Chemical interactions in diatoms: role of polyunsaturated aldehydes and precursors

Authors

  • Joséphine Leflaive,

    1. Laboratoire d’Ecologie Fonctionnelle (EcoLab), UMR 5245 CNRS/UPS/INPT, Université Paul Sabatier, bât 4R3, 118 route de Narbonne, 31062 Toulouse Cedex 09, France
    Search for more papers by this author
  • Loïc Ten-Hage

    1. Laboratoire d’Ecologie Fonctionnelle (EcoLab), UMR 5245 CNRS/UPS/INPT, Université Paul Sabatier, bât 4R3, 118 route de Narbonne, 31062 Toulouse Cedex 09, France
    Search for more papers by this author

Author for correspondence:
Joséphine Leflaive
Tel: +33 (0) 5 61 55 85 49
Email: leflaive@cict.fr

Summary

Chemicals produced by aquatic organisms, and especially micro-organisms, have received increasing attention in the last decade for their role in shaping interactions and communities. Several cases emphasize the fact that chemical signals or defence may modulate interspecific interactions. Notably, it has been shown that diatoms, unicellular algae and key primary producers in aquatic ecosystems produce a wide range of bioactive metabolites. Among these compounds, polyunsaturated short-chain aldehydes in vitro strongly impair the reproduction of various potential grazers. In the field, the relationship between aldehyde production and reproductive failure in copepods remains unclear. Recent studies have suggested that these putative defence compounds may also be involved in intercellular communication and in interactions with competitors. Potential effects of the aldehyde precursors on various organisms have also been described. This review presents an overview of various results obtained in the last decade that could help us to understand the role of polyunsaturated aldehydes and their precursors in the ecology of diatoms. It is focused on the dichotomy between freshwater and marine environments. Indeed, most of the results on anti-proliferative aldehydes concern marine planktonic diatoms, whereas they are also known to be produced by benthic and freshwater species.

Introduction

Aquatic organisms have been shown to produce a variety of bioactive compounds, mainly secondary metabolites. Aquatic chemical ecology, which is a dynamic and rapidly developing discipline (Poulson et al., 2009), studies the roles and impacts of secondary metabolites in ecosystems. There is now a body of evidence showing that chemical signals produced by aquatic organisms are involved in many processes, such as predator defence, competition, resource foraging and reproduction (Hay, 2002). Biologically active compounds, such as dimethylsulphide produced by phytoplankton, have indirect effects on the ecology and evolution of communities and ecosystems (McClintock & Baker, 2001; Hay & Kubanek, 2002; Paul et al., 2007; Flynn & Irigoien, 2009). Research in this field has been slowed down for some time by difficulties linked to the characterization of the molecules involved (Pohnert et al., 2007). Nevertheless, recent progress has been achieved in the description of chemical signals in marine and freshwater environments (Rittschof & Cohen, 2004; Leflaive & Ten-Hage, 2007).

As phototrophic micro-organisms are the basis of the aquatic food web, the factors controlling the specific composition and biomass of the phytoplankton have been studied extensively. Among these factors, chemical interactions have received increasing attention as the field of aquatic chemical ecology has emerged. It is not the aim of this review to provide an exhaustive report of phytoplankton defence or allelopathy, but rather to propose a comprehensive view of recent results obtained in both marine and freshwater environments for one type of chemical signal. This illustrates the difficulty of deciphering the role of some compounds in complex interactions with conspecifics, competitors and predators. In this perspective, we have chosen to consider some secondary metabolites produced by diatoms, that is polyunsaturated aldehydes (PUAs) and polyunsaturated fatty acids (PUFAs), which are their precursors (Fig. 1). Diatoms are highly productive microalgae, widespread in marine and freshwaters; they can be either free-living in the water column or can form biofilms fixed on a solid substratum. These highly grazed organisms produce chemical compounds involved in anti-predator defence mechanisms (Miralto et al., 1999). Interestingly, PUAs and PUFAs seem to play complex roles in intra- and interspecific interactions, in addition to predator defence processes (Fig. 2). Recently, several authors have reviewed the chemistry of PUAs and diatom–grazer interactions in marine environments (Paffenhöfer, 2002; Ianora et al., 2003; Paffenhöfer et al., 2005; Pohnert, 2005; Pohnert et al., 2007). These two aspects are examined, but the specificity of the present review is rather to address less advanced research domains, that is diatom–grazer interactions in freshwater environments, PUAs as infochemicals, and PUFAs and PUAs as allelopathic compounds.

Figure 1.

