Hazardous Effects of Curcumin on Mouse Embryonic Development through a Mitochondria-Dependent Apoptotic Signaling Pathway

In this study, we examined the cytotoxic effects of curcumin, the yellow pigment of Curcuma longa, on the blastocyst stage of mouse embryos, subsequent embryonic attachment, and outgrowth in vitro and in vivo implantation by embryo transfer. Mouse blastocysts were incubated in medium with or without curcumin (6, 12 or 24 μM) for 24 h. Cell proliferation and growth were investigated using dual differential staining, apoptosis was analyzed with terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL), and implantation and post-implantation development of embryos were measured by in vitro development analysis and in vivo embryo transfer, respectively. Blastocysts treated with 24 μM curcumin displayed significantly increased apoptosis and decreased total cell number. Interestingly, we observed no marked differences in the implantation success rates between curcumin-pretreated and control blastocysts during in vitro embryonic development through implantation with a fibronectin-coated culture dish. However, in vitro treatment with 24 μM curcumin was associated with decreased implantation rate and increased resorption of postimplantation embryos in mouse uterus, as well as decreased fetal weight in the embryo transfer assay. Our results collectively indicate that in vitro exposure to curcumin triggers apoptosis and retards early postimplantation development after transfer to host mice. In addition, curcumin induces apoptotic injury effects on mouse blastocysts through ROS generation, and further promotes mitochondria-dependent apoptotic signaling processes to impair sequent embryonic development.


Introduction
Curcumin, a common dietary pigment and spice, is a hydrophobic polyphenol derived from the rhizome of the herb Curcuma longa that is used as a traditional Indian medicine [1] for the treatment of wounds, liver ailments, hepatitis and urinary tract diseases, as well as a cosmetic compound [2].
Curcumin exerts a wide range of pharmacological effects, including anti-inflammatory, anti-carcinogenic, hypocholesterolemic and anti-infection activities [3][4][5][6][7][8]. As a potential antioxidant, curcumin displays anti-proliferative and anti-carcinogenic properties in a variety of cell lines and animals [8][9][10][11][12]. Moreover, the efficacy of curcumin in various diseases, including cancer, is well established [13]. Recent studies have shown that the anti-tumor activity of curcumin is attributed to its ability to induce apoptosis via caspase-3 activation [14,15]. Moreover, various in vivo animal assay models and human studies confirm that dietary curcumin is extremely safe and does not exert hazardous effects, even at high doses [16][17][18][19]. For example, three separate phase I clinical trials demonstrate that dietary curcumin administered at doses as high as 12 g per day is well tolerated [18][19][20]. Curcumin displays high pharmacological safety and efficacy, and is thus a potential candidate agent for the treatment and prevention of a wide range of human diseases. Importantly, a recent study by our group shows that curcumin inhibits methylglyoxal-induced reactive oxygen species (ROS) generation and various apoptotic biochemical events in embryonic stem cells and blastocysts isolated from pregnant mice [21]. Moreover, another study by our group focusing on the possible effects of curcumin on ROS generation, intracellular adenosine triphosphate (ATP) levels and cell death mode in osteoblast cells revealed that curcumin induces apoptosis or necrosis in a dose-dependent manner [15].
However, while multiple biological functions have been identified for curcumin, the ambiguous issue of its activity as an apoptotic inducer or inhibitor and the precise molecular mechanisms underlying these actions are yet to be fully determined. To date, virtually no studies have investigated the potential of curcumin as a cytotoxic agent against embryo development.
Apoptosis plays important roles in development and disease [22]. While apoptosis is an established contributor to normal embryonic development [23][24][25], several other studies have shown that mechanistically diverse teratogens induce excessive apoptosis in early embryos, leading to developmental impairment [21,[26][27][28][29][30]. Importantly, a recent investigation by our group revealed that curcumin induces apoptotic changes, including c-Jun N-terminal kinase (JNK) activation, caspase-3 activation, and cleavage of poly-(ADP-ribose) polymerase (PARP) and p21-activated kinase 2 (PAK2) at treatment concentrations less than 25 μM in human osteoblast cells. In contrast, 50-200 μM curcumin did not induce apoptosis but triggered necrotic cell death in human osteoblasts [15]. In a further study, the curcumin dosage was show to determine its possible effects on ROS generation, intracellular ATP levels, and apoptosis or necrosis in osteoblast cells [15]. These findings collectively indicate that curcumin promotes apoptosis or necrosis in a dose-dependent manner in human osteoblast cells. To our knowledge, the present report is the first to show that the curcumin dosage significantly influences the cell death mode of osteoblasts. These novel findings provide important insights into the impact of curcumin on other mammalian cell lines, particularly in terms of embryonic stem cells or embryonic development.
Here, we examined whether curcumin has cytotoxic effects and determined the related regulatory mechanisms in mouse embryonic development. Our results show that curcumin suppresses embryonic cell proliferation during the blastocyst stage predominantly by inducing apoptosis in the inner cell mass (ICM). We also monitored subsequent impairment of blastocyst development in vitro and following embryo transfer in vivo. However, the mechanisms underlying curcumin-induced apoptosis of mouse blastocysts remain to be determined.

