Pheromone signaling during sexual reproduction in algae



Algae are found in all aquatic and many terrestrial habitats. They are dominant in phytoplankton and biofilms thereby contributing massively to global primary production. Since algae comprise photosynthetic representatives of the various protoctist groups their physiology and appearance is highly diverse. This diversity is also mirrored in their characteristic life cycles that exhibit various facets of ploidy and duration of the asexual phase as well as gamete morphology. Nevertheless, sexual reproduction in unicellular and colonial algae usually has as common motive that two specialized, sexually compatible haploid gametes establish physical contact and fuse. To guarantee mating success, processes during sexual reproduction are highly synchronized and regulated. This review focuses on sex pheromones of algae that play a key role in these processes. Especially, the diversity of sexual strategies as well as of the compounds involved are the focus of this contribution. Discoveries connected to algal pheromone chemistry shed light on the role of key evolutionary processes, including endosymbiotic events and lateral gene transfer, speciation and adaptation at all phylogenetic levels. But progress in this field might also in the future provide valid tools for the manipulation of aquaculture and environmental processes.


The colloquial term ‘algae’ refers to the photosynthetic representatives of the various protoctist groups, and includes both unicellular eukaryotes (microalgae) and their multicellular descendants (seaweeds). Most algae belong to the super-groups Archaeplastida and Excavata (Adl et al., 2005, 2012), which are phylogenetically very distantly related and represent an enormous diversity in morphology, physiology, life cycles and ecology. Algae are distributed ubiquitously as phytoplankton, photosynthetic biofilms or macrophytic underwater vegetation in marine and freshwaters, but also occur widespread in terrestrial habitats, including soils, wet surfaces on rocks, buildings, snow fields and glaciers as well as the leaves of rainforest trees (Metting, 1981; Broady, 1996). As key primary producers within aquatic ecosystems, algae contribute ca. 50% to the worldwide carbon fixation and have a major effect on global biogeochemical cycles (Falkowski, 1998; Field, 1998; Sabine, 2004). Proliferations of algae can have adverse effects on ecosystems and cause enormous economic losses, e.g. due to biofouling or the occurrence of harmful algal blooms (Fusetani, 2004; Roy et al., 2013). Importantly, algae also emerge as interesting organisms for industrial exploitation due to their high reproduction rate and biochemical composition. In particular, intensive efforts are undertaken to evaluate the production of value products in large scale aquacultures (de Jesus Raposo et al., 2013; Murray et al., 2013; Skjanes et al., 2013).

As with many other aquatic organisms, a key feature of the life of algae is their interaction with other individuals of their own species and with other species via the production and perception of chemical cues. Interactions among conspecifics during life cycle transitions, and in particular sexual reproduction, are mediated by pheromones, while interspecific communication involves allelochemicals. Allelochemical interactions can play a role during competition by the production of compounds that suppress other species or can involve mutualistic relationships in which the release of metabolites benefits the growth of other species (Cembella, 2003; Pohnert et al., 2007; Paul et al., 2009; Vanelslander et al., 2012). Natural products from algae can also serve as a defense mechanism against herbivores or mediate the interaction with associated or pathogenic microorganisms (Ianora et al., 2011; Amin et al., 2012; Roy et al., 2013). As a consequence of this pervasive use of secondary metabolites in communication and defense, algae are known to be a proliferative source for natural products (Sieg et al., 2011; Roy et al., 2013). While research has largely focused on the identification and characterization of the functional role of compounds involved in interspecific interactions, much less information is known about signaling molecules employed during communication among individuals of the same species. Such intraspecific chemical communication may be involved in the induction of resting stage formation or in stress surveillance (Vardi et al., 2006), but the best known function is in the mediation of processes connected to sexual reproduction. Generally, chemical signals that affect individuals of the same species are termed pheromones, a word that is derived from the Greek word pherein, [to carry], and hórmon [excite] (Karlson and Lüscher, 1959). Unlike hormones that act within an organism, pheromones are secreted from the producing organisms into the environment and target conspecifics. As genetic recombination is one of the principal factors driving evolution in eukaryotes and facilitates the adaptation of species to changing environmental conditions, a deeper knowledge about regulative processes during sexual reproduction is essential to the understanding of very fundamental biological processes.

