Perception of Environmental Signals by a Marine Diatom

See allHide authors and affiliations

Science  30 Jun 2000:
Vol. 288, Issue 5475, pp. 2363-2366
DOI: 10.1126/science.288.5475.2363


Diatoms are a key component of marine ecosystems and are extremely important for the biogeochemical cycling of silica and as contributors to global fixed carbon. However, the answers to fundamental questions such as what diatoms can sense in their environment, how they respond to external signals, and what factors control their life strategies are largely unknown. We generated transgenic diatom cells containing the calcium-sensitive photoprotein aequorin to determine whether changes in calcium homeostasis are used to respond to relevant environmental stimuli. Our results reveal sensing systems for detecting and responding to fluid motion (shear stress), osmotic stress, and iron, a key nutrient that controls diatom abundance in the ocean.

The predominance of diatoms in marine ecosystems indicates that they possess sophisticated strategies for responding to environmental variation. However, little is known about these strategies at the cellular level because of the difficulties of monitoring key internal physiological processes in these organisms. The recent availability of procedures to generate transgenic diatoms (1–3) has opened up a range of techniques that can be used to address these questions.

We transformed the marine diatom Phaeodactylum tricornutum(4) with a construct containing the apoaequorin cDNA derived from the jellyfish Aequorea victoria (5). Transgenic cells displaying high levels of aequorin were selected (4) and used to analyze diatom responses to a range of different stimuli.

Transient changes in cytosolic calcium concentrations ([Ca2+]cyt), which are characteristic of the activation of signal transduction (6), could be observed in response to the simple addition of seawater [artificial seawater (ASW)] (Fig. 1A). A maximal increase in [Ca2+]cyt from 500 nM to 2 μM was observed after 1 to 2 s, which quickly disappeared within 10 s. This response weakened with declining stimulus strength (7). To exclude the possibility that chemical signaling may have been involved as a result of the addition of fresh medium to the cells, we confirmed that conditioned medium (medium in which the diatoms had been growing) was able to generate the same effects, as was the mechanical stimulation of the cell suspension with a needle (Fig. 1A). When these experiments were repeated with a pH microelectrode in the suspension, no changes in pH were detected after the treatments (8), thus excluding the possibility that the observed calcium responses were a consequence of external pH changes.

Figure 1

Influence of fluid motion on cytosolic calcium. (A) Approximately 104 cells suspended in 50 μl of ASW were stimulated by the addition of 50 μl of fresh medium (ASW) or conditioned medium, or by mechanical stimulation performed with a needle. Treatments were performed 3 s after the beginning of the trace. (B) Adaptive responses to shear. Cells (104 in 50 μl of fresh medium) were given two successive stimuli of the same intensity. Times between the two stimuli were 5 s, 15 s, 30 s, 1 min, 2 min, 5 min, 10 min, 15 min, and 30 min. The graph shows the differences between the maximal [Ca2+]cyt obtained with the second treatment (inj2) with respect to the [Ca2+]cyt in response to the first stimulus (inj1), normalized with respect to the first response [(inj2 − inj1)/inj1]. Standard deviations are shown (n = 5 repetitions).

When organisms are exposed to a stimulus for a long enough period, they typically lose their ability to respond with their original sensitivity. By this process of short-term adaptation or desensitization, a cell reversibly adjusts its sensitivity to the level of the stimulus. At the molecular level, the best understood examples of adaptation occur in bacterial chemotaxis (9) and in photoperception in the retina (10). To determine whether diatoms possess such sophisticated sensing systems for responding to fluid motion, we examined the effect of giving two successive stimuli to the same cells (Fig. 1B). If a second treatment immediately followed the initial stimulus (for example, after 5 s), no difference in the response was observed. However, when the time between the two treatments was increased (from 30 s to 5 min), the response was desensitized (Fig. 1B). This “recalcitrant period” lasted for approximately 5 min before the cells became progressively resensitized once more (after 10 to 15 min). The response to fluid motion is therefore rapid (seconds) and highly controlled, and the time scale for intracellular adjustments to changes in the cell boundary layer appears to be on the order of minutes. Furthermore, the observation that diatom desensitization develops slowly (up to 5 min) is unlike typical desensitization processes in higher eukaryotes, where no response is commonly observed if the second stimulus is applied shortly after the first.