Results and Discussion
To initially examine the possibility of curcumin-induced cytotoxicity, we treated mouse blastocysts with 6, 12 or 24 μM curcumin at 37 °C for 24 h, and monitored apoptosis using the TUNEL method. Curcumin clearly induced apoptosis in mouse blastocysts at a concentration of 24 μM ( Figure 1A).
Quantitative analysis revealed 7.5-fold higher apoptosis in curcumin-treated blastocysts versus untreated controls ( Figure 1B), confirming curcumin-induced apoptosis in mouse blastocysts. Values are presented as means ± SEM. *** P < 0.001 versus the control group. Apoptotic cells/embryo *** We further used differential staining for cell counting to examine proliferation in blastocysts treated with 6, 12 or 24 μM curcumin or left untreated for 24 h. The results revealed significantly lower total and ICM cell numbers in curcumin-treated blastocysts (24 μM) versus controls ( Figure 2A). Annexin V and PI staining were further used to detect the cell death modes in ICM or TE or both. We observed significantly higher numbers of Annexin V-positive/PI-negative (apoptotic) cells in the ICM of treated blastocysts versus controls, but no differences in the TE ( Figure 2B). Thus, it seems that curcumin induces significant apoptosis in the ICM but not TE of mouse blastocysts. These results further indicate that curcumin impairs the developmental potential of blastocysts by inducing ICM apoptosis.  To further determine the effects of curcumin on blastocyst development in vivo, we transferred mouse blastocysts subjected to curcumin pretreatment or left untreated, and examined the uterine content at 13 days post-transfer (day 18 post-coitus). The implantation ratio in the curcumin-pretreated group (24 μM) was significantly lower than that of the untreated control group ( Figure 4A). In addition, embryos that implanted but failed to develop were subsequently resorbed. Early resorption was visualized as small dark moles with no distinct structures, whereas late resorptions resembled placentas with no fetal structure. The proportion of implanted embryos that failed to develop normally (early and late resorption groups) was significantly higher in the 24 μM curcumin-pretreated group versus the control group ( Figure 4A). Interestingly, no differences in placental weight were evident between the curcumin-treated and untreated groups ( Figure 4B), but fetal weight was lower in the curcumin-treated group. Moreover, previous and recent experiments by our group reveal that 35-40% of fetuses weigh more than 600 mg, with the average weight of total surviving fetuses being ~600 ± 12 mg in the untreated control group at day 18 of pregnancy in a mouse embryo transfer assay [27,[31][32][33][34]. Fetal weight is an important indicator of developmental status, and the average fetal weight of the untreated control group was thus used as a key indicator for measuring the development status of curcumin-treated blastocysts. The results show that only about 11% of the fetuses in the 24 μM curcumin-pretreated group weighed more than 600 mg (an indicator of successful embryonic and fetal development), in contrast to 44% of control fetuses ( Figure 4C). Thus, curcumin exposure at the blastocyst stage poses a potential risk to postimplantation development. Bcl-2 are additionally relevant to the mitochondria-dependent apoptotic pathway [37,38], with high and low Bax/Bcl-2 ratios associated with lower and higher apoptotic thresholds, respectively.
Accordingly, we determined whether curcumin induces apoptosis via modulation of Bax and Bcl-2 expression. Immunostaining experiments revealed increases and decreases in the Bax and Bcl-2 levels, respectively, upon curcumin treatment of mouse blastocysts ( Figure 5B). Further investigation of the effects of curcumin on mitochondrial membrane potential (MMP) changes in mouse blastocyst cells revealed that treatment with 24 μM curcumin suppressed DiOC 6 (3) uptake into the mitochondria of mouse blastocyst cells, indicative of significant MMP loss ( Figure 5C). In addition, 24 μM curcumin significantly stimulated caspase-3 activation, an important indicator of apoptosis ( Figure 5D).  Figure 6A). Moreover, pretreatment with caspase-9 (Z-LEHD-FMK) and caspase-3 (Z-DEVD-FMK) specific inhibitors effectively blocked apoptosis, while the caspase-8 inhibitor Z-IETD-FMK had no effect ( Figure 6A). Importantly, treatment with 24 μM curcumin was associated with a lower implantation ratio, and failure of further development was effectively blocked by NAC and inhibitors of caspase-9 and caspase-3 by embryo transfer. The group pretreated with the caspase-8 inhibitor displayed no significant differences compared with the untreated control ( Figure 6B). In addition, fetal weight was lower in the 24 μM curcumin-treated group, which was effectively rescued upon pretreatment with NAC and specific caspase-9 and -3 inhibitors ( Figure 6C). Based on these data, we suggest that curcumin triggers ROS generation, in turn, activating mitochondria-dependent apoptotic processes in mouse blastocyst cells.
