Encysted embryos of the brine shrimp Artemia franciscana are able to survive prolonged periods of anoxia even when fully hydrated. During this time there is no metabolism, raising the question of how embryos tolerate spontaneous, hydrolytic DNA damage such as depurination. When incubated at 28°C and 40°C for several weeks, hydrated anoxic embryos were found to accumulate abasic sites in their DNA with k=5.8×10–11s–1 and 2.8×10–10s–1, respectively. In both cases this is about 3-fold slower than expected from published observations on purified DNA. However, purified calf thymus DNA incubated under similar anoxic conditions at pH 6.3, the intracellular pH of anoxic cysts, also depurinated more slowly than predicted(about 1.7-fold), suggesting that cysts may in fact accumulate abasic sites only slightly more slowly than purified DNA. Upon reoxygenation of cysts stored under N2 for 30 weeks at 28°C, the number of abasic sites per 104 bp DNA fell from 21.1±4.0 to 9.8±2.0 by 12 h and to 6.2±2.1 by 24 h. Larvae hatched after 48 h and 72 h had only 0.59±0.17 and 0.48±0.07 abasic sites per 104 bp,respectively, suggesting that repair of these lesions had largely taken place before hatching commenced. Thus, unlike bacterial spores, Artemiacysts appear to have no specific protective mechanism beyond what might be afforded by chromatin structure to limit spontaneous depurination, and rely on the repair of accumulated lesions during the period between reoxygenation and hatching.
Embryos of the brine shrimp Artemia franciscana are able to arrest development as gastrulae, form thick-walled cysts and enter a stage of diapause, which is characterized by low metabolic activity and high stress tolerance. Diapause can be broken by a number of environmental factors,including desiccation. Dry cysts display a remarkable resistance to ionising and non-ionising radiation, extremes of temperature and other environmental insults and high viability can be maintained for many years(Clegg and Conte, 1980). In this dehydrated state, damage to proteins, membranes and other macromolecular structures is prevented by the presence of a high level of trehalose and a number of stress proteins (MacRae,2003). This stress resistance is shared by a number of other anhydrobiotic animals, including rotifers and tardigrades; however, Artemia embryos have the almost unique ability to survive for extended periods in the fully hydrated state in the absence of molecular oxygen (Clegg, 1992; Dutrieu and Chrestia-Blanchine,1966; Stocco et al.,1972). Even after four years of continuous anoxia at 20–23°C, a hatch rate of at least 60% can be achieved when oxygen is restored. During this time, all metabolism appears to be at a complete but reversible standstill (Clegg,1997; Hontoria et al.,1994). This lack of metabolism means that any accumulated molecular damage, such as spontaneous molecular hydrolysis, could not be repaired by energy-requiring systems. This damage must either be prevented or repaired once oxygen is restored. When anoxic embryos are reoxygenated, the onset of hatching is delayed and the hatch rate reduced: the longer the period of anoxia, the greater the delay and the slower the rate. For example after four years of anoxia, hatching does not begin until about 120 h after reoxygenation, compared with the 16–20 h observed when dry cysts are rehydrated directly in oxygenated seawater(Clegg, 1997).