 Chemical structures of some polyunsaturated fatty acids (a) and α,β,γ,δ-unsaturated aldehydes (b) produced by diatoms. Curly lines represent the stereoisometry, E or Z, of the molecules. (c) Structure of the reactive Michael acceptor element found in the anti-proliferative compounds from diatoms (R, alkyl or alkenyl) (after Pohnert et al., 2002).

Figure 2.

 Schematic representation of the interactions between diatoms and competitors or predators (white arrows) that may be mediated by polyunsaturated aldehydes produced by diatoms, such as 2E,4E-decadienal (black arrows). The main effects of the aldehydes are shown in italics (after Pohnert, 2005; Vardi et al., 2006).

Chemicals

Several bioactive compounds belonging to the oxylipin family are produced by diatoms. This family of compounds groups several different metabolites having, as their origin, the oxygenase-catalysed oxidation of PUFAs. Diatom oxylipins include short-chain unsaturated aldehydes and hydroxy, keto and epoxyhydroxy fatty acid derivatives (Fontana et al., 2007). The present review focuses on short-chain PUAs and on polyunsaturated free fatty acids from which they are derived. As shown by the recent examination of 51 freshwater and marine diatom species (71 isolates), PUA production is species and strain dependent (Wichard et al., 2005). Thirty-six per cent of the investigated species release PUA on wounding in various concentrations (0.01–9.8 fmol per cell). The dominant bioactive PUAs released by diatoms are C10 2E,4E/Z-decadienal (2E,4E/Z-DD) and 2E,4E/Z,7Z-decatrienal (Miralto et al., 1999), but also C8 2E,4E/Z-octadienal and 2E,4E/Z-octatrienal and C7 2E,4E/Z-heptadienal (D’Ippolito et al., 2002; Wichard et al., 2005) (Fig. 1). Diatoms may produce either one PUA or a mixture of several PUAs (Wichard et al., 2005; Fontana et al., 2007).

The mechanism of volatile PUA production has not been described fully. It seems clear that intact diatom cells do not contain free PUFAs (Jüttner, 2001). Only on cell damage are phospholipids rapidly cleaved by phospholipases and galactolipases to release free fatty acids (Pohnert, 2002; Cutignano et al., 2006), which are further transformed into PUAs and other metabolites. The main enzyme activity responsible for the initiation of aldehyde production was characterized as a phospholipase A2 (PLA2) (Pohnert, 2002). Eicosanoic fatty acids [C20, eicosapentaenoic acid (EPA) and arachidonic acid] serve as precursors for the production of 2,4,7-decatrienal and 2,4-DD (Pohnert, 2000), whereas different substrates (C16) are required for the biosynthesis of C8 aldehydes (D’Ippolito et al., 2003). The production of PUFAs is activated seconds after cell damage (Pohnert, 2000). Production in intact cells is certainly prevented by compartmentalization, separating phospholipases from phospholipids. The free fatty acids produced through this pathway are transformed into intermediate hydroperoxides by lipoxygenase (LOX). The hydroperoxides may next be cleaved either by the action of LOX itself or by hydroperoxide lyases. The enzymes involved in the biosynthesis of PUAs (PLA2, LOX-hydroperoxide lyases) in the marine diatom Thalassiosira rotula remain active in seawater over several minutes (Pohnert, 2000, 2002), which leads to a locally high concentration of the metabolites produced. Recently, Ribalet et al. (2007b) have shown that PUA production in diatoms relies on the physiological and environmental conditions during growth. PUA production by Skeletonema marinoi cells was found to be higher when cultures were in the stationary growth phase, or nitrogen or phosphorus limited. The limiting factor for PUA production in diatoms may be the availability of free fatty acids under limiting conditions (Pohnert, 2002; Ribalet et al., 2007b) and the enzyme activity during normal growth conditions (Ribalet et al., 2007b).

The bioactivity of PUFA derivatives is linked to the presence of an α,β- or α,β,γ,δ-unsaturated aldehyde group, which is a structural element typical for lipid peroxidation products (Fig. 1). Molecules bearing this aldehyde group are potent Michael acceptors (Vollenweider et al., 2000), which can form covalent adducts with nucleophiles and may thus be toxic through interference with many cellular processes (Refsgaard et al., 2000). In addition to the presence of this element, the polarity of the side carbon chain plays a role in PUA toxicity. Bioassays performed with synthetic and stereospecific α,β,γ,δ-unsaturated aldehydes have demonstrated that the toxicity of the metabolites for sea urchin increases with chain length from C7 to C10 (Adolph et al., 2003). Both 2E,4E- and 2E,4Z-dienals showed the same activity. This suggests that nonspecific interactions of PUAs with various molecules are responsible for their toxicity (Adolph et al., 2004).

