A recent study has shown that in the budding tunicate Polyandrocarpa misakiensis, the mitochondrial respiratory chain (MRC) dramatically attenuates the gene activity during senescence. In this study, we examined the possible involvement of superoxide dismutase (SOD) in the attenuation of gene expression of cytochrome c oxidase subunit 1 (COX1) in aged zooids. By RT-PCR and in situ hybridization, Cu/Zn-SOD (SOD1) was found to be expressed in most cells and tissues of buds and juvenile zooids but showed a conspicuous decline in senescent adult zooids, except in the gonad tissue in which the cytoplasm of juvenile oocytes was stained heavily. This expression pattern of SOD1 was similar to that of COX1. In contrast to SOD1, Mn-SOD (SOD2) was expressed constitutively in both somatic and germline tissues of buds, juvenile zooids, and senescent adult zooids. Knockdown of SOD1 by RNAi diminished the gene activity of not only SOD1 but also of COX1. The resultant zooids had transient deficiencies in growth and budding, and they recovered from these deficiencies approximately 1 month later. Our results indicate that in P. misakiensis, SOD1 is a senescence-associated nuclear gene and that the experimental decline in SOD1 gene expression accompanies the attenuation of MRC gene activity. Although it is uncertain how SOD1 is downregulated during tunicate senescence, the decreased SOD1 activity could be one of the main causes of MRC gene attenuation during normal senescence.
Senescence inevitably occurs in all eukaryotes. In mammals, senescence accompanies a decrease in the activity of the mitochondrial respiratory chain (MRC) (Byrne et al. 1991; Cooper et al. 1992; Boffoli et al. 1994). Such a functional deterioration in the mitochondria is brought about by time-related accumulation of DNA damage (Hebert et al. 2010). Reactive oxygen species (ROS) are one of the main causes of such DNA damage (Liu et al. 2000), and they are inevitably produced in vivo by the MRC because of the reduction-oxidation reaction in the mitochondria.
Reactive oxygen species are neutralized by the superoxide dismutase (SOD) protein family, peroxidases, and catalases, of which SOD is involved in the disproportionation of superoxide anions into oxygen and hydrogen peroxide. CuZn-SOD (SOD1) is located in the cytosol and the intermembrane space of mitochondria, and Mn-SOD (SOD2) is located in the mitochondrial matrix. An extracellular SOD, SOD3, is also present. Human SOD1 can extend the normal lifespan of Drosophila by up to 40% (Parkes et al. 1998). SOD2-deficient mice show neonatal mortality due to a significant decrease in cell proliferation and abnormal signal transduction in the mutant fibroblasts (Zhang et al. 2010). SOD1, with a molecular weight of 17 kDa, has been purified from the hemocytes and blood plasma of the solitary tunicate, Halocynthia roretzi, and the protein plays a role in the defense mechanism of the tunicate (Abe et al. 1999).
In the case of the budding tunicate Polyandrocarpa misakiensis, the zooids live for 4–5 months. The zooidal senescence is accompanied by a decrease in the MRC gene activities and proliferating cell nuclear antigen (PCNA) immunoreactivity and by an increase in the activity of senescence-associated β-galactosidase (Kawamura et al. 2012a). The epidermis is a major tissue that shows such aging-associated events, with the epidermal senescence also being ultrastructurally evident. Interestingly, all these senescence-related events can be reset through budding (asexual reproduction), especially epidermal MRC gene activities, which can be restored repeatedly at every budding (Kawamura et al. 2012a). Tunicate cytostatic factor, TC14-3, is involved in such reactivation of mitochondrial genes via the activation of a nuclear gene Eed, a component of the Polycomb group (PcG) (Kawamura et al. 2012a,b). Therefore, it is probable that during senescence in P. misakiensis, the decrease in mitochondrial gene activities could be attributable to certain physiological conditions rather than irreversible gene mutation, although the underlying mechanism remains unclear.
The purpose of our study was to obtain evidence that SOD is involved in decreasing the transcriptional activity of the MRC. This is the first study on the gene expression of SOD1 and SOD2 homologues during the life span of tunicates. We have also shown the results of RNAi against PmSOD1, which is the first study to our knowledge to highlight the possibility of SOD influencing mitochondrial gene activities.
