Drp-1-Dependent Mitochondrial Fragmentation Contributes to Cobalt Chloride-Induced Toxicity in Caenorhabditis elegans

Abstract

Excess cobalt may lead to metallosis, characterized by sensorineural hearing loss, visual, and cognitive impairment, and peripheral neuropathy. In the present study, we sought to address the molecular mechanisms of cobalt-induced neurotoxicity, using Caenorhabditis elegans as an experimental model. Exposure to cobalt chloride for 2 h significantly decreased the survival rate and lifespan in nematodes. Cobalt chloride exposure led to increased oxidative stress and upregulation of glutathione S-transferase 4. Consistently, its upstream regulator skn-1, a mammalian homolog of the nuclear factor erythroid 2-related factor 2, was activated. Among the mRNAs examined by quantitative real-time polymerase chain reactions, apoptotic activator egl-1, proapoptotic gene ced-9, autophagic (bec-1 and lgg-1), and mitochondrial fission regulator drp-1 were significantly upregulated upon cobalt exposure, concomitant with mitochondrial fragmentation, as determined by confocal microscopy. Moreover, drp-1 inhibition suppressed the cobalt chloride-induced reactive oxygen species generation, growth defects, and reduced mitochondrial fragmentation. Our novel findings suggest that the acute toxicity of cobalt is mediated by mitochondrial fragmentation and drp-1 upregulation.

cobalt, toxicity, oxidative stress, mitochondria fragmentation, drp-1

As a component of vitamin B12, cobalt is an essential metal for cellular growth, differentiation, and development (Bumoko et al., 2015; Deshmukh et al., 2013). However, excessive bodily cobalt can cause toxicity (Fowler, 2016; LiSon et al., 2018). Excessive cobalt serum levels can arise from a variety of sources including natural erosion, diet, exposure during industrial fabrication, or from alloys used in tooth and hip joint replacements (Cheung et al., 2016). High levels of cobalt may lead to metallosis, characterized by sensorineural hearing loss, visual, and cognitive impairment and peripheral neuropathy (Leyssens et al., 2017). In Wistar rats, CoCl₂ overexposure results in several distinct defects, including hypoxia, lipid peroxidation, apoptosis, and the loss of learning, memory, and spatial-exploration (Akinrinde and Adebiyi, 2019; Zheng et al., 2019). The transmission of CoCl₂ toxicity from parents to offspring in Caenorhabditis elegans has been reported by Wang et al., affecting lifespan, development, chemotactic plasticity, and hsp16-associated stress response (Wang et al., 2007).

CoCl₂ induces hypoxia and expression of hypoxia-inducible factor-1 in various cell types, accompanying toxicity (Karovic et al., 2007). CoCl₂-induced hypoxia has been shown to alter the expression of angiogenesis and apoptotic genes, therefore contributing to tumorigenesis (Bahadori et al., 2019; Rana et al., 2019). A hypoxia-inducible factor-1-independent mediator BLMP-1 (named according to its mouse homolog B lymphocyte-induced maturation protein 1) has been identified and is necessary for response to CoCl₂-induced hypoxia in C. elegans (Padmanabha et al., 2015). In addition to hypoxia, we have previously shown that CoCl₂ induces oxidative stress (Zheng et al., 2019).

Oxidative stress causes disturbance in mitochondrial homeostasis (Elfawy and Das, 2019). Mitochondria homeostasis, also known as dynamics, refers to the variations of shape of mitochondria. The tubular, elongated, interconnected mitochondrial networks are generated by fusion, whereas fission leads to the generation of discrete fragmented mitochondria (Diaz and Moraes, 2008). The balance between mitochondrial fission and fusion is of great importance for optimal energy generation and supply, and thus cell viability. In addition, mitochondrial fission and fusion have been proposed to play key roles in autophagy and mitophagy. Damaged mitochondria have been shown to be separated by fission, recognized by phagophore, and in turn forming autolysosome, which proceeds through mitophagy (Yoo and Jung, 2018). On the other hand, disruption of normal mitochondrial fission and fusion can also generate reactive oxygen species (ROS), thus leading to oxidative stress (Meyer et al., 2017). The ROS-involving dysregulation of mitochondrial dynamics has been further suggested to contribute in metabolic regulation, type 2 diabetes, cardiovascular disease, cancer, and neurodegenerative diseases (Bhat et al., 2015; Kim and Song, 2016; Rovira-Llopis et al., 2017; Subramaniam and Chesselet, 2013; Vásquez-Trincado et al., 2016). One of the main mediators of mitochondrial fission is Dynamin-Related Protein 1 (DRP-1) (Praefcke and McMahon, 2004). DRP-1 is also proapoptotic and mediates autophagy (Breckenridge et al., 2008; Martinez et al., 2018). Caenorhabditis elegans is a powerful in vivo model to study environmental pollutant-induced toxicity and mitochondrial gene-environment interactions, as mitochondrial functions along with many pathways are well conserved with those in humans (Byrne et al., 2019; Maglioni and Ventura, 2016; Tsang and Lemire, 2003). Moreover, the in vivo study of mitochondrial structure and function is of considerable importance, as a large number of the regulating signals arise from diverse tissues and cell types, which are lost when studying in in vitro models (McBride et al., 2006).