Diatom–grazer interactions in marine waters

Marine diatoms have been considered for some time as high-quality food for sustaining the growth of zooplankton (Riley, 1947). Since the mid-1990s, numerous studies (reviewed by Ianora et al., 2003; Paffenhöfer et al., 2005; Pohnert, 2005) have challenged this traditional view and have led to the consideration of a ‘diatom–copepod paradox’ (Ban et al., 1997). One of the first experimental observations was that a diatom diet induced a decrease in egg hatching success in the copepod Temora stylifera (Ianora & Poulet, 1993). Diatom extracts could also induce the abnormal development of Calanus helgolandicus embryos (Poulet et al., 1994, 1995). Some of the bioactive compounds produced by diatoms were first isolated by Miralto et al. (1999) and identified as three PUAs: 2E,4E-DD, 2E,4E,7Z-decatrienal and 2E,4Z,7Z-decatrienal. Strikingly, these putative defence compounds have no effect on adult copepods, but on the + 1 generation, a reason why this effect is called ‘insidious’ (Miralto et al., 1999). The originality of the interaction between diatoms and copepods is the induction of reproductive failure rather than the more common poisoning, feeding deterrence or repellence.

The specific effects of PUA-producing diatoms have been studied extensively in vitro on various marine organisms (Table 1). Two types of experiment have been classically performed: feeding experiments in which grazers are fed with PUA-producing diatoms (e.g. Ianora et al., 2004), and incubation tests in which individuals are exposed to pure compounds (e.g. Adolph et al., 2004; Caldwell et al., 2004). Another method of delivery of PUAs is to employ dinoflagellates pre-incubated with the compound as a carrier (Ianora et al., 2004). More recently, Buttino et al. (2008) have employed giant aldehyde-encapsulating liposomes (same size as the phytoplankton cells eaten by copepods) in order to link DD ingestion and the reproductive success of two copepods. By this means, the authors showed that the effective concentrations of DD were one order of magnitude lower than those used in classical feeding experiments.

Table 1. The comparative effects of some polyunsaturated aldehydes and eicosapentaenoic acid (EPA) on various organisms
Target organismGroupCompoundExposure timeEffective concentration or tested concentration (μg ml−1)EffectsReference
  1. DD, 2E,4E/Z-decadienal; EC50, compound concentration giving a 50% reduction in growth rate; HD, 2E,4E/Z-heptadienal; LC50, lethal dose required to kill 50% of individuals; MIC, minimum inhibitory concentration; OD, 2E,4E/Z-octadienal.

Thalassiosira weissflogiiAlga (diatom)DD24 hEC50 = 0.29
0.5
Growth inhibition
Interference with cell cycle progression
Photosynthetic efficiency decrease
Apoptosis-like cell death
Casotti et al. (2005)
Thalassiosira weissflogiiAlga (diatom)DD2 h10Nitric oxide generationVardi et al. (2006)
Phaeodactylum tricornutumAlga (diatom)DD24 h
2 min
30 min
2 h
EC50 = 1.07
10
10
10
Growth inhibition
Increase in intracellular calcium
Nitric oxide generation
Cell death
Vardi et al. (2006)
Dunaliela tertiolectaAlga (chlorophyte)DD
OD
HD
DD, HD, OD
24 h
24 h
24 h
24 h
EC50 = 0.33
EC50 = 0.70
EC50 = 1.18
twice EC50
Growth inhibition
DNA degradation
Ribalet et al. (2007a)
Amphidinium carteraeAlga (dinoflagellate)DD
OD
HD
DD, HD, OD
24 h
24 h
24 h
24 h
EC50 = 0.25
EC50 = 0.65
EC50 = 0.99
twice EC50
Growth inhibition
Chromatin fragmentation
Ribalet et al. (2007a)
Staphyloccocus aureusBacteriaDD18 hMIC = 7.8Growth inhibitionBisignano et al. (2001)
Vibrio splendidusBacteriaDD24 h33.3 μg per discGrowth inhibition (agar diffusion assay)Adolph et al. (2004)
Listonella anguillariumBacteriaEPA24 h0.3Growth inhibitionDesbois et al. (2009)
Artemia salinaCrustacean (brine shrimp)DD24 hEC50 = 2.14Larval mortalityCaldwell et al. (2003)
Calanus helgolandicusCrustacean (copepod)DD72 h
1 h
1.5 (female exposure)
5
Teratogenesis and nauplii death - Induction of apoptosis in embryosIanora et al. (2004)
Romano et al. (2003)
Temora styliferaCrustacean (copepod)DD24 h1.5Hatching success decreaseMiralto et al. (1999)
Ceballos & Ianora (2003)
Tisbe holothuriaeCrustacean (copepod)DD24 hLD50 = 1.4Nauplii deathTaylor et al. (2007)
Sphaerechinus granularisEchinoderm (urchin)DD2–3 h1.5–3Cell cleavage blockage (eggs)
Cell divisions blockage (embryos)
Adolph et al. (2004)
Asterias rubensEchinoderm (sea star)DD
EPA
48 h
4 h
4 h
15 min
4 h
0.5
25
5
0.05
20
Hatching success decrease
Development inhibition
Fertilization success decrease
Decrease in sperm motility
No effect
Caldwell et al. (2002)
Caldwell et al. (2004)
Microbotryum violaceumFungiDD
OD
24 h
24 h
50 μg per disc
50 μg per disc
Growth inhibition (agar diffusion assay)
No effect
Adolph et al. (2004)
Crassostrea gigasMollusc (oyster)DD30 min0.3–7.6Modification of structure of cytoskeleton (haematocyte)
Apoptosis (haematocyte)
Inhibition of phagocytosis (haematocyte)
Adolph et al. (2004)
Arenicola marinaPolychaete (worm)DD
EPA
96 h
4 h
4 h
15 min
4 h
0.125
0.5
0.5
0.05
20
Hatching success decrease
Development inhibition
Fertilization success decrease
Decrease in sperm motility
No effect
Caldwell et al. (2002)
Caldwell et al. (2004)

