Cadmium-induced oxidative stress in Meretrix meretrix gills leads to mitochondria-mediated apoptosis

Cadmium (Cd) is one of the most important marine environmental pollutants that can cause oxidative damage and apoptosis in living organisms, and mitochondria are the key cell organelles affected by Cd toxicity. In this study, we investigated the effect of Cd on the mitochondria in the gill cells of the clam Meretrix meretrix and the underlying mechanism of mitochondria-mediated apoptosis following exposure to the metal. Exposure of the clams to artificial seawater containing 1.5, 3, 6 and 12 mg L−1 Cd2+ led to swollen mitochondria compared with the untreated clams. The mitochondria also became vacuolated at the higher Cd2+ concentrations. Biochemical assays showed that monoamine oxidase (MAO) activity and mitochondrial membrane potential (Δψm) increased at 1.5 mg L−1 Cd2+, but decreased at higher Cd2+ concentrations, while the activities of malate dehydrogenase (MDH) and cytochrome oxidase (CCO) and the scavenging capacities of anti-superoxide anion (ASA) and anti-hydroxy radical (AHR) all decreased with increasing Cd2+ concentrations. Significant increases in the levels of malondialdehyde (MDA) and H2O2 as well as in the activity levels of caspase-3, -8, and -9 were also observed in the Cd2+-treated clams. The results implied that Cd might induce apoptosis in M. meretrix via the mitochondrial caspase-dependent pathway.