Therefore, in the present study, we sought to determine the molecular and mitochondrial mechanisms of CoCl₂-induced toxicity in C. elegans. We posited that in C. elegans, CoCl₂ induces toxicity through oxidative stress and mitochondrial fragmentation, secondary to the activation of DRP1. CoCl₂-induced toxicity in survival and lifespan was first established, followed by oxidative stress and mitochondrial morphology analyses. The role of drp-1 in CoCl₂-induced toxicity was confirmed with drp-1 mutant worms.

MATERIALS AND METHODS

Culture and maintenance of C. elegans strains

Caenorhabditis elegans strains, Bristol N2, CL2166 (dvIs19 III. [Pgst-4p::GFP]), VP596 (dvIs19 III; vsIs33 V. [[Pgst-4p::GFP; Pdop-3::RFP]), SD1347 (ccIs4251 I. [Pmyo-3p::GFP + Pmyo-3p::mitochondrial GFP]), CU6372 [drp-1(tm1108)], and Escherichia coli OP50, were procured from Caenorhabditis Genetics Centre (University of Minnesota, Minnesota), grown on nematode growth medium (NGM) and cultured at 20°C (Brenner, 1974). A synchronized population of nematodes was obtained by bleach treatment. Synchronized nematode cultures were initiated by bleaching gravid young adults and sucrose separation to obtain eggs (day 0). Sucrose flotation was included to remove potential contaminations such as residual dead debris and bacteria. Collected eggs were allowed to hatch overnight (approximately 18 h after plating) on unseeded NGM plates. The hatched larval stage 1 (L1) nematodes were used for experiments the next morning (day 1). For larval stage 4 (L4) experiments, L1-arrested larvae were transferred to 15-cm NGM plates seeded with OP50 E. coli on the late afternoon (day 1). On day 3 morning (40–44 h after plating), those nematodes will reach L4 (with the iconic half circle in the ventral side representing the developing vulva) and ready for experiments.

CoCl₂ exposure

CoCl₂ was used in this study to examine the toxicity of cobalt for its high bioavailability (defined as 100%) compared with other cobalt compounds, including Co₃O₄ and CoS (in the range of 0.06%–0.1%) (Danzeisen et al., 2020). Moreover, CoCl₂ has been wildly used for cobalt toxicity determination in both in vivo and in vitro models (Bahadori et al., 2019; Guan et al., 2015; Rana et al., 2019; Wang et al., 2000, 2018). By choosing the same compound, our study afforded cross-species evaluation. Unless otherwise stated, all reagents were obtained from Sigma-Aldrich (St Louis, Missouri). A 2 M stock solution of CoCl₂ was prepared with ddH₂O. Because cobalt ion reacts with phosphate groups in M9 buffer and forms insoluble Co₃(PO₄)₂, 85 mM NaCl was used for CoCl₂ exposure. The 20× CoCl₂ exposure solutions of various doses were prepared using 85 mM NaCl. The exposure doses for nematodes were selected based on preliminary survival assays (0, 2.5, 5, 10, 25, and 50 mM). The doses are for acute liquid exposure (2 h), thus higher than the 0.1–4.76 mM in plates (solid exposure) for 12 h to 3 days previously reported in the literature (Chong et al., 2009; Padmanabha et al., 2015). After 2 washes with 85 mM NaCl, nematodes were exposed to CoCl₂ for 2 h. Due to the difference in sensitivity and size, 5000 L1 nematodes or 2500 L4 nematodes were exposed to CoCl₂ in 500 µl reaction system containing 25 µl of 20× CoCl₂ exposure solutions. This was followed by 2 washes with 85 mM NaCl to remove residual CoCl₂.

