Abstract
The tripeptide thiol antioxidant glutathione (GSH) has multiple physiological functions. Female mice lacking the modifier subunit of glutamate cysteine ligase (GCLM), the rate-limiting enzyme in GSH synthesis, have decreased GSH concentrations, ovarian oxidative stress, preimplantation embryonic mortality, and accelerated age-related decline in ovarian follicles. We hypothesized that supplementation with thiol antioxidants, N-acetyl cysteine (NAC), or α-lipoic acid (ALA) will rescue this phenotype. Gclm−/− and Gclm+/+ females received 0 or 80 mM NAC in drinking water from postnatal day (PND) 21–30; follicle growth was induced with equine chorionic gonadotropin (eCG) on PND 27, followed by an ovulatory dose of human CG and mating with a wild type male on PND 29 and zygote harvest 20 h after hCG. N-acetyl cysteine supplementation failed to rescue the low rate of second pronucleus formation in zygotes from Gclm−/− versus Gclm+/+ females. In the second study, Gclm−/− and Gclm+/+ females received diet containing 0, 150, or 600 mg/kg ALA beginning at weaning and were mated with wild type males from 8 to 20 weeks of age. α-Lipoic acid failed to rescue the decreased offspring production of Gclm−/− females. However, 150 mg/kg diet ALA partially rescued the accelerated decline in primordial follicles, as well as the increased recruitment of follicles into the growing pool and the increased percentages of follicles with γH2AX positive oocytes or granulosa cells of Gclm−/− females. We conclude that ovarian oxidative stress is the cause of accelerated primordial follicle decline, while GSH deficiency per se may be responsible for preimplantation embryonic mortality in Gclm−/− females.
Introduction
Premature ovarian failure (POF, also called primary ovarian insufficiency) is manifested by accelerated depletion of the ovarian follicle pool and poor oocyte quality. Clinically, POF is defined as cessation of ovarian function before the age of 40, while the normal age of menopause is 50 ± 4 years [1, 2]. Premature ovarian failure affects 1% of women. Although abnormalities of the X chromosome, various rare mutations, exposure to chemotherapy, and ionizing radiation are known causes of POF, the cause remains unknown in 90% of cases [2–4]. Oxidative stress has been implicated in POF caused by exposure to toxicants and ionizing radiation [5–7].
The tripeptide glutathione (GSH, γ-glutamyl cysteinyl glycine) is the most abundant intracellular nonprotein thiol, with intracellular concentrations in the millimolar range. Glutathione has numerous intracellular functions, including antioxidant defenses, reduction of disulfide bonds, and S-glutathionylation as a post-translational protein modification [8–12]. Glutathione can scavenge free radicals directly, participate in the reduction of hydrogen peroxide and lipid peroxides as a cofactor for glutathione peroxidases and peroxiredoxin, and detoxify electrophilic toxicants as a cofactor for glutathione transferase [8, 9, 13]. Glutathione is synthesized in two ATP-dependent reactions. The first rate-limiting reaction is catalyzed by glutamate cysteine ligase (GCL), a heterodimer composed of a catalytic (GCLC), and a modifier (GCLM) subunit [14, 15]. Glutamate cysteine ligase catalytic subunit is responsible for the catalytic activity of the enzyme, while binding of GCLM to GCLC increases GCL enzymatic activity [14, 15]. Gclc−/− mice die during embryonic development [16], whereas Gclm−/− mice survive to advanced age, despite greatly decreased tissue levels of GSH ([17, 18] and our unpublished observations). Our prior studies showed that Gclm−/− mice have low ovarian GSH concentrations, chronic ovarian oxidative stress, poor oocyte quality resulting in delayed male pronucleus formation in zygotes and early preimplantation embryonic mortality, and accelerated age-related decline in ovarian follicle numbers [19, 20]. The Gclm−/− mouse model thus recapitulates the decreased oocyte quality and early depletion of the ovarian reserve, which are key characteristics of idiopathic POF in women. Moreover, these prior findings support a role for ovarian oxidative stress as a driver of POF.
N-acetyl cysteine (NAC) increases GSH synthesis by providing a source of cysteine, the least abundant amino acid constituent of GSH. Glutathione itself is not taken up by cells [21, 22]. N-acetyl cysteine is orally bioavailable, non-toxic, and has been shown to be protective against acetaminophen-induced hepatotoxicity, which is caused by depletion of GSH and induction of mitochondrial oxidative stress by the reactive metabolite N-acetyl-p-benzoquinone imine [23]. N-acetyl cysteine is used clinically to treat acetaminophen hepatotoxicity in humans [24]. N-acetyl cysteine administered at 40 mM in the drinking water has been reported to increase pulmonary and blood, but not hepatic or brain, GSH levels and to protect against CNS oxygen toxicity in rats [25]. N-acetyl cysteine at 20 mM in drinking water administered to pregnant dams increased GSH levels in control mouse fetuses, but not in fetuses simultaneously exposed to buthionine sulfoximine, a specific inhibitor of GCL that depletes GSH by inhibiting its synthesis [26]. N-acetyl cysteine treatment did however protect against DNA damage caused by buthionine sulfoximine-induced GSH depletion in the fetuses, despite no increase in GSH concentrations [26]. Other studies also support antioxidant effects of NAC that are independent of its ability to enhance GSH synthesis [27, 28].
α-Lipoic acid (ALA) is a dithiol that is generated enzymatically in mitochondria from octanoic acid, where it is a necessary cofactor for α-ketoacid dehydrogenases [29]. α-Lipoic acid is also absorbed from dietary sources and is widely available as a nutritional supplement [29, 30]. In vitro, ALA is a direct acting antioxidant and metal chelator; however, concentrations required for these effects may not be achievable in vivo [29]. In vivo, ALA increases ascorbate and cysteine uptake from plasma and induces expression of numerous antioxidant genes, including Gclc and Gclm, by increasing de novo translation of the transcription factor Nuclear Factor, Erythroid 2 like 2 (NFE2L2, also known as NRF2) [29, 31, 32]. Dietary supplementation with ALA has been reported to protect against several conditions in which oxidative stress is thought to play a role, including diabetic neuropathy and insulin resistance, in humans [33–35]. Supplementation with dietary ALA protected C57BL/6 J mice against age-related hearing loss, which is caused by upregulation of proapoptotic genes in the cochlea by mitochondrial ROS [36]. Supplementation with intraperitoneal injections of ALA reportedly protected mice against depletion of the ovarian reserve by treatment with the anticancer drug methotrexate [37]. α-Lipoic acid supplementation decreased age-related mitochondrial oxidative damage and GSSG concentrations and increased mitochondrial GSH concentrations in the brains of rats [38]. Supplementation of media with ALA, NAC, and acetyl-L-carnitine alone or in combination during culture of mouse zygotes has been reported to improve development to the blastocyst stage [39].
If the hypothesis that oxidative stress is driving either the early embryonic mortality and/or the accelerated age-related follicle decline in Gclm−/− mice is correct, then providing these mice with an additional means of protection against oxidative stress should rescue these phenotypes. We first tested the efficacy of short-term supplementation with NAC during the later stages of oocyte maturation to ameliorate the reduced offspring production due to early embryonic mortality in Gclm−/− mice. We then tested whether long-term supplementation with ALA could rescue the decreased offspring production and/or slow the accelerated decline in ovarian follicles observed in Gclm−/− mice.
Materials and methods
Reagents
N-acetyl cysteine (# A7250) and ALA (# 62320) were obtained from Sigma Aldrich. α-Lipoic acid was sent to Bio-Serv (Flemington, NJ) to formulate three different diets: (1) control diet was Rodent Diet, AIN-93 M ½″ pellets (# F3155), (2) AIN-93 M diet supplemented with 150 mg/kg ALA, and (3) AIN-93 M diet supplemented with 600 mg/kg ALA. The caloric profile of the rodent diet was a total of 3.58 kcal/g, with 0.51 kcal/g as protein, 0.37 kcal/g as fat, and 2.70 kcal/g as carbohydrate. All other reagents and chemicals were obtained from Sigma Aldrich or Fisher Scientific unless otherwise noted.