One form of molecular damage that would clearly have to be prevented or repaired before development could properly resume is DNA damage. DNA is known to undergo a number of spontaneous hydrolytic reactions, including depurination (and to a lesser extent depyrimidination) and cytosine and adenine deamination (Lindahl,1993). Abasic sites (AP sites, apurinic/apyrimidinic sites) arise through the hydrolysis of the N-glycosylic linkage between the bases and sugars in DNA and RNA and are both potentially mutagenic and lethal(Boiteux and Guillet, 2004; Lhomme et al., 1999; Yu et al., 2003). They have been estimated to occur in mammalian cells at 37°C and pH 7.4 at a rate of up to 10,000 per cell per generation(Lindahl and Nyberg, 1972). Using data from that paper on the temperature and pH-dependence of DNA depurination, an empirical rate constant for depurination of 8.4×10–11 s–1 can be estimated for Artemia DNA at 23°C and pH 6.3, the intracellular pH of anoxic cysts (Busa et al., 1982). With a genome size of 2.9×109 bp(Rheinsmith et al., 1974),this corresponds to 40,000 bases lost per day or 0.25% of the total genomic code per year if uncorrected. This rises to 0.5% and 2.5% at 28°C and 40°C, respectively, temperatures that could be experienced at least temporarily by cysts in their natural habitat. In an actively metabolising cell, AP sites arising spontaneously and by enzymatic removal of damaged bases are efficiently and continuously repaired by the process of excision repair. However, this requires energy in the form of ATP for ligation and dNTPs for base replacement, both of which would be quickly depleted in the absence of restorative metabolism. Thus, to prevent a catastrophic genetic loss, Artemia embryos must somehow severely restrict depurination or else allow AP sites and other hydrolytic damage to accumulate and then rely on efficient post-anoxia repair. To investigate these alternatives, we have measured the number of AP sites in DNA purified from anoxic cysts stored at 28°C and 40°C for periods up to 36 weeks and from larvae hatched from these cysts.
MATERIALS AND METHODS
Preparation of anoxic Artemia cysts and calf thymus DNA
Premium-grade encysted embryos of Great Salt Lake Artemia franciscana (Kellogg) were obtained from ZMSystems, Winchester, UK. Portions (1.5 g) of dry cysts were hydrated in greased, ground-glass stoppered tubes in 12 ml 0.4 mol l–1 NaCl that had previously been bubbled with O2-free N2 gas for 6 h. N2bubbling was continued during hydration for 4 h, then the tubes were stoppered and sealed with paraffin wax after carefully flushing the air space with N2. Tubes were then incubated at 28°C or 40°C.
Calf thymus DNA (Sigma Chemical Co., St Louis, MO, USA) was dissolved in TE buffer (10 mmol l–1 Tris-HCl pH 7.5, 1 mmol l–1 EDTA) and adjusted to 0.8 mg ml–1. It was then dialysed extensively at 4°C against 20 mmol l–1MES-KOH pH 6.3, 0.15 mol l–1 KCl, 1 mmol l–1EDTA and sodium azide finally added to 0.1%. Portions (5 ml) were transferred to greased, ground-glass stoppered tubes and bubbled with N2 for 8 h. The tubes were then sealed and incubated as above. Periodically, the tubes were sampled (250 μl) under N2.
Preparation of DNA from Artemia cysts and larvae
A tube of hydrated, anoxic cysts was shaken, allowed to settle and any floating material removed by aspiration. The remaining cysts were decapsulated as previously described (McLennan and Prescott, 1984) and finally collected by vacuum filtration. Each tube yielded six 0.5 g portions, which were frozen at –20°C. Each portion of frozen cysts was transferred to a pre-cooled mortar and liquid N2 added. After grinding the cysts to a fine powder, 5 ml DNAzol(Invitrogen, Carlsbad, CA, USA) was added and homogenization continued briefly. The contents were transferred to a tube, 100 μl proteinase K(Sigma Chemical Co., 10 mg ml–1 in water) added and the tube incubated with rolling at room temperature (RT) for 2–3 h.
After centrifugation (13,000 g, 5 min), 2.5 ml of 100%ethanol was added to the supernatant, the mixture shaken gently for 1 min and the DNA (heavily contaminated with orange lipid) removed by spooling into 2.8 ml 10 mmol l–1 Tris base. Once the DNA had dissolved, it was extracted three times with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1 saturated with TE buffer, pH 8.0). 1/10 volume 3 mol l–1 sodium acetate pH 5.2 was added to the final aqueous layer followed by 2.5 volumes of cold 100% ethanol. After gentle mixing, the DNA was allowed to precipitate at –20°C for at least 10 min. The DNA was washed twice in 70% ethanol, once in 100% ethanol, dissolved in 1 ml TE buffer and adjusted to 200 μg ml–1 with TE. The final yield was typically 0.4 mg DNA per 0.5 g portion of hydrated cysts.
