Abstract
Hyperthermia or heat stress (HS) occurs when heat dissipation mechanisms are overwhelmed by external and internal heat production. Hyperthermia negatively affects reproduction and potentially compromises oocyte integrity and reduces developmental competence of ensuing embryos. Autophagy is the process by which cells recycle energy through the reutilization of cellular components and is activated by a variety of stressors. Study objectives were to characterize autophagy-related proteins in the ovary following cyclical HS during the follicular phase. Twelve gilts were synchronized and subjected to cyclical HS (n = 6) or thermal neutral (n = 6) conditions for 5 days during the follicular phase. Ovarian protein abundance of Beclin 1 and microtubule associated protein light chain 3 beta II were each elevated as a result of HS (P = 0.001 and 0.003, respectively). The abundance of the autophagy related (ATG)12–ATG5 complex was decreased as a result of HS (P = 0.002). Regulation of autophagy and apoptosis occurs in tight coordination, and B-cell lymphoma (BCL)2 and BCL2L1 are involved in regulating both processes. BCL2L1 protein abundance, as detected via immunofluorescence, was increased in both the oocyte (∼1.6-fold; P < 0.01) and granulosa cells of primary follicles (∼1.4-fold P < 0.05) of HS ovaries. These results suggest that ovarian autophagy induction occurs in response to HS during the follicular phase, and that HS increases anti-apoptotic signaling in oocytes and early follicles. These data contribute to the biological understanding of how HS acts as an environmental stress to affect follicular development and negatively impact reproduction.
Introduction
Heat stress (HS) occurs when an animal's core temperature rises above biologically imprinted set points because inherent thermoregulatory mechanisms can no longer sufficiently cope with elevated environmental temperature [1]. Attempts to maintain euthermia during HS can lead to perturbations in physiological processes, such as redistribution of blood flow, losses in feed efficiency, growth performance, reproductive ability, and altered body composition [2,3]. These detrimental effects of HS create an annual financial burden on animal agriculture. St. Pierre et al. [4] predicted that the combined effects of HS accounted for an approximately $2 billion loss in US livestock industries, and the economic burden created by HS on the US swine industry alone was estimated to be $900 million per year [5].
Though HS is well known to compromise follicular and early embryo development in a variety of animal models and agriculturally important species [6–10], little is known about how HS affects the integrity of follicular cells. Heat stress impairs gonadotropin receptor function and steroidogenesis in the granulosa cells (GC) [11–13]. Our group has shown that HS decreases the maturation rate for in vitro matured oocytes, and the quality of subsequent embryos from heat-stressed oocytes [14]. This suggests that mitigating stress within ovarian cells could partially alleviate the decrease in reproductive competency caused by HS. One potential inherent mechanism to decrease cellular stress is autophagy.
Autophagy is a key regulator of cellular homeostasis [15], which includes three major types of autophagy: chaperone-mediated autophagy, microautophagy, and macroautophagy. Macroautophagy accounts for the highest turnover of cellular components of the three different types [16]. Macroautophagy (hereafter called autophagy) is the sequestration of cytoplasm into a double-membraned cytosolic vesicle, the autophagosome, which fuses with a lysosome to form an autolysosome for degradation by lysosomal hydrolases [17]. The steps of autophagy are induction, autophagosome formation, autophagosome-lysosome fusion, and degradation [18]. These processes are marked by the formation of large protein complexes, and regulation can occur at the post-translational level via protein–protein interactions [19,20].
Beclin 1 (BECN1) plays an important role in the nucleation of autophagosomes via an association with phosphatidylinositol 3-kinase catalytic subunit type 3 [21,22]. During autophagy, there are two ubiquitin-like conjugation pathways involved with the extension of the autophagosomal membrane. One pathway includes the formation of the autophagy related (ATG)12-ATG5-ATG16 complex, where ATG7 acts like an E1-actiating enzyme to conjugate ATG12 to ATG5 [23,24]. The second ubiquitin-like conjugation pathway results in the cleavage of microtubule associated protein light chain 3 (LC3) alpha/beta, exposing a glycine residue at the C-terminal end. This process results in the conjugation of LC3 with phosphatidylethanolamine, ultimately forming LC3-II [25]. Generation of LC3-II is a well-described marker of mammalian autophagy [26,27].
