Dose-dependent effects of silver nanoparticles on cell death modes in mouse blastocysts induced via endoplasmic reticulum stress and mitochondrial apoptosis

Abstract

In view of the rapidly expanding medical and commercial applications of silver nanoparticles (AgNPs), their potential health risks and environmental effects are a significant growing concern. Earlier research by our group uncovered the embryotoxic potential of AgNPs, showing detrimental impacts of these nanoparticles on both pre- and post-implantation embryonic development. In the current study, we showed that low (50–100 μM) and high (200–400 μM) dose ranges of AgNPs trigger distinct cell death programs affecting mouse embryo development and further explored the underlying mechanisms. Treatment with low concentrations of AgNPs (50–100 μM) triggered ROS generation, in turn, inducing mitochondria-dependent apoptosis, and ultimately, harmful effects on embryo implantation, post-implantation development, and fetal development. Notably, high concentrations of AgNPs (200–400 μM) evoked more high-level ROS generation and endoplasmic reticulum (ER) stress-mediated necrosis. Interestingly, pre-incubation with Trolox, a strong antioxidant, reduced ROS generation in the group treated with 200–400 μM AgNPs to the level induced by 50–100 μM AgNPs, resulting in switching of the cell death mode from necrosis to apoptosis and a significant improvement in the impairment of embryonic development. Our findings additionally indicate that activation of PAK2 is a crucial step in AgNP-triggered apoptosis and sequent detrimental effects on embryonic development. Based on the collective results, we propose that the levels of ROS generated by AgNP treatment of embryos serve as a critical regulator of cell death type, leading to differential degrees of damage to embryo implantation, post-implantation development and fetal development through triggering apoptosis, necrosis or other cell death signaling cascades.

Graphical Abstract

Graphical Abstract.

Introduction

Silver nanoparticles (AgNPs) are widely used in daily products, including textiles, detergents, and a range of household items. Additionally, several biomedical materials, biomedicines, medical imaging, drug delivery technologies, diagnostics and medical devices contain or utilize AgNPs.^1–3 Owing to their widespread daily usage, these nanoparticles readily enter the human body through various pathways (for instance, skin absorption, inhalation, or medical injection) and are transported to the major organs, including liver, kidney, lung, and brain, through the circulatory system.^4,5 Several research papers have elucidated the distinct sizes of silver nanoparticles that can be released from AgNP-containing products or devices in daily use^6,7; the physical and chemical properties of these nanoparticles have also been documented in previous investigations.^8,9 Moreover, AgNPs can continually release silver ions – an effect that may be considered one of the mechanisms by which these particle induce cell death.¹⁰ Uptake of free silver ions into cells can cause deactivation of mitochondrial respiratory enzymes, triggering generation of reactive oxygen species (ROS) and interrupting ATP production.¹¹ In recent years, several innovative synthetic methodologies have been developed and the physical and chemical properties of novel nanomaterials elucidated, which may lead to a further increase in daily usage of nanomaterials in biomedicine and other products in the near future.^12–16 Accordingly, the potential hazardous effects of nanomaterials on human health and their underlying regulatory mechanisms are critical issues that require comprehensive investigation.

Several toxic effects of AgNPs on various mammalian cell types have been documented.^17–22 Data from both in vitro and in vivo studies suggest that the regulatory mechanisms of AgNP-triggered cytotoxicity involve an increase in intracellular ROS, mitochondrial dysfunction, and ER stress, in turn, inducing multiple biological and physiological responses, including tissue inflammation and/or cell death (apoptosis, necrosis, or autophagy).^4,5 Several studies have demonstrated that initiation of apoptotic processes or other cell death modes in pre-implantation stage embryos (zygotes to blastocysts) can severely impair both pre- and post-implantation embryonic development.^23–26 The collective findings indicate that AgNPs function as apoptotic inducers of biological or physiological impairment.^27,28 In-depth investigation of the embryotoxic potential of AgNPs by our research team, both in vitro and in vivo, has revealed that these nanoparticles trigger apoptotic processes in mouse blastocysts with deleterious effects on pre- and post-implantation embryonic development.²⁹ Furthermore, AgNPs exert hazardous effects on mouse oocyte maturation, fertilization, embryo implantation, and subsequent embryonic development.³⁰