To prepare DNA from larvae, anoxic cysts (1.5 g wet mass) were hatched and the swimming larvae separated from unhatched cysts and other material by attraction to a light source in a separator box(Persoone and Sorgeloos,1972). After collection by vacuum filtration through a small piece of cheesecloth, larvae were frozen in liquid N2 and weighed. DNA was then prepared as described above, with appropriate volume adjustments.
Preparation of depurinated DNA standards
Depurinated calf thymus DNA was prepared as previously described(Asaeda et al., 1998; Mohsin Ali et al., 2004). Briefly, RNAase A (Sigma Type II) was added to a 0.8 mg ml–1solution of calf thymus DNA (Sigma Chemical Co.) in TE buffer to final concentration of 100 μg ml–1 and incubated for 1 h at 37°C. Existing abasic sites were removed by addition of NaBH4to 100 mmol l–1 and incubation for 1 h at RT. DNA was then purified by extracting three times with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1 saturated with TE buffer, pH 8.0). The DNA was dialysed extensively at 4°C against 10 mmol l–1 sodium citrate pH 5.0, 100 mmol l–1 NaCl then heated at 70°C for various times up to 90 min. Samples were removed every 9 min, purified by ethanol precipitation and redissolved in TE buffer at 200 μg ml–1. DNA heated in this way for 9 min contains two AP sites per 104 bp, etc.(Lindahl and Nyberg, 1972; Mohsin Ali et al., 2004).
Assay for AP sites
AP sites were assayed using a modification of the aldehyde reactive probe(ARP) assay previously described (Asaeda et al., 1998; Mohsin Ali et al.,2004). This assay tags the free aldehyde group of AP sites with biotin and these are then detected with high sensitivity using peroxidase-conjugated streptavidin. ARP-DNA samples were prepared from cyst,anoxic calf thymus and depurinated calf thymus DNA by incubating 50 μl (10μg) DNA in TE buffer with 50 μl 10 mmol l–1 ARP(Dojindo, Rockville, MD, USA) for 2 h at 37°C. Unreacted ARP was removed by sequential dilution and concentration three times using Microcon 30 centrifugal concentrators (Millipore, Watford, UK). The final ARP-DNA was adjusted to 1 μg ml–1 with TE buffer.
ARP-DNA samples (200 μl) were added to the wells of a protamine-coated 96-well EIA plate (Bio-Rad, Hercules, CA, USA) and incubated at 37°C for at least 1 h (Asaeda et al.,1998). Excess DNA was removed by suction and the wells washed five times with 350 μl PBS–0.1% Tween (137 mmol l–1 NaCl,2.7 mmol l–1 KCl, 4.3 mmol l–1Na2HPO4 7H2O, 1.4 mmol l–1KH2PO4, 0.1% Tween 20). Wells were blocked with 350μl SuperBlock buffer (Thermo Fisher, Loughborough, UK) for 30 min at RT,then this was replaced by 100 μl of a 1:2000 dilution of streptavidin–peroxidase polymer (Sigma Chemical Co.) in SuperBlock buffer containing 0.5% Tween. After 30 min at RT, the conjugate was removed and the wells washed 10 times with 350 μl PBS–0.5% Tween as follows:4×1 min; 1×5 min; 4×1 min; 1×5 min, followed by one wash with 350 μl PBS–0.1% Tween.
Peroxidase substrate was freshly made by adding 20 μl of 35 mg ml–1o-phenylenediamine and 0.5 μl 30%H2O2 per ml of buffer (51 mmol l–1Na2HPO4, 24 mmol l–1 citric acid). 160μl of this was added to each well and incubated for 30 min at 28°C with occasional shaking. After adding 40 μl 4 mol l–1H2SO4, the absorbance was measured at 495 nm.