Animals
Gclm−/− mice were generated by disrupting the Gclm gene by replacing exon 1 with a β-galactosidase/neomycin phosphotransferase fusion gene and backcrossing onto a C57BL/6 J genetic background [18, 40]. Mice for these experiments were bred by crossing Gclm+/− males and females in our colony. Mice were housed in an American Association for the Accreditation of Laboratory Animal Care-accredited facility on a 14:10 h light-dark cycle, with temperature maintained at 69–75 °F. Mice had free access to deionized water and soy-free laboratory chow (Harlan Teklad 2919), except when placed on experimental diets described below. The experimental protocols were carried out in accordance with the Guide for the Care and Use of Laboratory Animals [41] and were approved by the Institutional Animal Care and Use Committee at the University of California Irvine.
Experimental protocols
For all experiments, Gclm+/+ and Gclm−/− females were randomly assigned to treatment groups using a random number table, avoiding assignment of littermates of the same genotype to the same experimental group. All endpoints were analyzed by investigators blind to experimental group.
Experiment 1: Effect of short-term NAC supplementation during later stages of oocyte maturation on fertilization and zygote development. Gclm+/+ and Gclm−/− mice were given ad libitum access to NAC-supplemented drinking water (80 mM, pH adjusted to 6.0 with fresh NaOH) or unsupplemented water beginning at 21 days of age, then injected with equine chorionic gonadotropin (eCG) to induce follicle development at 27 days of age, followed by injection of hCG to stimulate ovulation (while continuing to receive NAC in the drinking water) 46 h later. The NAC concentration was chosen to be higher than concentrations found to be effective in previous studies [25, 26] because we hypothesized that the Gclm−/− mice would require higher concentrations to use NAC directly as an antioxidant due to limited ability to upregulate GSH synthesis despite increased cysteine. N-acetyl cysteine-supplemented drinking water was prepared fresh every 2 days [25]. Water consumption and body weight were monitored throughout the study. Females were mated with wild type males the evening of the day of hCG administration. Females were euthanized by CO2 inhalation, and zygotes were collected from the oviducts 20 h after hCG, as previously described [19]. Numbers of oocytes without evidence of fertilization, oocytes with single sperm or multiple sperm without pronuclei, zygotes with one pronucleus, zygotes with two pronuclei, and fragmented oocytes were counted using DIC microscopy with a Zeiss Axiovert microscope.
Experiment 2: Effect of long-term ALA supplementation on fertility and ovarian follicle numbers. In a preliminary experiment, we found no differences in the amounts of ALA-supplemented diet consumed compared with control diet (data not shown). Therefore, we concluded that pair-feeding was not necessary. Gclm+/+ and Gclm−/− females were fed control AIN-93M chow or AIN-93M chow supplemented with 150 or 600 mg/kg chow of the antioxidant ALA from postnatal day (PND) 21 until euthanasia at approximately 5.5 months of age (N = 8/group for ovarian endpoints, except Gclm+/+, 600 mg/kg group N = 7; one female removed because of incorrect genotype assignment). Throughout this experiment, body weights and feed consumed were measured biweekly.
From 2 months of age, each female was individually housed with a proven breeder wild type male. Females were checked each morning for 12 weeks (84 days) for abdominal distention indicative of pregnancy and for births. When a birth was observed, the number of live and dead pups, and sex and weight of live pups were recorded. Newborn pups were euthanized by hypothermia followed by decapitation. At the end of 12 weeks of the breeding study or after the last litter was born if that occurred after 12 weeks, vaginal cytology was performed for at least 2 weeks to monitor estrous cycles, and mice were euthanized on the day of metestrus of the cycle by CO2 inhalation. One ovary was fixed in Bouin’s solution for follicle counts and the other was fixed in 4% paraformaldehyde (PFA) for immunostaining. For the breeding study endpoints, one female was removed from the Gclm+/+, 600 mg/kg group (N = 6 for breeding endpoints) and one from the Gclm−/−, 150 mg/kg ALA group (N = 7 for breeding endpoints) because they were inadvertently not checked for birth for several days so pup number and weights at birth could not be ascertained.
Experiments 3 and 4: Effects of NAC or ALA supplementation on ovarian GSH concentrations. To determine if NAC administration increases ovarian GSH concentrations in Gclm−/− and Gclm+/+ mice, 2-month old Gclm−/− and Gclm+/+ female mice were given ad libitum access to NAC-supplemented (80 mM) or unsupplemented drinking water for 7 days, timed based on the mouse’s estrous cycle so that they were euthanized on the day of metestrus. Water consumption and body weight were monitored throughout the study. Mice were euthanized by CO2 inhalation, and ovaries plus attached oviducts were homogenized in six times volume in microlitre per milligram tissue weight 20 mM Tris, 1 mM EDTA, 250 mM sucrose, 2 mM l-serine, 20 mM boric acid (TES-SB) for GSH and GSSG assay as we have previously described [42, 43]. Briefly, after removal of a 10 μL aliquot of homogenate for protein assay, remaining homogenates were acidified with one quarter volume 5% sulfosalicylic acid; the homogenates were processed directly for total GSH assays (GSH plus GSSG). For GSSG assay, GSH in the homogenates was first conjugated with 2-vinylpyridine and extracted with chloroform, and the aqueous phase was then assayed. Triplicates of samples or standards were combined with 33 μL metal free water and incubated for 10 min at 30 °C. The samples were mixed with 140 μL of 0.3 mM NADPH, 20 μL of 6 mM 5,5′-dithiobis (2-nitrobenzoic acid) (DTNB), and 2 μL of 50 units/mL GSH reductase. The rate of thiobis (2-nitrobenzoic acid) (TNB) formation from DTNB is proportional to the total GSH concentration in each sample. Thiobis (2-nitrobenzoic acid) formation was monitored by measuring the absorbance at 412 nm for 5 min every 10 s using a VersaMax microplate spectrophotometer (Molecular Devices, Sunnyvale, CA). The concentrations of total GSH in the samples were calculated from a standard curve generated from the slopes of the standards.
To determine if ALA supplementation increases ovarian GSH concentrations, 21-day old Gclm−/− and Gclm+/+ female mice were given ad libitum access to control or ALA supplemented diets. Beginning at 8 week of age, vaginal cytology was performed daily for at least two estrous cycles, and mice were euthanized by CO2 on the day of metestrus. Both ovaries were processed for total GSH assay as above.
Estrous cycling
For evaluation of estrous cycling, mice were individually housed, and vaginal lavage was performed every morning using 0.9% NaCl for at least 14 days. Cells in the lavage fluid were examined by light microscopy immediately after collection, and the predominant cell types present in the fluid were recorded as previously described [44].
Ovarian follicle counts
The ovaries from ALA-supplemented or control mice were fixed in Bouin’s solution overnight at 4 °C and embedded as a group in blocks (five ovaries per block) of glycol methacrylate resin (Technovit 8100; Heraeus Kulzer GmbH, Wehrheim, Germany) to maximize sectioning and counting efficiency. Each block was serially sectioned at 20 μm followed by hematoxylin and eosin staining. Every fifth section of each ovary was used to estimate the total number of primordial and primary follicles using stereological software, Stereo Investigator 2017 (MBF Bioscience Inc., Williston, VT, USA) in conjunction with an Olympus BX-40 microscope fitted with 4× PLAN FLUO and 60× (NA = 1.4) U PLAN FLUO oil immersion objectives and a motorized stage controlled via the software. A 2D counting frame (45 μm × 45 μm; 2025 μm2) was generated and superimposed over each ovary section by the Optical Fractionator probe for systematic random sampling and then shifted progressively along the XY-axis in a sampling grid that subdivided the sections into 90 μm × 90 μm (8100 μm2) squares. Oocytes with nucleus of each follicle were counted if they fell entirely or partially within the inclusion boundaries of the counting frame and if they came into sharp focus within the middle 8 μm of the section (Z-axis). The upper (2 μm) and lower (2 μm) portions of the sections were excluded from counting to avoid sectioning artifact. Raw counts from each ovary were then converted to an estimate of the total number in the entire ovary by multiplying times the reciprocal of the counting fractions: raw count × section interval (5) × grid frame area/counting frame area (8100/2025) × average section thickness/counting frame height × 12/8 [45, 46]. Secondary and antral follicles were counted using an Olympus BX-60 microscope by following through every section of the ovary to avoid double counting. Follicles were classified as primordial, primary, secondary, and antral and as healthy or atretic as previously described [20].