Apoptosis and necrosis are distinct cell death types with different biochemical, morphological and pathological features and regulatory mechanisms.^31–33 Previous studies have revealed that the initial stimulus strength or magnitude, but not stimulus type, plays a key role in determining whether cell death occurs via the necrotic or apoptotic mode.^34–37 Notably, apoptosis and necrosis also have common features and regulatory properties, including secondary messengers, inhibitors, and activators.^38–40 These findings imply that the cell death processes are not independent of each other and potentially share common signaling pathways. However, to our knowledge, no studies to date have focused on whether the treatment dosage of AgNPs differentially affects hazardous impacts on embryonic development and the underlying regulatory mechanisms, which are addressed in the current investigation.

Materials and methods

Chemicals and reagents

Silver nanoparticles (10 nm; product number 795925), dimethyl sulfoxide (DMSO), pregnant mare serum gonadotropin (PMSG), bovine serum albumin (BSA), polyvinylpyrrolidone (PVP), bisbenzimidine, 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCF-DA), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), human chorionic gonadotropin (hCG), and sodium pyruvate were purchased from Sigma-Aldrich (St. Louis, MO, USA). The TUNEL in situ cell death detection kit (product number 11684817910) was obtained from Roche (Mannheim, Germany) and CMRL-1066 medium from Thermo Fisher Scientific (Waltham, MA, USA). Silver nanoparticles were dissolved in deionized water and stored at 4 °C for experimental use.

Embryo collection

ICR mice were obtained from the National Laboratory Animal Center (Taipei, Taiwan). All experimental procedures were approved by the Animal Research Ethics Board of Chung Yuan Christian University (Taiwan). Mouse breeding and care were conducted according to the Guide to The Care and Use of Experimental Animals (Canadian Council on Animal Care, Ottawa, 1993; ISBN: 0–919,087–18-3). Embryo collection was performed in keeping with previously published reports by our group.^29,41 In brief, 6 week-old female ICR mice were administered 5 IU PMSG via intraperitoneal injection (IP), followed by IP injection with 5 IU hCG after 48 h. A pair of mice (one injected female and one fertile male) were housed in each cage for mating overnight. Next morning, the vaginal plug of each mated female was examined and plug-positive cases defined as day 0 of gestation. Mice were sacrificed via cervical dislocation and each uterine horn flushed with medium (CMRL-1066 supplemented with 1 mM glutamine and 1 mM sodium pyruvate) on day 4 for blastocyst collection. Embryos collected from different mice were pooled, mixed, and randomly distributed for subsequent experimental procedures.

Assessment of cell numbers of embryos

Cell numbers of blastocysts were counted via Hoechst 33342 staining. Briefly, embryos were cultured in M2-BSA and co-incubated with M2 containing 20 μM Hoechst 33342 at room temperature. After 30 min of incubation, blastocysts were washed three times with M2 and nuclei counted under a fluorescence microscope (Olympus BX51, Tokyo, Japan). The number of nuclei counted represents the embryonic cell number.

TUNEL analysis of embryonic cell apoptosis

Apoptotic cells of blastocysts were evaluated using the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay according to the manufacturer’s instructions in conjunction with previous reports.^29,41,42 Briefly, embryos were fixed in 4% paraformaldehyde (PFA) at room temperature for 2 h and permeabilized via treatment with 20 μL TUNEL reaction mixture at 37 °C for 30 min. After three washes with PBS containing 0.3% (w/v) BSA, embryos were incubated with converted peroxidase (POD) solution (20 μL) at 37 °C for 30 min, followed by treatment with 20 μL 3,3′-diaminobenzidine (DAB) substrate solution for 2 min. Apoptotic cells were distinguished as black spots under a fluorescence microscope (Olympus BX51).