We also analyzed diatom responses to osmotic stress. Many organisms, such as bacteria and plants, have evolved mechanisms to adjust their internal osmolarity in response to osmotic stress, based on the intracellular accumulation of ions or other osmolytes (11,12). Diatoms were given a hypo-osmotic stress by treatment of ASW-grown cells with diluted ASW (Fig. 2A). Again, the [Ca2+]cyt response was rapid and transient, and reflected quantitatively the amount of stress imposed on the cells (Fig. 2B). In general, the response was stronger than the shear response [for example, the [Ca2+]cyt peak was around 4 μM in 80% ASW (Fig. 2A)]. To determine whether this response was due to an osmotic response or to the dilution of a particular ion, diatoms were incubated in a sucrose solution with an osmolarity equivalent to that of 100% ASW (Fig. 2C). Reductions in osmolarity by dilution of sucrose were found to provoke responses identical to those seen with dilution of ASW, thus confirming that this response was due to an osmotic sensing mechanism. The possibility that the responses to hypo-osmotic stress were an artefact of cell lysis was excluded by determining diatom uptake of Evans blue (13).

Figure 2

Influence of osmotic stress. (A) Diatom responses to hypo-osmotic shock, fluid motion, and low- and high-temperature treatments. Diatom cells (104) in 100 μl of fresh medium (100% ASW) were stimulated with 25 μl of distilled water to generate a hypo-osmotic shock (80% ASW final); cells were also stimulated with ice-cold and warm (37°C) media. (B) Effect of osmolarity differences on hypo-osmotic shock–induced cytosolic Ca2+ elevations. Osmotic shocks were generated by injecting 50 μl of different dilutions of fresh media (ASW) to 104 cells in 100% ASW. (C) The same osmolarity differences were also generated by injection of different concentrations of sucrose solutions to the cells suspended in a sucrose solution with an osmolarity equivalent to that of 100% seawater. Arrows show the osmolarity of 100% ASW and in the corresponding sucrose solution (940 mosmol). The osmotic pressure of solutions was measured with a micro-osmometer (Roebling). (D) Diatom cells in 100 μl of fresh medium were stimulated by two successive treatments of distilled water that generated the same hypo-osmotic shock (80% ASW, final salt concentration). Times between the two stimuli were 5 s, 15 s, 30 s, 1 min, 2 min, 5 min, 10 min, 15 min, and 30 min. The graph shows the differences between [Ca2+]cyt peaks obtained with the second treatment (inj2) with respect to the [Ca2+]cyt in response to the first stimulus (inj1), normalized with respect to the first response [(inj2 − inj1)/inj1]. Standard deviations are shown (n = 5 repetitions).

We tested desensitization to hypo-osmotic shock by performing two treatments separated by different time intervals. In contrast to shear, no desensitization was observed in response to hypo-osmotic shock (Fig. 2D). This result therefore indicated substantial differences between the signal transduction pathways for the two different stimuli, which could correspond to distortion and dilution, respectively, of the cell boundary layer.

Although reductions in osmolarity yielded a quantitatively appropriate response that was proportional to the reduction, a hyper-osmotic shock produced no change in cytosolic calcium above the shear signal (Fig. 2B). However, it was possible to adapt the cells to a particular osmolarity (for example, 1140 mosmol), and the response to a reduction of 100 mosmol was identical to that observed in cells adapted to 940 mosmol and then shifted to 840 mosmol (14). Tidal diatoms are likely to be exposed regularly to both hypo- and hyper-osmotic shocks, so one might suppose that a different sensing mechanism (one that does not rely on calcium) is used to detect and quantify hyper-osmotic conditions.

In contrast with higher plant cells (15,16), alterations in temperature did not result in any changes in [Ca2+]cyt above the basal shear response (Fig. 2A). This may reflect a divergent importance of temperature changes between terrestrial and marine ecosystems; that is, temperature variability on land exhibits a much greater range and more rapid variability than in the ocean.

External nutrient concentrations are key regulators of phytoplankton growth. In the marine environment, nutrients such as nitrate, silicate, and phosphate are extremely important, and strong evidence also implicates dissolved iron as being a limiting resource for phytoplankton growth, particularly in high-nutrient, low-chlorophyll (HNLC) regions of the contemporary ocean (17, 18). The “iron hypothesis” has recently been tested and supported, at least for short time scales, through mesoscale iron fertilization experiments in HNLC regions (19, 20). Consequently, new methods have been developed to determine iron deficiency in phytoplankton, such as the analysis of photochemical quantum efficiency based on variable fluorescence measurements (21) and the use of immunological probes (22, 23).