A body of evidence exists for the anti-proliferative effect of PUAs on diverse organisms that are potential consumers of diatoms (Table 1). PUAs inhibit sperm motility in sea urchin embryos (Caldwell et al., 2004), reduce fertilization, embryogenesis and hatching success in polychaetes and echinoderms (Caldwell et al., 2002) and block the fertilization current in ascidian oocytes (Tosti et al., 2003). The effective concentrations of DD, often used as a model of PUA, range between 0.05 and 30 μg ml−1. In copepods, PUAs have been shown to induce apoptosis in maturing oocytes (Poulet et al., 2007a), during embryo development (Romano et al., 2003) and in newly hatched nauplii (Ianora et al., 2004). The consequences are a reduced hatching success and the induction of malformations in larval stages, leading to low survival rates. The concentrations used in the different tests (c. 2 μg ml−1) seem relevant. Indeed, after Ianora et al. (2004), on the basis of the diatom ingestion rates determined by Turner et al. (2001), the quantity of DD ingested by copepods at 1.5 μg ml−1 is below the quantity that they can ingest in the field.

In the field, observations show more contrasts. Several studies have suggested that copepods are negatively affected by the massive presence of diatoms. The egg hatching success of Acartia clausi was only 12% during a diatom bloom, compared with 90% in post-bloom conditions (Miralto et al., 1999). The reproductive success of the copepods Pseudocalanus newmani and Calanus pacificus was transitorily reduced during blooms of Thalassiosira spp. (Halsband-Lenk et al., 2005; Pierson et al., 2005). By contrast, a few studies have shown no impact of diatoms on grazers in the field. Indeed, a 7-yr survey in the English Channel showed no effect of diatoms on Calanus helgolandicus egg production (Irigoien et al., 2000). Similarly, no relationship has been found between the diatom biomass and hatching success of copepods in 12 globally distributed areas (Irigoien et al., 2002). More recently, the compositions of phytoplankton communities in mesocosms have been manipulated via the Si : N ratio (Sommer, 2009). This ratio strongly influenced the proportion of diatoms and the hatching success was highly variable, but there was no link between the ratio and copepod reproduction.

With regard to studies performed in various areas, it seems well established that diatoms can induce reproductive failure in copepods in the field (e.g. Laabir et al., 1995; Ianora et al., 2004, 2008; Poulet et al., 2006, 2007b; Vargas et al., 2006). In some cases, a direct correlation between PUA production and the reduced hatching success of a copepod has been found (Halsband-Lenk et al., 2005; Horner et al., 2005). Nevertheless, the link between PUAs and reproductive failure is not very clear. The effective role of PUAs has been challenged by recent studies that failed to correlate PUA production and reproductive failure in the field (Poulet et al., 2006, 2007a; Wichard et al., 2008) and in the laboratory (Dutz et al., 2008).

