Diapause termination and development of encysted Artemia embryos: roles for nitric oxide and hydrogen peroxide

The brine shrimp Artemia avoids predation and competition by residing in high salinity habitats where they frequently experience drying, anoxia, food depletion and temperature fluctuation. To survive environmental stress these crustaceans undergo two different developmental pathways, with females consequently releasing either swimming larvae (nauplii) or encysted gastrulae (cysts) (MacRae, 2003). The larvae undergo several moults to reach adulthood and then reproduce, but cysts enter diapause, a physiological condition where development stops, metabolism is greatly reduced and stress tolerance is high (Drinkwater and Clegg, 1991; Clegg, 1997; MacRae, 2003; MacRae, 2005). Resistance to stress depends on the rigid, semi-permeable cyst shell (Anderson et al., 1970; Morris and Afzelius, 1967; Van Stappen, 1996), trehalose (Clegg and Jackson, 1998) and the accumulation of molecular chaperones such as p26, ArHsp21, ArHsp22 and artemin which prevent irreversible protein denaturation and inhibit apoptosis (Liang and MacRae, 1999; Villeneuve et al., 2006; Sun et al., 2006; Chen et al., 2007; Qiu and MacRae, 2008a; Qiu and MacRae, 2008b).

The structure and stress tolerance of Artemia cysts are relatively well characterized, with both contributing to diapause maintenance, but information on diapause induction and termination has been slower to emerge. Artemia embryos presumably enter diapause in response to a cue from the female but the signal and its origin are unknown. Several up-regulated genes have been identified in diapause-destined embryos at 2 days post-fertilization, and one of these encodes a homologue of the mammalian transcription cofactor p8, a protein with the potential to regulate cell growth, development, apoptosis and stress tolerance (Qiu et al., 2007; Qiu and MacRae, 2007). Diapause termination in Artemia has yet to be examined systematically at the molecular level although the proteome of A. sinica diapause cysts has been investigated (Zhou et al., 2008), as have changes in the proteome of post-diapause A. franciscana cysts (Wang et al., 2007). Exposure to specific environmental stimuli such as light, desiccation and cold promotes resumption of cyst development and metabolism, and these conditions are habitat specific with variation among cyst populations (Drinkwater and Crowe, 1987; Van Der Linden et al., 1988; Drinkwater and Clegg, 1991; Nambu et al., 2008). For example, A. franciscana from the San Francisco Bay, a highly variable environment, terminate diapause in response to several cues including either cold or drying whereas A. franciscana from the Great Salt Lake required both drying and cold, although there is contrary evidence regarding this latter point (Nambu et al., 2008). Artemia monica, found in Mono Lake, a large Alpine lake, terminates diapause only after a long cold period but not in response to drying. Additionally, several empirically developed techniques which are strain/batch dependent and of varying effectiveness have been employed to terminate Artemia cyst diapause, including dehydration/rehydration, freezing, cold storage, decapsulation and exposure to hydrogen peroxide (H₂O₂) (Van Stappen, 1996; Van Stappen et al., 1998). In this report the effect of nitric oxide (NO) and H₂O₂ on the development of Artemia embryos was investigated, confirming that H₂O₂ ends diapause and showing for the first time that both H₂O₂ and NO modulate cyst development. The results suggest cell/molecular mechanisms that control diapause termination and promote post-diapause development, poorly understood processes in many animal species. Furthermore, the work has implications in aquaculture because Artemia is used extensively in the diets of crustaceans and fish (Sorgeloos, 1980), as well as in forestry and agriculture, where the ability to undergo diapause increases the destructive potential of pest insects.