Knowledge and understanding of intraspecific signaling of algae not only provide fundamental insight into life cycle regulation and physiology but may also open new avenues to manipulate the growth of algae in aquaculture and artificial surfaces. In aquaculture, for example, the ability to control sexual reproduction would be highly beneficial in breeding programs to improve strains and to prevent reduced productivity resulting from sexual reproduction occurring in algal production facilities. Further mating disruption based on targeted pheromone delivery might be envisaged to control biofouling on artificial surfaces. Algal life cycles are very diverse both with respect to the ploidy and duration of the asexual phase, mating system and the size of the gametes. Depending on the dominant asexual generation, algal life cycles may be haplontic, diplontic and haplodiplontic (or biphasic), with mitosis/vegetative growth occurring during the haploid, diploid or both haploid and diploid phases, respectively. Sexual reproduction in unicellular and colonial microalgae usually involves similar patterns: two specialized, sexually compatible cells – the haploid gametes – are produced by meiosis (in the case of diplontic species) or by activation of a sexual development program (in the case of haplontic species and the diploid pennate diatoms (see below)) in gametangial cells. In multicellular algae, the gametangium can be more complex structurally and develops on a gametophyte thallus. During mating, the gametes or the gametangia establish physical contact and fuse, either by random encounters or directed by chemical signals. If successful mating can occur between gametes of the same clonal population, the mating system is termed homothallic. In heterothallic mating systems, pairing does not occur within the self-sterile clones, but requires sexually compatible and genetically different mating types. In some groups, mixed strategies are common with clones behaving preferentially but not exclusively, as homothallic or heterothallic and vice versa. Finally, gametes may differ greatly in morphology or behavior. Isogametes are identical physiologically and morphologically. In contrast, anisogamous fertilization is mediated by physiologically and/or morphologically different anisogametes. An extreme case of anisogamy is known as oogamy, in which the immobile female gametes (eggs) attract male gametes, called sperm. In many algal groups, species with either isogamous, anisogamous and oogamous sexual reproduction occur, whereby the latter two modes are considered to be derived from the more primitive isogamous reproduction. In many of these systems, observations and/or experimental evidence indicate that pheromones play a central role in the synchronization of sexual events and increase mating efficiency.

This review aims to summarize our current understanding of the nature and role of these pheromones involved in sexual reproduction of algae. Earlier comprehensive reviews on algal pheromones date back to Jaenicke and Boland (1982), Maier (1993), and Maier and Müller (1986). Since then, progress has been made and for several groups excellent specialized reviews have been published (see in the respective sections below). In this paper, we aim to provide a comparative overview on the chemical diversity of the pheromones and signaling systems employed by different groups of algae. We conclude that, although progress has been relatively slow and our knowledge of pheromone signaling during sex in algae is still very incomplete, this situation may rapidly change in the future thanks to the development of new technologies and molecular resources. In part, the slow progress is due to the fact that elucidating pheromones during sexual reproduction and their functional role is complicated, and this case is due to several reasons. First of all, a good knowledge of the physiology of species is required, in particular with respect to cultivation conditions, induction of sexual behavior, synchronization of cultures etc. For many species, however, detailed knowledge on their life cycle is either missing or sexuality cannot be reliably triggered in laboratory conditions. This situation results in a scenario in which no dependable sources for pheromone extraction are available. In addition, bioassays are complicated by erratic responses. As a consequence, bioassay-guided structure elucidation, which is the traditional approach in pheromone research, is not feasible for those species. Secondly, active concentrations of the compounds released into the environment are often very low and many analytical methods are not sensitive enough to permit easy detection. In addition, due to the minute amounts of pheromones released by the producers, a direct structural elucidation is often not possible, requiring cultivation of large amounts of biomass of cells residing in the proper physiological condition. This outcome can be particularly challenging, as some pheromone classes like the divinylcyclopropanes from brown algae are very labile with half-lives of only minutes in aqueous media (Pohnert and Boland, 2002). Thirdly, the chemical nature of algal pheromones is highly diverse and active structures may range from small non-polar hydrocarbons to high-molecular-weight polar glycoproteins. This structural diversity prevents the establishment of standard methods that can be readily applied to new, unstudied groups, therefore often necessitating the application of unselective extraction and tedious fractionation protocols (Prince and Pohnert, 2010). Final confirmation of the validity of structural suggestions requires often complex synthesis of the pheromones. Therefore, successful projects typically require a close collaboration between several laboratories from different fields of expertise.

By focusing on algae with elucidated or partially characterized pheromones this review mainly deals with green algae, diatoms, and brown algae. These topics will be treated in the following sections before a comparative overview will be introduced.

Sex Pheromones in Green Algae

Sexual reproduction is only known from relatively few species among the Chlorophyta, which together with the Streptophyta constitute the major groups of green plant-like organisms. Although chlorophytes generally have a haploid life cycle, several members of the Ulvophyceae have a diplontic life cycle with alternating free-living gametophyte and sporophyte phases. Sexual reproduction and associated pheromone signaling has been best studied in the volvocine green algae, which comprises both unicellular (Chlamydomonas reinhardtii) as well as colonial and multicellular species (Volvox carterii) (Hallmann, 2011).

