Activation of p53 in anoxic freshwater crayfish, Faxonius virilis

Northern crayfish (Faxonius virilis) are a species of freshwater invertebrates native to North America, where they live at the rocky bottoms of streams (Taylor and Schuster, 2004; Taylor et al., 2007). High pollution and eutrophication in streams or stagnant pools cause severe oxygen depletion. Additionally, during winter, ponds and streams become ice-locked, limiting oxygen transfer and eventually causing decreased oxygen levels. Unlike other crustaceans that move to the surface or leave the water to breathe, F. virilis has a well-developed anoxia tolerance that enables these organisms to endure severe oxygen deprivation (Morris and Callaghan, 1998).

Several survival strategies both physical and molecular have been studied that are adopted by crayfish in response to oxygen depletion (Broughton et al., 2017; Gorr et al., 2010; McMahon, 2001). Studies have shown that under oxygen-deprived conditions, comparable species drop their metabolic rate by approximately 90% (Storey and Storey, 1990). Anoxia tolerance relies on a coordinated regulation between the activation of pro-survival and protective pathways and the suppression of non-essential functions for efficient use of the limited amount of energy available. Cell death pathways such as autophagy and apoptosis need to be prevented to extend viability. The importance of this over the period of months during which the animal is living in a hypometabolic and energy-restricted state can be appreciated. A critical decision on the maintenance or sacrifice of cells must be made for the survival of the organism.

The tumor suppressor protein p53 is a key transcription factor that regulates apoptosis, cell cycle checkpoints, cellular senescence and DNA repair, in response to stress stimuli such as DNA damage, nutrient deprivation and low-oxygen conditions, among others (Aubrey et al., 2018; Pitolli et al., 2019; Sionov and Haupt, 1999). The cellular mechanisms in response to hypoxia/anoxia have not been significantly investigated in invertebrates, but have been extensively studied in several vertebrate models under various stresses. Felix-Portillo et al. (2016) found the initiation of apoptosis during hypoxia in shrimps, independent of p53 and metallothionein. Nevertheless, in mice, p53 suppresses growth by regulating the transcription of numerous target genes. Under normal cellular conditions, p53 is ubiquitinated by the negative regulator E3 ubiquitin–protein ligase MDM2/MDM4 complex, and is maintained at low levels (Brooks and Gu, 2006; Wu and Prives, 2018). In response to stress signaling, p53 undergoes post-translational phosphorylation, and is released from the MDM2/MDM4 complex, freeing it for translocation into the nucleus (Fig. 1). In the nucleus, p53 binds to the DNA sequence in the promoter region of stress-specific genes and regulates their expression to promote either apoptosis or cell cycle arrest.

During DNA damage, amongst the multiple signaling cascades that are initiated, phosphorylation is a critical step for the stability of p53 (Fig. 1). Hyperproliferative signals activate p19-ARF, which blocks the nucleocytoplasmic shuttling of MDM2 and sequesters it to nucleoli (Weber et al., 1999). This inhibits the function of MDM2, hence preventing the degradation of p53. Activation and stabilization of p53 are further mediated under genotoxic stress by ataxia telangiectasia and Rad3-related protein (ATR) and ataxia telangiectasia mutated protein (ATM) via their substrate checkpoint kinase 1 (Chk1) and Chk2 (Pauklin et al., 2005). Chk1 and Chk2 phosphorylate serine residues S37 and S15 on p53, which prevents it from binding to MDM2, improving its function and stability (Murphy and Levine, 1998). This is considered the first step in the p53 activation process (Pauklin et al., 2005). Translocation of phosphorylated p52 (p-p53) to the nucleus regulates gene activity, contributing to cellular activities. Etoposide-induced 2.4 kb transcript (ei24) – also known as p53-induced gene 8 (PIG8) – is highly induced by p53 in response to DNA damage. Overexpression of this gene induces apoptosis and suppresses cell growth.

DNA damage is one of the risks for organisms exposed to low-oxygen conditions. Studies have shown that hypoxia induces a cascade of signaling reactions, particularly ATM- and ATR-dependent signaling pathway, and activates Chk2 to protect cells against apoptosis and DNA damage (Blackford and Jackson, 2017; Hammond et al., 2004). Additionally, hypoxia also induces DNA-dependent protein kinase activation that induces the repair of double stranded breaks through the non-homologous end joining pathway (Bouquet et al., 2011). Considering that p53 is activated in response to fluctuations in environmental oxygen levels and that it plays an important role in preventing DNA damage, studying the p53 signaling pathway and its activation via ARF/ATM/ATR kinases is important. The current study explored the regulation the p53 DNA damage response in the tail muscle and hepatopancreas of normoxic and anoxic crayfish.