These contrasting results may be partly explained by the large variability of PUA production by diatoms (Wichard et al., 2005), by the variability of copepod sensitivity (Ianora et al., 2008; Sommer, 2009) and by the impact of diatom nutritional status on their deleterious effects (Jones & Flynn, 2005). Several hypotheses have also been formulated: the production of active compounds different from PUAs (Fontana et al., 2007; Ianora et al., 2008), food limitation caused by the decrease in available essential fatty acids, which are PUA precursors (Wichard et al., 2007), and the possibility of detoxification by certain species of copepod (Ianora et al., 2008; Wichard et al., 2008). The first hypothesis is supported by the fact that other products from the LOX cascade, such as fatty acid hydroperoxide or highly reactive oxygen species, can affect grazer reproduction (Fontana et al., 2007; Ianora et al., 2008). For example, in the absence of PUAs, hydroxy and keto derivatives of EPA and docosahexaenoic acid from a phytoplankton bloom presumably caused a low fecundity of copepods (Ianora et al., 2008). Sommer (2009) also proposed that the negative effects of diatoms on copepods may be compared with the effects of toxic plants on grazers in grasslands. Toxic plants may locally and in particular circumstances represent a problem for grazers, but are not important on a larger scale. He proposed a few reasons to explain this: the presence of alternative food in the field, the presence of nontoxic diatoms in phytoplankton assemblages, the presence of resistant copepods and the short toxicity of diatoms in the life cycle of copepods. An alternative hypothesis was proposed by Dutz et al. (2008), who suggested that the negative impact of diatoms may be linked to nutritional deficiency induced by the incomplete digestion of diatom cells. They rejected the idea of a role of PUAs in the induction of reproductive failure.

The results cited above may be summarized as follows. In vitro, PUAs produced by diatoms undoubtedly impair the reproduction of several organisms by whom they are potentially consumed, and this happens at relevant concentrations. In the field, copepod reproduction is dependent on their food and may be impaired by diatoms. In vitro and in the field, the negative impact of diatoms on copepod reproduction is not always correlated with PUA production. The recent advances in this domain challenge the view of PUAs as a chemical defence against grazers. This is supported by Flynn & Irigoien (2009), who showed via modelling that PUA production does not confer any advantage to diatoms. In this context, PUAs may be secondary metabolites of no evolutionary selective advantage. The factors affecting copepod reproduction and the role(s) of PUAs in diatom–grazer interactions are still unclear.

Diatom–grazer interactions in freshwaters

The formation of unsaturated aldehydes has been described for some time in freshwaters, notably in planktonic chrysophytes (Jüttner, 1981), benthic diatoms (Wendel & Jüttner, 1996) and diatom-dominated biofilms (Jüttner & Dürst, 1997; Jüttner, 2005). Considering that, in freshwaters, Daphnia are key zooplankton grazers, comparable with copepods in seawaters, several authors have studied the effects of PUAs on these organisms. Daphnia pulicaria showed similar ingestion rates on PUA-producing and nonproducing diatom strains, indicating that PUAs are not feeding deterrents for this species (Carotenuto & Lampert, 2004). A subsequent study indicated that the juvenile growth rate of D. pulicaria was not correlated with the PUA content of the diatom diet (Carotenuto et al., 2005). Similarly, several field studies found no correlation between PUA production and reproductive failure in marine copepods (Dutz et al., 2008; Wichard et al., 2008). Egg hatching success was affected by DD release only for late clutches that do not contribute significantly to population parameters (Carotenuto et al., 2005). These results suggest that PUAs may not be active against Daphnia. This hypothesis was moderated by the study of Watson et al. (2007), who observed that Daphnia magna could perceive and react to decatrienal at micromolar concentrations by a short-term increase in swimming. As for marine copepods, a high variability in the sensitivity of Daphnia species to PUAs may be expected. Another study performed with a natural community of grazers (cladocerans and copepods) provided evidence that PUAs could be repellent for these crustaceans (Jüttner, 2005). Both pure compounds and extracts from diatom-dominated biofilms with decatrienal and octadienal as major compounds exhibited repellent activity. It was shown that a major part of EPA was retained in damaged cells, whereas PUAs were released into the medium. The EPA : PUA molar ratio was 80 in biomass and 0.28 in lake water. It has been shown previously that EPA is toxic for a freshwater grazer Thamnocephalus platyurus (Jüttner, 2001). Therefore, Jüttner proposed that PUAs may form a robust microzone around damaged cells, signalling the presence of toxic EPA and preventing further attack of intact cells from grazers. Interestingly, in most studies, EPA was found to be nontoxic against marine organisms (e.g. Miralto et al., 1999; Caldwell et al., 2003, 2004). To our knowledge, only one study reported the haemolytic activity of this compound (Fu et al., 2004). These various results suggest that PUFAs, with EPA as a model compound, and PUAs, with DD as a model compound, may have slightly different biological and ecological functions in the planktonic marine environment and benthic/planktonic freshwater environment (Fig. 3). Further research is needed to achieve this comparison and to better understand the impact of PUFAs and PUAs on freshwater micro- and macro-organisms. This would be particularly relevant for benthic organisms. Indeed, biofilms are particularly interesting from the viewpoint of chemical defences because they are under intense grazing pressure (Feminella & Hawkins, 1995) and because the algal density is very high (Jüttner, 1999).