Materials and methods
Asexual strains of P. misakiensis were cultured on glass slides in culture boxes, settled in Uranouchi Inlet near the Usa Marine Biological Institute, Kochi University. Growing buds, 2–3 week-old juvenile zooids, and senescent adult zooids were used for studies (Fig. 1A) (Kawamura et al. 2012a).
Poly(A)+ RNA was extracted and purified from buds, juvenile zooids, and senescent adult zooids by using the biotinyl magnet method according to the manufacturer's protocol (Roche, Mannheim, Germany). Single-stranded DNA, complementary to poly(A)+ RNA, was synthesized at 42°C for 1 h by using StrataScript reverse transcriptase (Stratagene, CA, USA). Polymerase chain reaction (PCR) was performed in two steps: one cycle for sense strand synthesis (30 s at 94°C, 2 min at 52°C, and 2 min at 72°C); and 20–30 cycles of denaturation at 94°C for 30 s, annealing at 52°C for 60 s, and extension for 90 s at 72°C. β-actin cDNA was used as an internal standard and was amplified using PCR.
For quantification, PCR products were separated by agarose gel electrophoresis and stained with ethidium bromide, after which they were scanned and quantified using ImageJ, a public domain software developed by the National Institutes of Health.
In situ hybridization
The animals were fixed overnight on ice in 4% paraformaldehyde in phosphate-buffered saline containing 0.1% Tween 20 (PBST) as described elsewhere (Sunanaga et al. 2007). After the specimens were treated with 10 μg/mL proteinase K for 30 min at 37°C, they were hybridized with a digoxigenin (Dig)-labeled antisense ribonucleotide probe for 12–14 h at 56°C. Blocking was carried out in 1% skimmed milk in Tris-buffered Tween 20-containing salt solution (TBST) for 6 h on an ice bath. Then, the samples were treated overnight on ice with anti-Dig monoclonal antibody labeled with alkaline phosphatase (Roche, Mannheim, Germany). The samples were then stained with a color development solution, dehydrated, and embedded in Technovit 8100 resin (Heraeus Kulzer, Germany). Serial sections (thickness, 2 μm) of the resin-embedded samples were cut with glass knives by using an ultramicrotome (Kawamura et al. 2012a,b).
Senescence-associated β-galactosidase assay
Specimens were fixed, washed, and pre-incubated twice for 10 min in 20 mmol/L Na-citrate buffer (pH 6.0) as described elsewhere (Kawamura et al. 2012a). Samples were incubated in a staining solution (3 mmol/L K-ferrocyanide, 3 mmol/L K-ferricyanide, 150 mmol/L NaCl, 1 mmol/L MgCl2, and 0.1% x-gal in the same buffer) overnight at 37°C.
Three different siRNAs were designed using PmSOD-1 mRNA and were purchased from SIGMA-PROLIGO (Tokyo, Japan). Their oligonucleotide sequences were as follows: siRNA-1, 5′-AAUUUCACAAGUGCCGCCGUU-3′ and 5′-CGGCGGCACUUGUGAAAUUUC-3′; siRNA-2, 5′-UGCUAUCUGGAAAUCCACCUU-3′ and 5′-GGUGGAUUUCCAGAUAGCAAG-3′; and siRNA-3, 5′-UGCAUGUUUAUGGAUUACCUU-3′ and 5′-GGUAAUCCAUAAACAUGCAAU-3′. The siRNA oligonucleotides were dissolved in filtered seawater at a concentration of 50 μmol/L. Immediately before use, the same volumes of the three solutions were mixed with one another and incubated for more than 20 min in seawater containing Lipofectamine 2000 (10:1; Invitrogen, Carlsbad, CA, USA) at a final concentration of 10 μg/mL.
Growing buds and juvenile zooids were injured with razor blades to make it easy for siRNA to infiltrate into animals and treated with the liposome solution for 30 min in moist chambers. Then, growing buds were cultured in natural seawater for 3 days until RT–PCR and in situ hybridization experiments. Juvenile zooids were allowed to heal and grow for 2 weeks before analyses.