The freshwater crayfish, Faxonius virilis (Hagen 1870), were bought from Britannia Bait (Ottawa, ON, Canada). The crayfish were acclimatized at 15°C for 7 days prior to experimentation in aerated freshwater. The water for the crayfish in the control group was aerated with oxygen using a bubbler, whereas the water for the anoxic condition group was bubbled with 100% nitrogen gas to deoxygenate the water and induce acute anoxia. For this experiment, oxygen levels in the anoxic water were less than 1% O₂/2 Torr following 1 h of nitrogen gas bubbling. After 1 h of aerated water treatment, the control crayfish were sampled, while the crayfish for anoxic conditions were sampled after 20 h of exposure to deoxygenated water. The animal was killed via rapid decapitation and all relevant organs were flash frozen in liquid nitrogen and stored at −80°C until further use.

Total protein extracts from hepatopancreas and tail muscle tissues for control and 20 h anoxic crayfish were prepared as previously described (Gupta and Storey, 2020). Briefly, the samples were crushed in liquid nitrogen and mixed with chilled homogenization buffer (20 mmol l⁻¹ Hepes, pH 7.5, 200 mmol l⁻¹ NaCl, 0.1 mmol l⁻¹ EDTA, 10 mmol l⁻¹ NaF, 1 mmol l⁻¹ Na₃VO₄ and 10 mmol l⁻¹ β-glycerophosphate), 1 µl of protease inhibitor cocktail (BioShop, Burlington, ON, Canada, cat. no. PIC001.1), and a few PMSF crystals. Samples were immediately homogenized using a polytron PT10 homogenizer (Kinematica) for about 20 s and later centrifuged for 15 min at 10,000 g at 4°C. The supernatant containing total proteins was collected and the pellet was discarded. The protein concentration of the collected samples were determined using the Coomassie method with Bio-Rad prepared reagent (BioRad Laboratories, Hercules, CA, USA, cat no. 500-0006). The samples were standardized to 10 µg µl⁻¹ using homogenization buffer. An equal volume of 2× sodium dodecyl sulphate (SDS) buffer (100 mmol l⁻¹ Tris-HCl, 20% v/v glycerol, 4% w/v SDS, 0.2% w/v Bromophenol Blue and 10% v/v β-mercaptoethanol) was added to make a working concentration of 5 µg µl⁻¹. The samples were boiled for 10 min and immediately chilled on ice for 10 min, then stored at −80°C until further use.

Frozen samples of hepatopancreas and tail muscle were weighed and ∼75 mg of each was placed in tubes pre-chilled in liquid nitrogen (n=4 biological replicates). Samples were homogenized 1:5 w/v in cytoplasmic homogenization buffer A (10 mmol l⁻¹ Hepes, pH 7.9, 10 mmol l⁻¹ KCl, 10 mmol l⁻¹ EDTA and 20 mmol l⁻¹ β-glycerophosphate) with 10 μl of 100 mmol l⁻¹ dithiothreitol (DTT) and 10 μl of Protease Inhibitor Cocktail (BioShop, cat. no. P1Coo1.1) added per ml. Homogenization was performed using a Dounce homogenizer. Samples were incubated on ice for 25 min and then centrifuged at 12,000 g for 15 min at 4°C. The supernatant was transferred to new pre-chilled tubes and stored as the cytoplasmic fraction.

The pellets were re-suspended 1:5 w/v (based on original sample weights) in cytoplasmic homogenization buffer B (100 mmol l⁻¹ Hepes, 2 mol l⁻¹ NaCl, 5 mmol l⁻¹ EDTA, 50% v/v glycerol, 100 mmol l⁻¹ β-glycerophosphate, pH 7.9, and 100 mmol l⁻¹ DTT) and protease inhibitor cocktail at 1:1000. The samples were sonicated using a Polytron PT1000 homogenizer (Brinkmann Instruments, Rexdale, ON, Canada) at 50% of full power for 5 s and then incubated on ice for 10 min followed by centrifugation at 14,000 g for 15 min at 4°C. The supernatant obtained was transferred to new pre-chilled tubes and stored as the nuclear fraction.