We have previously shown that curcumin induces osteoblast apoptosis at doses of 12.5-25 μM and necrosis at doses greater than 50 μM [15]. Interestingly, curcumin is able to both stimulate and inhibit apoptotic signaling. For instance, curcumin induces apoptosis in human melanoma (30-60 μM for 24 h) [39], human leukemia HL 60 (10-40 μM for 16-24 h) [40,41], AK-5 tumor (10 μM for 18 h) [14,42] and MCF-7 breast cancer cells (25 μM for 24 h) [43]. Conversely, curcumin (10 μM for 12 h) inhibits dexamethane-induced apoptosis in rat thymocytes, chemotherapy-induced apoptosis in breast cancer cells [44,45], and methylglyoxal-triggered apoptosis in mouse embryonic stem cells [21]. Our novel results, along with previously published findings, indicate that the treatment protocol (i.e., (C) treatment period and dosage) determines the effects of curcumin on various cell types. Moreover, during the complex and precisely orchestrated process of embryonic development, chemical or physical injury can affect normal development and lead to malformation or miscarriage of the embryo.
Thus, it is important to examine the possible teratogenic effects of various agents, including natural chemical compounds contained in food that are significant potential apoptotic inducers, such as curcumin. In the present study, we investigated whether curcumin adversely affects the blastocyst stage of mouse embryos and subsequent early postimplantation embryonic development. We report for the first time that treatment with 24 μM curcumin for 24 h induces apoptosis in mouse blastocysts ( Figure   1). Based on this finding, we further analyzed the effects of curcumin on embryonic development by incubating blastocysts in medium containing 6, 12 or 24 μM curcumin for 24 h. Our results show that curcumin suppresses cell viability in mouse blastocysts via apoptosis (Figures 1 and 2). Dual differential staining further showed that this curcumin-induced cell loss and apoptosis occurs primarily in the ICM (Figure 2).  [46], indicating that reduction in the TE cell lineage may reduce implantation and embryonic viability [47,48]. However, in our experiments, curcumin induced cell apoptosis only in the ICM and not TE, and did not have deleterious effects on embryonic attachment and outgrowth in vitro. Importantly, we observed a lower implantation rate of curcumin-pretreated embryos in the mouse uterus compared with the untreated control group in vivo using the embryo transfer assay (Figures 2-4). Previous studies have shown that a ~30% or greater reduction in the number of cells in the ICM is associated with high risk of fetal loss or developmental injury, even in cases where the implantation rate and TE cell numbers are normal [49]. In addition, the ICM cell number is essential for proper implantation, and reduction in this cell lineage may decrease embryonic viability [47,48,50]. While apoptosis is responsible for eliminating unwanted cells during normal embryonic development, this process does not normally occur at the blastocyst stage [51,52].
Excessive apoptosis before or during the blastocyst stage is likely to delete important cell lineages, affecting embryonic development and potentially leading to miscarriage or embryonic malformation [53]. Thus, in view of the observation that curcumin reduces the cell number and increases apoptosis specifically in the ICM of mouse blastocysts, we investigated the possibility that curcumin causes embryonic implantation and mortality and/or developmental delay in postimplantation mouse embryos in vitro and in vivo (Figure 2). Our results show that curcumin-treated blastocysts suffer from decreased embryonic development and increased embryonic death in vitro and implantation in vivo (Figures 3 and 4).
Mechanistically, our data emphasize that curcumin directly evokes intracellular oxidative stress ( Figure 5A), leading to ROS-mediated apoptosis in mouse blastocyst cells ( Figure 6A). In addition, these effects seem to involve the mitochondria-dependent apoptotic pathway, as evident from curcumin-induced changes in the intracellular levels of Bcl family members (Bax and Bcl-2) and loss of mitochondrial membrane potential ( Figure 5B). Our findings are consistent with previous data showing that curcumin induces apoptosis via activation of the mitochondrial pathway in human osteoblasts [15]. Given that recent studies have shown that the addition of specific compounds to commonly used cell culture media triggers generation of ROS, such as hydrogen peroxide [54,55], we co-incubated curcumin and culture medium, and measured ROS using the ferrous iron oxidation-xylenol orange method [54]. No artifactual ROS generation was detected under these conditions (data not shown). Importantly, a well-known ROS scavenger, NAC, effectively prevented curcumin-induced apoptosis in mouse blastocysts ( Figure 6A). We further determined the precise mechanisms of curcumin-triggered apoptosis in mouse blastocysts (Figures 5 and 6). Our results collectively demonstrate that curcumin triggers apoptosis in ICM cells of mouse blastocysts, leading to impairment of embryo development via ROS generation, which stimulates downstream processes through the mitochondria-dependent pathway.