Expression of PmSOD1 and PmSOD2 during the life span of P. misakiensis
The cDNA for PmSOD1 (AB591737) was 1092 bp long, with the polypeptide consisting of 154 amino acid residues (Fig. S1). The cDNA for PmSOD2 (AB591738) was 821 bp long, and the polypeptide consisted of 225 amino acid residues (Fig. S2). On the molecular phylogenic tree, PmSOD1 formed a cluster different from the group to which PmSOD2 belongs (Fig. S3).
The results of RT–PCR showed that the expression of PmSOD1 was conspicuously lower in senescent adults than in the buds and juvenile zooids, whereas the PmSOD2 expression was relatively constant during the entire life span (Fig. 1B). A previous study has shown that mitochondrial genes encoding cytochrome c oxidase subunit 1 (PmCOX1) and NADH dehydrogenase subunit 1 (PmND1), similar to PmSOD1, exhibit attenuated activities during senescence (Kawamura et al. 2012a). In this study, NADH dehydrogenase subunit 4 (PmND4) showed similar results (Fig. 1B).
The results of in situ hybridization showed that in growing buds including the bud primordium, PmSOD1 expression was moderate in the epidermis (Fig. 2A,B), coelomic cells (Fig. 2B,C), and the atrial epithelium (Fig. 2C). In the juvenile zooids of P. misakiensis, gene expression was maintained in the atrial epithelium and coelomic cells and looked stronger in the epidermis (Fig. 2D). On the ventral side of the zooids, the coelomic cells associated with the atrial epithelium strongly expressed PmSOD1 (Fig. 2E), and they might belong to the primordial gonad (Sunanaga et al. 2007). The pharynx also showed a strong PmSOD1 signal (Fig. 2F). In contrast, the PmSOD1 signal almost disappeared from the body walls of senescent adult zooids (Fig. 2G,H). Among the differentiated tissues, the endostyle, pharynx, esophagus, and stomach stained weakly (Fig. 2I–K). The cytoplasm of developing oocytes showed the strongest staining, and the male germ cells were stained moderately (Fig. 2L).
In contrast to PmSOD1, PmSOD2 was not observable in the bud epidermis (Fig. 3A,B), but the atrial epithelium and associating coelomic cells expressed PmSOD2 in moderate levels (Fig. 3A,C). In juvenile zooids, PmSOD2 expression was seen to be moderate in the epidermis, coelomic cells, pharynx, and gonad rudiments (Fig. 3D–F). In senescent adult zooids, moderate signals were observed in the epidermis and atrial epithelium (Fig. 3G). The pharynx showed stronger signals (Fig. 3H). In the gonad of P. misakiensis, the cytoplasm of young oocytes stained heavily, and dot-like signals were observed in the germinal vesicle (Fig. 3I arrowheads). Fully-grown oocytes and testis showed weak staining.
Influence of PmSOD1 knockdown on mitochondrial gene expression
As PmSOD1 gene expression was age-dependent, similar to the MRC genes, we examined the relationship between PmSOD1 and PmCOX1 in buds by using RNAi against PmSOD1. In the control treated with siRNALacZ, the gene expression pattern of PmSOD1 was the same as that in intact buds (Fig. 4A–C). The buds treated with siRNASOD1 showed a remarkably diminished signal of PmSOD1 from the epidermis, atrial epithelium, and coelomic cells (Fig. 4D,E), although morphogenesis occurred normally (Fig. 4F). This attenuation of gene expression was observed in seven specimens out of 15, and in the eight remaining specimens, the atrial epithelium still showed strong staining. In contrast to situ hybridization, RT–PCR showed that both the amounts and increasing curves of PCR products were very similar between siRNASOD1 and siRNALacZ (Fig. 4G–I).