Protein concentrations in both cytoplasmic and nuclear supernatants were measured and standardized to 5 µg µl⁻¹ following the aforementioned procedure. Protein sample integrity and purity were checked by running both fractions on SDS-PAGE gels each with equal concentrations of cytoplasmic and corresponding nuclear fractions. The proteins were transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, Etobicoke, ON, Canada, cat. no. IPVH07850, 45 μm pore). The membranes were probed with anti-histone H3 (Cell Signaling, Dancers, MA, USA, cat. no. 9715) and alpha-tubulin antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Histone H3 served as the nuclear marker and alpha-tubulin as the cytoplasmic marker.

The samples were run as previously described (Gupta and Storey, 2020). Briefly, 10%, 12% or 15% SDS PAGE resolving gels, depending on the molecular weight of the protein of interest, with 5% stacking gels were used. PiNK Plus Prestained Protein Ladder (Froggabio, cat no. PM005-0500) was used as the ladder in the first lane and 20 µg of each of the sample protein (aerobic control condition and 20 h anoxic condition) was loaded in the following lanes. A positive control sample from a mammalian species (Ictidomys tridecemlineatus) was run alongside the crayfish samples to confirm the appropriate band on interest. The gels were run at 180 V. Proteins were transferred to PVDF membrane by electroblotting in 1× transfer buffer (25 mmol l⁻¹ Tris pH 8.5, 192 mmol l⁻¹ glycine, 20% methanol) at 160 A for 90 min. Blocking was done using 3% milk w/v in TBST (20 mmol l⁻¹ Tris base, pH 7.6, 140 mmol l⁻¹ NaCl, 0.05% v/v Tween-20) for 30 min. Membranes were washed and incubated overnight at 4°C with protein-specific polyclonal primary antibody. The antibodies for p53 (cat. no. 2524), phospho-p53 (Ser15) (cat. no. 9284), phospho-p53 (S37) (cat. no. 9289) and p19-ARF (cat. no. 77184) were purchased from Cell Signaling (Danvers, MA, USA). The antibodies for ATM (cat. no. A01222) and ATR (cat. no. A01253) were purchased from Genscript (Piscataway, NJ, USA), and the MDM2 antibody (cat. no. SC-7918) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). As F. virilis has not yet been sequenced, protein sequence alignments were performed with several species (Homo sapiens, Mus musculus, I. tridecemlineatus, Xenopus laevis, Daphnia pulex, Penaeus vannamei and Drosophila melanogaster) to ensure that the antibody epitope was conserved amongst different species. Next, the membranes were washed 3×5 min in TBST and incubated with primary antibody-specific HRP-linked secondary antibody for 30 min. The secondary antibody used in the current study was anti-rabbit IgG secondary antibody conjugated with horseradish peroxidase (cat. no. APA007P.2, BioShop). Membranes were visualized using hydrogen peroxide and luminol.

The total proteins were extracted from frozen tissues of tail muscle and hepatopancreas (control and 20 h anoxia) as previously described for use in a transcription factor binding ELISA (TF ELISA) (Gupta and Storey, 2020). Briefly, a pre-chilled lysis buffer cocktail and protease inhibitor (BioShop, cat. no. PIC001) were mixed with the frozen tissue samples and homogenized using a homogenizer. The samples were then incubated on ice for 30 min and then centrifuged at 14,000 g for 20 min at 4°C. The supernatant was collected and stored at −80°C until further use. The protein concentration as measured using the Bio-Rad assay as mentioned above and protein integrity was checked by running samples on SDS-PAGE gels.