Figure 3.

 Comparison of the known effects of polyunsaturated fatty acids (PUFAs) and polyunsaturated aldehydes (PUAs) produced by diatoms in marine and freshwater environments. Both are usually produced when diatom cells are damaged, the first essentially remaining in the cells and the second forming a microzone of diffusion outside the cells. Eicosapentaenoic acid (EPA) and 2E,4E/Z-decadienal (DD) are often employed as models for PUFA and PUA, respectively. The information reported in the figure was derived from several studies (Miralto et al., 1999; Jüttner, 2001, 2005; Pohnert et al., 2002; Carotenuto & Lampert, 2004; Watson et al., 2007).

PUAs as infochemicals: role in cell–cell communication

New ideas have emerged concerning the biological and ecological functions of bioactive PUAs produced by diatoms. In particular, it has been suggested that they may play a role in phytoplankton cell–cell communication (Watson, 2003). PUAs may thus be infochemicals, that is chemicals that convey information in an interaction between two individuals, evoking in the receiver a behavioural or physiological response that is adaptive to either one of the interactants or both (Dicke & Sabelis, 1988). Casotti et al. (2005) were the first to suggest a role of PUAs as chemical signals of unfavourable conditions. They demonstrated that DD could induce in Thalassiosia weissflogii a degenerative process leading to cell death. DD could reduce T. weissflogii growth by 50% at 0.29 μg ml−1. The authors noticed that dying diatom cells exhibited morphological and biochemical features close to those of mammalian apoptotic cells. Apoptosis is a sequence of morphological characters seen in dying/dead metazoan cells during/after the activation of programmed cell death (PCD) (Franklin et al., 2006). Such observations suggest that DD could induce PCD, that is cell death resulting from gene expression. In the past 10 yr, several studies have demonstrated the existence of PCD in photosynthetic microorganisms (Berges & Falkowski, 1998; Vardi et al., 1999), but its interpretation is still debated (Bidle & Falkowski, 2004; Lane, 2008).

The mechanism of cellular reaction to DD has been studied in detail in the marine diatoms T. weissflogii and Phaeodactylum tricornutum (Fig. 4), two species that do not produce PUAs (Wichard et al., 2005). As observed by Casotti et al. (2005), DD could induce cell death, with a threshold concentration of 3 μg ml−1 for P. tricornutum. It induced a rapid accumulation of nitric oxide (NO) in a subcellular compartment close to the nucleus. Notably, NO is known to be involved in the defence response in plants (Delledonne et al., 1998). DD-induced NO production is dependent on an NO synthase-like activity, which is calcium dependent. Vardi et al. (2006) have demonstrated that NO is directly involved in the cell death cascade. An infochemical function of PUAs is supported by the additional result that the fate of DD-exposed cells is dose dependent: sublethal doses of DD may trigger signalling phenomena, leading to induced resistance, whereas higher doses induce cell death (Fig. 4). The authors also presented evidence that DD-induced stress could be transmitted to nonexposed cells by a diffusible signal from stressed cells. These results suggest the presence of a stress surveillance system in diatoms. A cascade reaction to DD exposure has been further described by Vardi et al. (2008). Using a functional genomics approach in P. tricornutum, they identified a novel protein (PtNOA) belonging to a subfamily of guanosine triphosphate-binding proteins thought to play a role in NO generation (Guo et al., 2003). This protein was upregulated in the presence of DD. Transformed cells overexpressing this protein were hypersensitive to sublethal doses of DD. This cell line also showed an altered expression of superoxide dismutase and metacaspases, both being involved in stress and death pathways (Wolfe-Simon et al., 2006; Bidle & Bender, 2008). This suggests a role of PtNOA in the activation of PCD in marine diatoms, notably in response to DD exposure.

Figure 4.

 Cellular reactions and consequences resulting from the exposure of a diatom cell to the polyunsaturated aldehyde decadienal (after Casotti et al., 2005; Vardi et al., 2006, 2008). MnSOD, superoxide dismutase; NO, nitric oxide; NOS, nitric oxide synthase; PtNOA, protein of the marine diatom Phaeodactylum tricornutum belonging to a conserved subfamily of guanosine triphosphate-binding proteins; PUA, polyunsaturated aldehyde.