Next, we used juvenile zooids for RNAi experiments. All control zooids (siRNALacZ) developed into adults and began budding (Fig. 5A) (Table 1). Experimental zooids (siRNASOD1) developed into adults (Fig. 5B, #1, #3), or they did not grow well or failed to produce buds (Fig. 5B, #2, #4) (Table 1). We examined whether the growth deficiency of zooids accompanied the increasing activity of senescence-associated β-galactosidase (SA-Gal) and found that the SA-Gal activity was very weak in both well-grown zooids and growth-deficient zooids (Fig. 5B). The growth deficiency was transient, and approximately a month later, experimental zooids were fully grown and commenced budding (data not shown).
Table 1. Effect of RNAi of PmSOD1 on zooid growth and budding in Polyandrocarpa misakiensis
Only growth-deficient animals were examined by in situ hybridization and RT–PCR. As compared with the control treated with siRNALacZ (Fig. 6A), the PmSOD1 signal of siRNASOD1-treated zooid was very faint in the epidermis, coelomic cells, atrial epithelium, and pharynx, except the esophagus that exhibited strong signals (Fig. 6B–D). The results of RT–PCR showed that siRNASOD1 caused a conspicuous decrease in the increasing curve of PmSOD1 products, as compared with the control (Fig. 6E). The effects of siRNASOD1 on PmCOX1 expression were also examined. In contrast with the control (Fig. 6F), signals for PmCOX1 virtually disappeared from most tissues, including the epidermis, atrial epithelium, coelomic cells, pharynx, and endostyle (Fig. 6G–I). PmND1, like PmCOX1, conspicuously decreased in gene activity in the epidermis, coelomic cells, atrial epithelium, and pharynx after RNAi of PmSOD1 (Fig. S4D–F) in contrast with the siRNALacZ control (Fig. S4A–C). By RT–PCR, the increasing curve of PmCOX1 products in the experiment (siRNASOD1) was lower than that in the control (siRNALacZ), the difference comparable to at least two PCR cycles delay (Fig. 6J).
Gene expression pattern of PmSOD1 and PmSOD2 in relation to aging
Superoxide dismutase plays a role in the disproportionation of superoxide anions into oxygen and hydrogen peroxide, which would protect, in collaboration with peroxidase and catalase, biological materials, in particular, mitochondrial DNA from oxidative damages. The present study has shown that both PmSOD1 and PmSOD2 are expressed widely in somatic and gonadal tissues, similar to the ubiquitous expression of mammalian SOD1 (Schisler & Singh 1985). In P. misakiensis, the SOD expression was especially prominent in juvenile germ cells, which likely provides protection against oxidative stress to the germ cells that would otherwise be vulnerable to ROS, leading to genetic mutation and cell death (Liu et al. 2000).
In the colonial tunicate Botryllus schlosseri, SOD appears to serve as a scavenger of ROS that causes cytotoxicity when two incompatible colonies encounter each other and a rejection reaction is initiated (Ballarin et al. 1998, 2002). In the solitary tunicate H. roretzi, SOD1 protein (HrSOD1) is found in the coelomic cells and blood plasma (Abe et al. 1999). The HrSOD1 in the plasma enhances the phagocytic activity of coelomic cells, thus playing a role in the defense mechanism.
Interestingly, we found that PmSOD1 was active in buds and juvenile zooids and became inactive in senescent adult zooids, especially in epidermal cells. This change of gene activity was similar to those of MRC genes such as PmCOX1, PmND1, and PmND4 (Kawamura et al. 2012a and this paper), indicating that PmSOD1 is a novel senescence-associated gene in P. misakiensis. This gene expression pattern of PmSOD1 was similar to that of mammalian SOD2 rather than SOD1. Mammalian SOD1 is expressed constantly and irrespective of the age (Oh-Ishi et al. 1995), and the enzymatic activity of protein rather increases with age in the muscles (Hollander et al. 2000), lungs (Hass & Massaro 1987), and liver (Carrillo et al. 1992). Mammalian SOD2, on the other hand, significantly lowers transcriptional and enzymatic activities with ageing in muscle (Hollander et al. 2000) and liver (Santa Maria et al. 1996). In P. misakiensis, results of RT–PCR showed that the overall expression of PmSOD2 did not change remarkably during senescence. The epidermis maintained low transcriptional levels throughout zooidal life, and the pharynx of senescent adults showed a rather stronger signal than that of juvenile zooids.