The number of AP sites was measured in DNA purified from cysts stored under N2 for up to 36 weeks at 28°C or up to seven weeks at 40°C. The number of AP sites per 104 bp DNA was estimated by comparison with heat-depurinated calf thymus DNA standards prepared under conditions where the level of depurination has previously been calculated(Lindahl and Nyberg, 1972; Mohsin Ali et al., 2004). Figs 1 and 2 show that AP sites did accumulate with time at both temperatures, with calculated rate constants for depurination of 5.8×10–11 s–1(28°C) and 2.8×10–10 s–1(40°C). In each case, the rate was 3-fold slower than the predicted rates of 1.7×10–10 s–1 (28°C) and 8.4×10–10 s–1 (40°C), suggesting that there might be some form of protection in the cyst. However, when samples of purified calf thymus DNA stored under N2 at physiological ionic strength and pH 6.3, the intracellular pH of anoxic cysts(Busa et al., 1982), were compared, they accumulated AP sites with k=9.4×10–11s–1 (28°C) and 5.2×10–10s–1 (40°C), 1.6–1.8-fold slower than predicted in both cases. Given the likely differences in the precise conditions employed here and in the study of Lindahl and Nyberg upon which the calculated rates are based (Lindahl and Nyberg,1972), the agreement between both is remarkably good for purified DNA. Thus, the cyst DNA may only be accumulating AP sites at a slightly lower rate (1.7-fold) than expected at both temperatures.
The % viability of anoxic embryos stored at 28°C (but not 40°C) was tested by measuring the hatch rate 64 h after reintroduction to oxygenated seawater. These were found to be 85, 83, 75 and 67% after 0, 8, 20 and 36 weeks anoxia. Although impressive, these rates are considerably lower than those reported by Clegg for San Francisco Bay cysts, which maintained almost full hatchability for two years (Clegg,1997). However, the cysts were not dried before hatching, which may improve the hatch by helping to break the anoxia-induced diapause, so the unhatched cysts may not be dead but still locked in diapause(Abatzopolous et al., 1994; Clegg, 1994). To find evidence that accumulated DNA damage is repaired before hatching, the number of AP sites was measured in samples of cysts and in larvae hatched from cysts stored under N2 for 30 weeks at 28°C followed by incubation in oxygenated seawater for up to 72 h. A marked reduction in AP sites per 104 bp from 21.1±4.0 to 9.8±2.0 was observed in cysts 12 h after reoxygenation, with a further reduction to 6.2±2.1 by 24 h,when 4% of the cysts had hatched (Fig. 3). However, as only 71% of the cysts hatched by 72 h, and this number did not increase significantly beyond this time, the level of DNA damage remaining in `24 h' cysts (30% of that in `0 h' cysts) may represent that specifically in dead and diapause-arrested cysts and so the damage in viable, hatchable cysts may have been fully repaired by this stage. Unfortunately, insufficient larvae had hatched by 24 h to test this directly;however, the level of AP sites per 104 bp present in larvae hatched after 48 and 72 h was much lower (0.59±0.17 and 0.48±0.07,respectively) and was comparable with that found in 48 h larvae hatched from cysts that had not been exposed to anoxia (0.60±0.23 per 104bp, N=6), indicating that full repair had taken place by 48 h.
These results clearly show that ametabolic, encysted Artemiaembryos accumulate AP sites in their DNA during storage in the hydrated state under anoxic conditions and that most, and possibly all, of this damage is repaired in the lag period between reoxygenation and hatching. It is likely that other forms of spontaneous DNA damage accumulate and that these and the probable oxidative damage induced by reoxygenation itself induce cell cycle arrest to permit post-anoxia repair before DNA synthesis can resume(Freiberg et al., 2006). The slightly lower than predicted rate (1.7- to 3-fold) may simply reflect a mild protective effect of chromatin structure. Although one measurement of AP site generation in live cells has yielded a rate very similar to that found for naked DNA of 9000 per day per generation(Nakamura et al., 1998), a 2.3-fold reduction in the rate of acid depurination of chromatin compared with DNA has also been reported (Duijndam and Vanduijn, 1975). Furthermore, although the intracellular pH of anoxic cysts has been measured at 6.3 (Busa et al., 1982), the value used to predict the rate of depurination,this may not truly represent the microenvironment of the chromatin. Thus, a small reduction in the expected rate of depurination cannot be interpreted as a specific protective effect.