Vardi et al. (2008) also observed that, in the PtNOA-overexpressing line, the attachment strength of P. tricornutum oval cells to the substrate was markedly reduced compared with the wild-type line. This is consistent with the results from Thompson et al. (2008), who proposed the involvement of NO in diatom adhesion. The addition of an NO donor to the culture of the pennate diatom Seminavis robusta did not affect its growth, but significantly reduced its adhesion to the substrate. Further research is needed to determine whether DD can affect diatom adhesion. If PUAs could induce the detachment of benthic diatoms, this may be interpreted as a method of escape from an unfavourable environment. Such a hypothesis is consistent with Casotti et al.’s (2005) theory of PUAs as chemical signals of unfavourable conditions.

The role of PUAs as infochemicals is also supported by the recent study of Vidoudez & Pohnert (2008), who showed that PUAs fulfil three requirements for infochemicals: they are released into the medium, they are present for a short period and they have an effect on cultures in specific stages. The authors cultured Skeletonema marinoi for 30 d and detected low but significant concentrations of heptadienal (up to 0.032 μg ml−1) and octadienal (up to 0.010 μg ml−1) in the late stationary phase. A short – 5 d – burst of PUAs occurred before cell lysis, indicating for the first time that these compounds may be released by intact cells. Because of their reactivity, PUAs do not accumulate in the medium, which make them reliable time-limited signals. Naturally occurring concentrations of heptadienal and octadienal had no effect on diatom growth, except when the cells were exposed in the late stationary phase. This suggests a time-dependent susceptibility to PUAs.

Although data on the environmental concentration of PUAs are lacking, these studies suggest a potential role of PUAs in cell–cell communication in phytoplankton communities. They may be signals of unfavourable conditions (senescent population, presence of grazers) that trigger a complex cellular response. Because of the differential sensitivity of algal species to PUAs (Ribalet et al., 2007a), these compounds may participate in the control of species’ successions during blooms or in bloom termination.

Allelopathic activity of PUAs and PUFAs – interactions between diatoms and competitors

Phototrophic micro-organisms, such as diatoms, are under intense competitive pressure for resources, such as light, nutrients or space. In this context, they have developed mechanisms to outcompete other algae, bacteria or even fungi. One strategy is allelopathy (see reviews by Legrand et al., 2003; Leflaive & Ten-Hage, 2007). Allelopathy is a process by which secondary metabolites produced by plants or microorganisms influence the growth or development of competitors (Rice, 1984). Allelopathic interactions are likely to play important roles in species’ successions (Keating, 1977) and in the occurrence of blooms (Takamo et al., 2003). In field and laboratory studies, diatoms have been shown to produce allelopathic compounds (Takamo et al., 2003; Yamasaki et al., 2007). Several studies have suggested that PUAs and PUFAs may be allelopathic compounds. Until recently, this hypothesis seemed far-fetched as allelopathic compounds are supposed to be actively released into the medium. It became more plausible when Vidoudez & Pohnert (2008) described the release of PUAs by intact cells (see above). By contrast, PUFAs still seem to be released only on cell lysis (Jüttner, 2001, 2005).

Several studies have shown the negative effects of PUAs on various microorganisms. Marine diatoms are negatively affected by the presence of DD (Casotti et al., 2005; Hansen & Eilersten, 2007; see above). Nevertheless, Hansen & Eilersten (2007) found no negative correlation between the presence of the PUA-producing haptophyte Phaeocystis pouchetii and the abundance of sensitive diatoms in the natural environment. In addition to diatoms, others phytoplanktonic species belonging to different taxonomic groups have been shown to be inhibited by DD, octadienal or heptadienal (Ribalet et al., 2007a). In the study by Ribalet et al. (2007a), the effective concentrations ranged between 0.15 and 0.377μg ml−1 for DD. These concentrations are much higher than those of heptadienal and octadienal detected in cultures of Skeletonema marinoi (Vidoudez & Pohnert, 2008). It is noticeable that algae seem to be as, or even more, sensitive to PUAs than do animals. For example, the concentrations affecting crustaceans ranged between 0.18 and 2.35 μg ml−1 (Caldwell et al., 2003). This may be counteracted by the fact that PUAs are directly produced inside the digestive tract of grazers (Wichard et al., 2008). Interestingly, although data on freshwater micro-algae are scarce, they seem to be less sensitive to PUAs than do marine micro-algae (Adolph et al., 2004; Leflaive et al., 2008). For all the phototrophic organisms tested, the observed effects of PUAs were similar to apoptosis, as described for animals (Ianora et al., 2004; Poulet et al., 2007a). This is shown by cell shrinkage, chromatin condensation and the degradation of nuclear DNA to nucleosomal size fragments.