Green algae often include sex in their life cycle to survive unfavorable environmental conditions. In Chlamydomonas reinhardtii, reduced nitrogen supply triggers the induction of gametogenesis (Starr et al., 1995). This induction includes the synthesis of sex-specific glycoproteins located on the flagella, called agglutinins (Adair, 1985; Snell, 1985; Musgrave and van den Ende, 1987; Coleman, 2012). When flagellae of two compatible gametes come by chance into contact, the gametes can pair (Starr et al., 1995) (Figure 1). This event initiates the production of an enzyme that results in the shedding of the cell wall and flagellar collar, ultimately leading to total gamete fusion and the formation of a quadriflagellar swimming cell that subsequently forms a non-motile zygote. Whereas gamete attraction appears to be lacking in C. reinhardtii and mating thus depends on successful random encounters between compatible cells, chemotactic behavior of gametes has been observed in other species of Chlamydomonas (Tsubo, 1961; Maier, 1993). The plastoquinone-related attraction pheromone lurlenic acid (Figure 2) is produced by the motile MT gametes of Chlamydomonas allensworthii (Jaenicke and Marner, 1995; Starr et al., 1995; Mori and Takanashi, 1996b) and attracts the motile MT+ gametes at concentrations as low as 1 pm. Other isolates of the same morphological species employ lurlenol, a derivative of lurlenic acid (Figure 2), as attraction pheromone acting at picomolar concentrations as well (Jaenicke and Starr, 1996; Mori and Takanashi, 1996a). Mating appears to be impossible between strains that utilize these differing pheromone variants. Phylogenetic comparison based on sequencing of the ITS 1 and 2 regions confirmed the existence of at least two cryptic lineages within this species, corresponding to the subdivision into two pheromone response types (Coleman et al., 2001).

Figure 1.

Mating in Chlamydomonas. (a) Schematic illustration of the sexual processes in Chlamydomonas. Nitrogen starvation induces gamete production from the vegetative cells. These possess sex-specific agglutinins on the flagella. Dependent on the species, gametes encounter either by chance or can be directed by the attraction pheromones of the lurlene family, secreted by the females. The agglutinins cause sexual adhesion of the gametes, which fuse, forming the zygote. (b) Three Chlamydomonas cells, adhered by flagella, the lower two just fused to form a zygote, the upper displaying an activated mating structure (arrow). Picture with permission of Goodenough et al. (2007).

Figure 2.

Structures of selected pheromones.

Another example of the formation of environmentally resistant structures as a product of pheromone chemistry can be found in Volvox. For instance, the heterothallic species Volvox carteri f. nagariensis lives in ponds that dry out during the summer. Only dormant zygotes can survive the drought (Hallmann et al., 1998). Heat shock causes the production of an inducer molecule by male clones that affects both male and female colonies (Kirk and Kirk, 1986). In the sub-groups Merrillosphaera and Janetosphaera, this pheromone inducer was identified as a large-molecular-weight glycoprotein acting at concentrations around 10−16 m, controlling the sexualization process by initiating gametogenesis (Starr, 1974; Sumper et al., 1993). This protein is at first produced by somatic cells (Kirk and Kirk, 1986) that induce the production of sperm cells in male and eggs in female individuals, respectively. Subsequently, sperm cells also produce this substance that now acts as a pheromone to synchronize all individuals of the population for the mating process (Figure 3). The sensitivity of the system is demonstrated impressively by the fact that one male sperm cell produces enough pheromone to convert all other individuals within one liter volume (Hallmann et al., 1998). The gene that encodes the pheromone could be found (Tschochner et al., 1987) and the corresponding protein contains 208 amino acids with a molecular mass of 22 kDa (Mages et al., 1988). Production through heterologous expression in yeast and mammalian cells and bioassays confirmed its function, although the native Volvox pheromone acted at lower concentrations (Haas and Sumper, 1991; Jaenicke et al., 1993). Related species appear to rely on similar pheromone chemistry as demonstrated by the finding that the Volvox carteri f. weismannia pheromone can sexually cross-induce Volvox carteri f. nagariensis males. Inducer-specificity differs in different Volvox carteri sub-groups with some of these reacting to all inducers from other species, over those that only react to some inducers up to completely specifically reacting strains that are only responsive to their own pheromone (Al-Hasani and Jaenicke, 1992).

Figure 3.

Sexual reproduction in Volvox. (a) Schematic illustration of the sexual processes in Volvox. Heat shock triggers the production of the sex-inducing pheromone in the somatic cells, which induce alterations of the extracellular matrix as well as sperm pack and egg production in the gonidia of male and female colonies, respectively. The sperm packs enter the female colony and released sperm fertilize the eggs, yielding the zygote. (b) Asexual Volvox colony, arrows show somatic cells. (c) Vegetative cell with chloroplast, arrow shows the colonial boundary of the extracellular matrix, arrowhead indicates a cellular envelope (chloroplast (ch), Golgi body (G) and pyrenoid (P)). (d) Sperm packet. (e) Male gamete bearing a cytoplasmic protrusion (arrowhead) near the base of the flagella. (f, g) Sperm packet (arrowheads) attaches to colony. (h) Sperm (arrow) attaching to lateral anterior portion of egg before fertilization. Micrographs with permission of Nozaki et al. (2006) (b–e), and Nozaki (1996) (f–h).