Encysted embryos (cysts) from a parthenogenetic strain of Artemia were harvested in Bolshoye Yarovoy, Russia, during 2005–2006 [Artemia Reference Center (ARC) code number BY 1706] (Baitchorov and Nagorskaja, 1999; Van Stappen et al., 2009). These cysts, which were received dry (water content 10.1%) and stored vacuum packed at 4°C, were used in this study as an alternative to the more commonly studied A. franciscana because they were readily available, slow to break dormancy during storage and yielded only 25% hatching when incubated in non-supplemented sea water in capped tubes. The tendency to remain in dormancy provided a constant supply of uniform cysts over a long time period and a broad range over which to examine breakage of dormancy because hatching of control cyst populations was low. Cysts, stored in the dark before use, were incubated at 26–28°C with constant illumination in 25 ml of sea water under varying experimental conditions as detailed below. Cyst development was the same with Instant Ocean® (Belcopet, Brugge, Belgium) artificial sea water and with 0.22 μm filtered sea water from the Northwest Arm, Halifax, NS, Canada, both at 32 g l⁻¹ salinity, and with mixing by rotation or on a reciprocating shaker (P>0.9).

Additionally, a population of A. franciscana Kellogg 1906 cysts approximately 85% in diapause was obtained from the Great Salt Lake in UT, USA, as a gift from Dr Brad Marden, and used to test the effects of H₂O₂ and NO on diapause termination (see below).

Tightly capped 50 ml plastic tubes containing 95.0±0.5 mg of Artemia cysts in Instant Ocean® artificial sea water supplemented separately with the NO donors 3-(2-hydroxy-2-nitroso-1-propylhydrazino)-1-propanamine (Papa NONOate; half-life 15 min), N-[4-[1-(3-aminopropyl)-2-hydroxy-2-nitrosohydrazino]butyl]-1,3-propanediamine (Spermine NONOate; half-life 39 min) and 3-morpholine syndnonimine (Sin-1 chloride; half-life 20 h), or with H₂O₂ (Sigma-Aldrich, St Louis, MO, USA) were incubated with rotation. To quantify larval emergence and hatching (Fig. 1) (Go et al., 1990; Rafiee et al., 1986), which served as measures of cyst development, 6 samples of 250 μl from each of three 50 ml tubes were mixed individually with 250 μl of sea water and 2 drops of Lugol's solution (Van Stappen, 1996) prior to counting with the aid of a dissecting microscope. Two drops of NaOCl [14% (technical) active chlorine] and NaOH (32% w/v) were then added to dissolve cyst shells, revealing non-hatched embryos for counting. The extent of development was determined as the percentage of developed cysts, which was calculated as either the number of hatched larvae or the number of hatched and emerged larvae, obtained from 100 cysts (represented by the sum of hatched larvae, emerged larvae and undeveloped cysts containing embryos).

To determine whether development was dependent on NO generation three capped 50 ml plastic tubes containing 95.0±0.5 mg of cysts in Instant Ocean® artificial sea water supplemented individually with 0.5 μmol l⁻¹ Papa NONOate, 6.0 μmol l⁻¹ Spermine NONOate and 6.0 μmol l⁻¹ Sin-1 chloride were incubated with rotation for 48 h after addition of the NO scavenger 2-phenyl-4,4,5,5,-tetramethylimidazoline-1-oxyl 3-oxide (PTIO) (Sigma-Aldrich) to 300 μmol l⁻¹. In related experiments, a volume of 0.88 mol l⁻¹ H₂O₂ sufficient to give a final concentration of 0.18 mmol l⁻¹ upon addition to culture tubes was exposed to bovine liver catalase (Sigma-Aldrich) at 0.0024 mg ml⁻¹ in 0.05 mol l⁻¹ potassium phosphate buffer, pH 7.0, for 60 min at 25°C. The H₂O₂–catalase mixture was then put in four capped 50 ml plastic tubes containing 95.0±0.5 mg of cysts in Northwest Arm sea water and these were incubated on a reciprocating shaker for 48 h. The percentage of developed cysts was calculated as described above.