Introduction
In recent years, pollution caused by toxic metals has become a serious environmental problem that continues to increase in magnitude due to rapid industrialization and the massive use of fertilizers (Chen et al. 2020;Zhang and Gao 2015). Cadmium is one of the most concerned toxic metals and it is commonly found in aquatic environments Chen et al. 2020). In normal water, the level of Cd is generally below the first-class seawater quality standard of China (1 µg L −1 ). However, levels that exceed 10 µg L −1 have also been recorded in some coastal areas contaminated by toxic metals (Meng et al. 2013;Chan and Wang 2018). In addition, a number of shellfishes in the contaminated coastal water have been exposed to Cd, and these include Crassostrea rivularis Gould, Meretrix meretrix, Perna viridis Linnaeus, Chlamys nobilis in Guangxi Beibu Gulf (average cadmium content is 2.79 mg kg −1 ) (Jiang et al. 2013) and cultivated shellfishes (Chlamys nobills, Ruditapes philippinarum, Perna viridis, Haliotis sp., Paphia sp., Saccostrea sp.) in Dongshan Bay, Fujian (average cadmium content is 0.65 mg kg −1 ) (Yang et al. 2019), and the level of Cd in these shellfishes were found to exceed the safety level required for pollution-free aquatic products to be considered safe for consumption in China (0.1 mg·kg −1 fresh weight) (GB 18406.4-2001). Such evidence seems to confirm the historical contamination of Cd in aquatic environments, and the threats of Cd to marine organisms certainly deserve serious attention. Cd is highly toxic to aquatic organisms even at low concentrations because it is non-biodegradable and it has a long biological half-life (Capriello et al. 2019;Park et al. 2020), and therefore, it can accumulate in the food chain, causing tissue injury and threatening the health of aquatic animals (Chen et al. 2020;Jing et al. 2019;Pan et al. 2018). At the cellular level, Cd also has the potential to affect cellular antioxidant defenses and damage the DNA repair systems, indicating that Cd plays a vital role in apoptosis (Lin et al. 2017;Park et al. 2020;Xia et al. 2016).
Mitochondria are the key intracellular targets for Cd stress (Bhansali et al. 2017), and Cd exposure can damage the integrity of the mitochondria in aquatic organisms, both structurally and functionally (Adiele et al. 2011). Ji et al. (2019) found that Cd can induce the reconstruction of energy homeostasis, stress resistance and apoptosis in Ruditapes philippinarum, and suggested that mitochondria are the key target of Cd toxicity in clams. Monoamine oxidase (MAO), malate dehydrogenase (MDH) and cytochrome oxidase (CCO) are the key enzymes of mitochondria. Cd can affect the activities of all these three enzymes (Achard-Joris et al. 2006;Goswami et al. 2014;Wu et al. 2015). Mitochondrial ROS is a major source of oxygen radicals in a cell, and these oxygen radicals include species such as superoxide radical (O 2 •− ), hydroxyl radical (•OH) and hydrogen peroxide (H 2 O 2 ) (Bhansali et al. 2017). Cadmium has been shown to facilitate the formation of reactive oxygen species (ROS) by interfering with the electron transport chain in the mitochondria, consequently stimulating lipid peroxidation via the oxidation of polyunsaturated fatty acids in organisms such as the bivalves (Goswami et al. 2014;Ji et al. 2019). Through the electron transport chain of the mitochondria, O 2 •− generates H 2 O 2 , which is then split by catalase into water and oxygen or is partially reduced to hydroxide radical (•OH) in a Fenton reaction, causing a wider cellular oxidative damage (Feng et al. 2014). A high level of •OH, H 2 O 2 , or O 2 •− in mitochondria can result in the attack of membrane phospholipids by free radicals. The loss of mitochondrial membrane potential (MTP, Δψm) can cause the alteration of mitochondrial membrane permeability, leading to the release of cytochrome c (Cyt c) and subsequent stimulation of intracellular caspase-8 and caspase-9, followed by the induction of caspase-3 and initiation of apoptosis (Achard-Joris et al. 2006;Đukić-Ćosić et al. 2020;Liu et al. 2013;Zhao et al. 2015). The activation of caspases can be considered a crucial mechanism of Cd-induced apoptosis. Wallace et al. (2019) suggested that a decrease in mitochondrial membrane permeability in hepatic cells exposed to Cd, as well as an increase in both caspase-3 and caspase-9 activities, can directly lead to programmed cell death. Cadmium can interfere with apoptotic pathways, which converge on the activation of caspase-3, eventually culminating in cellular fragmentation and cell death (Đukić-Ćosić et al. 2020).
Bivalves are widely distributed in the water ecosystem. As these animals (e.g., clams, oysters and mussels) have poor locomotion ability and are filter feeders that can consume about 1 × 10 5 cell mL −1 algae daily (Zhang et al. 2021), they are rather sensitive to the contamination of the aquatic environment by toxic metals, making them a group of important animals for aquatic toxicology research and the evaluation of pollution caused by toxic metals (Chandurvelan et al. 2015;Chen et al. 2020;Ji et al. 2019). Meretrix meretrix is one of the economically and ecologically important bivalves in China, and it is widely cultivated in the different coastal regions of China. Meretrix meretrix has been gradually used as a suitable bioindicator of toxic metal pollution in recent years (Alyahya et al. 2011;Meng et al. 2013;Wan et al. 2018). Bivalve gills are the main target organs for the accumulation of metals, and they usually show the highest level of pollutants because their gills are in direct contact with the heavy metals in the water (Jing et al. 2019;Trevisan et al. 2014). We have previously compared the levels of accumulated Cd in the gills, foot, hepatopancreas and mantle of M. meretrix, and found the highest level in the gills (Huang et al. 2020). Jing et al. (2019) suggested that dissolved metals are absorbed mainly by direct adsorption occurring on the animal surface (gill, mantle, and foot), while particulate metals tend to be absorbed by the digestive organs along with the food. The gill is also the main barrier against environmental pollutant-mediated injuries and pathological agents, making it the first site where the injury caused by toxic metals is likely to occur (Huang et al. 2020;Marasinghe Wadige et al. 2017). Oxidative damage and apoptosis in clams resulting from Cd toxicity have been confirmed by numerous morphological and biochemical studies (Achard-Joris et al. 2006;Huang et al. 2020;Ji et al. 2019;Xia et al. 2016). However, the mechanism associated with the apoptotic process that is triggered by Cd-induced oxidative stress was not elucidated in these studies. We, therefore, speculated that in M. meretrix, mitochondria in the gill cells might play an important role in the intrinsic apoptotic pathway under extreme environmental cadmium exposure. Such speculation could be further investigated by comparing the changes in mitochondrial structure, anti-free radical capacity and apoptosis-related factors between untreated and Cd-treated M. meretrix individuals. This study, therefore, aimed to explore the effect of Cd on cellular oxidative damages in the mitochondria of M. meretrix gills, and to identify the underlying mechanism of the mitochondria-mediated apoptotic pathway in the gills following induction by Cd.