Chemicals
Curcumin, Pregnant mare's serum gonadotropin (PMSG), Bovine serum albumin (BSA), sodium pyruvate and puerarin were purchased from Sigma (St. Louis, MO, USA). Human chorionic gonadotropin (hCG) was obtained from Serono (NV Organon Oss, the Netherlands). The TUNEL in situ cell death detection kit was obtained from Roche (Mannheim, Germany) and CMRL-1066 medium was from Gibco Life Technologies (Grand Island, NY, USA). Z-DEVD-FMK, Z-LEHD-FMK and Z-IETD-FMK were from Calbiochem (La Jolla, CA, USA). Expanded blastocysts from different females were pooled and randomly selected for experiments.

Curcumin Treatment and TUNEL Assay
Blastocysts were incubated in medium containing with or without 6, 12, or 24 μM curcumin for 24 h. For apoptosis detection, embryos were washed in curcumin-free medium, fixed, permeabilized and subjected to TUNEL labeling using an in situ cell death detection kit (Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer's protocol. Photographic images were taken under brightfield illumination using a fluorescence microscope.

Curcumin Treatment and Cell Proliferation
Blastocysts were incubated with or without culture medium containing indicated concentrations of curcumin (6-24 μM) for 24 h. And then, blastocysts were washed with curcumin-free medium and dual differential staining was used to facilitate counting of cell numbers in the inner cell mass (ICM) and trophectoderm (TE) [47]. Blastocysts were incubated in 0.4% pronase in M 2 -BSA medium (M 2 medium containing 0.1% bovine serum albumin) for removal of the zona pellucida. The denuded blastocysts were exposed to 1 mM trinitrobenzenesulphonic acid (TNBS) in BSA-free M 2 medium containing 0.1% polyvinylpyrrolidone (PVP) at 4 °C for 30 min, and then washed with M 2 medium (Sigma) [56]. The blastocysts were further treated with 30 μg/mL anti-dinitrophenol-BSA complex antibody in M 2 -BSA at 37 °C for 30 min, and then with M 2 medium supplemented with 10% whole guinea-pig serum as a source of complement, along with 20 μg/mL bisbenzimide and 10 μg/mL propidium iodide (PI), at 37 °C for 30 min. The immunolysed blastocysts were gently transferred to slides and protected from light before observation. Under UV light excitation, the ICM cells (which take up bisbenzimidine but exclude PI) appeared blue, whereas the TE cells (which take up both fluorochromes) appeared orange-red. Since multinucleated cells are not common in preimplantation embryos [57], the number of nuclei was considered to represent an accurate measure of the cell number.