Lactate dehydrogenase assay

The CytoTox 96® non-radioactive cytotoxicity assay kit (Promega, Madison, WI, USA) was utilized to measure lactate dehydrogenase (LDH) activity for assessment of LDH release from the cytosol to culture supernatant during cell lysis. LDH activity in culture medium was evaluated as an index of necrosis.⁴³ Embryo culture medium (100 μL) was added into 96-well microtiter plates and LDH activity measured according to the manufacturer’s protocol. Absorbance values at 490 nm were evaluated using an ELISA reader (Ultrospec 2,100 Pro, Amersham Pharmacia Biotech, Amersham, UK), with that of culture medium alone used as a blank sample.

Assessment of embryonic development via embryo transfer assay

For the embryo transfer assay, thirty female pseudopregnant recipient mice were used as dams. The protocol was conducted in accordance with published reports by our group.^23,25 For assessment of embryo implantation rates into uteri of dams and postimplantation embryo developmental status, AgNP-treated and untreated control blastocysts were washed in culture medium three times and immediately transferred into the uterine horn of individual recipients (eight embryos per horn) for implantation. After 13 days of pregnancy, all transfer dams were sacrificed and day 13 post-transfer (day 18 post-coitus) embryo developmental status evaluated based on the number of embryo implantation sites, resorption sites, and placental and fetal weights. Embryo resorption sites were formed upon failure of implanted embryos to further develop that resulted in white spots in the uterine horn. After sacrifice of animals, placentae and fetuses were weighed immediately. The ratios of implanted, resorbed or surviving fetuses per total transferred embryos were calculated based on multiplying the number of implantations, resorptions, or surviving fetuses/total transferred embryos by 100.

Assessment of ROS levels

Embryos were incubated with 20 μM H₂DCF-DA fluorescence dye for measurement of intracellular ROS levels. Fluorescence intensity was recorded under a fluorescence microscope (Olympus BX51) and quantified using Image J software.

Measurement of caspase activity

Caspase-9 activity was assessed using a colorimetric caspase-9 assay kit (Calbiochem, San Diego, CA, USA) and caspase-3 activity evaluated using the Z-DEVD-AFC fluorogenic substrate according to previously published protocols.^44–46

Measurement of PAK2 activity via immunoprecipitation

Cell extracts of 100 embryos from each experimental group were mixed with lysis solution to a final volume of 0.5 mL. For immunoprecipitation of PAK2, cell extracts were incubated with anti-PAK2 (C15) antibody (200 μg/mL) at 4 °C for 1.5 h. Immunoprecipitated mixtures were further treated with 40 μL Protein A-Sepharose CL-4B (30% v/v; Pharmacia, Uppsala, Sweden) with shaking at 4 °C. After 1.5 h, mixtures were centrifuged at 1000 g for 5 min. Immunoprecipitated pellets were collected and washed with 1 mL solution A (20 mM Tris/HCl, pH 7.0, 0.5 mM DTT) containing 0.5 M NaCl for 3 times. After washing, centrifugation pellets were collected and resuspended in 40 μL solution A. Myelin basic protein (MBP; 0.1 mg/mL) and 0.2 mM [γ-p³²]ATP were used as the substrates for assessment of immunoprecipitated PAK2 activity. Phosphorylation of MBP protein with ³²P was measured according to previous protocols.^47,48

Quantitative real-time PCR analysis of gene expression

The mRNA levels of glucose-regulated GRP78 and GRP94 were measured via quantitative real-time PCR. Briefly, blastocyst cells were lysed with TRIzol reagent (Life Technologies, CA, USA) and total RNA purified with the RNeasy Mini kit (Qiagen, MD, USA) in keeping with the manufacturer’s protocol. Quantitative mRNA levels were assessed via real-time RT-PCR using the ABI 7,000 Prism Sequence Detection System (Applied Biosystems, CA, USA). β-Actin mRNA served as the endogenous control for quantitative normalization. Sequences of the primers used for RT-PCR were as follows: GRP78 (Forward 5’-CAT GGT TCT CAC TAA AAT GAA AGG-3′ and Reverse 5’-GCT GGT ACA GTA ACA ACT G-3′), GRP94 (Forward 5’-ACTG TTG AGG AGC CCA TGG AG G-3′ and Reverse 5’-GCT GAA GAG TCT CGC GGG AAA C-3′), and β-Actin (Forward 5’-CGT ACC ACA GGC ATT GTG ATG-3′ and Reverse 5’-CTT CTA GGA CTG GCT CGC AC-3′).