Immunohistochemistry
All PFA fixed ovaries from the ALA experiment were cryoprotected in 30% sucrose in PBS at 4 °C, embedded in optimal cutting temperature compound, and stored at −80 °C before being serially sectioned using a cryostat at a thickness of 10 μm. Immunohistochemistry was performed using the Vectastain ABC kit (PK-4001; Vector Laboratories, Burlingame, CA, USA). Briefly, sections were thawed and heated for 15 min at 95 °C in a 10 mM citrate buffer (pH 6.0). The primary antibodies—rabbit anti-cleaved caspase-3 Asp 175 (1:100; Cell Signaling #9664, Beverly, MA, USA), rabbit anti-Ki67 (1:500; Abcam #15580, Cambridge, MA, USA), and rabbit anti-γH2AX (1:200; Cell Signaling #9718)—were detected using biotinylated goat anti-rabbit secondary antibody in 5% normal goat serum. All immunostaining procedures included avidin/biotin and 3% hydrogen peroxide blocking steps. Peroxidase activity was visualized using 3,3′-diaminobenzidine as substrate (Roche). Sections were counterstained with hematoxylin. The following negative controls were included in every experiment: secondary antibody without primary antibody; primary antibody without secondary antibody; and primary antibody replaced by rabbit IgG with secondary antibody. Both negative and positive follicles for each marker were counted and quantities expressed as fractions were used for statistical analyses.
The numbers of Ki67, cleaved caspase-3, and phosphorylated histone 2AX (γH2AX)-positive and -negative follicles or oocytes were counted in one section per slide for 3 or 4 slides per endpoint distributed throughout the ovary by an investigator (JL, LL, and SA) blind to Gclm genotype and ALA treatment using an Olympus BX-60 microscope with 10, 20, and 40× U PLAN FLUO objectives equipped with a Retiga 2000R cooled CCD digital camera system with Q-Capture Pro software. The percentages of positive primordial and primary follicles (containing one or more positive granulosa cells per largest cross-section or containing positive oocytes) and secondary and antral follicles (containing three or more positive granulosa cells per largest cross-section or containing positive oocytes) were calculated and used for data presentation and analysis.
Statistical analysis
All values are presented as mean ± SEM in figures and tables. Two-way ANOVA was used to test the effects of genotype and antioxidant supplementation on continuous endpoints (follicle numbers, number of oocytes ovulated, and GSH concentrations). Quantities expressed as fractions were subjected to arcsine square root transformation prior to ANOVA [47] or analyzed by nonparametric Kruskal–Wallis test followed by Mann–Whitney tests for intergroup comparisons. If statistically significant effects or interactions were observed in the two-way ANOVA, we performed one-way ANOVA followed by LSD test for intergroup comparisons. Statistical analyses were performed using SPSS 25 for Mac OS X (IBM Software).
Results
N-acetyl cysteine supplementation does not rescue zygote development in Gclm−/− females
This experiment was designed to test whether short-term treatment with the thiol antioxidant NAC during the later stages of oocyte maturation could rescue the decreased ability of fertilized eggs from Gclm−/− females to reach the two pronucleus stage by 0.5 days post-coitum [19]. Consistent with our prior findings [19], Gclm+/+ and Gclm−/− females ovulated similar numbers of oocytes in response to exogenous gonadotropin stimulation, but at 0.5 days post coitum, Gclm−/− females had significantly higher percentages of zygotes with one pronucleus (P = 0.010, effect of genotype) and significantly lower percentages of zygotes with two pronuclei (P < 0.001, effect of genotype) compared with Gclm+/+ females (Table 1). Supplementation with 80 mM NAC in drinking water for 1 week prior to superovulation failed to rescue the low rate of second pronucleus formation in zygotes from Gclm−/− females compared with zygotes from Gclm+/+ females (Table 1). Both Gclm+/+ and Gclm−/− females supplemented with NAC tended to ovulate fewer oocytes than females drinking unsupplemented water (P = 0.050, effect of NAC).
Effects of NAC supplementation on ovulation and early embryonic development
| 0 mM NAC | 80 mM NAC | |||
|---|---|---|---|---|
| Gclm+/+ (N = 5) | Gclm−/− (N = 5) | Gclm+/+ (N = 8) | Gclm−/− (N = 5) | |
| # oocytes ovulateda | 54.8 ± 4.2 | 58.4 ± 2.7 | 40.4 ± 10.7 | 32.0 ± 12.3 |
| % unfertilized or fragmentedb | 19.7 ± 5.4 | 29.5 ± 3.4 | 59.0 ± 9.0 | 26.2 ± 9.9 |
| % fertilizedc | 80.3 ± 5.4 | 70.5 ± 3.4 | 41.0 ± 9.0 | 73.8 ± 9.9 |
| % zygotes, 1 pronucleusd | 1.1 ± 0.8 | 34.2 ± 4.5 | 9.5 ± 6.6 | 42.1 ± 15.2 |
| % zygotes, 2 pronucleie | 53.5 ± 11.8 | 7.2 ± 3.6 | 21.6 ± 9.7 | 9.6 ± 6.3 |
| 0 mM NAC | 80 mM NAC | |||
|---|---|---|---|---|
| Gclm+/+ (N = 5) | Gclm−/− (N = 5) | Gclm+/+ (N = 8) | Gclm−/− (N = 5) | |
| # oocytes ovulateda | 54.8 ± 4.2 | 58.4 ± 2.7 | 40.4 ± 10.7 | 32.0 ± 12.3 |
| % unfertilized or fragmentedb | 19.7 ± 5.4 | 29.5 ± 3.4 | 59.0 ± 9.0 | 26.2 ± 9.9 |
| % fertilizedc | 80.3 ± 5.4 | 70.5 ± 3.4 | 41.0 ± 9.0 | 73.8 ± 9.9 |
| % zygotes, 1 pronucleusd | 1.1 ± 0.8 | 34.2 ± 4.5 | 9.5 ± 6.6 | 42.1 ± 15.2 |
| % zygotes, 2 pronucleie | 53.5 ± 11.8 | 7.2 ± 3.6 | 21.6 ± 9.7 | 9.6 ± 6.3 |
aNS, effects of genotype and NAC.
bP = 0.099, effect of NAC; P = 0.017, genotype × NAC.
cP = 0.099, effect of NAC; P = 0.017, genotype × NAC.
dP < 0.001, effect of genotype.
eP = 0.010, effect of genotype.
Effects of NAC supplementation on ovulation and early embryonic development
| 0 mM NAC | 80 mM NAC | |||
|---|---|---|---|---|
| Gclm+/+ (N = 5) | Gclm−/− (N = 5) | Gclm+/+ (N = 8) | Gclm−/− (N = 5) | |
| # oocytes ovulateda | 54.8 ± 4.2 | 58.4 ± 2.7 | 40.4 ± 10.7 | 32.0 ± 12.3 |
| % unfertilized or fragmentedb | 19.7 ± 5.4 | 29.5 ± 3.4 | 59.0 ± 9.0 | 26.2 ± 9.9 |
| % fertilizedc | 80.3 ± 5.4 | 70.5 ± 3.4 | 41.0 ± 9.0 | 73.8 ± 9.9 |
| % zygotes, 1 pronucleusd | 1.1 ± 0.8 | 34.2 ± 4.5 | 9.5 ± 6.6 | 42.1 ± 15.2 |
| % zygotes, 2 pronucleie | 53.5 ± 11.8 | 7.2 ± 3.6 | 21.6 ± 9.7 | 9.6 ± 6.3 |
| 0 mM NAC | 80 mM NAC | |||
|---|---|---|---|---|
| Gclm+/+ (N = 5) | Gclm−/− (N = 5) | Gclm+/+ (N = 8) | Gclm−/− (N = 5) | |
| # oocytes ovulateda | 54.8 ± 4.2 | 58.4 ± 2.7 | 40.4 ± 10.7 | 32.0 ± 12.3 |
| % unfertilized or fragmentedb | 19.7 ± 5.4 | 29.5 ± 3.4 | 59.0 ± 9.0 | 26.2 ± 9.9 |
| % fertilizedc | 80.3 ± 5.4 | 70.5 ± 3.4 | 41.0 ± 9.0 | 73.8 ± 9.9 |
| % zygotes, 1 pronucleusd | 1.1 ± 0.8 | 34.2 ± 4.5 | 9.5 ± 6.6 | 42.1 ± 15.2 |
| % zygotes, 2 pronucleie | 53.5 ± 11.8 | 7.2 ± 3.6 | 21.6 ± 9.7 | 9.6 ± 6.3 |
aNS, effects of genotype and NAC.
bP = 0.099, effect of NAC; P = 0.017, genotype × NAC.
cP = 0.099, effect of NAC; P = 0.017, genotype × NAC.
dP < 0.001, effect of genotype.
eP = 0.010, effect of genotype.