Autophagy is a mechanism by which dysfunctional intracellular machinery is disposed of, thus autophagy represents a potential molecular mechanism by which the ovary and oocyte attempt to mitigate the detrimental effects of HS. Our working hypothesis is that HS upregulates the autophagy pathway and the objective of this study was to characterize specific autophagy proteins in response to HS. To test this hypothesis, sexually mature female pigs were subjected to either thermal neutral (TN) or HS conditions during the follicular phase of the estrous cycle, and autophagy-related proteins were characterized in the collected ovaries.
Materials and methods
Animals and experimental design
All animal procedures were approved by the Iowa State Institutional Animal Care and Use Committee. Twelve gilts were reproductively synchronized using 15 mg altrenogest (Matrix; Merck Animal Health, Madison, NJ), a progesterone analog, which was administered orally for 14 days to ensure that gilts were entering the follicular phase immediately following withdrawal. Beginning immediately after altrenogest withdrawal and concurrent with synchronized follicular development, gilts were subjected to cyclical HS (n = 6) or TN (n = 6) conditions for 5 days. During the 5 days of temperature treatment, the six gilts in the TN room experienced a constant temperature of approximately 20°C ± 0.5°C. The gilts in the HS room experienced steadily increasing temperature at approximately 0.5°C per hour to a maximum ambient temperature of approximately 31°C at 2000 h. The HS room was then allowed to cool off during the night. This diurnal pattern of cyclical HS treatment was used to mimic environmental HS experienced at a commercial level. Rectal temperatures were taken on all 12 gilts at 0800 h and then every hour between 1200 and 2000 h for all 5 days up until sacrifice. At 124 h following altrenogest withdrawal, gilts were sacrificed and ovaries were harvested and protein extracted for western blot analysis. Due to an effective synchronization program, ovaries were collected at the late follicular stage of the estrous cycle preceding expected ovulation.
Ovary collection
After each gilt was sacrificed and ovaries were excised, the number of dominant follicles was counted on both ovaries and the diameter of each dominant follicle was measured with digital calipers (KD Tools). Dominant follicles were defined based on morphology and a diameter over 6 mm [28]. One ovary was then flash-frozen in liquid nitrogen and the other was fixed in 4% formaldehyde, alternating left and right between gilts. Frozen ovary samples were broken into pieces under liquid nitrogen using a mortar and pestle and then mechanically homogenized in 10 mM phosphate buffer containing 2% SDS. Protein concentrations were estimated using the Pierce bicinchoninic acid assay Protein Assay Kit (Thermo Scientific), and each whole ovary protein sample per animal was diluted to 5 mg/mL. Fixed ovaries were processed, paraffin embedded, sectioned to 5 μm, and mounted onto slides at the University of Iowa Comparative and Histology Lab.
Transmission electron microscopy
Sections of ovary previously fixed in 4% paraformaldehyde in phosphate-buffered solution (PBS) were rehydrated, cut to approximately 2 mm cubes, then and fixed in cacodylate buffer containing 2% paraformaldehyde, 2% glutaraldehyde, and 0.1 M sodium for 24 h at 4°C. The sections were then washed three times for 15 min in 0.1 M cacodylate buffer, and postfixed in 1% osmium tetroxide in 0.1M cacodylate buffer for 1 h. The sections were washed again three times for 10 min each in deionized water, en bloc stained with 2% uranyl acetate for 2 h, washed once for 10 min in deionized water, and dehydrated in 50% graded ethanol for 30 min. Sections were then dehydrated through graded ethanol series (70%, 85%, 95%, 100%) and transitioned into 100% acetone. Samples were infiltrated and embedded into EPON epoxy resin (Electron Microscopy Sciences, Hatfield, PA). Thick and thin sections were made using a Leica UC6 ultramicrotome (Mager Scientific, Dexter, MI). Light microscope images were taken after staining with 1% toluidine blue-o with an Olympus BX-40 compound microscope. Electron microscope images were taken using a JEOL 200kV 2100 scanning and transmission electron microscope (Japan Electron Optics Laboratories, Peabody, MA). Sections of ovarian cortex from three gilts that experienced TN conditions and three gilts that experienced HS conditions (n = 6) were used to visualize the effect of temperature treatment on follicular cell morphology via transmission electron microscopy (TEM).