The sex-inducing pheromone also induces a remodeling of the hydroxyproline-rich extracellular matrix, which causes Volvox cells to stick to each other (Kirk et al., 1986; Sumper and Hallmann, 1998). Upon induction with the pheromone, additional proteins, called pherophorins, are produced with a similar structure compared with the sex-inducing pheromone itself. These chemicals might act as a mechanism for signal amplification (Sumper et al., 1993; Godl et al., 1995). Additionally, somatic cells start to incorporate sulfated glycoproteins in the matrix, which might collect or transport the positively charged sex-inducing pheromone to the gonidia (Wenzl and Sumper, 1986). Another effect of the pheromone is the induction of an extracellular glycoprotein that consists almost exclusively of hydroxyproline (Ender et al., 1999). Interestingly, the very same protein is induced by wounding of Volvox. Wounding, as well as pheromone induction, thus triggers the expression of the same genes, suggesting a molecular link between environmental stress, wound-healing and sexual reproduction (Amon et al., 1998).

Sexual induction has been also observed in other Volvox species, however different signal molecules are involved (Starr, 1974). V. capensis seems to rely on L-glutamic acid that induces development of gametes at concentrations as low as 68 nm (Starr, 1980). Other Volvox species do not respond to L-glutamic acid (Jaenicke, 1982), but seem to use other glycoproteins for gamete induction. However, glutamic acid is rather an exotic signal in this context as a survey of Volvox species reveals the prevalence of glycoproteins in sexual induction (Coleman, 2012). While the developmental and physiological processes are under the tight control of pheromones in Volvox, encounters between gametes appear to occur by chance and, in contrast with many other micro- and macroalgae, no attraction pheromones are produced. This case is for example in species of the heterothallic green algae Oedogonium. Chemoattraction of male gametes by a pheromone produced by the female oogonia containing the female gametes has been demonstrated (Machlis et al., 1974; Hill et al., 1989). The pheromone is a polar molecule with a molecular mass between 500–1500, but no further structural characterization is available.

Among the Streptophyta, sexual reproduction and pheromone chemistry has been extensively studied in the zygnemaphycean Closterium peracerosum-strigosum-littorale complex (Sekimoto et al., 2012). MT cells secrete a glycoprotein pheromone that induces another glycoprotein in MT+ gametes. Subsequently, this protoplast-release-inducing protein (PR-IP) induces the release of protoplasts in MT gametes prior to cell fusion (Sekimoto et al., 1990) (Figure 4). The amino acid sequence of the 18 kDa PR-IP inducer (Nojiri et al., 1995) resembles a sex pheromone with a similar function identified in Closterium ehrenbergii (Fukumoto et al., 1997, 2002). Functional similarity was further supported by the finding that heterologous expression of the pheromone of C. ehrenbergii showed biological activities in both species (Tsuchikane, 2005). An additional pheromone, an extracellular 20 kDa protein produced by MT of C. ehrenbergii induces the gametangial cell migrating activity in MT+ prior to pairing (Fukumoto et al., 1998). The production of this protein depends, once more, on environmental conditions and is enhanced in nitrogen-deficient medium.

Figure 4.

Mating in Closterium. (a) Schematic illustration of the action of protoplast-release-inducing protein (PR-IP) and the PR-IP inducer, which control gamete development in the Closterium peracerosum-strigosum-littorale complex. (b–h) Stages of sexual reproduction in Closterium. (b) Vegetative cells (arrow) and gametes (arrowhead). (c) Gametes are produced from vegetative cells by sexual cell division. (d) Pairing of gametes, which is mediated by a pheromone in C. ehrenbergii. (e, f) Protoplast release from MT+ and MT−- gametes. (g) Fusion of protoplasts released from MT+ and MT− gametes. (h) A zygote. Scale bar: 100 μm. Figures with permission of Imaizumi et al. (2007).

Gametogenesis in the green macroalga Ulva mutabilis is under the control of chemical factors. Blade cells of this alga excrete regulatory factors that are essential for the maintenance of the vegetative state. A cross-linked glycoprotein acts as sporulation inhibitor that suppresses the differentiation of blade cells into gametangia (Stratmann et al., 1996). Production of this factor gradually ceases during thallus maturation until a threshold inhibitory concentration below 10−14 m is reached. While this factor is released into the environment, a second low-molecular-weight sporulation inhibitor is located at the inner space between the two blade cell layers. This factor controls probably the spatial distribution of developing gametangia within the blade with highest concentration at the basal sectors (Stratmann et al., 1996; Wichard and Oertel, 2010). Early after induction of gametogenesis an additional factor, the so-called swarming inhibitor is produced. The highly biologically active compound with a mass of 292 atomic mass units inhibits gamete release from U. mutabilis and U. lactuca. This metabolite might additionally function as synchronization factor by acting as a cell-cycle regulator (Wichard and Oertel, 2010).

An additional factor that influences mating success is the phototaxis exhibited by gametes of many green and brown algae. It has, for example, been proven experimentally for Monostroma angicava that positive phototactic behavior of the gametes increases the rate of gametic encounters (Togashi et al., 1999).