Survival assay

After CoCl₂ exposure from 0 to 100 mM and washed with 85 mM NaCl, approximately 30–40 nematodes were then transferred to 35 mm NGM plates freshly seeded with OP50 in triplicates. After incubation at 20°C for 24 h, nematodes were scored as alive or dead with a stereomicroscope (Stemi 2000, Zeiss, Germany). In some cases, dead nematodes were confirmed by touching the head region with the point of a platinum picker.

Measurement of ROS

After CoCl₂ exposure, 2′,7′-dichlorodihydrofluorescein diacetate/2′,7′-dichlorofluorescein (H₂DCFDA) was used to determine ROS in C. elegans. A previously published detailed method (Yoon et al., 2018) was followed. Green fluorescence was measured by fluorescent plate reader (FluoStar OPTIMA, BMG LabTech) for up to 6 h.

Oxidative stress reporter assay

Activation of skn-1, the worm homolog of nuclear factor (erythroid-derived-2)-like 2 (Nrf2), was measured using CL2166 and VP596 strains, which expresses GFP under the control of the promoter for the skn-1 target GSH S-transferase 4 (gst-4). VP596 nematodes also express RFP under the dop-3 promoter, serving as the loading control. After CoCl₂ exposure, GFP fluorescence of CL2166 nematodes was determined by a Leica SP8 Confocal Microscope and quantified by Fiji (Schindelin et al., 2012). VP596 nematodes, on the other hand, were transferred to a 96-well black plate. Levels of RFP and GFP florescence were measured (RFP: excitation/emission 544/590 nm and GFP: 485 /520 nm). GFP florescence was then divided by RFP florescence to normalize the data to worm number.

RNA isolation and quantitative real-time polymerase chain reactions

Ten thousand L1 nematodes or 2500 L4 nematodes per group were adequate for quantitative real-time polymerase chain reaction (qRT-PCR). mRNA was isolated with TRIzol following manufacture’s protocol (Life Technologies, USA) with minor modifications. In brief, nematodes went through 3 freeze and thaw cycles with liquid nitrogen to breakdown cuticles and release RNA. The concentration and purity of the isolated RNA were determined by a NanoDrop 2000 spectrophotometer (Fisher, USA). cDNA was then reverse transcribed from RNA using High Capacity cDNA Reverse Transcription kit (Applied Biosystems, USA). Gene expression was detected by Taqman gene-expression assay probes (Thermo Fisher Scientific). Probes used in this study were Ce02412618_gH (tba-1); Ce02484980_g1 (egl-1); Ce02452076_g1 (ced-9); Ce02446175_g1 (ced-4); Ce02466776_m1 (ced-3); Ce02463990_m1 (bec-1); Ce02433594_g1 (lgg-1); Ce02407440_g1 (drp-1); Ce02433121_g1 (fzo-1); and Ce02463146_g1 (miro-1). The target gene expression was normalized to the relatively stable expression gene tba-1 (Zhang et al., 2012). The relative mRNA levels were determined with the 2^−△△CT method (Livak and Schmittgen 2001).

Mitochondrial imaging

Nematodes with mitochondria-tagged GFP proteins in body wall muscle (ccIs4251 I.) were used to assess alterations in mitochondrial morphology upon CoCl₂ exposure. Synchronized L4 larvae were treated with CoCl₂ for 2 h, followed by 1 h recovery. Worms were then anesthetized in 1 mM levamisole and then mounted onto a microscope slide containing a 4% agarose pad. Images were taken immediately after slide preparation to avoid artifact on a Leica SP8 Confocal Microscope with a 63× oil lens with excitation/emission wavelengths at 488/520 nm for GFP.

In additional experiments, nematodes were labeled with MitoTracker prior to imaging. Upon pilot experiments, we discovered that MitoTracker staining of the mitochondria and CoCl₂ exposure cannot be performed at the same time. When exposed to 10 mM of CoCl₂, only half of the nematodes had their mitochondria labeled. As the dose of CoCl₂ increased, the dye failed either to enter the worms or to target to the mitochondria (data not shown). Therefore, we stained the mitochondria prior to the CoCl₂ exposure. Nematode growth medium plates were freshly seeded with OP50 mixed with 5 µM MitoTracker red CMXRos (Invitrogen, Carlsbad, California). Synchronized L4 nematodes were grown on the plates overnight, followed by CoCl₂ exposure. Afterward, nematodes were recovered in clean seeded plates (without MitoTracker dye) for 1 h to remove all residue dyes in the gut. Fluorescence images were captured using a Nikon ECLIPSS 80i fluorescence microscope with a 60× oil lens.