To ascertain the duration of NO and H₂O₂ exposure required to achieve maximal development four capped 50 ml plastic tubes containing 95.0±0.5 mg of cysts in Northwest Arm sea water supplemented with either 0.5 μmol l⁻¹ Papa NONOate or 0.18 mmol l⁻¹ H₂O₂ were incubated on a reciprocating shaker for varying times. The cysts were then washed three times with sea water and incubated in non-supplemented sea water such that the combined incubation time in the presence and absence of either NO or H₂O₂ was 24 h. The percentage of developed cysts was calculated as described above.

In addition to experiments done in capped tubes, four Petri plates (8.5 cm diameter) containing 0.10±0.05 mg of cysts in non-supplemented Northwest Arm sea water or in sea water containing either 0.5 μmol l⁻¹ Papa NONOate or 0.18 mmol l⁻¹ H₂O₂ were incubated concurrently in the absence of agitation for 60 h with removal of hatched larvae every hour. The percentage of developed cysts was calculated by determining the absolute number of larvae that hatched from embryo-containing cysts in each plate. To obtain diapause cysts, Petri plates containing non-supplemented Northwest Arm sea water and 0.10±0.05mg of Artemia cysts were incubated for 7 days with hatched larvae removed periodically. Cysts remaining at the end of 7 days were harvested and incubated in stationary Petri plates containing Northwest Arm sea water supplemented with either Papa NONOate or H₂O₂ in varying concentrations. The percentage of developed cysts was calculated as described above.

The data in Fig. 2 were analysed by a non-linear regression used for the scrutiny of bell shaped concentration–response curves and which employed Prism version 5.00 for Windows (GraphPad Software, San Diego, CA, USA). Data are expressed as the mean ± standard error (s.e.) of three replicates and if smaller than the symbol s.e. is not shown. All curves were computer generated.

Fulfillment of the assumptions of single classification analysis of variance (ANOVA) for Figs 3, 4, 5 and 6 was verified prior to statistical analysis. Normality was tested using the D'Agostino–Pearson test and the homogeneity of variance was assessed with Bartlett's test. Data were reported as the mean percentage of developed cysts + s.e. for each experiment and were compared using single classification ANOVA (P<0.05) followed by Tukey's post-hoc test of multiple comparisons when a difference between groups was indicated. If the experimental data did not meet the assumption of normality, a Kruskal–Wallis non-parametric test of variance was completed (P<0.05). When differences between groups were demonstrated Dunn's post-hoc test was performed. All statistical analyses were executed with SYSTAT 10.0 (Statistical Product and Service Solutions, Chicago, IL, USA).

Incubation for 48 h with any of the three NO generators used in this study yielded almost equal amounts of hatched larvae with few emerged larvae remaining. However, Papa NONOate promoted development most effectively at 0.63 μmol l⁻¹ whereas 4.0 μmol l⁻¹ Spermine NONOate and Sin-1 chloride were required for maximum development of cysts into larvae (Fig. 2A–C). NO generators at concentrations of 40 or 60μmoll⁻¹ tended to reduce the overall extent of development achieved after 48 h, an effect most prominent with Papa NONOate (Fig. 2A–C). Moreover, after 24 h, Papa NONOate inhibited hatching almost completely at 40 and 60μmoll⁻¹, an effect overcome by 48 h, whereas the effects of Spermine NONOate and Sin-1 chloride at these concentrations were less dramatic, but obvious (Table 1). These were the only significant differences noted when the extent of hatching was compared at 24 and 48 h (P>0.1). By comparison, after 48 h of incubation, cyst development was promoted most effectively by 0.18 mmol l⁻¹ H₂O₂, whereas emergence and hatching were reduced at higher concentrations (Fig. 2D, Table 1). As shown for Papa NONOate, high concentrations of H₂O₂ inhibited hatching after 24 h, but unlike the situation with Papa NONOate, this was not reversed at 48 h. Maximum cyst development obtained with NO generators was approximately 64%, compared with 82% for H₂O₂. Addition of 300 μmol l⁻¹ PTIO to tubes containing sea water supplemented individually with 0.5μmoll⁻¹ Papa NONOate, 6.0μmoll⁻¹ Spermine NONOate and 6.0 μmol l⁻¹ Sin-1 chloride inhibited cyst development, as did incubation of H₂O₂ with 10 units of catalase prior to use (Fig. 3). Compared with controls the respective inhibition of NO generators and H₂O₂ imposed by PTIO and catalase was complete (P<0.001).