Statistical analysis

Statistical analyses were conducted using SPSS 23.0 software (SPSS Inc., Chicago, IL, USA). Data were statistically analyzed using one-way ANOVA, followed by Dunnett’s post hoc test for multiple comparisons between experimental groups. Data are expressed as means ± standard deviation (SD). Differences were considered significant at P < 0.05.

Results

Dose dependence effects of AgNPs on mouse blastocysts

To investigate the dosage effects of AgNPs on embryos, mouse blastocysts were treated with 0–400 μM AgNPs for 24 h and cell numbers measured via Hoechst staining. Our results showed a significant dose-dependent decrease in cell number upon pre-treatment with 50–400 μM AgNPs (Fig. 1A). Treatment with low-dosage AgNPs (within a range of 50–100 μM) triggered apoptosis, leading to a reduction in cell number. Interestingly, no apoptotic events were observed (Fig. 1B and C), while necrosis was detected in groups treated with high-dosage AgNPs (200–400 μM; Fig. 1D). The effects of high- and low-dose AgNPs on embryo development were further investigated via the embryo transfer assay. Following treatment with AgNPs, embryo implantation in the uterus was impaired to variable extents in a dose-dependent manner. Embryo implantation rates of all groups treated with 50–400 μM AgNPs were significantly lower than those of untreated control groups (Fig. 2A). Importantly, increasing rates of embryo resorption (implanted embryos that failed to further develop) and decreasing numbers of surviving fetuses were detected in 50–400 μM-treated groups in a dose-dependent manner (Fig. 2B). Moreover, the weights of placentae and surviving fetuses of AgNP-treated groups were markedly lower relative to those of vehicle-treated control groups (Fig. 2C and D). Our results clearly indicate that low-dose (50–100 μM) and high dose (200–400 μM) AgNP treatments trigger cell death modes that induce deleterious effects on embryo implantation and post-implantation development of mouse blastocysts.

Fig. 1.

Dosage effects of AgNPs on cell death of mouse blastocysts. Mouse blastocysts were exposed to various concentrations of AgNPs, as indicated, for 24 h. Vehicle-treated embryos were incubated with deionized water as the control group, designated 0 μM AgNPs. (A) Cell number was determined via Hoechst staining (20 μM). (B) Apoptosis was assessed through TUNEL staining and apoptosis-positive cells distinguished as black spots in the staining reaction. (C) Apoptosis-positive cells per blastocyst were counted. (D) Cell necrosis was analyzed according to activity of lactate dehydrogenase (LDH) released from cytoplasm (used as a necrotic cell index) in culture medium. Maximal LDH activity was determined following total cell lysis. Data are presented as a percentage of the maximal level (max). Values are expressed as means ± SD of six independent determinations. Different symbols indicate significant differences at P < 0.05. Scale bar = 20 μm.

Fig. 2.

Dosage effects of AgNPs on embryonic development of mouse blastocysts in vivo. Mouse blastocysts were incubated with AgNPs (0–400 μM) or control (deionized water) for 24 h and embryo implantation, fetal resorption, and fetal survival measured via the embryo transfer assay, as described in materials and methods. (A) The percentage of implantation was calculated as the number of implanted embryos in uteri per number of transferred embryos × 100. (B) The percentage of resorption or survival was calculated as the number of resorbed or surviving fetuses per number of transferred embryos × 100. (C) Placentae of 30 pseudopregnant recipient mice (dams) were weighed and the average weight calculated. (D) Weights of day 18 (day 13 post-transfer) surviving fetuses were determined. Different symbols indicate significant differences at p < 0.05.