Sex Pheromones in Heterokontophyta

Sex pheromones in brown algae

The first brown algal pheromone was identified by Müller et al. (1971) from Ectocarpus siliculosus. As in most brown algae, its haplodiplontic life cycle comprises male and female haploid gametophytes and diploid sporophytes, respectively. The diploid sporophyte produces zoospores through meiosis, which develop into the haploid gametophytes. The male and female gametophytes produce motile gametes, which fuse into the diploid zygotes that develop into the next sporophyte generation (Coelho et al., 2012). The first male-attracting substance was characterized as ectocarpene, a very apolar, fatty acid derived hydrocarbon (Müller et al., 1971). However, later re-investigation revealed that this compound is formed after thermal rearrangement of an initially released precursor (Figure 2) (Boland et al., 1995). Ectocarpene itself shows only a moderate activity at 10 mm compared with the thermolabile pheromone pre-ectocarpene, which is active at 5 pm concentrations (Boland et al., 1995). Interestingly, the inactivation process, which takes place with half lifes of several minutes, is entirely controlled by temperature, not requiring any additional enzymatic activity. It has been hypothesized that such a spontaneous inactivation prevents misguidance of male gametes to aged pheromone sources. Ectocarpene has been also found to function as a chemoattractant in other species of the same genus but also in other brown algae genera, including Sphacelaria rigidula and Adenocystis utricularis (Müller and Gassmann, 1980; Müller et al., 1985). Whether ectocarpene or its precursor is also the native pheromone in these species has not been addressed to date. Gamete recognition and fusion is presumably mediated by membrane-associated lectin–glycoprotein complexes (Schmid, 1993). In addition to the role of pheromones in brown algae gamete attraction, chemical compounds known as ‘release’ factors induce the release of gametes from the gametophyte. In Laminaria digitata, for example, an epoxidized hydrocarbon that is related structurally to ectocarpene, synchronizes the release of male gametes (Müller et al., 1979; Maier and Müller, 1982). A comprehensive overview pheromones in brown algal species can be found in Pohnert and Boland (2002).

Interestingly, ectocarpene and related compounds that are identical to the pheromones found in brown algae like hormosirene (Figure 2) have also been reported from the related heterokont diatoms. In fact, the investigation of the biosynthesis of this compound class was nearly entirely based on work on diatoms, as brown algal gametes were rarely available in sufficient quantities. While detailed mechanistic information on the formation of the hydrocarbon pheromones from fatty acids is available from the work with diatoms, the few studies on brown algae indicate striking similarities in biosynthetic pathways with identical precursor fatty acids and with lipoxygenases of identical positional specificity (Stratmann et al., 1992; Pohnert and Boland, 2002; Rui and Boland, 2010). Curiously, the function of these metabolites in diatoms was never connected to sexual reproduction, but was linked to chemical defense (Pohnert and Boland, 1996; Hay et al., 1998; Schnitzler et al., 2001).

Sex pheromones in diatoms

Diatoms are one of the most species-rich groups of algae, with an estimated 30 000–100 000 extant species (Mann and Vanormelingen, 2013) and contribute around 40% of the total oceanic primary production. In contrast with many other algal groups, diatoms have a diplontic life cycle. The hallmark of diatoms is their rigid silica cell wall that offers a mechanical defense against enemies (Hamm et al., 2003). Based on their morphology, diatoms are grouped in the radially symmetric centrics and the bilateral symmetric pennates. Similar to a Petri dish, the cell wall consists of two unequally-sized halves termed thecae, and a series of interlinking structures, the girdle bands. During cell division, each daughter cell inherits one parental theca and one of the two wall elements that are newly formed within the constraints of the parental cell. This process leads to a gradual reduction in the mean cell size of asexually reproducing diatom populations which ultimately leads to cell death unless cell size is restored via vegetative cell enlargement or – more typically – sexual reproduction (Round et al., 1990; Chepurnov and Mann, 2004). Sexual reproduction only occurs below a species-specific sexual size threshold (SST), which is often around 30–40% of the original size (William and Lewis, 1984). The zygote that results from the fusion of gametes expands into an auxospore, which upon germination forms initial cells that mitotically divide to produce a new generation of large vegetative cells (Figures 5 and 6).

Figure 5.

Life cycle and mating signaling in Seminavis robusta. Through asexual reproduction, diatoms reduce in cell size. Once cell size reaches the sexual size threshold (SST), the attracted mating type MT+ secretes a sex-inducing pheromone (green triangle), which initiates (a) the production of the chemoattractant diproline (green circle) in the attracting mating type MT. Additionally, MT secretes a different sex-inducing pheromone (red triangle), which induces (b) a putative diproline-receptor (red semi-circle) in MT+. After chemoattraction, the cells pair, form gametes through meiosis which fuse to yield zygotes developing to auxospores and finally cells with restored fully cell size. Microscopic figures with permission of: zygotes, Chepurnov et al. (2002); pairing cells, auxospores and initial cells, Chepurnov et al. (2008).

Figure 6.