Incubation of cysts for 1 h in sea water containing 0.5 μmol l⁻¹ Papa NONOate followed by washing and incubation in non-supplemented sea water for 23 h promoted development compared with cysts not exposed to Papa NONOate (P<0.05) (Fig. 4A). However, to achieve cyst development equivalent to that obtained with continuous exposure to 0.5 μmol l⁻¹ Papa NONOate for 24 h required incubation with Papa NONOate for 4 h followed by 20 h in non-supplemented sea water (P>0.9) (Fig. 4A). Development was slightly greater for Papa NONOate exposures of 8 h as opposed to 4 h but the difference was not significant (P>0.6). After 4 h incubation in sea water containing 0.18 mmol l⁻¹ H₂O₂ followed by washing and incubation in non-supplemented sea water for 20 h cyst development was equivalent to that achieved by contact with 0.18 mmol l⁻¹ H₂O₂ for 24 h (P>0.9) (Fig. 4B). The earliest effect was apparent after a 0.5 h exposure to H₂O₂ followed by 23.5 h in non-supplemented sea water (P<0.05) (Fig. 4B). Incubation in H₂O₂ for 8 h before washing and incubation for an additional 16 h in non-supplemented sea water consistently reduced the extent of development compared with 4 or 16 h H₂O₂ exposures followed by incubation in non-supplemented sea water to a combined incubation time of 24 h (P<0.05) (Fig. 4B). The decline experienced upon an 8 h H₂O₂ exposure was reversed at 16 h. Regardless of the observed differences, NO and H₂O₂ both initiated developmental processes that continued maximally in the absence of either chemical after 4 h of exposure, although H₂O₂ effects were initiated more quickly.

Approximately 25% of cysts hatched after 48 h in capped tubes containing non-supplemented sea water (Fig. 2). By comparison, the extent of development in non-supplemented sea water in stationary Petri plates was higher than in capped tubes and very similar to that obtained when NO was present, either in capped tubes or in Petri plates (Figs 2 and 5). Moreover, cysts incubated for 4 h in plates with non-supplemented sea water and then transferred to tubes for an additional 20 h achieved the same level of hatching as cysts experiencing uninterrupted incubation in plates for 24 h. Cyst hatching was, however, greatest in both Petri plates and capped tubes when incubation was in the presence of H₂O₂ (Figs 2 and 5). These results reveal a cyst sub-population that developed upon exposure to H₂O₂ but not when incubated in non-supplemented sea water or with NO. In support of this proposal approximately 50% of cysts failed to develop when maintained at 26–28°C in Petri plates containing non-supplemented sea water for up to 7 days. These cysts, considered to be in diapause, were almost completely refractory to NO in Petri plates (Fig. 6A) (P>0.1), whereas approximately 50% hatched when exposed to 0.18 mmol l⁻¹ H₂O₂ for 24 h (Fig. 6B). Hatching was reduced at higher H₂O₂ concentrations, although it was not significantly different from the control (P>0.1), because development stalled at emergence, as was observed in capped tubes (Table 1). When the experiment was repeated in capped tubes 42.2% of the recovered (diapause) BY 1706 cysts hatched in the presence of H₂O₂ whereas none of the cysts hatched when exposed to NO. To further test the effects of H₂O₂ and NO on diapause termination a population of A. franciscana cysts approximately 85% in diapause was employed (see Materials and methods). When incubated with H₂O₂ in capped tubes hatching reached 75–80%, whereas in the presence of NO hatching was maximally 11.7%, only marginally higher than in non-supplemented sea water (Table 2).