Materials and methods
Clam collection, treatment and gill sample collection The clams (M. meretrix) used in the experiment were collected from Lingkun aquafarm in Zhejiang Province. Lingkun is a small town without an industrial zone. It is famous for breeding M. meretrix, Scylla serrata, Litopenaeus vannamei and other high-quality seafood. The level of Cd in the seawater used in the clam aquafarm was 0.54 μg L −1 , which is within the Chinese standard for firstclass seawater quality (Cd ≦ 1 μg L −1 ) (GB 3097-1997). The average length, width and height of these clams were 52.72 ± 2.20 mm, 26.95 ± 1.26 mm, and 43.41 ± 1.88 mm, respectively, and their average weight was 39.06 ± 4.81 g. The clams were acclimated without food for 2 days in 15‰ artificial seawater (pH 7.9) inside an aquarium (67 × 46 × 36.5 cm) equipped with auto-temperature control. The temperature of the water was kept at 22 ± 1°C. After acclimation, 525 healthy clams were then randomly assigned to a control group and four CdCl 2 -treated groups. Each group consisted of 105 clams, and these were kept in 3 separate aquariums, with 35 clams per aquarium. The clams in the CdCl 2 -treated groups were exposed to 1.5, 3, 6 and 12 mg L −1 CdCl 2 , which corresponded to 1/10, 1/5, 1/2.5, 1/1.25 of the LC 50 (15.01 mg L-1) previously determined for a 96 h exposure (Xia et al. 2016). The water in the aquarium of the control group was replaced daily with 15‰ artificial seawater only while the water in each Cd 2+ -treated group was replaced daily with 15‰ fresh artificial seawater containing the same Cd 2+ concentration. All other experimental conditions were kept the same as during acclimation. All the clams were collected and dissected after 5 days of Cd 2+ exposure, and the gills were quickly removed. The gills collected from each aquarium were divided into three parts: one part was immediately immersed in liquid nitrogen before being stored in a −80°C freezer for biochemical indexes analysis, one part was fixed in 4% (v/v) paraformaldehyde for histopathological analysis, while the remaining part was fixed in 2.5% glutaraldehyde for transmission electron microscopy analysis.

Morphological analysis by histology and transmission electron microscopy
In order to determine the histological morphology of the gill filaments, the gills from the control and exposed clams were fixed in 4% (v/v) paraformaldehyde. The specimens were then immersed in the same buffer for 24 h followed by dehydration in ethanol baths with increasing concentrations of ethanol (75%, 85%, 90%, 95%, and 100%). The specimens were then embedded in paraffin and placed at room temperature to allow the paraffin to solidify. The paraffinembedded specimens were cut into slices (thickness of 6-8 μm) using a microtome (YD-315, China). The slices were then dewaxed, dehydrated, stained with haematoxylin and eosin (H.E.), and finally sealed with neutral balsam. These slices were observed with a light microscope (Olympus, Japan).
For electron microscopy, the gills from the control and exposed clams were immediately cut into approximately 1-mm 3 pieces and fixed in 2.5% glutaraldehyde (4°C) for 2 h and then in 1% osmium tetroxide for 2 h. The fixed samples were contrasted with 2% uranyl acetate in 10% ethanol for 4 h, dehydrated in an ascending series of acetone, embedded in Epon 812 Resin for 24 h, and then incubated at 70°C for 24 h. After that, the samples were cut into 60-nm thick slices using an ultramicrotome (LKB-2088) and stained with uranyl acetate and lead citrate. These slices were observed with a transmission electron microscope (Hitachi −7500, Japan).

Sample preparation for the determination of relevant parameters
The gill tissue extracted from each clam was homogenized on ice in 9 vol (w v −1 ) of 0.9% ice-cold physiological saline using a high throughput tissue grinder (SCIENTZ-48, China). The homogenate was then centrifuged at 4000 × g (4°C) for 20 min, and the supernatant was retained, yielding the gill extract. The levels of MDA, H 2 O 2 , ASA and AHR, and the activity levels of CCO, MAO, MDH and caspases in the gill extract were then measured as described below. The protein content in the gill extract was determined followed by the method of Bradford (1976).

MDA, H 2 O 2 content and Caspase-3, -8, -9 activity assay
The contents of H 2 O 2 , MDA (Jiancheng Biotechnology, Nanjing, China) and the activities of Caspase-3,-8,-9 (Beyotime Biotechnology; Jiangsu, China) in the gill extract were assayed with commercial assay kits as described by Lin et al. (2017). The concentrations of H 2 O 2 and MDA were expressed as nmol mg protein −1 whereas the activities of caspase-3, -8, and -9 were expressed as U mg protein −1 .

CCO, MAO and MDH activities assay
Cytochrome oxidase (CCO), Monoamine oxidase (MAO) and malate dehydrogenase (MDH) activities in the gill extract were assayed using commercial assay kits (Jiangcheng Biotechnology, Nanjing, China). CCO assay is based on the disappearance of reduced cytochrome c (Thibault et al. 1997) and MAO assay is based on the oxidation of benzylamine to benzaldehyde (Ashafaq et al. 2014), while MDH assay is based on the reduction of oxaloacetic acid to malic acid (Childress and Somero 1979). The activities of CCO, MAO and MDH were expressed as U mg protein −1 .