Mitochondrial morphology was assessed in at least 30 animals analyzed for each condition and blindly scored. The morphological categories of mitochondria were defined according to Momma et al. (2017) with slight modification: (1) tubular: a majority of mitochondria were interconnected and elongated like tube shape, (2) intermediate: a combination of interconnected and fragmented mitochondria, and (3) fragmented: a majority of round or short mitochondria in the image taken.

Lifespan assay

L4 nematodes were exposed to 10 mM CoCl₂ for 2 h, then moved to fresh 3.5-cm NGM plates seeded with OP50 E. coli (lifespan experiment “day 0”). To prevent offspring from nematodes under study from reaching adulthood, fluorodeoxyuridine was added to OP50-spread NGM plates to a final concentration of 120 µM (Sutphin and Kaeberlein, 2009). A hundred and twenty nematodes were studied per strain, divided on two 35 mm NGM plates. Mortality was confirmed by poking nematodes lightly with a platinum wire (3 times) and monitoring them for at least 5 min; nematodes that did not move after stimulation were scored as dead and removed from the plate. Nematodes that died due to protruding/bursting vulva, bagging, or crawling off the agar were marked as “censored.” In group WT-0, WT-10, drp1 (tm1108)-0, and drp1 (tm1108)-10, censored worms were 4, 5, 2, and 3, respectively. Therefore, the exact total N value for each group was > 114. Median, mean, and maximal lifespan were determined for each strain relative to concurrently studied wildtype nematodes.

Statistical analysis

Statistical analysis was performed using GraphPad PRISM, Version 8 (California, USA). Each experiment was repeated at least 3 times and normalized to unexposed groups. Analysis of variance followed by Tukey’s or Dunnett’s (where comparison was only to control) post hoc tests were utilized. For all experiments, p < .05 was deemed to be statistically significant. If not specifically stated, data are expressed as mean ± SD.

RESULTS

CoCl₂-Impaired Survival and -Induced Oxidative Stress in C. elegans

Upon CoCl₂ exposure, the survival of wildtype L1 C. elegans was reduced in a dose-dependent manner, with an LD₅₀ (50% lethal dose) of 20.69 mM (Figure 1). A similar lethality pattern was also found in L4 nematodes (LD₅₀: 29.98 mM) (Figure 5A). At higher doses where survival was significantly reduced compared with controls, a portion of worms displayed health defects, their bodies straightened and/or failing to move (extreme sluggishness) (Supplementary Figure S2).

Figure 1.

CoCl₂ exposure induces lethality and oxidative stress in Caenorhabditis elegans. A, L1 nematodes were exposed to CoCl₂ at various doses for 2 h. N = 3 with more than 30 animals per plate. Individual experiment was repeated for 5 times. B, L1 animals exposed to CoCl₂ (0, 2.5, 5, and 10 mM), were detected by 2′,7′-dichlorodihydrofluorescein diacetate/2′,7′-dichlorofluorescein assay. Fluorescence was followed every 0.5 h for 6 h, with representative 2 and 4 h shown here (N = 3). Individual experiments were independently repeated for 3 times. *p < .05 compared with 0 mM CoCl₂. ^#p < .05 compared with 5 mM Co group.

Figure 2.

Oxidative stress defense system is turned on by CoCl₂. A, Worms with oxidative stress inducible gst-4p::GFP were exposed to CoCl₂ for 2 h and detected by confocal microscopy. Bar indicates 50 µm. B, Fluorescent intensities of at least 30 animals were quantified by Fiji. C, Skn-1 mRNA, the activator of gst-4, is increased (N = 3 with 10 000 animals per sample). Individual experiments were independently repeated for 4 times. *p < .05 compared with 0 mM CoCl₂.

CoCl₂-induced oxidative stress in C. elegans was detected with the H₂DCFDA assay. H₂DCFDA, the cell permeant fluorescent probe, is rapidly oxidized in the presence of ROS to highly fluorescent DCF. Therefore, an increase in the fluorescence signal is deemed to positively correlate with increased intracellular ROS generation (Labuschagne and Brenkman, 2013). Although 5 and 10 mM CoCl₂ failed to alter survival (Figure 1A), both doses significantly induced ROS generation (Figure 1B). These results establish that CoCl₂ induced both lethality and oxidative stress in C. elegans.