Biotin labelled oligonucleotides were used for p53. The sequence of the biotin-conjugated probe was 5′-Biotin-TACCCGGGCATGTCTAAGCATGCTG-3′, and the complementary sequence was 5′-CAGCATGCTTAGACATGCCCGGGTA-3′. Considering that p53 is a highly conserved protein, DNA–protein binding ELISAs and EMSAs conducted on other animals were referred to find the DNA binding site of p53 as used in the current experiment. A standardized protocol from previous work was followed (Gupta and Storey, 2020). Probes (biotinylated and complementary probes) were reconstituted in double distilled water to a final concentration of 500 pmol μl⁻¹. Forward and the reverse probes were mixed in a 1:1 ratio at 95°C for 10 min in a thermocycler. The double-stranded probes were diluted 10× to obtain a working concentration of 50 pmol μl⁻¹ using PBS (10 mmol l⁻¹ Hepes, pH 7.9, 50 mmol l⁻¹ KCl, 0.5 mmol l⁻¹ EDTA, 3 mmol l⁻¹ MgCl₂, 10% v/v glycerol, 0.5 mg ml⁻¹ BSA and 0.05% NP-40); 50 μl aliquots containing 40 pmol of diluted probe were added to each well, which had been pre-coated with streptavidin (R&D Systems, Minneapolis, MB, Canada, cat. no. CA73521-134). The plate was incubated for 1 h at room temperature of 20–22°C. Aliquots of protein extracts containing 20 μg of protein were combined with 50 μl of 1× protein binding buffer [PBS, 1 μg salmon sperm DNA (BioShop), 0.5 mmol l⁻¹ DTT and 40 mmol l⁻¹ NaCl]. The plate was incubated at RT with mild agitation for 75 min and then washed 4 times with wash buffer. Aliquots of 60 μl of the primary antibody for phospho-p53 (S15) (Cell Signaling, cat. no. 9284) were diluted 1:250 in PBS, then added to the wells, and the plate was incubated for 3 h. The plate was then washed as above and 60 μl of anti-rabbit IgG-HRP (Bio-Shop, cat. no. APA007P.2) diluted 1:2000 in PBS was added and incubated for 1 h. Following this, the plate was washed as above and 60 μl of tetramethylbenzidine (TMB; BioShop, cat no. TMB333.100) was added. Once color had developed, the reaction was stopped by the addition of 1 mol l⁻¹ HCl and optical density was read at 450 nm with a reference wavelength of 655 nm using a Multiskan Spectrum (Thermo Electron Corporation, Waltham, MA, USA).

RNA was isolated from frozen samples of hepatopancreas and tail muscle tissues for control and 20 h anoxia-exposed crayfish; 50 mg of the samples was homogenized in 1 ml Trizol (BioShop, cat. no. TRI118.100) using a polytron PT10 homogenizer. cDNA synthesis was carried out on the samples using the protocol outlined in our previous study (Gupta and Storey, 2021a).

Forward and reverse primers for ei24 were designed from conserved regions based on DNA alignment of the sequences from both vertebrate and invertebrate species for use in quantitative reverse transcription PCR (RT-qPCR). Primers were designed using Primer Blast by the National Center for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov/tools/primer-blast/). Primers used for ei24 and tubulin (control) were as follows: tubulin forward 5′ GACATTTGTTCATGGTATGC 3′, reverse 5′ TGTCGGTGGAATAGGGAAAG 3′; ei24 forward 5′ TATAGCATGTGCACTGTGCC 3′, reverse 5′ GTCTGTACTCCGGGAACAAC 3′. The PCR program consisted of 7 min at 94°C, followed by 35 cycles of 1 min at 94°C, 1 min at 58°C for ei24 and 53°C for tubulin, and 1.5 min at 72°C. The final step was 72°C for 10 min. PCR products were separated on a 1% agarose gel stained with ethidium bromide, visualized using the ChemiGenius imaging system (Syngene, Frederick, MD, USA) under UV light and quantified using the GeneTools program. The bands from the most dilute cDNA sample that gave visible product were used for quantification to ensure that the products had not reached amplification saturation. PCR products were sequenced by DNA Landmarks (Saint-Jean-sur-Richelieu, QC, Canada) and sequences were verified as encoding the correct genes using the program BLASTN (http://www.ncbi.nlm.nih.gov/blast).

Immunoblot protein bands were quantified by densitometric analysis using ChemiGenius BioImaging System and the associated GeneTools software (Syngene). To control for irregularities in SDS-PAGE gel loading, band intensity in each lane was normalized against the corresponding density of Coomassie-stained protein bands in the same lane. Immunoblot data for each condition are expressed as means±s.e.m., with n=4 samples from different animals. Statistical analysis used one-way ANOVA, with Dunnett's post hoc test, using the RBioPlot software, with P<0.05 accepted as significant (Zhang and Storey, 2016).