Gametes of diatoms. (a, b) Centric Ditylum brightwellii (a) Antheridium with rudimentary thecae (arrow); (b) egg (arrow), sperm (arrowhead), and auxospore (double arrow). Figures with permission of Koester et al. (2007). (c–f) Pennate raphid Cylindrotheca closterium, production and fusion of gametes to zygotes. (c) Paired gametangia, (d) rouded gametes, (e) gamete fusion, (f) zygotes lying between the empty gametangia. Figures with permission of (Vanormelingen et al., 2013). (g, h) Pennate araphid Pseudostaurosira trainorii. (g) Male gamete with branched thread; (h) Amoeboid/pseudopodium-like structure of the male gamete (arrowhead indicates finer projection). Figures with permission of Sato et al. (2011).

The predominantly planktonic centric diatoms are oogamous and often have a homothallic mating system. Once the cell size drops below the SST, cells differentiate into gametangia, which respond to environmental cues to undergo meiosis and produce either egg cells or uniflagellate sperm (Figure 6) (Mann, 1993; Chepurnov and Mann, 2004; Kaczmarska et al., 2013). Environmental cues are diverse and may involve light, salinity, or even biotic interactions with bacteria (Nagai and Imai, 1998) but, in only a few cases, nutrient stress has been reported to trigger sexual reproduction (Chepurnov and Mann, 2004). Although it is thought generally that oogamy is associated with pheromone signaling to maximize reproductive success, no signaling compound is known. However, a Sexually Induced Gene 1 (Sig1) in Thalassiosira spp. was shown to have high divergence, both within and between species (Armbrust, 1999; Armbrust and Galindo, 2001). Although its function remains unknown, it is hypothesized that the polypeptides encoded by these genes might play a role in sperm–egg recognition.

In contrast with the centrics, most pennate diatom species have a benthic life style and show a wide diversity in reproductive characteristics, ranging from isogamy to extreme anisogamy, resembling oogamy and exhibit both homothallic and heterothallic mating systems (Chepurnov and Mann, 2004; Chepurnov et al., 2004; Quijano-Scheggia et al., 2009; Kaczmarska et al., 2013). Whereas environmental cues initiate sex in centric diatoms, a key feature of sexual reproduction in many pennate diatoms is their complex signaling mechanisms that involve the communication between diploid gametangial cells. Although numerous observations have suggested the involvement of chemoattractants during the mating process (discussed in (Chepurnov et al., 2004)), the first experimental evidence for the involvement of pheromones was provided only recently in the araphid pennate Pseudostaurosira trainorii (Sato et al., 2011). This heterothallic species shows extreme anisogamy and produces motile male gametes that can move by means of microtubule based structures called ‘threads’ to fertilize the non-motile female gamete. Behavioral observations clearly suggest that pheromones are involved in this process. A ‘female’ sex pheromone secreted by vegetative cells below the SST, induces meiosis and thus sexualization of male vegetative cells. Subsequently, these male gametangial cells and/or the gametes produced within them release a different pheromone, which stimulates sexualization of the females. Finally, when a male gamete comes (randomly) within close range of a female gamete, it becomes amoeboid and moves directly towards the female. Thus, a putative third pheromone could direct movement of males and serves as an attractant (Figure 5). Apparently, this complex signaling cascade in P. trainorii operates in a light-independent manner as successful fertilization also appears to take place in the dark.

The first diatom pheromone structure was identified recently in the heterothallic pennate Seminavis robusta (Gillard et al., 2013). As in P. trainorii, this diatom employs an elaborate, multi-stepped sequence of chemical signaling involved in the different phases of the auxosporulation process. However, a key difference is that in S. robusta meiosis and gamete formation only take place following successful pairing of the gametangia of both opposing sexes. Moreover, interaction between the gametangia is strictly light dependent. In cells below the SST an attracting mating type MT secretes a sex-inducing pheromone that induces a putative pheromone receptor in the attracted mating type MT+. MT+ itself also secretes a sex-inducing pheromone, which triggers MT to produce an attractant that finally guides MT+ to MT cells (Figure 5). The production of this pheromone is dependent on the cell cycle as, in dark-synchronized cultures (Gillard et al., 2008), its release in the medium starts after illumination but when cells are still in early G1 phase (Gillard et al., 2008, 2013). This attracting pheromone was identified as a cyclic dipeptide derived from two proline moieties (Figure 2). Structure elucidation of this diketopiperazine was enabled by a comparative metabolomics approach, which dramatically reduces the work load compared with traditional approaches for pheromone identification as bioassay-guided fractionation (Prince and Pohnert, 2010). The successful identification of the pheromone was achieved by a metabolomics-based comparison of extracellular metabolites produced by MT cells that were kept in the presence and absence of a sex-inducing pheromone. This sex-inducing pheromone was delivered in form of the crude medium of the opposite mating type. Signals in the metabolome that were up-regulated in the presence of the sex-inducer were identified as pheromone candidates. Indeed, the most significantly up-regulated peak could be attributed to the dipeptide pheromone. Providing that reliable cultures of mating systems are available, this emerging method generally has the potential to overcome a major bottleneck in pheromone identification and hence enable the discovery of further pheromones in all groups of microalgae. In mating cultures of S. robusta, the lowest detectable concentration of diproline was 1.5 nm. As mating also occurred when diproline was below the analytical detection limit, active concentrations must be in the picomolar ranges. The activity of synthetic diproline was in a comparable range (Gillard et al., 2013). Interestingly, the synthetic d-prolin-derived diprolin enantiomer showed comparable activity as the natural pheromone, thereby pointing towards unusual properties of the involved chemoreceptors, structures that usually only recognize specific enantiomers.