Annexin V Staining
Blastocysts were incubated in 0, 6, 12 or 24 μM curcumin for 24 h, washed with curcumin-free culture medium, and then stained using an Annexin V-FLUOS staining kit (Roche), according to the manufacturer's instructions. Briefly, the blastocysts were incubated in M 2 -BSA for removal of the zona pellucida, washed with PBS plus 0.3% BSA, and then incubated for 60 min with a mixture of 100 μL binding buffer, 1 μL fluorescein isothiocyanate (FITC)-conjugated Annexin V and 1 μL PI. After incubation, the embryos were washed and photographed using a fluorescence microscope under fluorescent illumination. Cells staining Annexin V+/PI-were considered apoptotic, while those staining Annexin V+/PI+ were considered necrotic.

Morphological Analysis of Embryonic Development
Blastocysts were cultured according to a modification of the previously reported method [58].
Briefly, embryos were cultured in 4-well multidishes at 37 °C. For group culture, four embryos were cultured per well. The basic medium consisted of CMRL-1066 supplemented with 1 mM glutamine and 1 mM sodium pyruvate plus 50 IU/mL penicillin and 50 mg/mL streptomycin (hereafter called culture medium). For treatments, the embryos were incubated with 0, 6, 12 or 24 μM curcumin for 24 h. Thereafter, the embryos were cultured for 3 days in culture medium supplemented with 20% fetal calf serum, and for 4 days in culture medium supplemented with 20% heated-inactivated human placental cord serum, for a total culture time of 8 days from the onset of treatment. Embryos were inspected daily under a phase-contrast dissecting microscope, and developmental stages were classified according to established methods [59,60]. Under these culture conditions, each hatched blastocyst attached to the fibronectin and grew to form a cluster of ICM cells over the trophoblastic layer via in a process called TE outgrowth. After a total incubation period of 96 h, morphological scores for outgrowth were estimated. Growing embryos were classified as either 'attached' or 'outgrowth', with the latter defined by the presence of a cluster of ICM cells over the trophoblastic layer. As described previously [61], ICM clusters were scored according to shape, ranging from compact and rounded ICM (+++) to a few scattered cells (+) over the trophoblastic layer.

Blastocyst Development Following Embryo Transfer
To examine the ability of expanded blastocysts to implant and develop in vivo, the generated embryos were transferred to recipient mice. ICR females (white skin color) were mated with vasectomized males (C57BL/6J; black skin color; from National Laboratory Animal Center, Taiwan, ROC) to produce pseudopregnant dams as recipients for embryo transfer. To ensure that all fetuses in the pseudopregnant mice came from embryo transfer (white color) and not from fertilization by C57BL/6J (black color), we examined the skin color of the fetuses at day 18 post-coitus. To assess the impact of curcumin on postimplantation growth in vivo, blastocysts were exposed to 0, 6, 12 and 24 μM curcumin for 24 h, and then 8 embryos were transferred in parallel to the paired uterine horns of day 4 pseudopregnant mice. The surrogate mice were killed on day 18 post-coitus, and the frequency of implantation was calculated as the number of implantation sites per number of embryos transferred. The incidence rates of resorbed and surviving fetuses were calculated as the number of resorptions or surviving fetuses, respectively, per number of implantations. The weights of the surviving fetuses and placentae were measured immediately after dissection.

Immunofluorescent Embryo Stain
Mouse blastocyst cells were fixed by formaldehyde, permeabilized by Triton X-100, blocked by bovine serum albumin (5 mg/mL in PBS), and incubated with anti-Bax or anti-Bcl-2 antibodies (40 mg/mL) at room temperature for 3 h. After washing three times with PBS, embryos were incubated with second antibody conjugated with FITC or Rhodamine (TRITC) (1:100) at room temperature for 1 h and then observed under a fluorescence microscope.

Statistics
The data were analyzed using one-way ANOVA and t-tests and are presented as the mean ± SEM, with significance at P < 0.05.

Conclusions
In summary, we show for the first time that curcumin induces apoptosis only in the ICM of mouse blastocysts through ROS-and mitochondria-dependent pathways, leading to decreased embryonic development and viability. These findings imply that curcumin is a potential hazardous risk factor for normal embryonic development.