In mammals, senescence is related to unrepaired oxidative damages against genetic materials (Lee & Wei 2012). It is, therefore, reasonable to assume that SODs can regulate the longevity of cells. In fact, human SOD1 can extend the normal lifespan of Drosophila by up to 40% (Parkes et al. 1998). SOD1 knockout in mice causes fibrosis, increased apoptosis, and other deteriorative events (Kojima et al. 2012).
However, there are contradictory studies indicating that the situation may be more complicated. As mentioned, SOD is involved in the disproportion of superoxide anions into O2 and H2O2, and human SOD1 mediates pulmonary fibrosis by augmenting the production of H2O2 (He et al. 2011). In nematodes, SOD knockout has no effect on the longevity of the animals (Doonan et al. 2008). The total loss of mitochondrial SOD does not significantly increase oxidative damage to mitochondrial DNA (Gruber et al. 2011). These findings lead to an argument against the notion that the superoxide radical is a major determinant of aging (Doonan et al. 2008).
Possible relationship between PmSOD1 and MRC genes
In P. misakiensis, MRC genes are downregulated during senescence and reactivated during budding (Kawamura et al. 2012a). TC14-3 is a budding-specific cytostatic factor (Matsumoto et al. 2001) and appears to be involved in PmCOX1 reactivation via the activation of the nuclear gene, PmEed, a component of the Polycomb group (Kawamura et al. 2012b). However, the mechanism that causes the aging-related downregulation of the MRC genes was yet to be elucidated. In this study, to investigate the effects of siRNASOD1 on PmCOX1, we used buds at first, and the results of RT–PCR were inconsistent with those of in situ hybridization. We supposed that the inconsistent results might be caused by the wrong method of sample collection, in which both RNAi-affected and RNAi-unaffected buds were included. Therefore, we changed the experimental material from buds to juvenile zooids since RNAi-affected zooids were discernible from unaffected zooids with ease when the zooids were allowed to grow for 2 weeks following RNAi treatment. Consequently, both quantitative and qualitative data indicated that growth-deficient zooids by RNAi had lower activities of not only PmSOD1 but also of PmCOX1. It should be noted that although the attenuation of MRC gene activity is one of senescence signals (Kawamura et al. 2012a), the growth-deficient zooids did not show the increasing activity of SA-Gal, suggesting that the dwarf phenotype of zooids would not completely reproduce the tunicate senescence.
The RNAi results in our study suggest that PmCOX1 gene expression requires normal activity of PmSOD1. The relationship between nuclear and mitochondrial genes is known in mammals where glutathione peroxidase, a ROS scavenger related to SOD, can regulate mitochondrial function (Handy et al. 2009). If MRC activities change in synchrony with the activity of ROS scavenger, the dysfunction of SOD would accompany the decrease in MRC activities and does not necessarily cause the increase in ROS amounts, which can explain why the SOD knockout in nematodes has little influence on mitochondrial DNA and longevity (Doonan et al. 2008; Gruber et al. 2011). Although we do not know yet how PmSOD1 is downregulated during zooid senescence, it is reasonable to assume that in P. misakiensis, the decrease in PmSOD1 activity is involved in the aging-related low activity of the mitochondrial genes.
To our knowledge, this study is the first to show the gene expression pattern of SOD1 and SOD2 during the life span of P. misakiensis; furthermore, we found that PmSOD1 is a senescence-related gene. This study also suggests that the gene activity of PmSOD1 has the relationship to MRC. Some feedback regulation would likely occur between PmSOD1 and MRC in order for the ROS sensors to suppress excess ROS production by lowering the mitochondrial gene activity in PmSOD1-insufficient senescent zooids. It is interesting to ask whether PmSOD1 and MRC are involved in zooidal longevity in P. misakiensis.
We thank the staff of Usa Marine Biological Institute of Kochi University for providing facilities for the culture of animals. This work was supported in part by KAKENHI (No. 19570208, 21570227, 21116507) from the Ministry of Education, Sport and Culture, Japan.
K.K., 80% (gene cloning, in situ hybridization, text preparation); T.S., 20% (RT–PCR).