ASA, AHR and MTP (Δψm) levels assay
The anti-superoxide anion (ASA) capacity and anti-hydroxy radical (AHR) capacity in the gill extract were determined with a superoxide anion free radical detection kit and hydroxyl free radical detection kit (Keygen Biotech, China), respectively. The ASA capacity of the tissue was expressed as U g −1 protein. One ASA unit was defined as the quantity of superoxide anion free radicals required to scavenge 1 mg of tissue protein for 40 min at 37°C, which is equal to one gram of vitamin C-scavenging under the same condition. Tissue AHR capacity was expressed as U mg −1 protein.
One unit was defined as the amount that decreased 1 mmol L −1 of H 2 O 2 within 1 min per milligram of tissue protein (Jiang et al. 2009).
The mitochondrial membrane potential (MTP, Δψm) of the gill cells was detected using the 5, 6 -Dichloro-1, 1′, 3, 3′-tetraethyl-imidacarbocyanine iodide (JC-10) assay kit (Keygen Biotech., China) according to the manufacturer's instructions. In brief, 50 mg of gill tissue was placed in a 1.5-mL tube and gently cut into fragments with a small sterile scissor. This was followed by the addition of 0.25% trypsin-EDTA solution to a total volume of 500 μL. Next, the sample was incubated at 20°C for 25 min and then centrifuged at 800 × g for 5 min at 4°C to collect the cells, which were then incubated in 0.5 mL of 1:500 diluted JC-10 dye solution at 37°C in the dark for 30 min. After that, the cells were again collected by centrifugation at 600 × g for 4 min at 4°C and washed with JC-10 dye buffer twice. About 3 × 10 5 cells were immediately measured with a fluorescence spectrophotometer (Hitachi F-4600, Japan) using an excitation wavelength of 488 nm and an emission wavelength of 535 nm.

Statistical analysis
Statistical analyses were performed using the SPSS statistical software (Version 20.0, Chicago, USA). The data were subjected to a one-way ANOVA analysis after being assessed for homogeneity of variance using Levene's test. All data were presented as means ± standard errors (SE). Differences among experimental groups were determined using Tukey's Multiple Comparison Test at a p < 0.05 level of significance. The least significant difference test (LSD) was used to perform multiple comparisons among different treatment groups.

Results
Morphological changes in gill epithelium of M. meretrix induced by Cadmium The gill of M. meretrix is formed by two-gill lamella (outer and the inner) which are connected via the interlamellar junction, with the gill water tube occupying the middle space of the two interlamellar junctions (Fig. 1 A). Each gill lamella consists of a series of parallel gill filaments (Fig.1B,  C). The gill filaments are joined to each other by the interfilamentar junctions present preferentially in the regions where the interlamellar junctions occur (Fig. 1B). The surface of the gill filament is completely covered by cilia, called frontal, latero-frontal and lateral cilia and two or more gill filaments share the same haemolymph vessel (Fig. 1C). Contact between the external environment and the interlamellar space is achieved by occasional pores called ostrium (Fig. 1B, C), located between the gill filaments. In the control group, the clams exhibited a normal health status as indicated by the organized lamellar epithelium. The gill filaments were slender and tightly arranged and were well separated from the haemolymph sinus ( Fig. 1  A, B). There were also abundant cilia on the surface of the gill epithelial cells (Fig. 1C, D). In the group of clams treated with 1.5 mg L −1 Cd 2+ , no significant alteration in gill structure was revealed by histological examination (Fig. 1E, F). However, the cilia of the gill epithelial cells were relatively scattered, and the number of cilia was slightly reduced (Fig. 1G, H). Exposure of the clams to increasing Cd 2+ concentrations resulted in the lamellar structure of the gills exhibiting signs of structural deformation, which included the fusion of gill filament, dilated hemolymphatic sinus and degeneration of cilia ( Fig. 1I-T). Clams that were exposed to 3 mg L −1 Cd 2+ exhibited loose interlamellar junctions in the gills (Fig. 1I) and an increasing number of hemocytes in the hemolymphatic sinus that probably caused the deformed morphology of the gill filament. There was a slight fusion between the hemolymphatic sinus and gill filaments (Fig. 1J). The number of epithelial cells decreased (Fig. 1K, double arrow), some of the gill filaments were broken and the shedding of cilia from the gill epithelial cells was more obvious (Fig. 1L, arrow and black triangle). For the clams that were exposed to 6 mg L −1 and 12 mg L −1 Cd 2+ , greater histopathological changes were observed for the gills as revealed by the regression in the epithelium thickness or as an epithelium detachment (Fig. 1M-T). There was a clear fusion between the hemolymphatic sinus and gill filaments (Fig. 1M, N, Q, R, *) as well as the presence of swelling endothelial cells. Furthermore, reduced epithelium thickness, some detached epithelia with shedding cilia, and vacuolation in the gill filament were also visible ( Fig. 1O, P, S, T, double arrow and black triangle).
In the control clams, the epithelial cells of the gill filaments showed typical morphology, with intact cell membrane, nucleus and mitochondrial cristae as revealed by transmission electron microscopy. Moreover, the cilia also displayed well-preserved microtubules ( Fig. 2A, B). In the clams exposed to 1.5 mg L −1 Cd 2+ , the morphology of the microvilli and the nucleus of the gill epithelial cells were basically intact, however, some mitochondria in the gill epithelial cells showed a loss of mitochondrial cristae Fig. 1 Histological analysis of M. meretrix gills following exposure of the clams to different Cd 2+ concentrations. A-D Control group; E-T Cd 2+ -exposed groups. These groups were exposed to Cd 2+ at 1.5 mg L −1 (E-H); 3 mg L −1 Cd 2+ (I-L); 6 mg L −1 (M-P); 12 mg L −1 (Q-T).  (Fig. 2C, D). A number of vesicles were also evident in the cytoplasm, but these were absent in the gill epithelial cells of the control clams (Fig. 2C). Damage to the cell structures increased with increasing Cd 2+ concentrations (Fig. 2E-J), such as swollen mitochondria and a more conspicuous loss of mitochondrial cristae in the clams exposed to 3 mg L −1 Cd 2+ (Fig. 2E, F). These clams exhibited numerous vesiculated particles in the gill epithelial cytoplasmic space, with electron-dense particles present inside the vesicles (Fig. 2E, white double arrow). In addition, a disorganized ciliary orientation and a lack of typical microtubules were also observed in the cilia (Fig. 2E). For the clams that were treated with 6 and 12 mg L −1 Cd 2+ , the disruption of mitochondrial cristae, shrinkage and vesiculated cytoplasm, and the condensation of chromatin became noticeable ( Fig. 2G-J). Again, it was possible to detect fragmentation and highly electron-dense particles inside these vesicles ( Fig. 2G-J, white double arrows). Progressive vacuolization of the nucleoplasmic content, swelling nuclear envelope ( Fig. 2G-J, arrows) and progressive disappearance of organelles were also detected in the clams exposed to a higher Cd 2+ concentration. Moreover, the number of lysosomes also increased in the cells (Fig. 2G, I) and the number of microvilli was greatly reduced (Fig. 2G-J, black double arrows) in 6 and 12 mg L −1 Cd 2+ -treated clams.