Mechanisms of dose-dependent cell death modes in AgNP-treated embryos

To further elucidate the regulatory mechanisms contributing to the dose-dependent ability of AgNPs to trigger different cell death modes, oxidative stress levels were measured in 100 and 400 μM AgNP-incubated embryos. Treatment with both 100 and 400 μM AgNPs evoked significant intracellular ROS generation in blastocysts (Fig. 3A and B). Notably, preincubation with 100 μM Trolox effectively blocked ROS generation by 100 μM AgNPs but only partially prevented 400 μM AgNP-evoked ROS generation to a level comparable to that induced by 100 μM AgNPs (Fig. 3C and D). Moreover, prevention of 100 μM AgNP-induced intracellular ROS generation effectively inhibited the onset of apoptosis (Fig. 4A). Interestingly, pretreatment with 100 μM Trolox promoted a shift from necrotic to apoptotic cell death in blastocysts exposed to 400 μM AgNPs (Fig. 4A and B). Our results imply that AgNP-evoked intracellular ROS levels serve as a critical regulator of the cell death mode, triggering either apoptosis or necrosis. Further investigation of the mechanisms underlying AgNP-induced cell death revealed that treatment with 100 μM but not 400 μM AgNPs induced loss of mitochondrial membrane potential (MMP) and activation of caspase-9, caspase-3, and PAK2 in mouse blastocysts. These apoptotic events could be effectively prevented by pretreatment with 100 μM Trolox (Fig. 5A–D). Interestingly, preincubation of blastocysts exposed to 400 μM AgNPs with 100 μM Trolox effectively led to a shift from necrotic to apoptotic cell death, as evident from apoptotic events, such as loss of MMP as well as activation of caspase-9, caspase-3, and PAK2 (Fig. 5A–D). Expression levels of GRP78 and GRP94, two markers of ER stress in cells, were additionally increased in 400 μM AgNP-treated but not 100 μM AgNP-treated groups (Fig. 6A and B). Both ER stress markers evoked by 400 μM AgNPs were effectively blocked upon pre-incubation with 100 μM Trolox (Fig. 6A and B). Clearly, 100 μM Trolox could effectively suppress deleterious effects on embryo implantation, post-implantation development, and fetal development in the 100 μM AgNP-treated group through complete inhibition of ROS generation and apoptosis (Fig. 7A–D). In the 400 μM AgNP-treated group, pre-incubation with 100 μM Trolox partially attenuated ROS generation and induced a shift in the cell death mode, which also effectively reduced impairment of embryo implantation, post-implantation development, and fetal development. The specific effects of AgNPs were correlated with the amount of ROS generated and cell death type (Fig. 7A–D).

Fig. 3.

Dosage effects of AgNPs on ROS generation in treated blastocysts. (A, B) Mouse blastocysts were incubated with AgNPs (0–400 μM) or vehicle (deionized water) for 24 h. ROS generation was evaluated by staining with H₂DCF-DA (20 μM) and recorded via fluorescence microscopy (A). Quantification of intracellular ROS in each group via image J software (B). (C, D) Blastocysts were preincubated with or without Trolox (100 μM) for 1 h and treated with AgNPs (0–400 μM) or vehicle (deionized water) for 24 h. ROS generation was measured (C) and quantified (D). Values are expressed as means ± SD of five independent determinations. Different symbols indicate significant differences at P < 0.05.

Fig. 4.

Effects of ROS levels on AgNP-triggered cell death types. Blastocysts were preincubated with or without Trolox (100 μM) for 1 h followed by various concentrations AgNPs, as indicated, for 24 h. Blastocysts were incubated with vehicle (deionized water) as the control. Apoptosis (A) and necrosis (B) were measured as described for Fig. 1.

Figure 5.

Dosage effects of AgNPs on mitochondrial membrane potential and caspase-9, caspase-3, and PAK2 activities in mouse embryos. Blastocysts were preincubated with or without Trolox (100 μM) for 1 h followed by various doses of AgNPs, as indicated, or vehicle (deionized water) for 24 h. (A) Mitochondrial membrane potential changes were measured by incubation of embryos with 40 nM DiOC6(3) at 37 °C for 1 h. fluorescence intensity was quantified using a fluorescence ELISA reader. (B) Caspase-9 activity was determined via colorimetric analysis. (C) ELISA of caspase-3 activity using Z-DEVD-AFC as the substrate. (D) Kinase activity of immunoprecipitated PAK2 was evaluated via in vitro phosphorylation analysis using myelin basic protein as the substrate. Data were obtained using 100 blastocysts per group from three independent experiments. Different symbols indicate significant differences at P < 0.05.