Dehydration/rehydration, cold, freezing/thawing and light, alone or in combination, terminate Artemia embryo diapause and promote cyst development (Drinkwater and Crowe, 1987; Van Der Linden et al., 1988; Drinkwater and Clegg, 1991; Nambu et al., 2008), but the intracellular molecular changes driven by these physical factors are unknown. In addition, environmental chemicals modulate diapause termination and post-diapause development; however, access for Artemia embryos to most molecules is restricted by the cyst shell, a multi-layered chitinous structure (Anderson et al., 1970; Morris and Afzelius, 1967; Clegg, 1986; Clegg et al., 1996). Nonetheless, water and gases do penetrate the shell, and with this in mind the effect of NO and H₂O₂ on encysted Artemia embryos was examined. NO and H₂O₂, whose activities are often integrated (Bright et al., 2006; Bian et al., 2006; Zhang et al., 2007; Neill et al., 2008; Forman et al., 2008), were also chosen because they influence physiological and developmental processes in many organisms (Stone and Yang, 2006; Bright et al., 2006; Giorgio et al., 2007; Zhang et al., 2007; Covarrubias et al., 2008; Zhao and Shi, 2009). As one example, these compounds promote the germination of seeds (Neill et al., 2002a; Neill et al., 2002b; Neill et al., 2003; Bethke et al., 2004; Bethke et al., 2006; Hancock et al., 2006; Sarath et al., 2007; Oracz et al., 2007; Bailly et al., 2008), biological structures that share characteristics with Artemia cysts.

As demonstrated by their development in Petri plates, but not in capped tubes, the cysts examined in this study contained a sub-population of individuals in a state of dormancy termed quiescence. These cysts failed to hatch when incubated in tubes, where for an unknown reason development was either interrupted or failed to initiate even though diapause was broken. NO, a versatile gaseous free radical signalling molecule typically converted rapidly into NO₃⁻ and NO₂⁻ by nitrogen dioxide (Neill et al., 2003; Forman et al., 2008), promoted post-diapause development of quiescent Artemia embryos, a process also enhanced by H₂O₂, but it failed to terminate diapause. NO acted at much lower levels than H₂O₂, perhaps due to more efficient penetration of the cyst shell. For crustaceans NO has been studied mainly for its influence on neural plasticity/function, sensory activity, heart action and bacterial resistance (Scholz et al., 2002; Scholz et al., 1998; Christie et al., 2003; Yeh et al., 2006; Ott et al., 2007). This report shows, for the first time to the best of our knowledge, that NO promotes post-diapause development of a crustacean embryo.

NO may advance cyst development by acting as a reactive nitrogen species which drives the formation of NO–metallo linkages in haem-containing proteins (Villalobo, 2006; Forman et al., 2008), the reversible post-translational S-nitrosylation of proteins at cysteine thiol moieties (Bogdan, 2001; Ahern et al., 2002; Villalobo, 2006; Forman et al., 2008) and nitrotyrosine creation by processes either directly or indirectly mediated by H₂O₂ (Bian et al., 2006; Villalobo, 2006; Hancock et al., 2006; Forman et al., 2008). Additionally, NO sparks guanylate cyclase activity, increasing the second messenger cyclic 3′,5′-guanosine monophosphate (cGMP), a regulator of protein kinases, phosphatases and ion channels (Aherm et al., 2002; Kim et al., 2004; Eddy, 2005; Villalobo, 2006). Collectively, NO-driven protein changes may affect cell structure, metabolism and the expression of genes required to enhance post-diapause embryo development (Bogdan, 2001). Employing NO donors, as in this study, short-circuits the need for intracellular NO generation and/or substitutes for external sources of NO in terrestrial and aquatic niches (Bethke et al., 2004; Eddy, 2005).