Regulation of Sexual Events

Although much information remains to be discovered, a few generalizations about the functional roles of pheromones in sexual reproduction of algae are possible. Generally processes related to sexual reproduction are highly synchronized by environmental cues or endogenous factors, or a combination of both. As gametes are often released into the environment this synchronization is essential to warrant gamete encounters and to guarantee sufficient mating success. In many cases in which environmental stress (imposed by e.g. reduced nutrient supply rates, increased salinity or heat) triggers sexual reproduction and the formation of resistant zygotes, sex not only has a role in generating new genetic diversity but also functions as a means to overcome adverse conditions. In other cases in which either no resistant zygotes are produced or when it is not clear to what extent environmental cues incur physiological stress to the alga, synchronization by external factors still contributes to maximizing reproductive efficiency.

A good example of this type of environmental cues is light (Agrawal, 2012), in terms of intensity, quality and photoperiod, as seen for example in diatoms (Chepurnov and Mann, 2004; Mouget et al., 2009), green algae (Starr, 1980; Fukumoto et al., 1998; Hoham et al., 2000) and brown algae (Lüning, 1981). In addition to an apparent overall control of the sexual process, light-dependent pheromone production can be very sophisticated through its control over cell-cycle progression as was shown for the pennate diatom S. robusta (Gillard et al., 2008, 2013). Sometimes it is not a single factor but rather a precise and subtle combination of abiotic conditions that determine the synchronization of sexual reproduction. Thus, for example, gamete release in fucoid brown algae such as Fucus ceranoides depends not only on light in semilunar cycles but additionally the tides and turbulence of the water control the production of sexual cells (Brawley, 1992; Pearson, 2006; Agrawal, 2012).

Pheromone signaling systems involved in sexual reproduction can vary greatly in complexity, both within and between algal groups. In the simplest systems, environmental cues directly induce the formation of gametes that engage in mating via random encounters (e.g. Coleman, 2012; Shim et al., 2012). One extreme case is the isogamous Chlamydomonas reinhardtii in which pheromones are absent altogether and fusion of gametes only occurs as a result of passive encounters induced by turbulence. Only gamete recognition prior to fusion is then controlled by species-specific compounds associated with the surface of gametes (Figure 1). In several, well documented cases, however, environmental cues lead to the production of pheromone inducers of sexual developmental programs in the other sex/mating type. In addition, for several groups and species, additional pheromones are produced that act as chemoattractants for motile gametes thereby enhancing encounter rates. In general, investment in attraction pheromones is mainly observed in algae with an anisogamous or oogamous sexual reproduction. The most sophisticated regulatory systems that control sexual reproduction known to date are found in the desmid genus Closterium and especially in pennate diatoms. In the latter, mating appears to be under a strong endogenous control as demonstrated by the highly complex and multi-step signaling during mating of the isogamous pennate diatom S. robusta. In addition to a strict cell size-controlled sexual maturation of cells, different pheromones are involved in synchronization and, in addition, attraction pheromones are observed. Furthermore, there is evidence for a significant density dependency of the mating process as well as light dependency, further contributing to synchronization of sexual events in the sexually mature portion of local diatom populations (Gillard et al., 2013). We hypothesize that this investment in potentially costly pheromone signaling is compensated by the avoidance of gamete loss.

It is clear from the above synthesis that the pheromone chemistry in algae is highly diverse. Pheromones are produced by several different pathways that include ribosomal protein production, fatty acid catabolism, terpenoid pathways and, most likely, the involvement of non-ribosomal peptide synthases. The resulting structural diversity is remarkable, and the molecular weight range extends from small hydrocarbons to large protein complexes. Surprisingly, also the polarity of signals employed in the water is highly diverse. While readily water-soluble polar amino acids and proteins have been identified, also highly apolar nearly water-immiscible hydrocarbons play a key role in signaling (Figure 2). But also medium polar plastoquinone derivatives and cyclic dipeptides were identified as pheromones. This situation requires a broad range of recognition sites of the involved receptors but to date no molecular details on algal pheromone receptors have been reported.

