragraph), the exact types of stressors that may be relevant in the context of the findings of this study are not clear.

As was shown in our previous studies, diverse environmental stressors lead to similar oxidation patterns as observed in response to H2O2 (see the newly added Figure 1)1,2, or may lead to H2O2 production3–7. More specifically, high light, nitrogen limitation, cyanogen bromide (a toxic infochemical produced by some diatoms) were shown to cause roGFP oxidation in the chloroplast in P. tricornutum. To further address this important issue of environmental relevance we examined the effects of high light, an important stressor for photosynthetic microbes in the aquatic ecosystems, as shown in Figures 8, Figure 8—figure supplement 1 and Figure 8—figure supplement 2, and in the adequate section in the Results (subsection “High light induces variability in chloroplast ROS accumulation and subsequent cell survival”). High light can lead to H2O2 accumulation in the chloroplast through photoreduction of O2 to superoxide, which in turn is rapidly converted to H2O2 by superoxide dismutase (SOD)4–6. Indeed, following exposure to high light a distinct oxidized subpopulation emerged, similarly to the response to external addition of H2O2 (Figure 7A-B and Figure 7—figure supplement 2A-F). The survival of the oxidized subpopulation gradually decreased over time of exposure to high light, exhibiting a similar pre-commitment phase to that observed in response to H2O2 addition, during which cells were still able to recover upon change of conditions (i.e. by sorting into fresh media, Figure 7D). Taken together with the increased sensitivity to oxidized stress observed under dark conditions, these results demonstrate that light regime plays an important role in determining cell fate and sensitivity to additional stressors in marine diatoms. Light fluctuations are frequent in marine environments due to mixing, shading, diel cycle and tidal oscillations (in coastal and intertidal regions), and therefore these results are highly relevant to environmental scenarios. We modified the Discussion accordingly and added in Figure 8 high light and infochemicals as possible stressors.

Reviewer #3:

[…] While I find this paper to be a rigorous examination of this phenomenon (with impressive single cell data!), I do wonder how influential the observation of this phenomenon is and how this paper drives its field forward. I can evaluate this paper based on my own expertise in the field of phenotypic heterogeneity, but have little ability to comment on diatom physiology beyond my knowledge of bacterial physiology. Simply showing that within population heterogeneity exists after a perturbation is no longer a significant contribution from my perspective (see Ackermann, 2015). I would be more inclined to see it as exciting if the authors were able to demonstrate the mechanisms underlying the cause of the heterogeneity or that this heterogeneity has important consequences for the organism or its ecosystem. The authors address this in the Discussion and state that they don't have much evidence for the causes of this heterogeneity or its ecological consequences (Discussion, fifth and last paragraphs). I may be missing something regarding this study’s findings and its relevance to what is known about programmed cell death, diatom physiology, or general marine microbiology. If this is the case I am open to learning more about this, but the paper doesn't frame this point as its central contribution to the field.

We thank the reviewer for acknowledging the rigorosity of the work, but we respectfully disagree with the reviewer summarizing the paper as simply another paper describing phenotypic heterogeneity. We feel that the main problem is the knowledge gap between the well-established bacterial community, which over the last 2 decades has demonstrated the huge importance of phenotypic heterogeneity among bacterial model systems, and the very young field of cell biology and genetics of marine phytoplankton. The subject of phenotypic variability has been studied extensively in bacterial, yeast and mammalian model systems17–19, mainly in model systems with little ecological relevance to the marine environment. In contrast, very little is known or reported for marine microbes in general, and specifically in marine phytoplankton. Especially, there was never specific effort to track single cell response in high resolution in order to better understand how unicellular algae sense, respond and communicate stress and chemical cues. We believe that our paper is a first step towards an important discussion about the relevance of phenotypic heterogeneity in aquatic ecosystems, and a fascinating research question to specifically explore in the context of phytoplankton bloom dynamics. Importantly, to date little is known about the genetic programs and signal transduction pathways that mediate perception of abiotic and biotic stresses that phytoplankton are typically being subjected to during natural bloom succession. We propose that our approach will enable the discovery of essential genes that mediate acclimation strategies, and will include genes involved in activation of PCD and in cellular defense and recovery. Current understanding of phytoplankton responses to environmental stressors is based primarily on bulk analyses, neglecting cell-to-cell heterogeneity. Here, we established a novel system for studying redox-based heterogeneity in diatoms, and it can be used in the future to study other ecologically relevant microorganisms. We believe that our manuscripts greatly contributes to the field, as shown by the major important novel findings:

- This is the first study to track redox-based heterogeneity in the chloroplast in response to oxidative stress and high light, leading to differential survival within isogenic populations.