Effects of Cd on H 2 O 2 and MDA contents
The levels of MDA in the gills of each group of Cd 2+exposed clams exhibited a significant increase (F = 61.16, df = 4, 25, p < 0.05) compared with the control group, and the increase was dependent on Cd 2+ concentration (Fig. 3). The maximum level of MDA was found in the group exposed to the highest Cd 2+ concentration (12 mg L −1 ) and this level was also significantly higher than the levels found in all the other Cd 2+ -exposed groups. However, the level of MDA found in the group exposed to 3 mg L −1 Cd 2+ did not differ significantly from that found in the group exposed to 6 mg L −1 Cd 2+ (Fig. 3).
As for the level of H 2 O 2 , a gradual increase caused by Cd 2+ exposure was also evident, and the increase was dependent on Cd 2+ concentration (Fig. 3). The highest level of H 2 O 2 was found in the group exposed to the highest Cd 2+ concentration, which was significantly (F = 18.31, df = 4, 25, p < 0.05) higher than the control group and all other Cd 2+ -exposed groups. However, there were no significant differences among the control group, the group exposed to 1.5 Cd 2+ and the group exposed to 3 mg L −1 Cd 2+ (Fig. 3). Thus, the levels of MDA and H 2 O 2 found in the gills of M. meretrix exposed to Cd 2+ were consistent with the fact that Cd 2+ could induce the production of MDA and H 2 O 2 as part of its toxic effect.

Effects of Cd on CCO, MAO and MDH activities
Cadmium inhibited the activity of CCO in the gills of clams that were exposed to this metal. Except for the group exposed to 1.5 mg L −1 Cd 2+ , all the other Cd 2+ -exposed groups exhibited a significant decrease (F = 19.87, df = 4, 25, p < 0.05) in the level of COO activity in the gill compared with the control group. The largest decrease was found in the group exposed to the highest Cd 2+ concentration, which was also significantly lower than the levels found in all the other Cd 2+ -exposed groups (Fig. 4a).
The effect of Cd 2+ on MAO activity in the gills of the clams seemed to vary, depending on the concentration of Cd 2+ that the clams were exposed to. The highest level of MAO activity was found in the gills of the group treated with 1.5 mg L −1 Cd 2+ , while the lowest level was found in the gill of the group treated with 12 mg L −1 Cd 2+ (Fig. 4b).
MDH activity was also significantly suppressed in the clams exposed to Cd 2+ compared with the control clams (Fig. 4c). The minimum level of MDH activity was found in the group exposed to the highest Cd 2+ concentration, which was significantly (F = 14.22, df = 4, 25, p < 0.05) different from the control group and the other Cd 2+ -exposed groups except for the group exposed to 6 mg L −1 . Meanwhile, no significant differences in MDH activity were observed among the groups exposed to 1.5, 3 and 6 mg L −1 Cd 2+ (Fig. 4c).
Taken together, the results suggested that higher Cd 2+ concentrations (3-12 mg L −1 ) may significantly reduce the activities of the mitochondrial marker enzymes (CCO, MAO, MDH), but the effect of Cd 2+ on these enzymes tended to vary depending on the enzyme.