Relative protein levels of p53 were detected using immunoblotting and a single band around 53 kDa was observed in both hepatopancreas and tail muscle. The results showed no change in the protein levels of p53 in both hepatopancreas and tail muscle under anoxic conditions when compared with control (Fig. 2). The levels of Ser37 phosphorylated p53 (p-p53 S37) were measured using a polyclonal anti-p-p53 S37-specific antibody. It was observed that levels in the hepatopancreas significantly increased by 1.4(±0.05)-fold in anoxic conditions and by 1.71(±0.12)-fold in anoxic tail muscle relative to control levels (Fig. 2).

Changes in the levels of ATM and ATR were observed in response to anoxia. Relative levels of ATM increased significantly during anoxia in both hepatopancreas and tail muscle by 3.07(±0.31)-fold and 1.33(±0.02)-fold, respectively, relative to control (Fig. 2). Similarly, a significant increase was observed in protein levels of ATR during anoxia in both tissues; in the hepatopancreas, levels increased by 2.37(±0.19)-fold relative to control, and in tail muscle, an increase of 1.49(±0.14)-fold relative to control was observed (Fig. 2).

In response to low oxygen levels, p19-ARF acts as an activator and MDM2 acts as an inhibitor to p53. Therefore, relative protein levels of both activator and inhibitor were measured. Under anoxic conditions, there was a significant increase of 1.95(±0.25)-fold in p19-ARF expression relative to control in the hepatopancreas, and a 1.22(±0.02)-fold increase relative to control in tail muscle (Fig. 2). The level of MDM2 decreased significantly during anoxia to 57.2±9.2% compared with control in hepatopancreas, and 59.4±6.4% compared with control in tail muscle (Fig. 2).

Upon phosphorylation at Ser15, p53 detaches from MDM2 and translocates to the nucleus, resulting in stress-specific transcriptional activation of numerous genes. Therefore, along with the total cellular protein expression changes in p-p53 S15, the protein levels in cytoplasmic and nuclear fractions were also assessed. The phosphorylation profile in hepatopancreas total protein levels showed a significant increase in p-p53 S15 by 1.95(±0.26)-fold during anoxia relative to control values, whereas no significant change was observed in anoxic tail muscle (Fig. 2). Cytoplasmic levels of p-p53 S15 showed a significant increase during anoxia in both tissues relative to control – 2.15(±0.11)-fold in hepatopancreas and 1.50(±0.10)-fold in tail muscle – but no significant change was observed in nuclear protein expression for either tissue (Fig. 3).

In response to anoxia, changes in the DNA binding activity of p-p53 S15 were assessed using a TF ELISA. The results showed a significant decrease in the binding levels of p-p53 S15 during anoxia to 38±6% compared with control levels in the hepatopancreas, whereas a significant increase of 2.23(±0.23)-fold relative to control was observed in anoxic tail muscle (Fig. 4).

ei24 is one of the downstream targets of p53 that is linked with apoptosis. Changes in the transcript levels of ei24 in response to anoxia were measured using RT-qPCR. No significant change was observed in the mRNA levels of the gene in anoxic hepatopancreas, while ei24 transcript levels increased significantly by 1.80(±0.02)-fold relative to control in anoxic tail muscle (Fig. 5).

Many organisms encounter environmental stress in their natural habitats brought on by seasonal temperature changes (desiccation in high temperatures, freezing in low temperatures), changes in oxygen levels (hypoxia/anoxia), or a drop in nutrient availability (caloric restriction) (Storey, 2015; Storey and Storey, 1990). Many organisms adopt a similar strategy to survive hypoxic and anoxic conditions in their regular stages of life, including turtles (Breedon et al., 2021; Gupta et al., 2021b), brine shrimp (Storey and Storey, 2011), crayfish (English et al., 2018; Green and Storey, 2016) and wood frogs (Gupta and Storey, 2021b, 2022; Gupta et al., 2020; Wu et al., 2018), among many others (Gupta et al., 2021a; Krivoruchko and Storey, 2010; Tessier et al., 2017). Many stress-tolerant organisms undergo metabolic rate depression (MRD), entering a hypometabolic state and reprioritizing their energy expenditure to favor pro-survival pathways (Gupta et al., 2021b; Storey, 1998, 2004; Storey and Storey, 2005), including DNA damage repair. The northern crayfish is a good model for studying hypometabolism during anoxic conditions, allowing us to examine how they regulate p53 to circumvent apoptosis and promote DNA repair to sustain long-term viability under anoxic conditions.