Figure 6.

Dosage effects of AgNPs on GRP78 and GRP94 expression in mouse embryos. Blastocysts were preincubated with or without Trolox (100 μM) for 1 h followed by various concentrations AgNPs, as indicated, or vehicle (deionized water) for 24 h. The gene expression levels of GRP78 (A) and GRP94 (B) were quantified in relation to the housekeeping gene, β-actin, via the comparative CT method (DDCT) of real-time PCR. Data were obtained from 100 blastocysts per group. Different symbols indicate significant differences at P < 0.05.

Fig. 7.

Effects of Trolox pretreatment on embryonic development of AgNP-treated mouse blastocysts in vivo. Blastocysts were preincubated with or without Trolox (100 μM) for 1 h followed by various concentrations AgNPs, as indicated, or vehicle (deionized water) for 24 h. Embryo development was assessed using embryo transfer assay. (A) The percentage of implantation was calculated as the number of implanted embryos in uteri per number of transferred embryos × 100. (B) The percentage of resorption or survival was calculated as the number of resorbed or surviving fetuses per number of transferred embryos × 100. (C) Placentae of 30 pseudopregnant recipient mice (dams) were weighed and the average weight calculated. (D) Weights of day 18 (day 13 post-transfer) surviving fetuses were determined. Different symbols indicate significant differences at P < 0.05.

Our data indicate that AgNP-evoked ROS levels serve as a critical switch to activate cell death pathways that impair embryonic development. AgNPs at a concentration of 100 μM induce mitochondria-dependent apoptotic processes in embryos, in turn, causing cell death. Enhancement of ROS levels with 400 μM AgNPs triggers ER stress, causing cell death through necrotic processes. Remarkably, a decrease in 400 μM AgNP-evoked ROS generation from high to moderate levels via preincubation with Trolox has the potential to induce a switch in the type of cell death from ER stress-related necrosis to mitochondria-dependent apoptosis.

Activation of PAK2 is critical in apoptosis and impairment of embryo development in AgNP-treated mouse blastocysts

The potential functions of PAK2 in AgNP-triggered cell death and impairment of embryo development were further investigated. To this end, mouse blastocysts were pre-transfected with PAK2-targeted siRNA nucleotides, which led to effective and significant attenuation of PAK2 mRNA expression as well as activation and apoptosis in 100 μM AgNP-treated groups (Fig. 8A–C). Notably, 100 μM AgNP-induced impairment of embryo implantation was significantly prevented upon pre-transfection with PAK2 siRNA oligonucleotides in an embryo transfer assay (Fig. 8D). The resorption ratio was also significantly reduced and conversely, the ratio of surviving fetuses was significantly increased in the group pre-transfected with PAK2 siRNA and treated with 100 μM AgNPs relative to the group treated with 100 μM AgNPs only (Fig. 8E). Furthermore, poorer fetal development (<400 mg) and a lower percentage of fetuses weighing >600 mg were observed in the 100 μM AgNP-treated than the vehicle-treated group, which were significantly rescued by pre-transfection with PAK2 siRNA (Fig. 8F). Based on these findings, we propose that PAK2 serves as a key regulator of low-dose AgNP-triggered mitochondria-dependent apoptotic processes that cause impairment of embryo development but is not involved in high-dose AgNP-induced ER stress-related necrosis and consequent hazardous effects on embryo development.

Fig. 8.

Effects of siRNA-mediated knockdown of PAK2 on AgNP-induced apoptosis and impairment of embryo development in blastocysts. Mouse blastocysts were transfected with or without mouse PAK2-targeted small interfering RNAs (siPAK2) or control siRNA (siControl) for 72 h. Embryos were treated with AgNPs (100 μM) or vehicle (deionized water) for 24 h and their development evaluated via the embryo transfer assay. (A) PAK2 mRNA levels were quantified via RT-PCR based on five independent experiments. (B, C) Activation of PAK2 (B) and cell apoptosis (C) were assessed as described in Fig. 1 and 5. (D) The percentage of implantation represents the number of implanted embryos in uteri per number of transferred embryos × 100. (E) The percentage of resorption or survival represents the number of resorbed or surviving fetuses per number of transferred embryos × 100. (F) Weight distributions of day 18 (day 13 post-transfer) surviving fetuses from embryo transfer. Different symbols indicate significant differences at P < 0.05.