Effects of Cd on ASA, AHR and Δψm levels
Cadmium appeared to cause a reduction in the levels of ASA and AHR in the gills of the clams after exposure to the metal. The levels of ASA and AHR in all Cd 2+exposed groups were lower than those of the control group. However, significant differences in both ASA (F = 5.43, df = 4, 25, p < 0.05) and AHR (F = 5.71, df = 4, 25, p < 0.05) levels were observed between the group exposed to 12 mg L −1 Cd 2+ , which exhibited the lowest levels, and the control group (Fig. 5a, b). As for the effect of Cd 2+ on Δψm, there appeared to be a significant (F = 86.42, df = 4, 15, p < 0.05) reduction in Δψm for the groups exposed to 3, 6 and 12 mg L −1 Cd 2+ compared with the control group, with the group exposed to 12 mg L −1 Cd 2+ exhibiting the greatest reduction (Fig. 5c). Overall, Cd 2+ seemed to reduce the antioxidant capacity and membrane potential of the mitochondria in the gill cells of M. meretrix exposed to the metal, with the extent of the impact being proportional to the concentration of Cd 2+ .

Effects of Cd on caspase-3, -8, -9 activities
The effect of Cd 2+ on caspases was investigated by comparing the activity levels of caspases-3, -8, -9 in the gills of Cd 2+ -exposed clams with those in the gill of control individuals. All Cd 2+ -exposed groups showed significant increases in caspase-3 (F = 536.33, df = 4, 15, p < 0.05), caspase-8 (F = 523.62, df = 4, 15, p < 0.05) and caspase-9 (F = 526.64, df = 4, 15, p < 0.05) activities compared with the control group, with the highest level occurring at the highest Cd 2+ concentration (12 mg L −1 ) (Fig. 6). The levels of caspase-3, -8, -9 activities in the control group were significantly lower than those found in all the Cd 2+ -exposed groups. The results, therefore, suggested that exposure of M. meretrix to Cd 2+ in the range of 1.5-12 mg L −1 could cause the activation of caspase-3, -8 and -9 in the gills and the extent of activation is dependent on the concentration of the metal.