H₂O₂ is a diffusible, ubiquitously distributed reactive oxygen species (ROS) which modifies intracellular redox potential and is readily formed and destroyed during aerobic metabolism. H₂O₂ was shown in this study to terminate Artemia cyst diapause, thus reflecting earlier work (Van Stappen et al., 1998), and to promote development of quiescent cysts. In order to affect development under normal circumstances H₂O₂ may be generated within Artemia cysts in response to external cues through the action of peroxidases or NADPH oxidase, the latter susceptible to regulation by Rho-like small G proteins sensitive to environmental signals (Neill et al., 2002b). H₂O₂, which modifies targets directly or by way of intermediate compounds (Winterbourn and Hampton, 2008), oxidizes thiol protein residues and functions as a second messenger via enhancement of tyrosine phosphorylation and dephosphorylation (Neill et al., 2002a; Neill et al., 2002b; Hancock et al., 2006; Bian et al., 2006; Forman et al., 2008). Mitogen-activated protein kinases (MAPKs) are stimulated by exogenous H₂O₂ and redox-controlled transcription factors have been identified (Stone and Yang, 2006; Hancock et al., 2006; Giorgio et al., 2007; Covarrubias et al., 2008). Thus, H₂O₂, by its oxidative properties and through its function as a second messenger, may reversibly modify proteins post-translationally. These modifications either activate or inhibit regulatory, metabolic and structural proteins, suggesting how H₂O₂ influences diapause termination and subsequent development in Artemia cysts.

When taken together the data indicate that H₂O₂ terminates diapause whereas NO does not, although both compounds promote post-diapause development. NO and H₂O₂ modify proteins post-translationally, perhaps at identical residues (Hancock et al., 2006), and they function as signalling molecules. Knowing this allows interpretation of experimental results and, as elaborated above, the generation of mechanistic models describing Artemia diapause termination and post-diapause development. For example, H₂O₂ or another ROS may terminate diapause while promoting NO production during post-diapause development. Such a proposal explains why H₂O₂ and NO, with the latter appearing not to terminate diapause and thus to function downstream of H₂O₂, both promote development of cysts, while clearly portraying H₂O₂ as the key to diapause termination and post-diapause development. Linear relationships between abscisic acid, H₂O₂ and NO, where NO activity depends on H₂O₂, regulate stomatal closure in Arabidopsis (Bright et al., 2006) and the activation of antioxidant defence in maize leaves, requiring up-regulation of MAPK and antioxidant enzymes (Zhang et al., 2007). Moreover, as based on their activities in other organisms, the inhibition of Artemia hatching by higher concentrations of Papa NONOate and H₂O₂ results from inappropriate post-translational protein modification. A parallel situation is observed in seeds where high levels of an NO generator delay germination and reduce root growth (Bethke et al., 2004; Bailly et al., 2008). Changes to essential proteins could have an immediate effect, hindering cyst development and interrupting emergence, the time when larvae become available to external molecules. Overcoming this inhibition may require the action of intracellular protective mechanisms, reflecting the results shown in Table 1.

To summarize, H₂O₂ terminates Artemia diapause whereas NO lacks this activity but was shown for the first time to influence post-diapause development in a crustacean embryo. Speculation concerning NO and H₂O₂ function provides conceptual frameworks upon which to build future investigations of diapause termination, work now in progress. Moreover, a better understanding of diapause termination has practical value because larvae derived from stored Artemia cysts are used commercially as feed in aquaculture. Diapause must be terminated prior to hatching and it may be possible to improve this process, an outcome with economic and social significance when the increasing worldwide importance of aquaculture is considered.

Financial support for this work included a Natural Sciences and Engineering Research Council of Canada Discovery Grant to T.H.M. The authors thank Mr Christ Mahieu for excellent technical assistance and Dr Brad Marden, Great Salt Lake Artemia, Mt Green, UT, USA, for the gift of diapause A. franciscana cysts.