Discussion

AgNP-based nanomaterials are widely used in everyday consumer products and commercial applications.^3,49 However, accumulating evidence suggests that AgNPs promote ROS generation and trigger several detrimental cellular responses and toxic reactions, including induction of apoptosis or necrosis.^50,51 A recent study demonstrated that AgNPs exert distinct deleterious effects on peripheral organs based on the variable toxic tolerance levels, sensitivities and responses of distinct sets of differentiated tissue cells.⁵² Nanoparticles are readily transported and systemically distributed to multiple tissues and organs in the body through the circulatory system. However, physical and chemical characteristics of AgNPs, such as particle size, surface chemistry properties, exposure method and exposure time, affect absorption and tissue distribution/accumulation in various organs.⁵³ Common potential exposure routes of AgNPs contained in a plethora of daily products and commercial applications are dermal contact, oral administration (via the gastrointestinal tract), inhalation (via the respiratory tract), and blood circulation (via intravenous injection).^53–55 The variable concentrations of AgNPs accumulating in tissues and organs are related to their deleterious and toxic effects. However, the physiological metabolism and concentrations of AgNPs in specific tissues or organs remain to be established. Although the exposure doses of AgNPs in this study do not cover those relevant to all commercial AgNP-containing products, we believe that attention should be paid to their potential hazardous effects, in view of the potential of these nanoparticles to accumulate in the human body. Early-stage embryonic development processes are precisely orchestrated and developmental processes at this stage are susceptible to interference from external chemical, physical and biological teratogens or other factors that exert hazardous effects, including long-term impairment of fetal health. However, the regulatory mechanisms of action and cytotoxicities of AgNPs and their secondary metabolites in vivo in various peripheral organs after entry into the circulation are not fully understood at present.

Excess production of ROS or decreased or limited ability of the antioxidant defense system to eliminate intracellular excessive oxidative stress are important contributors to cytotoxicity, including damage or death (apoptosis or necrosis) of embryo cells.^56,57 Excessive ROS generation is clearly implicated in AgNP-induced hazardous effects.^17,19,58 Previous research has demonstrated that chemical compounds induce distinct cell death types depending on treatment intensity or stimulus strength.^36,37,59 Low doses of antimycin A, curcumin or ethanol predominantly trigger cell death through apoptotic processes, while necrotic processes are increasingly favored in response to higher doses of these compounds.^36,37,59 Similarly, our group showed that AgNPs promote ROS generation, triggering apoptotic processes and causing injury to oocyte maturation, fertilization, pre- and post-implantation embryo development and fetal weight upon exposure to doses higher than 50 μM in vitro or intravenous injection of 5 mg/kg body weight in vivo.^29,30 In this study, we further demonstrated that low doses (25–100 μM) and high doses (200–400 μM) of AgNPs evoke distinct intracellular ROS levels, resulting in different cell death modes in mouse blastocysts. Exposure to low-dose (25–100 μM) AgNPs elicited ROS generation to promote apoptotic processes (Figs. 3–5) while high-dose (200–400 μM) AgNPs induced more high-level ROS generation and promoted cell death via necrotic processes in mouse blastocysts (Figs. 3, 4, 6). Intravenous administration of 5 mg/kg (body weight) AgNPs is associated with upregulation of HO1, GPX and SOD1 mRNA levels in mouse liver and kidney.⁶⁰ The collective results clearly suggest that AgNPs evoke varying levels of intracellular ROS in a dose-dependent manner that trigger distinct regulated cell death mechanisms in mouse blastocysts.