Some marine and nonmarine bacteria, and some fungi, have also been shown to be sensitive to PUAs (Bisignano et al., 2001; Adolph et al., 2004; Ribalet et al., 2008), but the effective concentrations are one to two orders of magnitude higher than those having an effect in algal cells. Sensitivity to PUAs has been shown to be species and strain dependent. Adolph et al. (2004) suggested that reduced cell permeability to PUAs may be one of the reasons for the resistance of certain species. They observed that a wild-type Saccharomyces cerevisiae was insensitive to DD, whereas a genetically modified strain with increased cell permeability was strongly inhibited by this compound. Nevertheless, no relationship was found between the cell wall characteristics of bacteria (Gram-positive or Gram-negative) and their sensitivity to PUAs (Bisignano et al., 2001). This suggests that the penetration of DD into bacterial cells may not be linked to the general ultrastructure of the cell wall, but to species-specific characteristics, such as the presence of porines. Strikingly, one DD-resistant marine bacterial strain belongs to the genus Sulfitobacter, which is often associated with phytoplankton blooms (Ribalet et al., 2008). This suggests that PUAs may play a role in bacterial successions during blooms by favouring resistant species.

Although less studied than PUAs, PUFAs produced by diatoms have been shown to inhibit the growth of various Gram-positive and Gram-negative bacteria (Benkendorff et al., 2005; Desbois et al., 2009). EPA was active against marine strains and against potential human and animal pathogens at micromolar concentrations (Desbois et al., 2009). EPA and arachidonic acid also inhibited the growth of fish pathogens (Benkendorff et al., 2005). In addition to these compounds, two other PUFAs produced by P. tricornutum, palmitoleic acid (C16) and hexadecatrienoic acid (C16), showed bactericide activity (Desbois et al., 2008). The mode of action of fatty acids has not been elucidated fully, but some authors have suggested the existence of a peroxidative process involving hydrogen peroxide (Wang & Johnson, 1992).

Conclusions and future directions

Historically, a frontier subsists between research in marine and freshwater habitats. Although PUAs and PUFAs are known to be produced in both environments, their roles as oxylipins, infochemicals and toxins have not been described fully in freshwater environments. We believe that the study of PUAs and PUFAs would improve our understanding of their role in the freshwater environment, especially in biofilms. In addition to the marine/freshwater dichotomy, another difficulty that should be overcome in the coming years is the transposition of laboratory results to the field. This will be helped by the improvement of detection methods to determine the in situ concentrations of compounds (Vidoudez & Pohnert, 2008).

This review has focused on diatoms, but other micro-organisms, such as the haptophyte Phaeocystis pouchetii (Hansen et al., 2004) and the benthic green alga Ulva conglobata (Akakabe et al., 2003), are also able to produce PUAs. The production of toxic aldehydes and their potential consequences on biotic interactions may thus be explored in additional groups of algae. This would help to determine whether these compounds, which are present across the phylogenetic and environmental barriers, may play a role in the ecology of their producers. This also raises the question of the adaptation and co-existence of PUAs among producing and nonproducing diatoms, as well as their interactions with their competitors (i.e. dinoflagellates) and consumers (i.e. copepods). Flynn & Irigoien (2009) stated that the production of PUAs confers no selective advantage in terms of defence against grazers. Further studies may compare the fitness of PUA-producing and nonproducing strains, and their ability to outcompete other phytoplanktonic species. Experiments with complex phytoplankton communities with short exposure to PUAs may also help to determine whether these molecules may partly control species’ successions, as hypothesized by Vardi et al. (2006).

With regard to diatom–grazer interactions, results from recent studies question the role of PUAs in the induction of decreased egg viability by diatoms (e.g. Dutz et al., 2008; Wichard et al., 2008). To better understand the interaction between diatom and copepod, several directions may be explored. A very high variability is observed with regard to the sensitivity of copepods to PUAs, possibly linked to detoxification mechanisms (Ianora et al., 2008). Such mechanisms deserve further study. In addition, diatom metabolites other than PUAs have been shown to be active against copepods (Fontana et al., 2007; Ianora et al., 2008). Additional studies should focus on these compounds, on their anti-proliferative activity and on their distribution among diatom species. It will be necessary to evaluate the production of these compounds, together with PUA production, during phytoplankton successions in the field, in parallel with the measurement of parameters describing copepod reproduction. As emphasized by Paffenhöfer et al. (2005), these points will necessitate a multidisciplinary approach involving classical methods and new tools of genetics, as data on the diatom genome are available (Armbrust et al., 2004; Bowler et al., 2008).

Ancillary