Skn-1-Involved Oxidative Stress Defense System Is Activated by CoCl₂

Skn-1, the mammalian homolog of nuclear factor E2-related factor 2 (Nrf2), is a master regulators of oxidative stress defense systems. Because gst-4 is a main downstream target of skn-1, we used a gst-4::GFP strain (CL2166) to evaluate skn-1 activation (using GFP readout). Upon CoCl₂ exposure, fluorescence was significantly upregulated (Figure 2A). Fluorescent intensities were significantly increased in worms treated with 5 and 10 mM CoCl₂ (Figure 2B). Corroborative evidence of gst-4-induced GFP expression upon CoCl₂ exposure can be found in Supplementary Figure S2, using the VP596 strain. qRT-PCR (Figure 2C) of skn-1 confirmed that oxidative stress defense in the nematode is activated upon CoCl₂ exposure.

Apoptotic, Autophagic, and Mitochondrial Structure Regulators Are Activated Upon CoCl₂ Exposure

Next, we addressed additional physiological changes at the gene-expression level of CoCl₂ exposure in C. elegans. As shown in Figure 3, CoCl₂ significantly increased the expression of apoptosis-associated genes, namely egl-1, ced-9, and ced-3 (homologous of mammalian Bh3, Bcl-2, and Caspase-9, respectively). Moreover, autophagic hallmarks such as the ATG6 ortholog bec-1, as well as the LC3 homolog, lgg-1, were significantly induced by CoCl₂.

Figure 3.

Apoptotic, autophagic, and mitochondrial morphologic regulators are activated upon CoCl₂ exposure. mRNAs were isolated from L1 animals exposed to CoCl₂ (0, 2.5, 5, and 10 mM) for 2 h. Apoptosis, autophagy, and mitochondrial morphology-associated genes were examined by quantitative real-time polymerase chain reaction. mRNA levels were normalized to tba-1 (N = 3, 10 000 animals per sample). Individual experiments were independently repeated for 4 times. *p < .05 compared with 0 mM CoCl₂. Experiments were performed in L4 animals with analogous results.

Mitochondrion is not only a critical compartment where ROS are generated but also a key organelle in mediating apoptosis and autophagy (Bloemberg and Quadrilatero, 2019; Kwon et al., 2018). Increased mitochondrial fission by elevated ROS production could facilitate autophagy (Huang et al., 2016). Thus, we determined whether mitochondrial fission and fusion were in responsible for the augmented autophagy. In response to CoCl₂ exposure, the mitochondrial fission regulator, drp-1, was upregulated, but mRNA levels of mitochondrial fusion, such as fzo-1 and miro-1, were indistinguishable from control (Figure 3).

Mitochondrial Fragmentation Upon CoCl₂ Exposure

Based on the expression changes of mitochondrial dynamic regulators, next, we examined whether the mitochondrial morphology was altered upon CoCl₂ exposure using a fluorescent-labeled mitochondria strain, SD1347. In order to better identify the mitochondrial morphology changes, L4 and young adult nematodes were used. In controls, the majority of the mitochondria assumed a tubular shape. CoCl₂ induced alterations in mitochondrial morphology at doses as low as 5 mM. Corroborating the upregulated drp-1 data, a greater number of fragmented mitochondria were observed as the dose of CoCl₂ increased (Figure 4).

Figure 4.

CoCl₂ induces mitochondrial fission. L4 nematodes were exposed to CoCl₂ at various doses for 2 h, then examined by fluorescent microscopy (N > 30). Mitochondrial morphology was classified into 3 categories: tubular, intermediate, and fragmented. All exposure groups are significantly different to control group (p < .01), examined by Chi-square test. Individual experiments were independently repeated for 3 times.

Drp-1 Suppression Alleviated CoCl₂-Induced Growth Defects

To address the causal relationship between drp-1-induced mitochondrial fragmentation and CoCl₂-induced damage, we utilized a drp-1 mutant strain. Upon 50 mM CoCl₂ exposure, a greater percentage of drp-1 mutants (43.13%) survived compared with WT nematodes (26.98%) (Figure 5A). For lifespan studies, a lower dose of CoCl₂ was selected to eliminate the acute effects and prolong survival. As shown in Figure 5B, lifespan in WT worms was shortened upon CoCl₂ exposure at the L4 stage. Median survival of WT control and WT-CoCl₂ was 12 and 10 days, respectively. In the absence of CoCl₂, the lifespan of drp-1 mutants was statistically indistinguishable from WT nematodes. However, upon CoCl₂ exposure, drp-1 mutants had significantly longer lifespan compared with WT nematodes (Figure 5B). This was supported by the median survival for drp1-CoCl₂: 11 days. Thus, drp-1 mutation reduces the CoCl₂-induced decrease in survival and lifespan.