- This is the first report to assign heterogeneity to light dependent metabolic processes.

- We demonstrate that chloroplast EGSH is involved in sensing stress cues and mediating cell fate determination in algae. Specifically, we show a link between ROS accumulation in the chloroplast and subsequent cell death. The role of the chloroplast in mediating activation and regulation of PCD is still an open question in photosynthetic organisms including plants20,21, in contrast to the well-established mammalian literature on the pivotal role of the mitochondria in executing the PCD signaling cascade. Our results shed light on the role of light-dependent reactions in the chloroplast on regulating cell fate in diatoms.

- We provide insights into the time-line of events in the response to oxidative stress and the PCD cascade in diatoms. Chloroplast EGSH oxidation changes within minutes and precedes cell fate determination, exposing the “pre-commitment” phase that ranges 30-60 min in most cells.

- In order to add an important ecological context to the main findings of the manuscript, regarding the bi-stable response to oxidative stress in marine diatoms, we examined the response of the cells to high light. We demonstrate that high light, a key environmental stress factor in the ocean, controls diatoms’ stress physiology. Specifically, we show that high light can induce the emergence of distinct subpopulations, differing in ROS accumulation in the chloroplast and in their subsequent survival. In addition, transition to the dark greatly enhanced the sensitivity to oxidative stress. Together, these results may have great ecological implications, as phytoplankton experience frequent light intensity fluctuations due to mixing, the diel cycle, shading, and tide (in intertidal and coastal regions). Therefore, the time of day and position within the water column may greatly affect diatoms’ biotic interactions with other organisms in the marine environment and their ability to cope with additional stressors as nutrient limitation, temperature gradients and more.

- It is important to emphasize that although there are numerous studies examining heterogeneity, many rely on gene expression as their readout. In the current work we track an important metabolic readout of redox changes in vivo in an organelle specific manner, similar to recent efforts to probe NADPH, glucose etc. Our unique readout is strongly associated with ROS metabolism and thus gives an important view on the cellular stress physiology at the single cell and organelle levels.

As was shown in previous studies, diverse environmental stressors lead to similar oxidation patterns as observed in response to H2O2 (see the added Figure 1)1,2, or may lead to H2O2 production3–7. More specifically, high light, nitrogen limitation and cyanogen bromide (a toxic infochemical produced by some diatoms) were shown to cause roGFP oxidation in the chloroplast in P. tricornutum (Figure 1). To further address the issue of environmental relevance we examined the effects of high light, as shown in the added section in the Results (subsection “High light induces variability in chloroplast ROS accumulation and subsequent cell survival”). High light can lead to H2O2 accumulation in the chloroplast through photoreduction of O2 to superoxide, which in turn is rapidly converted to H2O2 by superoxide dismutase (SOD)4–6. Indeed, following exposure to high light a distinct “oxidized” subpopulation emerged, similarly to the response to external addition of H2O2 (Figure 7A-B). The survival of the “oxidized” subpopulation gradually decreased over time of exposure to high light, exhibiting a similar pre-commitment phase to that observed in response to H2O2 addition, during which cells were still able to recover upon change of conditions (i.e. by sorting into fresh media, Figure 7D). As explained above, these results are highly relevant to environmental scenarios.

We fully agree with the reviewer that the next critical step must be to unravel the underlying molecular mechanism that will regulate cell fate decisions in marine diatoms. We are currently investigating this challenging subject using the exciting system developed here as the fundamental experimental setup. We therefore aim to develop single-cell RNAseq protocols adapted to marine phytoplankton, which will enable in the near future to expose the transctriptome makeup that defines the different subpopulations with contrasting cell fates. Nevertheless, our data is still very preliminary and beyond the scope of the current manuscript, which is already packed with data.

Outside of this big picture issue of influence/significance to a general audience, I do find the study to be a strong demonstration of phenotypic heterogeneity which excludes other possible sources of cell-to-cell variability. The methods used and analyses performed are strong. The data is persuasive that phenotypic heterogeneity is happening in response to this particular perturbation.

We greatly appreciate the reviewer encouraging feedback on the quality of the work.

1) Graff van Creveld, S., Rosenwasser, S., Schatz, D., Koren, I. and Vardi, A. Early perturbation in mitochondria redox homeostasis in response to environmental stress predicts cell fate in diatoms. ISME J. 9, 385–395 (2015).

2) Rosenwasser, S. et al. Mapping the diatom redox-sensitive proteome provides insight into response to nitrogen stress in the marine environment.