The most extensive inventory of pheromone structures is available in brown algae, which employ a common biosynthetic pathway based on the lipoxygenase-mediated transformation of fatty acids to generate a family of structurally diverse hydrocarbons. Nevertheless this family seems to be limited in the number of active pheromones. In fact, observations of different species and even genera employing the same pheromone have been reported (Pohnert and Boland, 2002). Apparently, there has been little selection for highly species-specific pheromone signaling during brown algae evolution, a finding that is surprising given their role as attraction pheromone produced by the female eggs to attract sperm. Clearly, this lack of specificity should be compensated by the presence of an efficient gamete recognition system and/or differences in habitat preference and species-specific sexual synchronization by, for example, light or tide dependency (Agrawal, 2012).

In green algae the structural diversity is significantly higher compared with brown algae, with examples of pheromones from different metabolic pathways, including proteins, amino acids and unidentified apolar small molecules (Wichard and Oertel, 2010). Structurally-related pheromones are, however, found within one genus as demonstrated by Closterium, in which homolog variants of the 18 kDa glycoproteinaceous sex pheromone are observed in separate species. This finding suggests the evolution of species-specific signaling from a single ancestral Closterium pheromone (Fukumoto et al., 2005). The existing degree of reproductive isolation in the Closterium peracerosum-strigosum-littorale complex can be explained partially by differences in molecular structures and physiological activities of sex pheromones (Tsuchikane et al., 2008). Conversely, highly specific pheromones have been reported from Chlamydomonas allensworthii that contribute to prezygotic reproductive barriers and the avoidance of interspecific hybridization.

One additional aspect of regulation is the influence of bacteria on algal sexual reproduction. In the simplest case, bacteria might be involved in pheromone degradation and thus prevent misleading by gametes/gamatangia by outdated pheromone signals in the environment, as might be the case in S. robusta. In other cases, their role seems to be more significant. In Oedogonium cardiacum, bacteria secrete a substance that promotes the development of gametangia (Machlis, 1973) and, in Coscinodiscus wailesii, sperm production seems to be influenced by bacteria (Nagai and Imai, 1998). In the multicellular green algae Ulva mutabilis (Chlorophyta), the post-gamete development depends on the presence of two different bacterial species. Without these species, the cells cannot differentiate and lack normal cell walls (Spoerner et al., 2012). However, a systematic investigation of the role of bacteria in sexual reproduction in protists has not been carried out to date.

Conclusion and Outlook

Strategies and signals involved in sexual reproduction of algae are highly diverse, as is to be expected given the diverse life cycles and very distant relationships between the groups discussed. It is apparent that our understanding of pheromones in algae sexual reproduction is still very strongly biased toward a few groups. For major algal groups, sometimes with great economic and ecological importance like, for example, the red algae, sex pheromones are unknown although a sex-specific lectin involved in gamete binding was identified recently (Shim et al., 2012). Other examples include the abundant and often toxic dinoflagellates, chrysophytes and coccolithophores, for whom much information is known about their complex life cycles but no pheromones were identified so far.

As outlined in the introduction several complicating factors have hampered the progress in algal pheromone research. But from the examples above it becomes obvious that progress in approaches and methods has been made that now opens the field for further exploration. This progress includes the establishment of stable culture systems with controllable sexuality that can be reliably manipulated. But also new models might emerge in the close future as the capacity for sexual reproduction is widespread and occurs in groups that were thought to be asexual (Blanc et al., 2010). The ability to dissect experimentally the sexual processes and to develop unequivocal bioassays to monitor the activity of (mixtures of) chemical compounds is, however, still a major challenge and has to be established fundamentally for each novel model under investigation. The advent of metabolomics approaches that overcome shortcomings of previous fractionation approaches will facilitate the further search for active structures.

The rapidly increasing number of strains and species for which complete genome sequences are becoming available are obviously preferred targets for dedicated studies of pheromone signaling. In addition, increased research on the life-cycle regulation of commercially grown algal species in the booming aquaculture sector is highly needed. Being able to control sexual reproduction experimentally could speed up breeding and selection of strains with desired traits and accelerate the domestication of algal strains enormously. In a micro-evolutionary context, population-level and phylogeographic studies can provide important data on the history and dynamics of sexual reproduction in natural populations (Casteleyn et al., 2010; Grimsley et al., 2010). These insights could be linked fruitfully to targeted pheromone studies to reveal the role of pheromone diversification in the evolution of (incipient) reproductive barriers. Furthermore, targeted analyses of the rapidly increasing metatranscriptomic and metagenomic resources from natural environments may help to reveal the nature of conditions that trigger sexual reproduction or life cycle transitions, while comparative genomics may allow the detection of gene families involved in sexual reproduction (Ning et al., 2013). Alternatively, identification of the sex-determining region in the genome of heterothallic species using mating type specific linkage maps, provides another avenue to elucidate the mating type dependent activation of signaling pathways (Vanstechelman et al., 2013). Once these pathways are known, we may begin to investigate to what extent they are conserved within and among algal groups and how they are linked with the mitotic and meiotic machinery of cells. Within this context, increased attention to pheromone research would thus be highly rewarding. In combination with the above approaches it can provide fundamental insights into the life cycle of algae and other protists and shed light on the role of key evolutionary processes, including endosymbiotic events and lateral gene transfer, speciation and adaptation at all phylogenetic levels.