ATM and ATR kinases respond to DNA breaks at different checkpoints via different stimuli; double-stranded breaks in DNA trigger ATM (Bakkenist and Kastan, 2003), whereas UV damage/stalled replication forks trigger ATR (Zou and Elledge, 2003). Interestingly, both proteins are activated during hypoxia and reoxygenation (Fallone et al., 2013; Hammond and Giaccia, 2004; Olcina et al., 2014). ATR is activated as a result of S-phase arrest during hypoxia (Fallone et al., 2013), while ATM can be activated via non-DNA damaging signals during hypoxia and by increased reactive oxygen species (ROS) during reoxygenation (Olcina et al., 2014). Therefore, we would expect to see increased expression of both proteins during anoxia, which was indeed observed, with a significant increase in both ATM and ATR in anoxic hepatopancreas and tail muscle relative to control values (Fig. 2), confirming the oxygen sensitivity of crayfish.

p53 is a nuclear phosphoprotein that responds to environmental stress, specifically low oxygen levels and genomic damage, under tight regulation. As such, the relative expression of total cellular protein was analyzed. Interestingly, no significant change was observed during anoxia in either the hepatopancreas or tail muscle (Fig. 2). A likely explanation for this observation is that p53 undergoes significant post-translational modifications in response to stress signaling/genomic damage, without significantly changing its overall cellular concentration, corresponding to our findings of increased ATM and ATR protein expression. Studies have reported that p53 is activated by post-translational modifications, including phosphorylation, acetylation and glycosylation, in a stress-specific manner (Ko and Prives, 1996). Upon sensing DNA damage, p53 is phosphorylated at serine residues 15, 20, 33 and 37 in the amino-terminal domain, which plays an essential role in p53 activity and stability (Ashcroft and Vousden, 1999). Several kinases including ATM and ATR have been reported to phosphorylate p53 in a stress-specific manner (Lakin and Jackson, 1999). A study showed that even though ATR phosphorylates p53 at S15 and S37, ATM does not play any role in modification at S37 (Tibbetts et al., 1999). Hence, this study focused on phosphorylation of p53 at S15 and S37. Phosphorylation of H2A histone family member X (H2AX) is one of the initial responses to DNA damage and repair. H2AX is phosphorylated at Ser129 to become γH2AX (Mah et al., 2010). Phosphorylation levels act as a very specific and sensitive molecular marker to detect the extent of DNA damage (Mah et al., 2010). Further research on γH2AX levels in crayfish in response to anoxia is warranted to study how crayfish minimize DNA damage during anoxia.

Phosphorylation of p53 at S15 and S37 sites stimulates transcriptional activity during cellular stress (Dohoney et al., 2004). ATM and ATR can phosphorylate p53 at S15 to enhance its transactivation activity (Banin et al., 1998). Interestingly, under anoxic conditions, p53 was activated by S15 phosphorylation in hepatopancreas, with a significant increase in p-p53 S15 during anoxic conditions compared with control, whereas no significant change was observed in the levels of p-p53 S15 in tail muscle (Fig. 2). Additionally, relative protein levels of p-p53 S15 increased significantly in the cytoplasmic fraction of both the hepatopancreas and tail muscle during anoxia. Although an increasing trend was observed in the relative levels of p-p53 S15 in the nuclear fraction of anoxic tail muscle, the change was not significant (Fig. 3).

Enhanced cytoplasmic expression of p53 during anoxia observed in this current study in both hepatopancreas and tail muscle suggests an important role in the cytoplasm (Fig. 3). This finding is supported by studies which suggest that the build-up of p53 in the cytoplasm is essential for triggering many extracellular activities (Chipuk et al., 2004; Mihara et al., 2003). Furthermore, its abundance in cytoplasm inhibits autophagy and triggers intrinsic apoptotic machinery (Green and Kroemer, 2009). A study by Sun et al. (2016) in river prawn highlighted the same results, showing the presence of high levels of p53 in hepatopancreas compared with other tissues (Sun et al., 2016). Additionally, they observed a buildup of oxidative stress in hypoxic prawns (Sun et al., 2016). Anoxic conditions activate pro-apoptotic Bcl2 proteins, a downstream target of p53, and caspases to initiate apoptosis (Fridman and Lowe, 2003). Therefore, activation of apoptosis and inhibition of autophagy, as suggested previously, is an organized way to clear up cells damaged as a result of anoxia without harming the surrounding cells. Being an ATP-dependent process, the accumulation of p53 to initiate apoptosis could also be a preparatory mechanism for recovery (Eguchi et al., 1997). This is because, during anoxic conditions, cells are in a hypometabolic state with limited energy availability, so processes such as apoptosis could be halted until aerobic recovery.