The ER is a key organelle involved in membrane-bound ribosome-directed protein translation and posttranslational modifications and folding. External physical factors or chemical stimuli can induce malfunction of translation processes, causing unfolding or misfolding proteins to accumulate in the ER lumen that trigger the unfolded protein response to ER stress.⁶¹ GRP78 and GRP94 are two critical protein chaperones that operate as master regulators for detection of unfolded proteins, initiating three distinct UPR branch responses.^61,62 ER stress is involved in distinct cell death modes, including apoptosis and necrosis.^63,64 In the current investigation, the highest dose of AgNPs (400 μM) evoked large-scale ROS generation (about a 10-fold higher relative to the control group) that promoted ER stress-dependent necrosis but not apoptotic cell death in mouse blastocysts (Figs. 3, 4, 6). An earlier study showed that ER stress triggers mitochondrial dysfunction that ultimately leads to apoptosis, which causes irreversible damage through specific signaling pathways, including disruption of MMP and activation of apoptosis-related caspases⁶⁵ . In our experiments, treatment with low-dose AgNPs (100 μM) induced 6-fold higher intracellular ROS generation relative to the control group, leading to the activation of mitochondria-dependent apoptosis but not necrosis (Figs 3-5). Notably, preincubation with Trolox, a classic antioxidant, effectively blocked ROS generation by 100 μM AgNPs, preventing apoptosis and sequent impairment of embryo development (Figs. 3, 4, 7). Interestingly, Trolox induced a significant reduction in ROS generation by 400 μM AgNPs to a level comparable to that of the 100 μM AgNP treatment group, with a consequent shift in cell death type from ER stress-related necrosis to mitochondria-dependent apoptosis that also exerted sequent deleterious effects on embryo development (Figs. 4–7). These results highlight that the levels of intracellular ROS generated in AgNP-treated embryos function as a key upstream determinant of cell death type (apoptosis vs. necrosis), with resulting detrimental effects on embryonic development.

Previous studies have demonstrated the involvement of PAKs in regulatory signaling pathways of cell transformation, proliferation, survival, and apoptosis.^66–68 Over the past few years, our research group has clearly demonstrated the significance of caspase-mediated PAK2 cleavage/activation in various physiological and chemical apoptotic inducers that trigger mitochondria-dependent apoptotic processes.^46,69–73 Data from the current study showed that PAK2 is activated in 100 μM AgNP-induced apoptosis, but not 400 μM AgNP-induced necrosis (Fig. 5). Downregulation of PAK2 mRNA by pre-transfection with siPAK2 effectively suppressed activation of PAK2 and caused a significant decrease in apoptosis and impairment of embryo implantation and sequent post-implantation embryo and fetal development in 100 μM AgNP-treated blastocysts (Fig. 8). Overall, these findings clearly indicate that AgNPs induce activation of PAK2, which plays a critical role in stimulation of apoptotic pathways in mouse blastocysts. Further research is warranted to establish the downstream substrates phosphorylated and regulated by PAK2 in apoptotic processes during embryonic development.

Conclusion

In summary, AgNPs serve as a trigger of distinct cell death modes in a dose-dependent manner. Our results showed that treatment with low-dose (50–100 μM) AgNPs induced ROS-dependent apoptotic processes while high-dose (200–400 μM) AgNPs evoked more high-level ROS production that triggered ER stress-mediated necrosis. Pre-incubation with Trolox suppressed ROS generation in 200–400 μM AgNP-treated groups to a level comparable to that induced by 50–100 μM AgNPs, with a consequent shift in the cell death mode from necrosis to apoptosis. In consideration of these findings, we propose that ROS levels accumulating in AgNP-treated embryos regulate the cell death program, leading to differential degrees of damage. Moreover, activation of PAK2 appears essential for AgNP-triggered apoptosis and sequent harmful effects on embryonic development. Overall, both apoptosis and necrosis induced by AgNPs exert detrimental effects on embryo and fetal development.

Author contributions

Wen-Hsiung Chan conceived and designed the experiments; Cheng-Kai Lee, Fu-Ting Wang and Chien-Hsun Huang performed the experiments; all authors analyzed and interpreted the data; Wen-Hsiung Chan wrote the paper; all authors contributed to and approved the final version of the manuscript.

Funding

This work was supported by a grant from the National Science Council of Taiwan, ROC (MOST 109–2311-B-033-001).

Conflict of interest statement. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.