Proc. Natl. Acad. Sci. U. S. A. 111, 2740–5 (2014).

3) Sheyn, U., Rosenwasser, S., Ben-Dor, S., Porat, Z. and Vardi, A. Modulation of host ROS metabolism is essential for viral infection of a bloom-forming coccolithophore in the ocean. ISME J. 10, 1–13 (2016).

4) Asada, K. Production and scavenging of reactive oxygen species in chloroplasts and their functions. Plant Physiol. 141, 391–6 (2006).

5) Waring, J., Klenell, M., Bechtold, U., Underwood, G. J. C. and Baker, N. R.

Light-induced responses of oxygen photoreduction, reactive oxygen species production and scavenging in two diatom species. J. Phycol. 46, 1206–1217 (2010).

6) Exposito-Rodriguez, M., Laissue, P. P., Yvon-Durocher, G., Smirnoff, N. and Mullineaux, P. M. Photosynthesis-dependent H2O2 transfer from chloroplasts to nuclei provides a high-light signalling mechanism. Nat. Commun. 8, 49 (2017).

7) Vardi, A. et al. Programmed cell death of the dinoflagellate Peridinium gatunense is mediated by CO2 limitation and oxidative stress. Curr. Biol. 9, 1061–1064 (1999).

8) Seymour, J. R., Amin, S. A., Raina, J.-B. and Stocker, R. Zooming in on the phycosphere: the ecological interface for phytoplankton–bacteria relationships. Nat. Microbiol. 2, 17065 (2017).

9) Stocker, R. Marine Microbes See a Sea of Gradients. Science 338, 628– 633 (2012).

10) Chrachri, A., Hopkinson, B. M., Flynn, K., Brownlee, C. and Wheeler, G. L. Dynamic changes in carbonate chemistry in the microenvironment around single marine phytoplankton cells. Nat. Commun. 9, 74 (2018).

11) Volpert, A., Graff van Creveld, S., Rosenwasser, S. and Vardi, A. Diurnal fluctuations in chloroplast GSH redox state regulate susceptibility to oxidative stress and cell fate in a bloom-forming diatom. J. Phycol. 341, 329–341 (2018).

12) Meyer, A. J. et al. Redox-sensitive GFP in Arabidopsis thaliana is a quantitative biosensor for the redox potential of the cellular glutathione redox buffer. Plant J. 52, 973–986 (2007).

13) Meyer, A. J. and Dick, T. P. Fluorescent protein-based redox probes. Antioxid Redox Signal 13, 621–650 (2010).

14) Creveld, S. G. Van, Ben-dor, S., Mizrachi, A., Alcolombri, U. and Hopes, A. A redox-regulated type III metacaspase controls cell death in a marine diatom. (2018). doi:https://doi.org/10.1101/444109

15) Graff van Creveld, S., Rosenwasser, S., Levin, Y. and Vardi, A. Chronic Iron Limitation Confers Transient Resistance to Oxidative Stress in Marine Diatoms. Plant Physiol. 172, 968–979 (2016).

16) Creveld, S. G. Van, Ben-dor, S., Mizrachi, A., Alcolombri, U. and Hopes, A. A redox-regulated type III metacaspase controls cell death in a marine diatom. (2018). doi:https://doi.org/10.1101/444109

17) Balaban, N. Q. Persistence: Mechanisms for triggering and enhancing phenotypic variability. Curr. Opin. Genet. Dev. 21, 768–775 (2011).

18) Raj, A. and van Oudenaarden, A. Nature, Nurture, or Chance: Stochastic Gene Expression and Its Consequences. Cell 135, 216–226 (2008).

19) Ackermann, M. A functional perspective on phenotypic heterogeneity in microorganisms. Nat. Rev. Microbiol. 13, 497–508 (2015).

20) Van Aken, O. and Van Breusegem, F. Licensed to Kill: Mitochondria, Chloroplasts, and Cell Death. Trends Plant Sci. 20, 754–766 (2015).

21) Lam, E., Kato, N. and Lawton, M. Programmed cell death, mitochondria and the plant hypersensitive response. Nature 411, 848–853 (2001).

[Editors' note: the author responses to the re-review follow.]