Dietary ALA-supplementation did not rescue the decreased fertility of Gclm−/− females. Female Gclm−/− and Gclm+/+ mice were fed diets supplemented with 0, 150, or 600 mg/kg ALA from 3 weeks of age until euthanasia. Females were bred continuously with a wild type male from 2 months of age and litter and offspring production were assessed for 12 weeks as described in Materials and methods. (A) Litter production was decreased in Gclm−/− females, with a statistically significant effect of genotype at 28, 56, and 84 days (P ≤ 0.02). The diet by genotype interaction was statistically significant at 84 days (P = 0.012), due to lower offspring production in Gclm+/+ females fed the 600 mg/kg ALA diet at 84 days. (B) Offspring production was significantly less in all Gclm−/− groups compared with Gclm+/+ controls (P < 0.001, effect of genotype at all-time points). There was a significant diet by genotype interaction at 84 days (P = 0.012), with Gclm+/+ females fed the 600 mg/kg ALA diet having significantly fewer offspring than Gclm+/+ controls. The graphs show means ± SEM cumulative number of litters or offspring produced by females of the indicated Gclm genotypes and ALA diet at 28, 56, and 84 days of continuous breeding. N = 8/group, except N = 6 Gclm+/+, 600 mg/kg ALA group; N = 7 Gclm−/− 150 mg/kg ALA group). *P < 0.05, compared with Gclm+/+ controls at the same time point.
ALA supplementation does not rescue the decreased offspring production of Gclm−/− females
To test whether long-term antioxidant supplementation could rescue the decreased litter size due to preimplantation embryonic mortality in Gclm−/− mice, we chose the thiol antioxidant ALA because it can be administered via the diet. Dietary supplementation was started at weaning when the mice were 21 days old and continued throughout a 12-week continuous breeding assay that began when the mice were 2 months old. We did not observe evidence of beneficial effects of dietary ALA supplementation on cumulative numbers of litters or offspring in Gclm−/− females, and we observed evidence of an adverse effect of the 600 mg/kg ALA-supplemented diet on fertility in Gclm+/+ females (Figure 1A and B). There were small, statistically significant effects of Gclm genotype on the cumulative number of litters produced by Gclm−/− compared with Gclm+/+ females at 28, 56, and 84 days of continuous breeding (P ≤ 0.02, effect of genotype, two-way ANOVA), with the Gclm−/− females in the control and 150 mg/kg ALA diet groups having 28 and 18% fewer litters at 84 days compared with Gclm+/+ females in the same diet group (Figure 1A). There was also a statistically significant interaction between genotype and diet at 84 days (P = 0.012), with the number of litters tending to increase with increasing dietary ALA in the Gclm−/− females, but decreasing with the 600 mg/kg ALA diet in the Gclm+/+ females.
Gclm−/− females produced significantly fewer offspring at each time point (P < 0.001, effect of genotype; Figure 1B). There was no statistically significant effect of diet on cumulative offspring at any time point, but the interaction between genotype and diet approached significance at 84 days, with decreased numbers of offspring in the Gclm+/+, 600 mg/kg ALA group compared with Gclm+/+ controls (P = 0.06).
The percentage of dead pups and the body weights of male and female pups on the day of birth did not differ significantly with genotype or diet (data not shown).
No Gclm genotype- or ALA diet-related differences in estrous cycling
At the end of the breeding study, dietary ALA supplementation continued while vaginal cytology was conducted for at least 14 days to assess estrous cycling. Most females of both Gclm genotypes on the various ALA diets—control, 150 mg/kg, and 600 mg/kg—had regular 4- to 5-day estrous cycles (Table 2). The percentages of days of metestrus or diestrus, characterized by leukocytic vaginal cytology, proestrus, characterized by nucleated and cornified cells, and estrus, characterized by cornified cells, were similar between genotypes and diets (P > 0.05 effects of genotype and diet on all estrous cycle endpoints; Table 2).
ALA supplementation attenuated the accelerated age-related decline in ovarian follicles in Gclm−/− mice
Mice from the breeding study were euthanized on metestrus of the estrous cycle, and ovarian follicles were enumerated. There was a statistically significant effect of Gclm genotype on total number of healthy follicles of all stages of development (P = 0.002, effect of genotype, P = 0.203, effect of diet; P = 0.748 genotype by diet interaction two-way ANOVA; Figure 2A). Intergroup comparisons showed that Gclm−/− mice fed the control diet had significantly fewer follicles than Gclm+/+ mice on the control diet (P = 0.020), while total follicle counts did not differ significantly between Gclm−/− mice on the 150 mg/kg ALA diet and Gclm+/+ mice on any of the diets (P ≥ 0.136) or between Gclm−/− and Gclm+/+ mice on the 600 mg/kg ALA diet (P = 0.065). Total follicle counts in Gclm−/− mice on the 600 mg/kg diet were significantly lower than total follicle counts in Gclm+/+ mice on the control (P = 0.005) or 150 mg/kg (P = 0.004) diets.
There were statistically significant effects of Gclm genotype on numbers of primordial follicles (P < 0.001, effect of genotype; P = 0.378, effect of diet; P = 0.748 genotype by diet interaction two-way ANOVA; Figure 2B) and primary follicles (P = 0.037, effect of genotype; P = 0.149, effect of diet; and P = 0.762 genotype by diet interaction; Figure 2C) per ovary, with Gclm−/− females having fewer primordial and primary follicles than Gclm+/+ females. Inter-group comparisons revealed that Gclm−/− mice fed control diet or 600 mg/kg ALA diet had 68% and 61% fewer primordial follicles, respectively, than Gclm+/+ females fed the same diets (P = 0.009 and P = 0.049), while Gclm−/− mice fed 150 mg/kg ALA chow had only 45% fewer primordial follicles than Gclm+/+ females fed the same diet (P = 0.061) or the control diet (P = 0.067) and only 29% fewer follicles than Gclm+/+ on the 600 mg/kg diet (P = 0.347; Figure 2B). Gclm−/− mice fed the 600 mg/kg ALA diet had significantly fewer primordial follicles than Gclm+/+ mice on the control (P = 0.005) and 150 mg/kg diets (P = 0.004). Intergroup comparisons showed that the ovaries of Gclm−/− mice fed the 600 mg/kg ALA diet had significantly fewer primary follicles than ovaries of Gclm+/+ mice fed the control or 150 mg/kg ALA diets (P = 0.015, P = 0.017, respectively; Figure 2C); no other intergroup comparisons were statistically significant for primary follicle numbers. There were no statistically significant effects of Gclm genotype or dietary ALA supplementation on numbers of secondary and antral follicles (Figure 2D and E).
Figure 3 shows the fraction of primordial (Figure 3A) and primary (Figure 3B) follicles out of the total number of healthy follicles. Primordial follicles make up a smaller fraction (P = 0.007, effect of genotype; P = 0.686, effect of diet; P = 0.279, diet × genotype interaction) and primary follicles make up a non-significantly larger fraction (P = 0.056, effect of genotype; P = 0.772, effect of diet; P = 0.725, diet × genotype interaction) of all healthy follicles in Gclm−/− mice compared with Gclm+/+ mice on the control diet. The fraction of primordial follicles increased from 0.27 in Gclm−/− mice fed control diet to 0.39 in Gclm−/− mice fed 150 mg/kg ALA diet. Intergroup comparisons revealed that the fraction of primordial follicles in Gclm+/+ mice on the control diet or the 600 mg/kg diet was greater than the fraction of primordial follicles in Gclm−/− mice on the control diet (P = 0.010 and P = 0.012, respectively), while the fraction of primordial follicles in the Gclm−/− mice on the 150 mg/kg diet did not differ from those of Gclm+/+ mice on any of the three diets (≥0.125). The fraction of primary follicles tended to be higher in Gclm−/− mice compared with Gclm+/+ mice overall, with minimal effect of diet (P = 0.056, effect of genotype; P = 0.772, effect of diet; P = 0.725 genotype by diet interaction). These data provide further support that supplementation with 150 mg ALA/kg diet shifted follicle dynamics in Gclm−/− mice toward those observed in Gclm+/+ mice.