In prokaryotes, DNA-binding proteins such as HU and the ferritin-like Dps proteins protect the bacterial DNA from radiation and oxidative damage(Nair and Finkel, 2004)whereas small, acid-soluble spore proteins are known to provide protection against wet heat and to reduce the rate of DNA depurination in Bacillus subtilis spores by a factor of 20(Setlow and Setlow, 1994; Setlow, 2007). We have found no evidence for a similar degree of protection in Artemia despite the existence of high levels of various diapause-specific proteins and chaperones in cysts that are believed to impart stress tolerance. The small heat shock/α-crystallin protein p26 translocates to the nucleus during anoxia and may stabilize proteins and nucleoprotein structures towards thermal and oxidative denaturation and aggregation(Clegg, 2007; Clegg et al., 1994; Collins and Clegg, 2004; Willsie and Clegg, 2001)whereas the ferritin homolog, artemin, has also been proposed as a protein and RNA protectant during diapause and encystment(Chen et al., 2007; Warner et al., 2004). Small heat shock proteins, ArHsp21 and ArHsp22 (nuclear)(Qiu and MacRae, 2008a; Qiu and MacRae, 2008b), and homologs of the plant anti-aggregation LEA proteins(Hand et al., 2007; Wang et al., 2007) have also recently been found in Artemia. However, none of these proteins has been reported to bind DNA. In Artemia and other eukaryotes, it is likely that histones are the major protectant from radiation and oxidative stress by simply acting as local competitors for photon and free radical attack (Enright et al., 1992)but they appear to be of limited value at excluding water from the DNA. By acting as water replacements and hydrogen bonding to the DNA, trehalose and glycerol might be expected to reduce hydrolytic base loss and the modest reduction seen could in part be due to these compounds. However, it appears that in fully hydrated cysts, this effect is equally limited. The increase in developmental defects seen in larvae after long periods of anoxia(Clegg, 1997) suggests that unrepaired lesions can have a detrimental effect, probably as a result of mutagenic transcription during the post-anoxia emergence period.
Dry cysts that have not been exposed to anoxic conditions also display a delay between rehydration and hatching, typically 16–20 h, known as pre-emergence development (PED). During this time there is continued transcription, translation and cellular differentiation but no DNA replication or cell division (McLean and Warner,1971; Tate and Marshall,1991). It is possible that hydrolytic DNA damage accumulates during the period of diapause when the cysts are arrested in a hydrated state before desiccation and that PED offers the opportunity to repair this before replication. The more prolonged pre-emergence period displayed by anoxic cysts would then reflect the increased amount of accumulated molecular damage. We have previously observed a loss of alkali-labile sites in DNA from normoxic cysts during PED, which is consistent with the repair of abasic lesions(Slater and McLennan, 1982). Studies are now under way to test this directly.
In conclusion, despite their extensive armory of factors dedicated to the prevention of molecular damage under conditions of extreme physiological stress, Artemia cysts do not appear to have mechanisms beyond those available to all eukaryotes to prevent spontaneous hydrolytic DNA damage. In the dehydrated state, water replacement by trehalose and glycerol will clearly limit this damage but when hydrated, compaction of the DNA into chromatin,which may exclude the possibility of DNA binding by more specific protective proteins, may be the only mechanism available. Therefore, developing embryos must rely on efficient pre-hatching systems to repair this damage before DNA replication can resume.
LIST OF ABBREVIATIONS