We measured [Ca2+]cyt changes in response to varying nutrient concentrations. No changes in cytoplasmic calcium in response to nitrate, nitrite, ammonium, silicate, or phosphate were observed in either nutrient-replete (Fig. 3) or depleted (8) conditions. However, a response to dissolved iron was detected (Fig. 3), and in this case the calcium signature was very different from that observed after shear or hypo-osmotic shock: the response was not evident until about 5 min after iron addition, and [Ca2+]cyt continued to rise for a considerable time afterward. This time course closely resembles that of iron internalization rates in diatoms (24), although it is not possible in these experiments to know whether calcium has a direct effect on iron uptake. However, aequorin has no specific affinity for Fe2+ and Fe3+ (25), thus excluding the possibility that the observed bioluminescence was an artefact due to iron-stimulated aequorin light emission.

Figure 3

Diatom responses to nutrients. [Ca2+]cyt elevations in response to the addition of different nutrients are shown. 104 cells in 50 μl of ASW were treated with enriched ASW containing 1 mM NaNO3, 1 mM NaNO2, 1 mM NH4Cl, 1 mM Na2SiO3, 1 mM NaH2PO4, or 10 μM FeCl3. The initial calcium spike seen in all the traces corresponds to the fluid motion response.

We grew diatoms in ASW containing iron concentrations ranging from 10 nM to 10 μM (26). In 10 nM FeCl3, cells were clearly growth-limited (Fig. 4A) and in response to the addition of 20 to 30 pM dissolved inorganic iron (Fe′) only these cells showed a response, which was slow and weak (Fig. 4B). However, after the addition of a larger amount of iron (Fe′ = 48 to 60 pM), cells from all cultures responded, and it was possible to observe differences in the timing and kinetics of the responses (Fig. 4C). These results indicate that diatoms control iron metabolism using calcium-mediated regulatory mechanisms and that one component of this regulation prevents iron assimilation at very low concentrations unless the cell is extremely starved. Less starved cells, on the other hand, activate iron uptake only at higher iron concentrations. Therefore, the relationship between bioavailable iron and cellular response is not linear but depends on the history and physiology of the cells. The complex kinetics shown inFig. 4 and the differences in response between the most iron-starved culture and the other three cultures may indicate that in the iron-starved case, the available iron had been completely exhausted from the solution [consistent with the growth rate and iron demand of this culture (Fig. 4A)] and that all the ligand sites on the cell membrane were free of iron. If so, our results suggest a thermodynamic rather than a kinetic control of iron uptake (27). Measurement of the calcium response could therefore be a very sensitive tool for kinetic studies and, more important, a good marker for iron starvation, because the dramatic changes observed may represent intracellular signals for the reactivation of cell division, the stimulation of photosynthesis, and the activation of mitochondrial ATP (adenosine triphosphate) production that are known to occur under conditions of favorable iron bioavailability (21, 28).

Figure 4

[Ca2+]cyt changes in response to FeCl3 addition. (A) Specific growth rate (per day) of cells grown in ASW containing the iron concentrations indicated at the top of the panel (26). Experiments were performed at day 4, when the cells containing 10 nM FeCl3 were clearly iron-limited. Before the addition of iron, cells were washed twice in ASW without FeCl3 to remove all detectable traces of iron and were subsequently treated with enriched ASW containing 5 μM FeCl3 (B) and 10 μM FeCl3 (C), respectively. The dissolved inorganic iron contents (Fe′) of the growth media after these additions are shown.

The historical paradigm that plankton are passive and incapable of resisting physical forces at any scale (29) has already been superseded by evidence for the existence of active responses controlling buoyancy, local fluid viscosity, and life cycle (30–32). Our data indicate that diatoms detect and respond to physicochemical changes in their environment using sophisticated perception systems based on changes in [Ca2+]cyt. Such processes are likely to improve algal adaptation to ubiquitous ocean processes such as mixing and to changes in chemical (for example, nutrient) gradients in time and space. Based on the knowledge of calcium signaling in other organisms, the physiological responses of diatoms to environmental changes are likely to be regulated by sense-process-respond chains involving specific receptors and feedback mechanisms, whose activity is determined by the previous history of the cell. Therefore, the frequent assumption (especially in modeling efforts) that physiological responses of phytoplankton vary monotonically with resource availability is not correct. This new perspective should be considered in future studies and simulations of plankton dynamics.

  • * To whom correspondence should be addressed. E-mail chris{at}


View Abstract

Navigate This Article