Figure 5.

Inhibition of mitochondrial fission regulator DRP1 rescues cell growth and extends life span in nematodes upon CoCl₂ exposure. L4 drp-1 (tm1108) null mutant strain was exposed to CoCl₂ of various doses for 2 h, in line with WT strain. A, Drp-1 inhibition rescued lethality in 50 mM CoCl₂ exposure group (N = 3, at least 30 animals per replicate). *p < .05 compared with unexposed nematodes. ^#p < .05 compared WT-CoCl₂ with drp-1-CoCl₂. B, Drp-1 suppression extended lifespan in response to 10 mM CoCl₂ exposure (N = 106, 105, 108, and 107 for WT-0, WT-10, drp-1-0, and drp-1-10, respectively). *p < .0001 compared with unexposed nematodes. ^#p = .0042 compared WT-CoCl₂ with drp-1-CoCl₂, examined by log-rank method (Mantel-Cox method). Individual experiments were independently repeated for 3 times for survival and 4 times for lifespan.

Drp-1 Is Essential for CoCl₂-Induced Mitochondrial Fission and ROS Generation

Finally, we determined whether drp-1 has a role in mitochondrial fission, given that drp-1 mutants were less vulnerable to CoCl₂-induced toxicity. Studies on lifespan and behavior of the drp-1 mutant strain have shown unchanged or extended lifespan (Byrne et al., 2019), but none has addressed lifespan in this mutant strain upon CoCl₂ exposure. We examined mitochondrial morphology in both WT and drp-1 mutants upon CoCl₂ exposure (Figure 6A). At the tested doses, drp-1 mutants displayed higher percentage of tubular to intermediate mitochondria in response to CoCl₂ exposure, whereas WT nematodes displayed more fragmented mitochondria (Figure 6B). Thus, drp-1 mutation negatively correlates with CoCl₂-induced mitochondrial fission.

Figure 6.

Reduced mitochondrial fission and elevated reactive oxygen species production in drp-1 (tm1108) mutant upon CoCl₂ exposure. A, WT and drp-1 mutant strains were stained with MitoTracker red CMXRos overnight, followed by CoCl₂ exposure of 0, 10, and 25 mM. Mitochondrial morphology was examined using fluorescent microscopy. Bar indicates 5 µm. The lower panel shows the enlarged areas as boxed in the upper panel within each strain (N > 30). B, Quantification of mitochondrial morphology to tubular, intermediate, and fragmented. WT-10 and WT-25 are significantly different to WT-0, whereas drp-1-25 is significantly different to WT-0 and drp-1-0 (p < .01), examined by Chi-square test. Individual experiments were independently repeated for 3 times. C, L4 animals exposed to CoCl₂ (0, 10, and 25 mM), were detected by 2′,7′-dichlorodihydrofluorescein diacetate/2′,7′-dichlorofluorescein assay. Fluorescence was followed with 2 h RLU data shown here (cobalt-untreated group was set to 1) (N = 3). Individual experiments were independently repeated in triplicates. *p < .05 compared with 0 mM CoCl₂.

Mitochondrial fission and fusion dysregulation has been suggested to result in ROS and thus mitochondrial toxicity (Meyer et al., 2017). To further determine the potential mechanism of drp-1 in CoCl₂-induced toxicity, we measured the ROS level in both WT and drp-1 mutants. Corroborating L1 result (Figure 1B), CoCl₂ induced ROS in L4 animals. However, the induction of ROS was not evident in drp-1 mutants (Figure 6C). Thus, our results suggest that drp-1 mutation suppresses ROS generation upon CoCl₂ exposure.

DISCUSSION

Here, we demonstrate for the first time, that in the nematode, acute CoCl₂ exposure induces toxicity (growth defects, oxidative stress, apoptosis, and autophagy) by mitochondrial fragmentation. The activation of drp-1, the inducer of outer mitochondrial membrane fission, leads to ROS production and mitochondrial toxicity.