The DNA binding activity of p-p53 S15 showed a relative increase in the level of binding in anoxic tail muscle when compared with control (Fig. 4). Therefore, increased binding of transcription factor p-p53 S15 to the promoter region of the DNA could result in increased transcript levels of genes under the regulation of p-p53 S15 in the nucleus during anoxia. The increase in the relative transcript levels of ei24 in anoxic tail muscle corresponded with the results obtained on the DNA binding ability of p-p53 S15 (Fig. 5). The tumor suppressor gene ei24 is a downstream target of p53 with a pro-apoptotic function to suppress cellular growth (Sung et al., 2013). This supports the presence of transcriptionally active p53 in the cells. ei24 negatively regulates cell growth and induces apoptosis (Gentile et al., 2001; Gu et al., 2000). The increased expression of ei24 is likely either in preparation for increased apoptosis during aerobic recovery, when ROS production would be significantly elevated as a result of the sudden influx of oxygen, or for the induction of apoptosis during anoxic conditions as a protective mechanism to prevent the spread of cytotoxins resulting from autophagy or necrosis. Future research on oxidative stress in anoxic crayfish could provide better insight into the interplay of p53 activation, apoptosis and autophagy.

Relative protein expression of p-p53 S37 showed a significant increase in both anoxic hepatopancreas and tail muscles (Fig. 2). Phosphorylation prevents the binding of p53 to MDM2, which allows accumulation and activation of p53 during DNA damage (Shieh et al., 1997; Tibbetts et al., 1999). Accumulation of p-p53 S37 further supports the increased concentration of p53 in the cytoplasm during anoxia in both tissues (Figs 2 and 3). Given the role that MDM2 plays in p53 inactivation, it is expected that in order to increase levels of activated p53, there would be a similar decrease in MDM2 expression (Ashcroft and Vousden, 1999). Indeed, MDM2 protein levels were significantly decreased in both the tissues during anoxia. Considering that the northern crayfish has not been sequenced and phosphorylation of p53 is not well studied in crustaceans, a detailed investigation and mass spectrometry analysis to find the exact phosphorylation sites would be required for confirmation of these results.

Another mechanism that stabilizes p53 involves the activation of p19-ARF during DNA damage. p19-ARF binds to MDM2 in a region that is different from the p53 binding domain. This interaction inhibits the ubiquitin ligase activity of MDM2, preventing the degradation of p53 by MDM2 without affecting the binding of p53 with MDM2 (Honda, 1999). The increased levels of p19-ARF under anoxic conditions in both tissues suggest the stabilization of p53 in response to stress (Fig. 2). Altogether, decreased MDM2 levels, increased p53 S37 phosphorylation and increased p19-ARF expression would result in increased levels of p53-mediated transcription under stress.

The tumor suppressor protein p53 plays a critical role in programmed cell death in response to cellular stress. p53 stability and activation is a complex mechanism that is regulated via inhibition by MDM2, and phosphorylation-mediated activation by ATM/ATR. The findings presented in the current study suggest that under anoxic conditions, p53 is activated and stabilized by increased S15 and S37 phosphorylation, increased ATM, ATR and p19-ARF levels, and decreased MDM2 levels. The stability of p53 was confirmed by the increased transcript levels of ei24 under anoxia, an important downstream target of p53 that is transcribed in response to DNA damage. The increased protein levels of p53 and transcript levels of ei24 suggest activation of a protective mechanism against DNA damage and autophagy during anoxia. Taken altogether, upon sensing DNA damage, p53 is activated via phosphorylation and/or by p19-ARF. p53 is then free to act in the nucleus where it can promote the transcription of ei24, while also accumulating in the cytoplasm as a mechanism to restrict autophagy and resulting DNA damage. Overall, the results presented herein suggest the promotion of apoptosis in contrast to autophagy, which could potentially cause deleterious cytotoxicity. The current study provides important insights into the signaling mechanism and response of p53 to DNA damage in crayfish. However, further studies focusing on the specific regulation of apoptosis and autophagy in anoxic crayfish are needed to confirm these findings.

The authors thank J. M. Storey for comments on the manuscript and B. Lant for experimental contribution.