Effects of ALA supplementation on estrous cycling
| Diet | Genotype | Regular cycle, % a | Diestrous % time | Proestrous % time | Estrous % time | Cycle length (days if cycling) |
|---|---|---|---|---|---|---|
| Control | Gclm−/− | 100 | 55 ± 4 | 20 ± 2 | 24 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 87.5 | 41 ± 3 | 24 ± 3 | 36 ± 3 | 5 ± 0.2 | |
| 150 mg/kg ALA | Gclm−/− | 100 | 54 ± 6 | 21 ± 4 | 25 ± 8 | 5 ± 0.4 |
| Gclm+/+ | 100 | 46 ± 2 | 24 ± 2 | 36 ± 3 | 5 ± 0.3 | |
| 600 mg/kg ALA | Gclm−/− | 87.5 | 55 ± 5 | 23 ± 3 | 21 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 62.5 | 63 ± 6 | 21 ± 3 | 16 ± 4 | 6 ± 1.0 |
| Diet | Genotype | Regular cycle, % a | Diestrous % time | Proestrous % time | Estrous % time | Cycle length (days if cycling) |
|---|---|---|---|---|---|---|
| Control | Gclm−/− | 100 | 55 ± 4 | 20 ± 2 | 24 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 87.5 | 41 ± 3 | 24 ± 3 | 36 ± 3 | 5 ± 0.2 | |
| 150 mg/kg ALA | Gclm−/− | 100 | 54 ± 6 | 21 ± 4 | 25 ± 8 | 5 ± 0.4 |
| Gclm+/+ | 100 | 46 ± 2 | 24 ± 2 | 36 ± 3 | 5 ± 0.3 | |
| 600 mg/kg ALA | Gclm−/− | 87.5 | 55 ± 5 | 23 ± 3 | 21 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 62.5 | 63 ± 6 | 21 ± 3 | 16 ± 4 | 6 ± 1.0 |
aPercent of mice with regular 4 or 5 day estrous cycle.
Effects of ALA supplementation on estrous cycling
| Diet | Genotype | Regular cycle, % a | Diestrous % time | Proestrous % time | Estrous % time | Cycle length (days if cycling) |
|---|---|---|---|---|---|---|
| Control | Gclm−/− | 100 | 55 ± 4 | 20 ± 2 | 24 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 87.5 | 41 ± 3 | 24 ± 3 | 36 ± 3 | 5 ± 0.2 | |
| 150 mg/kg ALA | Gclm−/− | 100 | 54 ± 6 | 21 ± 4 | 25 ± 8 | 5 ± 0.4 |
| Gclm+/+ | 100 | 46 ± 2 | 24 ± 2 | 36 ± 3 | 5 ± 0.3 | |
| 600 mg/kg ALA | Gclm−/− | 87.5 | 55 ± 5 | 23 ± 3 | 21 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 62.5 | 63 ± 6 | 21 ± 3 | 16 ± 4 | 6 ± 1.0 |
| Diet | Genotype | Regular cycle, % a | Diestrous % time | Proestrous % time | Estrous % time | Cycle length (days if cycling) |
|---|---|---|---|---|---|---|
| Control | Gclm−/− | 100 | 55 ± 4 | 20 ± 2 | 24 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 87.5 | 41 ± 3 | 24 ± 3 | 36 ± 3 | 5 ± 0.2 | |
| 150 mg/kg ALA | Gclm−/− | 100 | 54 ± 6 | 21 ± 4 | 25 ± 8 | 5 ± 0.4 |
| Gclm+/+ | 100 | 46 ± 2 | 24 ± 2 | 36 ± 3 | 5 ± 0.3 | |
| 600 mg/kg ALA | Gclm−/− | 87.5 | 55 ± 5 | 23 ± 3 | 21 ± 3 | 5 ± 0.4 |
| Gclm+/+ | 62.5 | 63 ± 6 | 21 ± 3 | 16 ± 4 | 6 ± 1.0 |
aPercent of mice with regular 4 or 5 day estrous cycle.
Dietary ALA supplementation partially rescued the decreased ovarian follicle counts in Gclm−/− mice. Ovaries were harvested from females in the breeding study described in Figure 1 and were processed for follicle counts as described in Materials and methods. (A) Means ± SEM of the sum of healthy follicles at all stages of development in Gclm−/− and Gclm+/+ ovaries at the indicated levels of dietary ALA supplementation (P = 0.002, effect of genotype; P = 0.203, effect of diet; P = 0.748 genotype by diet interaction by two-way ANOVA). (B) Means ± SEM of healthy primordial follicles in Gclm−/− and Gclm+/+ ovaries at the indicated levels of dietary ALA supplementation (P < 0.001, effect of genotype; P = 0.378, effect of diet; P = 0.748 genotype by diet interaction by two-way ANOVA). (C) Means ± SEM of healthy primary follicles (P = 0.037, effect of genotype; P = 0.149, effect of diet; P = 0.762 genotype by diet interaction by two-way ANOVA). (D) Means ± SEM of healthy secondary follicles. There were no statistically significant effects of genotype or diet. (E) Means ± SEM of healthy antral follicles. There were no statistically significant effects of genotype or diet. *P < 0.050, for the indicated intergroup comparisons. N = 6–8/group.
Supplementation with 150 mg/kg diet ALA increased the fraction of primordial follicles in Gclm−/− mice compared with Gclm−/− mice on the control diet. (A) Fractions ± SEM of primordial follicles of total number of healthy follicles (P = 0.007, effect of genotype; P = 0.686, effect of diet; P = 0.279 genotype by diet interaction by two-way ANOVA). (B) Fractions ± SEM of primary follicles of total number of healthy follicles (P = 0.056, effect of genotype; P = 0.772, effect of diet; P = 0.725 genotype by diet interaction by two-way ANOVA). *P < 0.050, #P = 0.088 for the indicated intergroup comparisons. N = 6–8/group.
ALA supplementation prevented the accelerated recruitment of primordial follicles in Gclm−/− mice
As quiescent primordial follicles are recruited into the growing pool, the granulosa cells begin to undergo mitosis [48]. To test whether the apparent beneficial effect of 150 mg/kg ALA diet on ovarian follicle numbers in Gclm−/− mice could be due to slowed recruitment of follicles into the growing pool, we used Ki67, a specific marker of mitosis [49], to identify recruited follicles based on granulosa cell proliferation. The average percentage of primordial follicles with Ki67-positive granulosa cells was below 15% in all groups and did not differ significantly by Gclm genotype or dietary ALA concentration (Figure 4C). In contrast, the percentages of primary follicles with Ki67-positive granulosa cells varied significantly with dietary ALA concentration, and Gclm genotype by diet interaction (P = 0.033, P = 0.018, respectively, by two-way ANOVA). Intergroup comparisons showed that, consistent with our previous observations [20], the percentage of Ki67-positive primary follicles was significantly higher in Gclm−/− ovaries from mice fed the 0 mg/kg ALA diet than in Gclm+/+ ovaries from mice fed the same diet (P = 0.016; Figure 4D). In contrast, the percentage of Ki67-positive primary follicles in Gclm−/− mice fed 150 mg/kg ALA diet did not differ from Gclm+/+ females fed the same diet (P = 0.232) or the 0 mg/kg ALA diet (P = 0.534). There was also no statistically significant difference between genotypes in the percentage of Ki67-positive primary follicles in mice fed the 600 mg/kg ALA diet (P = 0.236), but ovaries from mice with this level of ALA supplementation had significantly higher percentages of Ki67-positive primary follicles than Gclm+/+ mice fed the control diet (P ≤ 0.025). As expected, essentially all secondary and antral follicles had Ki67-positive granulosa cells regardless of experimental group (data not shown).