Discussion
Oxidative damage of toxic metal to gill mitochondria The gills of bivalves are suitable for histopathological analysis since they consist of a simple epithelium with various cell types, in which the effects caused by toxic metals in the water can be easily observed (Trevisan et al. 2014;Zhen et al. 2018). Exposure of M. meretrix to cadmium resulted in noticeable structural changes in the gills as detected by light microscopy (LM) and transmission electronic microscopy (TEM) (Figs. 1 and 2). The reduced number of microvilli found on the gill epithelial cells could lead to a reduction in the ability to filter microscopic food in the water. This could then lead to physiological changes such as nutrient absorption that need to be compensated by the organism. Additionally, TEM revealed a significant change in the structure of the gill cells, which was manifested as cytoplasmic shrinkage and active membrane blebbing and fragmentation into membrane-enclosed vesicles resembling apoptotic bodies (Fig. 2G-J). A previous study on Cd toxicity in the clam Ruditapes philippinarum has revealed that the mitochondria of gill epithelial cells might be the organelles targeted by Cd . This is consistent with the Cd-induced changes in mitochondrial structure in M. meretrix gill cells that we observed, which included the loss of cristae and increased vacuolation following exposure of the clams to 6 and 12 mg L −1 Cd 2+ (Fig. 2G-J), indicating serious damage to the mitochondria. These morphological changes were accompanied by various biochemical changes, such as decreased enzyme activities and anti-free radical ability. As it is known, mitochondria are responsible for ATP production through oxidative phosphorylation and they play an essential role in the oxidative damage sustained by the cells (Meyer et al. 2013). The morphology of mitochondria is associated with the energetic state and the survival of the cells. Thus, damage to the mitochondrial morphology and internal structure will greatly affect the activity of the mitochondria, and in the case of damage inflicted by toxic metals, the metals can accumulate in the cells, and the cells will die as a result of metabolic disorder (Li et al. 2017).
Among the diverse cellular responses triggered by Cd exposure, oxidative stress (which is part of the early biological response) has been demonstrated in aquatic animals (Goswami et al. 2014;Pan et al. 2018;Park et al. 2020;Xia et al. 2016). The damage to the mitochondrial structure of gill cells in M. meretrix induced by Cd 2+ will directly affect the function of the mitochondria and induce oxidative stress. Cadmium can cause an increase in the content of intracellular ROS (O 2 •− , H 2 O 2 , and •OH, etc.), which will lead to further reactions in the cascade, leading to the induction of apoptosis and the expression of apoptosisrelated genes, indicating that oxidative stress is one of the most important mechanisms of Cd toxicity (Lin et al. 2017;Pan et al. 2018). Kurochkin et al. (2011) have suggested that in Cd 2+ -exposed molluscs, there might be at least two potential mechanisms of ROS formation: 1) direct inhibition of electron transport chain enzymes and/or substrate transporters by Cd binding; and 2) oxidative damage to these enzymes or substrate transport system by Cd-induced ROS.
In Anadara subcrenata, the binding of Cd to the thiol groups of proteins and nitrogen-containing ligands was found to result in the overproduction of ROS, which subsequently led to oxidative stress, DNA damage, lipid peroxidation, and even apoptosis (Yan et al. 2019). Huang et al. (2020) suggested that in M. meretrix, antioxidant enzymes and chelation of Cd 2+ via cysteine-rich molecules such as GSH and MT probably play a role in the scavenging of ROS generated in response to the effect of toxic metals. However, the antioxidant capacity might not be adequate to rid the ROS induced by increasing Cd 2+ levels, resulting in a concomitant increase in MDA (an indicator of lipid peroxidation) level in different tissues (Huang et al. 2020;Lin et al. 2017;Xia et al. 2016). The oxidative damage in gill cells induced by excessive ROS was also manifested by an increase in MDA and H 2 O 2 and a decrease in ASA and AHR contents (Figs. 3 and 5). Cd 2+ has been reported to induce the production of ROS by interfering with the mitochondrial electron transport chain, consequently causing damage to proteins and lipids, thereby impairing mitochondrial function and increasing mitochondrial membrane permeability (Kim et al. 2015). Mitochondria are the main site of ROS generation, but at the same time, they are also the targets most sensitive to the action of ROS because of the sensitivity of mitochondrial enzymes to Cd-induced damage (Pan et al. 2018;Zorov et al. 2014). Once the mitochondria have been challenged by toxins, a series of responsive proteins, such as CCO, MDH and MAO, will be differentially expressed to safeguard the function of the mitochondria (Meyer et al. 2013;Valera-Alberni and Canto 2018). MDH, an essential enzyme of the tricarboxylic acid (TCA) cycle, is located in the mitochondrial matrix. It catalyzes the formation of malic acid into oxaloacetic acid, which can scavenge H 2 O 2 and effectively prevent •OH from damaging DNA (Li et al. 2017). MAO and CCO are located on the outer and inner membrane of the mitochondria, respectively (Achard-Joris et al. 2006). MAO is a flavoenzyme involved in the oxidative deamination of amine neurotransmitters (Abdelouahab et al. 2010), while CCO can shield against the metal-induced superoxide stress to maintain a redox balance in the mitochondrial chambers (Achard-Joris et al. 2006;Liu et al. 2013). The activities of MDH and CCO in the mitochondria of M. meretrix gill cells were inhibited by Cd 2+ in a concentration-dependent manner, while the level of MAO activity in the group exposed to 1.5 mg L −1 was significantly higher than those in the control and other Cd 2+ -treated groups (Fig. 4). Fraser et al. (2017) found that a non-monotonic dose-response of MAO activity was observed in male Mytilus edulis, with increased MAO activity for the groups treated with 0.01 and 10 nM Cd 2+ and decreased MAO activity in 0.1 nM Cd 2+treated group, whereas in females, decreased MAO activity was observed for the group treated with 0.1 and 10 nM Cd 2+ . These authors suggested that analyzing male and female MAO activity levels separately might be important. In our experiment, MAO activity was determined from a mixture of gills randomly taken from both male and female clams. The mixture probably contained more gills from males than from females. This might have led to an increase in MAO activity observed. Vandenberg et al. (2012) proposed a number of mechanisms underpinning the nonmonotonic dose effects produced by why endocrinedisrupting chemicals and these mechanisms include receptor selectivity, competition and down-regulation. Besides, low concentrations of endocrine-disrupting chemicals often disturb biological endpoints in a non-monotonic manner. Reduced levels of MDH, CCO and MAO activities at higher Cd 2+ concentrations (3-6 mg L −1 ) might, in turn, lead to reduced metabolic capacity and scavenging of free radicals (H 2 O 2 , •OH) generated by Cd 2+ (Das et al. 2018;Li et al. 2017). Zhang et al. (2001) suggested that CCO inhibitors can cause a rapid and severe depletion of cellular ATP content, resulting in acute cell death. Recently, it was suggested that activation of MAO plays a key role in initiating cell apoptosis through the generation of ROS and that MAO expression is activated in response to oxidative stress (Abdelouahab et al. 2010). In M. meretrix, Cd may induce the production of ROS in the gills by interfering with the electron transport chain, destroying the membrane structure of mitochondria, affecting mitochondrial function and ultimately leading to apoptosis. A decrease in enzyme activity levels of the mitochondrial respiratory chain is closely related to the energy metabolism of the mitochondria, which ultimately leads to apoptosis (Mamos et al. 2016).