Cobalt induces toxicity in various organisms and tissues. In patients with hip replacement, cobalt levels have been shown to be 1.8-fold higher than in controls, concomitant with neurodegenerative changes (Catalani et al., 2012; Leyssens et al., 2017; Wu et al., 2018). We have previously demonstrated that oxidative stress is a key mediator of the toxicity of cobalt and its nanoparticles, both in rats and PC12 cells (Zheng et al., 2019). Moreover, the activation of Nrf2, the mammalian homolog of skn-1, and its protective role in CoCl₂-induced toxicity, has been established. In rat liver, microarray and mass spectrometry analysis studies have shown that exposure to CoCl₂ causes transcriptomic and proteomic changes, including Nrf2-mediated response and glutathione (GSH) generation, consistent with an oxidative stress response (Permenter et al., 2013). Overexpression of Nrf2 in mesenchymal stem cells has been shown to prevent CoCl₂-induced apoptosis, maintaining stemness (Yuan et al., 2017). In human keratinocytes, Nrf2 has been shown in the antioxidant response upon CoCl₂ treatment (Yang et al., 2018). Downstream targets of the Nrf2/ARE pathway, such as mfat-1, have been shown to protect cultured adult neural stem cell against CoCl₂-mediated hypoxia injury (Yu et al., 2018a). Here, we found activation of skn-1 and its downstream target gst-4 upon CoCl₂ exposure in C. elegans, consistent with the involvement of oxidative stress and Nrf2 defense systems in the acute response to this metal. Our results provide in vivo evidence in support of the involvement of ROS and the activation of skn-1 in CoCl₂-induced toxicity.

Moreover, we have observed the upregulation of apoptotic activator egl-1 and proapoptotic ced-4 upon CoCl₂ exposure. In agreement, apoptosis induced by CoCl₂ has been noted in C. elegans germline cells and rats, concomitant with increased Bax/Bcl-2 ratio and activation of Caspase 9 (Chong et al., 2009; Wang et al., 2018; Zheng et al., 2019). In addition to apoptosis, our study has shown the induction of autophagy upon CoCl₂ exposure. A close link exists between apoptosis and autophagy upon cellular stress. For example, inhibition of autophagy enhances apoptosis in response to CoCl₂ in rats lung cells (Yu et al., 2018b). Our in vivo results further corroborate this link because both apoptosis and autophagy were affected by CoCl₂. Autophagy, especially mitophagy, a type of macroautophagy that selectively degrades damaged mitochondria, might be an outcome of mitochondrial fission (Yoo and Jung, 2018).

Here, we have observed mitochondrial fragmentation in response to CoCl₂ exposure, consistent with the upregulation of drp-1 mRNA. Mitochondrial homeostasis is regulated by 2 sets of genes, the fission inducer (drp-1) and the fusion mediator (mfn-1/2 and miro-1), the latter homologous to fzo-1 and miro-1 in C. elegans. Mitochondrial fragmentation is a signal for mitochondrial dysfunction, which is prone to apoptosis and cell death (Byrne et al., 2019; Momma et al., 2017; Sebastian et al., 2017). Fusion-deficient nematodes were more sensitive to toxicants, including aflatoxin B1, arsenite, cisplatin, paraquat, and rotenone, suggesting that mitochondrial fusion is essential for cellular responses to toxic effects (Hartman et al., 2019; Luz et al., 2017). Moreover, drp-1-induced mitochondrial fragmentation in response to ethanol exposure has been demonstrated to contribute to the activation of the mitochondrial unfolded protein response (Oh et al., 2020). Here, we found that when drp-1 was mutated, the CoCl₂-induced defects including lethality, reduced lifespan, and increased ROS generation were no longer observed, concomitant with reduced mitochondrial fragmentation in nematodes. Similarly, in rat astrocytoma C6 cells, inhibition of drp-1 reduced the number of apoptotic nuclei and preserved the integrity of the mitochondrial network, thus prevented Mn-induced cell death (Alaimo et al., 2014). In rhabdomyolysis-induced acute kidney injury, reduced drp-1 protein accumulation in mitochondria ameliorated mitochondrial injury and apoptosis (Tang et al., 2013). An opposite effect was noted in the rat hippocampus upon drp-1 inhibition by synthetic xenoestrogen bisphenol A (Agarwal et al., 2016). Moreover, overexpressing Opa-1, the fusion protein, prevented drp-1-dependent mitochondrial fragmentation in SH-SY5Y cells overexpressing α-synuclein, an in vitro model of Parkinson’s disease (Martinez et al., 2018). In human lung bronchial epithelial cells, the antioxidant quercetogetin inhibited cigarette smoke extract-induced mitochondrial dysfunction and mitophagy secondary to the inhibition of phospho-DRP-1 and PINK1 (Son et al., 2018), corroborating our findings that drp-1 mutants fail to show changes in ROS production upon CoCl₂ exposure.