Supplementation with 150 mg/kg diet ALA prevented the increased recruitment of follicles into the growing pool in Gclm−/− ovaries. Ovaries harvested from females in the breeding study described in Figure 1 were processed for immunostaining for the mitosis marker Ki67, and follicles with Ki67-positive granulosa cells were scored as described in Materials and methods. (A) Representative Ki67 immunostaining in section from ovary of Gclm−/− mouse fed the control diet. Ki67-positive primary follicles with brown stained granulosa cells (red arrows) and Ki67-negative primordial follicles (white arrows) are present. Also seen are portions of larger follicles with numerous Ki67 positive granulosa cells. (B) Technical negative control in which primary antibody was replaced with non-immune serum. The graphs show the means ± SEM percentage of Ki67 positive follicles. (C) There were no statistically significant effects of Gclm genotype, dietary ALA concentration, or genotype by diet interaction on percentages of Ki67-positive primordial follicles. (D) The percentages of primary follicles with Ki67-positive granulosa cells varied significantly with dietary ALA concentration and Gclm genotype by diet interaction (P = 0.033, P = 0.018, respectively, by two-way ANOVA). *P < 0.05, indicated intergroup comparisons. N = 4/group. Scale bars, 50 μm.
Effects of Gclm genotype- and ALA diet on ovarian follicle apoptosis
To determine whether the apparent beneficial effect of 150 mg/kg diet ALA supplementation on follicle numbers in Gclm−/− mice could be due to decreased follicle apoptosis, we performed immunostaining for the activated, cleaved form of the so-called executioner caspase, caspase-3 (Figure 5). Percentages of secondary follicles with cleaved caspase-3-positive granulosa cells differed significantly by ALA diet group (P = 0.028, effect of diet, two-way ANOVA), but not by genotype or genotype by diet interaction, with the percentages of secondary follicles with cleaved caspase-3 positive granulosa cells increasing with increasing ALA concentration in ovaries of both genotypes (Figure 5B). No statistically significant diet or genotype related differences were observed in the percentages of antral follicles with cleaved caspase-3 positive granulosa cells (P > 0.05; Figure 5C). Cleaved caspase-3 immunostaining was not observed in primary and primordial follicles (data not shown).
Effects of Gclm genotype- and ALA diet on activated caspase 3 in ovarian follicles. Cleaved caspase-3 immunostaining was used to assess apoptosis in ovarian follicles as described in Materials and Methods. (A) Representative cleaved caspase 3 immunostaining in section from ovary of Gclm−/− mouse fed the control diet. Caspase 3-positive follicles have brown stained granulosa cells (red arrows). (B) Technical negative control in which primary antibody was replaced with non-immune serum. The graphs show the means ± SEM percentage of cleaved caspase-3 positive follicles per ovary. (C) Percentages of cleaved caspase-3 positive secondary follicles increased with increasing dietary ALA concentration (P = 0.028); pairwise intergroup comparisons were not statistically significant. (D) There were no statistically significant effects of genotype or diet on antral follicle cleaved caspase-3 immunostaining. N = 5/group. Scale bars, 50 μm.
Dietary ALA supplementation rescued increased DNA damage in ovarian follicles of Gclm−/− mice. Nuclear γH2AX immunostaining in oocytes and granulosa cells was scored as described in Materials and methods. Representative image of (A) experimental ovary section from Gclm+/+ fed the 600 mg ALA/kg diet. Red and white arrows point to γH2AX-positive (dark brown staining) and -negative follicles. Red arrow heads point to γH2AX-positive ooctyes. (B) Technical negative control in which primary antibody was replaced with non-immune serum. The graphs show means ± SEM percentages of follicles of the indicated stage of development with γH2AX-positive oocytes or granulosa cells. (C) There were no statistically significant differences in the percentages of primary follicles with positive granulosa cells among groups. (D) There were statistically significant effects of genotype and genotype by diet interactions in the percentages of secondary follicles with γH2AX-positive granulosa cells (P = 0.004, genotype; P = 0.049, genotype × diet; P = 0.063, diet, two-way ANOVA). (E) There was a statistically significant genotype by diet interaction on percentages of antral follicles with γH2AX-positive granulosa cells (P = 0.010), while the main effects of genotype and diet were not statistically significant (P = 0.541 and P = 0.408, respectively). (F) Effects of dietary ALA supplementation and genotype by diet interaction on primordial follicle oocyte γH2AX immunostaining approached statistical significance (P = 0.054 and P = 0.058, respectively by two-way ANOVA and P = 0.050 by Kruskal–Wallis test). (G) Percentages of primary follicles with γH2AX-positive oocytes varied significantly with ALA by genotype interaction (P = 0.009, two-way ANOVA) and the effect of ALA alone approached significance (P = 0.086). There were no statistically significant differences in percentages of secondary (H) or antral (I) follicles with γH2AX-positive oocytes among groups. *P < 0.05, indicated intergroup comparison. N = 5/group. Scale bars, 50 μm.
Effects of NAC or ALA supplementation on ovarian concentrations of GSH and GSSG in Gclm+/+ and Gclm−/− females. Total and oxidized GSH were measured and the Eh for GSH/GSSG were calculated as described in Materials and Methods. (A–C) Gclm−/− mice have significantly decreased ovarian concentrations of total GSH and oxidized GSH (GSSG) and a more oxidized Nernst potential (Eh) of the GSSG/GSH redox couple compared with wild-type littermates regardless of NAC supplementation (P < 0.001, effect of genotype for all three endpoints). (A) Mean ± SEM of total ovarian GSH in NAC supplemented and control mice (P = 0.037, effect of NAC; P = 0.124, genotype by NAC interaction). (B) Mean ± SEM of ovarian GSSG concentrations in NAC supplemented and control mice (P = 0.248, effect of NAC; P = 0.433, genotype by NAC interaction). (C) Mean ± SEM of Eh for GSH/GSSG redox couple in NAC supplemented and control mice (P = 0.004, effect of NAC; P = 0.439, genotype by NAC interaction). *P < 0.02 versus 0 mM NAC of same genotype. N = 5–7/group. (D) Mean ± SEM of total ovarian GSH concentrations in Gclm+/+ and Gclm−/− mice fed diets containing 0, 150, and 600 mg/kg ALA from 3 to 10 weeks of age. Ovarian GSH concentrations were significantly lower in Gclm−/− mice compared with Gclm+/+ mice (P < 0.001, effect of genotype), but dietary ALA supplementation did not affect ovarian GSH concentrations (P = 0.933, effect of ALA; P = 0.509 genotype × ALA interaction). N = 5–7/group.
ALA diets rescued increased DNA damage in ovarian follicles of Gclm−/− mice
Histone H2AX is phosphorylated at sites of double strand DNA breaks; once phosphorylated, it is called γH2AX [50]. We used nuclear immunohistochemical staining for γH2AX as a marker of DNA damage in the ovaries (Figure 6A). There were no statistically significant differences in the percentages of primary follicles with positive granulosa cells among groups (Figure 6C). Primordial follicles very rarely had granulosa cells positive for γH2AX (data not shown). There were statistically significant effects of genotype and genotype by diet interactions in the percentages of secondary follicles with γH2AX-positive granulosa cells (P = 0.004, genotype; P = 0.049, genotype × treatment, two-way ANOVA; Figure 6D), while the main effect of diet approached statistical significance (P = 0.063). Intergroup comparisons showed that the percentage of secondary follicles with γH2AX-positive granulosa cells was increased in Gclm−/− mice compared with Gclm+/+ mice on control diet, but not on the ALA diets. There was a statistically significant genotype by diet interaction on percentages of antral follicles with γH2AX-positive granulosa cells (P = 0.010, Figure 6E), while the main effects of genotype and diet were not statistically significant (P = 0.541 and P = 0.408, respectively). Intergroup comparisons showed that the percentages differed between ovaries from Gclm−/− and Gclm+/+ mice fed the control diet, but not for mice fed either ALA diet.