Endogenous mitochondrial apoptosis pathways induced by toxic metal
Mitochondria are the site for apoptosis initiation and calcium signal regulation . Mitochondrial membrane potential (MMP, Δψm), caspase-9, -8, -3, and cytochrome c (Cyt c) are known to be the major factors in mitochondrial-associated apoptosis (Düssmann et al. 2017;Kim et al. 2015). Caspase-3 is the most well-known downstream effector caspase and it is activated through cleavage, which converts the enzyme into the active form (cleaved caspase-3) that facilitates apoptosis (Zhao et al. 2015). In this study, the activities of caspases -3, -8 and -9 in the gill cells increased significantly and dose-dependently following the exposure of the clams to Cd 2+ (Fig. 6), suggesting that the enhanced activation of caspase-3 by Cd 2+ caused the apoptosis observed in the gill of M. meretrix.
The results were consistent with the findings of Wang et al. (2012), which demonstrated that caspase activation is involved in the process of Cd-induced apoptosis in Sinopotamon henanense. Yuan et al. (2018) suggested that decreased Δψm, the release of Cyt c from mitochondria into the cytoplasm and increased caspase-3 and caspase-9 activities can directly lead to apoptosis. A collapsing Δψm may be an early event in the apoptotic process, and Δψm is the premise index for the detection of apoptosis (Gao et al. 2021;Suhaili et al. 2017;Xu et al. 2016;Yuan et al. 2018). The negative effect of Cd on Δψm has been frequently reported in cells or isolated mitochondria (Belyaeva et al. 2008;Liu et al. 2013). Srdic-Rajic et al.
(2011) also observed a significant loss of Δψm in human cervix carcinoma cells (HeLa) and human endothelial cells (EA. hy 926 cells) following 3 h of exposure to Cd 2+ and Zn 2+ complexes or Cd 2+ and Ni 2+ complexes. The Δψm in the gills of M. meretrix exposed to 1.5 mg L −1 Cd 2+ -was not significantly different from the Δψm in the gills of the control M. meretrix, suggesting that a low concentration of Cd 2+ may not cause apoptosis in the gill cells. This was clearly supported by the lack of obvious change in the ultrastructure of the gills (Figs. 1 and 2). However, a decrease in the Δψm of the gill cells was observed in M. meretrix exposed to higher concentrations (3-12 mg L −1 ) of Cd 2+ (Fig. 5), indicating that Cd possibly enhanced the release of Cyt c from the mitochondria to the cytoplasm, therefore, leading to mitochondrial dysfunction. Our data were consistent with previous data on Ruditapes philippinarum where Cd was found to induce an increase in the production of extracellular ROS along with a decrease in Δψm ).

Conclusions
In summary, the results presented in this report demonstrated that decreased Δψm and caspase-dependent apoptosis is one of the pathways by which Cd toxicity could induce apoptosis in M. meretrix gill. Our data also suggested changes in mitochondrial structure induced by Cd 2+ may be the main factor of cadmium-induced cellular cytotoxicity, accompanied by elevated MDA and H 2 O 2 levels, decreased ASA, AHR and Δψm, all of which would eventually affect the function of the mitochondria, activating caspase signaling and triggering apoptosis in the gill cells. Therefore, cadmium toxicity is linked to the generation of reactive oxygen species, which could consequently lead to the induction of apoptosis. Since the gill cells underwent apoptosis as a result of stress-induced by Cd 2+ -toxicity, apoptosis is considered to proceed via the intrinsic pathway, which is mediated by the B-cell lymphoma 2 (Bcl-2) family of proteins. To obtain a better understanding of the underlying mechanisms of apoptosis induced by Cd, further study will look at the implication of several transcription factors, such as Bax/Bcl-2, and their regulation of gene(s) involved in the apoptosis of M. meretrix gill cells.

Data availability
All authors guarantee that all data and materials support our published claims and the Data are available by contacting XP Ying (xpying2008@wzu.edu.cn).