Taken together, these results suggest that CoCl₂ induces drp-1-associated mitochondrial fragmentation, in turn, causing ROS production and detrimental defects (Figure 7). Mitochondrial dysfunction is correlated with neurodegenerative processes (Itoh et al., 2013), suggesting that CoCl₂-induced mitochondrial dysfunction may trigger neurodegenerative changes associated with joint replacements in patients.

Figure 7.

A model for CoCl₂-induced toxicity in C. elegans. CoCl₂ induces oxidative stress, followed by altered apoptosis, increased lethality, and reduced lifespan. Increased reactive oxygen species (ROS) generation leads to skn-1 (Nrf2 homolog) activation and its downstream target gst-4, corroborating the involvement of oxidative stress defense systems (in green). CoCl₂ upregulates drp-1 and induces mitochondrial fragmentation. Mitochondrial impairment is likely caused by oxidative stress from ROS generation. When drp-1 is inhibited, mitochondrial fragmentation and ROS generation are suppressed, rescuing the CoCl₂-induced growth defects.

Other possible mechanisms associated with drp-1 inactivation that extend lifespan should be considered; they include (1) the reduction of insulin signaling by drp-1 suppression (Yang et al., 2011) and (2) the proapoptotic function of drp-1 downstream of ced-3 (Breckenridge et al., 2008). Accordingly, future studies should address the role of drp-1 in regulating the link between apoptosis and autophagy. A reverse mutation assay by introducing drp-1 back to the drp-1 mutants would be profitable in addressing this issue. Moreover, additional studies are required to determine the role of altered mitochondrial function in mediating CoCl₂-induced neurodegenerative changes. Moreover, we would like to point out that the mild effects of drp-1 inactivation seen herein could be explained by the involvement of other mitochondrial dynamic regulators such as fzo-1 and eat-3 (fusion) or miro-1 (mitochondrial transport and localization). Mutations in fzo-1 or eat-3 have been previously shown to significantly reduce the median lifespan of worms (Byrne et al., 2019; Luz et al., 2017). The disruption of miro-1 reduces the number of mitochondria by half and extends life span (Shen et al., 2016). These studies indicate potential roles for fzo-1, eat-3, and miro-1 in CoCl₂-induced lifespan alterations. Combined with our mRNA results showing that fzo-1 and miro-1 remains unchanged, the proposed regulation of drp-1 is most likely at translational or posttranslational level. Further explorations of these genes upon CoCl₂ exposure are clearly warranted.

In summary, our novel findings show that in nematodes, CoCl₂ exposure is associated with increased lethality and oxidative stress, as well as activation of autophagy and induction of apoptosis. Furthermore, CoCl₂ exposure causes mitochondrial fragmentation in C. elegans via the activation of drp-1, the inducer of mitochondrial fission. Inhibition of mitochondrial fission by drp-1 mutation rescues the CoCl₂-induced growth defects and mitochondrial fission. Our work establishes the relevance of oxidative stress and mitochondrial homeostasis to the pathogenesis of cobalt-induced toxicity.

SUPPLEMENTARY DATA

Supplementary data are available at Toxicological Sciences online.

DECLARATION OF CONFLICTING INTERESTS

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

FUNDING

NCI Cancer Center support grant (P30CA013330) and a shared instrumentation grant (1S10OD023591-01) for the use of Leica SP8 Confocal Microscope; National Institutes of Health (Grant Nos. R01ES07331 and R01ES10563); the National Natural Science Foundation of China (Grant Nos. 81903352 and 81973083); the Provincial Natural Science Foundation of Fujian Province (Grant No. 2019J05081); the high-level personnel research startup funding of Fujian Medical University (Grant No. XRCZX2018002); Sailing Funding of Fujian Medical University (Grant No. 2017XQ1010); and the Joint Funds for the Innovation of Science and Technology, Fujian province (Grant No. 2017Y9105).

ACKNOWLEDGMENT

We like to thank Andrea Briceno (Analytical Imaging Facility, Albert Einstein College of Medicine) for her training of the microscope.