The effects of dietary ALA supplementation and diet by genotype interactions on percentages of primordial follicles with γH2AX-positive oocytes approached significance (P = 0.054 and P = 0.058, respectively, by two-way ANOVA; effect of group P = 0.050 by Kruskal–Wallis test; Figure 6F). Intergroup comparisons supported that the percentages of primordial follicles with γH2AX-positive oocytes differed between Gclm−/− and Gclm+/+ ovaries in mice fed the control diet, but not the ALA diets (Figure 6F). The percentages of primary follicles with γH2AX-positive oocytes varied significantly with ALA by genotype interaction (P = 0.009) and the effect of ALA alone approached significance (P = 0.086). Intergroup comparisons showed that percentages of primary follicles with γH2AX-positive oocytes were higher in ovaries of Gclm−/− compared with Gclm+/+ mice on the control diet (P = 0.002; Figure 6G), while the percentages did not differ between Gclm−/− and Gclm+/+ ovaries on the 150 or 600 mg/kg ALA diets. There were no statistically significant differences in percentages of secondary or antral follicles with γH2AX-positive oocytes among groups (Figure 6H and I).
Compared with the 150 mg/kg ALA diet, the 600 mg/kg ALA diet caused lesser decreases in the percentages of secondary and antral follicles with γH2AX-positive granulosa cells and primordial and primary follicles with γH2AX-positive oocytes in Gclm−/− ovaries, while it tended to increase the percentages in Gclm+/+ ovaries compared with ovaries of the same genotype on the control diet (Figure 6).
N-acetyl cysteine supplementation increased ovarian GSH concentrations in Gclm+/+, but not Gclm−/−, mice
Supplementation of 2-month old mice with 80 mM NAC in drinking water for 7 days before euthanasia on metestrus of the estrous cycle, slightly, but statistically significantly, increased ovarian total GSH concentrations in Gclm+/+ females, but had lesser effect on ovarian GSH concentrations in Gclm−/− females (P < 0.001, effect of genotype; P = 0.037, effect of NAC; P = 0.124, genotype by NAC interaction; Figure 7A). In contrast, NAC supplementation did not significantly affect ovarian GSSG concentrations, which were significantly lower in Gclm−/− mice compared with Gclm+/+ mice (P < 0.001, effect of genotype; P = 0.248, effect of NAC; P = 0.433, genotype by NAC interaction; Figure 7B). The redox potential of the GSH:GSSG redox couple was more reduced in Gclm+/+ ovaries supplemented with NAC compared with control Gclm+/+ ovaries, while NAC supplementation decreased ovarian GSH:GSSG redox potential to a lesser extent in Gclm−/− mice (P < 0.001, effect of genotype; P = 0.004, effect of NAC; P = 0.439, genotype by NAC interaction; Figure 7C).
ALA supplementation did not affect ovarian GSH concentrations
Supplementation with 150 mg/kg or 600 mg ALA per kg diet did not increase ovarian GSH concentrations in Gclm+/+ or Gclm−/− mice (Figure 7D).
Effects of NAC on water intake and weight gain or ALA supplementation on food intake and weight gain
There were no statistically significant effects of Gclm genotype or NAC supplementation from PND 21 to 30 on weight gain during the study, but mice given NAC-supplemented water consumed about 2 mL per day less water than those given control water (P = 0.04, effect of NAC; Supplemental Table S1). In contrast, there were no statistically significant effects of genotype or NAC supplementation for 7 days in 2 month-old mice on water consumption or body weight (Supplemental Table S2).
There were no significant differences among groups in the amount of ALA-supplemented diet consumed during the 5 weeks prior to breeding from 3 to 8 weeks of age (Supplemental Table S3). In contrast, there were statistically significant effects of genotype (P = 0.002) and genotype by ALA interaction (P = 0.010) on food intake between 3 and 10 weeks of age prior to euthanasia for ovarian GSH concentration assay, with the Gclm−/− mice consuming less food overall, but tending to consume 0.2 g/day more when on the ALA-supplemented diets, while Gclm+/+ mice tended to consume 0.1–0.2 g/day less on average on the ALA diets (Supplemental Table S4). There was also a statistically significant effect of genotype and a statistically significant genotype by ALA interaction on weight gain during that period, with Gclm−/− mice gaining less weight, but tending to gain 0.4–0.6 g more weight on average on the ALA-supplemented diets, while Gclm+/+ mice tended to gain 1.3–1.5 g less weight on average on the ALA diets (Supplemental Table S4). The differences in food consumption between the two experiments could be explained by the longer duration of the second experiment or could have arisen by chance.
Discussion
We investigated whether the accelerated age-related decline in ovarian follicles and the preimplantation embryonic mortality resulting in lower offspring production, which we previously reported in GSH-deficient Gclm−/− mice [19, 20], could be rescued by thiol antioxidant supplementation. Our results show that dietary ALA supplementation from weaning until 20 weeks of age with a daily dose similar to that used in humans partially rescued the greater age-related decline in primordial follicle numbers in Gclm−/− mice, while a fourfold higher dose was not beneficial. Further, our results support that the rescue of primordial follicle numbers was due to normalization (to levels comparable with Gclm+/+ mice) of recruitment of primordial follicles into the growing pool and normalization of follicular DNA damage in 150 mg ALA/kg diet supplemented Gclm−/− mice. These findings are important because premature depletion of the ovarian follicle reserve is associated with increased risk of osteoporosis, cardiovascular disease, and dementia [51–54], in addition to shortening the fertile lifespan. In contrast, dietary ALA supplementation did not rescue the decreased offspring production, nor did it alter ovarian GSH concentrations, of Gclm−/− female mice. Supplementation with the GSH precursor NAC for 8 days beginning 6 days before initiation of hormonal stimulation for superovulation did not rescue the low rate of second pronucleus formation in zygotes of Gclm−/− females at 0.5 days post-coitum, which we previously found was associated with high embryonic mortality prior to the blastocyst stage [19]. Finally, NAC supplementation slightly, but statistically significantly increased ovarian GSH concentrations in Gclm+/+ mice, while having minimal effect on ovarian GSH in Gclm−/− mice.
Gclm−/− females are born with a similar complement of oocytes as wild type littermates, but they experience an accelerated decline in primordial follicles as they age [20]. This appears to be due to increased recruitment of primordial follicles into the growing pool, rather than increased death of primordial follicles [20]. Consistent with our prior findings, we observed significantly decreased primordial and primary follicle numbers, decreased fraction of primordial follicles, and increased percentages of primary follicles with Ki67-positive granulosa cells in Gclm−/− mice fed the control diet, and these genotype-related differences were abolished in Gclm−/− mice fed the 150 mg/kg ALA diet. Our follicle classification scheme places transitional or activated follicles into the same group as primary follicles. Therefore, the abolition of the Gclm genotype-related differences in the fraction of all healthy follicles at the primordial stage and in percentages of Ki67-positive primary follicles in Gclm−/− mice fed the 150 mg/kg ALA diet indicates that the 150 mg/kg ALA diet prevents increased recruitment of primordial follicles into the growing pool in Gclm−/− mice. These data provide additional support for accelerated recruitment of primordial follicles into the growing pool in Gclm−/− ovaries and further support that chronic ovarian oxidative stress may lead to accelerated recruitment of follicles into the growing pool. Consistent with our observation that the 600 mg/kg ALA diet failed to rescue the decreased primordial follicle numbers in Gclm−/− mice, the percentages of Ki67-positive primary follicles were higher in ovaries of both Gclm genotypes fed that diet compared with ovaries of mice of the same genotype fed control diet or 150 mg/kg ALA diet.