Essential revisions:

There are a few important yet straightforward editorial changes that should be made prior to publication. The most important of these concern the presentation of the "death threshold" data as well as referencing assertions about environmental H2O2 concentrations and light intensities. Reviewer #1: […] The authors have addressed my previous comments well. I have one final major issue with the way the authors present the OxD% "death threshold" findings in Figure 3 and in the text (subsection “Early oxidation of chloroplast EGSH predicts cell fate at the single-cell level”, last paragraph; Discussion, fifth paragraph). The authors appear to use the same dataset to define a OxD% threshold for predicting live-dead cells, and then test the accuracy of using that threshold for predicting cell fate. This is obviously circular, and a true test of the accuracy needs to use a different dataset/experiment. If the authors did an independent experiment they need to state it. If not, the authors need to modify these statements from the text because they will mislead readers to think that independent validation experiments have been done. Conceptually, I think they have evidence for a death threshold, but they don't have enough data to estimate the true value of that threshold or if the threshold changes with different conditions/stressors. While the exact single cell experiments weren't done to test this for high light stress, it appears that OxD% thresholds leading to death aren't equivalent across at least high-light and H2O2 (Figures 1, Figure 2, 4, and 8). I would invite the authors to discuss the concept of the "death threshold", but withhold from declaring they have measured a threshold that can accurately predict cell fate (which implies independent validation). I would also suggest they discuss the possibility of the threshold being dependent on the stressor or ecological context.

We thank the reviewer for acknowledging the importance of the work and for appreciating the substantial improvement of the paper. We also thank the reviewer for the constructive feedback, and agree with his comments regarding the “death threshold”, as the same dataset was used for generating the threshold and for testing it. According to the reviewer suggestions, we now discuss the concept of a possible “death threshold”, which may be exposed by the chl-roGFP level of oxidation, and discuss the possibility that additional factors may influence this threshold (see subsection “Early oxidation of chloroplast EGSH is linked to cell fate determination at the single-cell level”, last paragraph and Discussion, fifth paragraph). Accordingly, we removed any declarations regarding actual measurements of a threshold and its application in cell fate predictions.

Reviewer #2:

The authors have significantly revised their manuscript in response to the reviewers' comments. The authors made several changes in response to my major concern regarding the environmental relevance of the study – these include new figures and new data demonstrating that the chloroplast redox response to H2O2 application is similar in magnitude to other environmentally relevant stimuli, such as high light. I appreciate the additional experiments, but I'm still not sure about the choice of the high light level (~2000 μmol/photons/m^2), which was said to be similar to values measured in surface seawater (subsection “High light induces variability in chloroplast ROS accumulation and subsequent cell survival”, first paragraph), yet this is not general knowledge and I couldn't find any support for this statement in the manuscript. I also accept the argument that H2O2 concentrations could be higher in microenvironments compared to bulk seawater, but without additional references, I am skeptical that these microenvironments could reach tens of μm H2O2. Overall, I am partially satisfied that these changes resolve my original concerns. For example, it seems pertinent to discuss what the typical H2O2 concentrations are within cells under "normal" and oxidative stress conditions, and how these concentrations compared to the levels of H2O2 used in the experiments (the authors allude to this in their response, but I couldn't find anything on this subject in the manuscript). I accept that not all of the environmental relevance question needs to be addressed in this single manuscript. For example, I agree with the authors that the ecological significance in the marine environment should be addressed in future studies (last line of the Discussion), but I still think there is a bit more that could be easily done to address the environmental relevance of the present work. I would support acceptance of this version of the manuscript, provided that the authors are encouraged to consider the above comments during the proofs stage.

We thank the reviewer for the positive feedback and for his comments, and we now further elaborated the explanation for the chosen high light (HL) conditions and H2O2 concentrations and their environmental relevance, as well as providing additional key references (see Introduction, subsection “High light induces variability in chloroplast ROS accumulation and subsequent cell survival”). The HL conditions used in this study (2,000 µmol photons m-2 s-1) are equivalent to full sunlight conditions in nature, and are likely experienced by phytoplankton in surface seawater (Long, Humphries, and Falkowski 1994). For example, in central subtropical oceans cells may be trapped at the surface layer for at least part of the day due to the diurnal thermocline, which reaches 2-5 m depth during the day (Long et al., 1994). In addition, the same HL conditions were used in a previous study of HL acclimation in P. tricornutum (Lepetit et al., 2013). Milder HL conditions may also elicit a similar response perhaps with a milder magnitude, that may lead to a smaller fraction of oxidized cells compared to the HL conditions used in this study (for example see the chl-roGFP oxidation in response to 700 µmol photons m-2 s-1 in bulk measurements in Figure 1). However, we focused on understanding whether high light could generate an equivalent oxidation level in the chloroplast as exogenous application of H2O2. Therefore, we chose to use harsher HL conditions that also occur in natural environments, and did not examine in this study the full range of light intensities, which can be further investigated in detail in future work.

Regarding the concentrations of exogenous application of H2O2 used in this study; it is important to note that our analysis revealed that during diverse environmental stressors, such as toxic infochemicals, different light intensities and nutrient limitation, the changes in chloroplast targeted roGFP degree of oxidation are comparable to the effect of H2O2 concentrations that were applied exogenously in this study (see Figures 1 and 8). This data suggests that the applied H2O2 concentrations are relevant to physiological stress conditions and have similar effects on the chloroplast EGSH pool. To demonstrate in a more accessible manner that the response to H2O2 as applied in this study is comparable with other environmental stressors, we now present this as Figure 1.

To the best of our knowledge, until now there are no direct in vivo measurements of absolute intracellular H2O2 quantification, in marine microbes (including diatoms) or in their microenvironment. This knowledge gap mainly derives from the lack of tools required for such H2O2 quantification, and methods as genetically encoded H2O2 reporters were only developed in recent years and to the best of our knowledge were not yet implemented in algae. Current knowledge is based mainly on bulk measurements, which mask local and temporal fluctuations, or based on relative or indirect measurements typically assessed by chemical fluorescent probes that do not provide absolute in vivo intracellular quantification and without the high spatial resolution that is required to allow measurements of the cell’s microenvironment.

The reviewer refers to 10’s of nM that were measured in the marine environment in a non-bloom, steady state, extracellular bulk measurements. We would like to stress the point that these measurements are completely different from the intracellular, organelle-specific H2O2 net production after taking into account the cellular antioxidant capacity. In addition, according to Seymour et al., the local concentrations of different compounds in the phycosphere of an individual cell could be several orders of magnitude higher than in bulk seawater (Seymour et al., 2017). Taken together, it is likely that algae cells experience H2O2 concentrations in the range of µM at least under certain stress conditions, either due to intracellular production (for example in the chloroplast under high light conditions), or due to local or temporal production by neighbor cells. For example, some harmful bloom-forming algae species were shown to produce extracellular H2O2 and superoxide (which can be rapidly converted to H2O2) in rates of tens and hundreds of fmol cell−1 h−1 (Diaz et al., 2018). As stated in the manuscript, the environmental relevance of the bi-stable phenotype is yet to be further investigate in future works, and we aim to address additional environmental stress conditions as well as the possible ecological implications in our future research.

https://doi.org/10.7554/eLife.47732.046

Article and author information

Author details

  1. Avia Mizrachi

    Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel
    Contribution
    Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Code development for image analysis, Project administration
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-7724-9275
  2. Shiri Graff van Creveld

    Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel
    Contribution
    Conceptualization, Data curation, Formal analysis, Visualization, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-3445-3046
  3. Orr H Shapiro

    Department of Food Quality and Safety, Institute of Postharvest and Food Sciences, Agricultural Research Organization, The Volcani Center, Rishon LeZion, Israel
    Contribution
    Visualization, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-3222-9809
  4. Shilo Rosenwasser

    1. Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel
    2. The Robert H. Smith Institute of Plant Sciences and Genetics in Agriculture, The Hebrew University of Jerusalem, Rehovot, Israel
    Contribution
    Conceptualization, Formal analysis, Visualization, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8565-9979
  5. Assaf Vardi

    Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel
    Contribution
    Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Writing—original draft, Project administration
    For correspondence
    assaf.vardi@weizmann.ac.il
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7079-0234

Funding

Israel Science Foundation (712233)

  • Assaf Vardi

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Dr. Ron Rotkopf from the Bioinformatics Unit at the Life Sciences Core Facilities, Weizmann Institute of Science for assisting with the statistical analysis. We thank Dr. Vicente I Fernandez and Prof. Roman Stocker for help with the cell tracking algorithm. We thank Dr. Uri Sheyn and the flow cytometry unit at the Life Sciences Core Facilities, Weizmann Institute of Science for technical help with FACS operation. We thank Jenny Mizrahi for proofing the manuscript. This research was supported by the Israeli Science Foundation (ISF) (grant #712233) awarded to AV.

Senior Editor

  1. Ian T Baldwin, Max Planck Institute for Chemical Ecology, Germany

Reviewing Editor

  1. Dianne K Newman, California Institute of Technology, United States

Publication history

  1. Received: April 16, 2019
  2. Accepted: June 18, 2019
  3. Accepted Manuscript published: June 24, 2019 (version 1)
  4. Version of Record published: August 5, 2019 (version 2)

Copyright

© 2019, Mizrachi et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 944
    Page views
  • 182
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)