We previously showed that the ovaries of Gclm−/− mice are in a state of chronic oxidative stress, with concentrations of GSH decreased to a greater extent than GSSG, resulting in decreased ratio of GSH:GSSG and oxidized Nernst potential of the GSH:GSSG redox couple, as well as increased oxidative lipid and protein damage [20]. The present study further demonstrates that Gclm−/− ovaries from mice on the control diet have increased DNA damage, detected by immunostaining for γH2AX, a marker of double strand DNA breaks, in oocytes and granulosa cells of follicles, compared with Gclm+/+ ovaries on the same diet. Oxidative DNA damage is a known cause of double strand DNA breaks [55]. Consistent with oxidative DNA damage as the cause of the increased γH2AX immunostaining, dietary supplementation with 150 mg ALA/kg diet normalized the increased γH2AX immunostaining in Gclm−/− ovaries. In contrast, the 600 mg/kg diet resulted in lesser decreases in γH2AX immunostaining in Gclm−/− ovaries and tended to increase γH2AX immunostaining in Gclm+/+ ovaries compared with the control diet. We previously reported that 150 mg/kg diet ALA supplementation was protective against induction of γH2AX in ovarian follicles by irradiation of mice with heavy iron ions [7], but provided no protection against induction of γH2AX by irradiation with heavy oxygen ions [6]. Similarly, dietary ALA supplementation decreased induction of γH2AX in livers of rats treated with N-diethylnitrosamine and thioacetamide in a two-stage model of hepatic carcinogenesis [56]. α-Lipoic acid supplementation by intraperitoneal injection was also protective against DNA damage induced by drinking water administration of potassium dichromate measured using single cell gel electrophoresis (COMET assay) in peripheral blood mononuclear cells and bone marrow cells [57].
In contrast to partial prevention of the accelerated ovarian follicle decline in Gclm−/− females by dietary ALA supplementation, we did not observe amelioration by NAC supplementation of the decreased percentage of zygotes with two pronuclei at 0.5 days post-coitum or by ALA of the decreased offspring production in Gclm−/− females. Specifically, supplementation with NAC beginning 6 days before eCG injection did not rescue zygote pronuclear formation, and dietary ALA supplementation from 21 days of age did not rescue decreased offspring production in a 12-week continuous breeding assay. We note that NAC treatment did not occur throughout follicular development, beginning only during the later stages of secondary follicle development, which suggested that it may be necessary to initiate antioxidant supplementation beginning at the primordial follicle stage to rescue the increased early embryonic mortality observed during pregnancies of Gclm−/− females. Therefore, we conducted a study of long-term supplementation with ALA, which can be administered via the diet. However, ALA supplementation from weaning through euthanasia at 5.5 months of age also failed to rescue this phenotype of Gclm−/− mice. Together the current results and our prior work [19] show that embryos from Gclm−/− mothers progress to the 2-pronuclei stage at much lower rates by 0.5 dpc and have very low rates of maturation to the blastocyst stage by 3.5 dpc than embryos from Gclm+/+ females. Our prior work further showed that this is due to effects of Gclm deletion in the oocyte itself because oocytes from Gclm−/− females fertilized in vitro with Gclm+/+ sperm and cultured for 4 days also failed to reach the blastocyst stage [19]. Moreover, work of earlier investigators demonstrated that mature mouse and hamster oocytes contain GSH concentrations of 8–10 mM, higher than GSH concentrations in germinal vesicle or pronuclear stage oocytes [58, 59]. Using biochemical inhibition of GSH synthesis, prior studies demonstrated that GSH is required for sperm nuclear decondensation and subsequent male pronucleus formation in murine, hamster, and bovine eggs [58–60]. This is because GSH reduces the disulfide cross-linking of the sperm protamines, which is required for sperm nucleus reactivation to occur [61]. Although ALA could theoretically take the place of GSH in reducing disulfide bonds in the sperm nucleus, dietary supplementation with ALA does not achieve sufficiently high tissue concentrations of ALA because it is rapidly catabolized [29]. Therefore, we conclude that the disulfide reducing activity of GSH, as opposed to its antioxidant activity, is likely required to rescue the preimplantation mortality observed in Gclm−/− mice.
Based on their average daily food consumption and body weights, mice in our study consumed about 17 mg ALA/kg body weight on the 150 mg/kg ALA diet and 68 mg ALA/kg body weight on the 600 mg/kg diet. Human therapeutic oral doses reported in the literature range from 200 to 1800 mg/day (3–30 mg/kg for a 60 kg woman). Thus, the 600 mg ALA/kg diet dose is more than twofold higher than the highest doses used in humans. There is limited data in the literature concerning potential toxicity of ALA. A 14-year old girl died of multiorgan failure after purposeful ingestion of at least 6000 grams of ALA [62]. A 20-month old boy who accidentally ingested 226 mg/kg ALA developed metabolic acidosis and seizures, but survived [63]. A study in rats showed that intraperitoneal doses of 4.3 or 16.6 μmol/kg ALA both lowered the activity of hepatic biotin-dependent carboxylase enzymes by competing with biotin, to which ALA is structurally similar [64]. Feeding a biotin-deficient diet for 12 weeks has been reported to destroy primary, secondary and antral follicles, disrupt estrous cycling, and increase serum estradiol and progesterone concentrations in mice [65]. In contrast, feeding mice a biotin-deficient diet for 42 days did not disrupt estrous cycling, but increased the incidence of spindle abnormalities in ovulated oocytes (18% in controls versus 59% in deficient), which was not reversed in mice fed a biotin sufficient diet for 20 days after 22 days on a biotin-deficient diet (48%) [66]. While 600 mg/kg diet ALA supplementation would not be expected to mimic severe biotin deficiency, it is plausible that the relatively mild effects on offspring production and lack of beneficial effects on ovarian follicle decline, which we observed in mice fed the 600 mg/kg ALA diet, could be due, at least in part, to competition by ALA with biotin.
Prior in vivo studies have reported that ALA supplementation increases tissue concentrations of GSH by upregulating nuclear translocation of the transcription factor NRF2, a master regulator of expression of multiple antioxidant genes, including Gclc and Gclm. Intraperitoneal administration of ALA increased concentrations of cysteine and GSH in the brains and livers, but not the hearts, of aged male rats [31, 32]. The ratio of reduced GSH to GSSG was increased in all three tissues [31, 32]. Nuclear levels of the transcription factor NRF2 and its antioxidant response element binding activity were increased in livers of ALA treated rats [32]. Glutamate cysteine ligase catalytic protein levels and GCL enzymatic activity were increased in livers, but not brains, by ALA administration [31, 32]. Intraperitoneal ALA supplementation also increased concentrations of GSH in mitochondria isolated from brains of young and old male rats [38]. Together with our data demonstrating no effects of dietary ALA supplementation on total ovarian GSH concentrations, these prior studies suggest that ALA has varying effects on GSH synthesis in different tissues. It is also possible that ALA supplementation increased the ratio of GSH to GSSG in the ovaries in our study, but we had insufficient sample to measure GSSG. In addition, our observation that the 150 mg/kg ALA supplementation partially reversed the accelerated age-related decline in ovarian follicles without increasing GSH concentrations in the ovaries of Gclm−/− mice is consistent with upregulated expression of other antioxidant defense genes by ALA supplementation due to increased NRF2 signaling [29, 67].
In conclusion, dietary supplementation with the antioxidant ALA beginning at weaning, before ovarian follicle numbers have begun to diverge in Gclm+/+ and Gclm−/− mice, partially rescued the accelerated age-related decline in primordial ovarian follicles in Gclm−/− mice. Our results support that both increased follicular oxidative DNA damage and accelerated recruitment of primordial follicles into the growing pool may mediate the accelerated decline in ovarian follicles in Gclm−/− mice. In contrast, supplementation with the antioxidant NAC beginning during the secondary follicle stage or supplementation with ALA throughout postnatal follicle development from the primordial stage onward failed to rescue the abnormal pronucleus formation and resultant decreased offspring production of Gclm−/− females. Together with previous studies [58–61], the latter results support a requirement for the disulfide reducing activity of high GSH concentrations in the oocyte for normal early embryonic development.
Acknowledgments
The authors thank Dr. TJ Kavanagh, University of Washington, for the gift of the original breeding pair of Gclm+/− mice. We thank Johnny Nguyen, Angelica del Rosario, Ryan Ripperdan, and Lizette Nunez Magana for assistance with monitoring of mice during these experiments.
Conflict of interest
The authors have declared that no conflict of interest exists.
Footnotes
†Grant Support: Research reported in this publication was supported by the National Institute of Environmental Health Sciences of the National Institutes of Health (NIH) under Award Number R01ES020454, by a NIH Office of Dietary Supplements Supplement to R01ES020454 (to UL) and by University of California Irvine Summer Undergraduate Research Program fellowships (to ESN and SA).