Western blot analysis
Extracted protein samples were loaded into a 4%–20% Tris glycine gel (Lonza PAGEr Gold Precast Gels) using 50 μg per lane, where 12 lanes contained extracted whole ovarian protein from each animal (6 TN and 6 HS). The BioRad Mini PROTEAN Tetra System was used to run the gel at 60 volts (V) for 30 min followed by 120 V for 90 min. Following separation, the proteins were transferred to a nitrocellulose membrane using the iBlot 2 Dry Blotting System (Life Technologies), with 20 V for 1 min, 23 V for 4 min, and then 25 V for 2 min. Membrane blocking was conducted using 5% milk in PBS for 1 h at room temperature. Antibodies procured from Cell Signaling Technology were used: rabbit anti-BECN1 (C3495), rabbit anti-LC3B (3868), rabbit anti-ATG5 (12994), rabbit anti-ATG12 (4181), rabbit anti-phosphorylated BCL-2 (Thr56; 2875), rabbit anti-BCL2 (2970), rabbit anti-BCL2 like 1 (BCL2L1; 2764), or rabbit anti-caspase 3 (CASP3, 9665) was added (1:1000 dilution) to the membrane in 0.5% milk in PBS overnight at 4°C. A membrane using normal rabbit IgG in place of the primary antibody was used as a negative control. Following primary antibody incubation, the membranes were washed with PBST (PBS with 0.1% Tween) three times at room temperature for 10 min each. Donkey anti-Rabbit IgG (Amersham ECL NA934) was incubated (1:2000) with the membrane for 1 h at room temperature. The membrane was then washed three times for 10 min each at room temperature. Horseradish peroxidase substrate (Millipore, Billerica, MA) was added to the membrane for 1 min in the dark, and it was then exposed to X-ray film and developed for visualization. Average pixel intensity for the band corresponding to the target of each primary antibody was compared for each blot using Image Studio Lite (Li-Core). Loading discrepancies between samples of every blot were corrected for each protein by normalization to β-actin, and all graphs presenting western blot data are after normalization to β-actin. For every blot done, a negative control of rabbit IgG instead of primary antibody was done, and no signal was detected in these negative controls.
Immunofluorescence staining
Ovarian tissues were paraffin embedded, sectioned 5 μm thick, and mounted on microscope slides. Slides underwent two 5-min washes in CitriSolv, were hydrated by two incubations of 100% ethanol for 3 min: a 1-min incubation in 95% ethanol followed by a 1-min incubation in 80% ethanol. Antigen retrieval was accomplished by a 40-min incubation at 95°C in Tris-EDTA buffer (10 mM Tris Base, 1 mM EDTA solution, 0.05% Tween 20, pH 9). Tissue sections were blocked for 30 min in bovine serum albumin (BSA; Sigma-Aldrich), followed by an incubation in primary antibody in 5% BSA (1:100) overnight at 4°C. The following day, slides were washed for 10 min in PBS three times, incubated in fluorescein linked goat anti-rabbit IgG secondary antibody (Life Technologies) in 5% BSA (1:500) for 1 h, and then subjected to three additional 10-min PBS washes. Images of ovarian sections were captured with an inverted microscope at ×200 and ×400 magnification. Negative controls included exclusion of primary antibody, as well as replacing the primary antibody with normal rabbit IgG. ImageJ software (NCBI) was used to quantify signal difference between treatments for fluorescence microscopy [29]. Two ovarian sections per animal were used to quantify oocyte-specific BCL2L1, BECN1, or LC3B abundance, where at least three oocytes of either secondary or early tertiary follicles per gilt, or ten primary follicles per gilt, were quantified. The intensity of fluorescence signal presented is oocyte specific signal intensity per area of individual oocytes or the primary follicle specific signal intensity per area of primary follicles per image.
Statistical analysis
Statistical analysis of rectal temperature differences, ovary follicle number, ovary follicle type, western blot data, and immunofluorescence signal was conducted using PROC MIXED in SAS with random effects, where a standard Student t-test was used to compare statistical differences. Statistical significance was determined when P values were less than or equal to 0.05.
Results
Heat stress had no effect on ovarian follicle size or number of follicles
The average room temperature for each day was 20.3 ± 0.5°C for TN conditions, and the HS room temperature ranged from 26°C to 32°C to mimic a diurnal pattern. During the maximal heat load for each day, the HS pigs had increased (P = 0.001) average rectal temperatures (39.8 ± 0.2°C) compared to the TN pigs (38.8 ± 0.2°C; Figure 1A). There was no treatment effect detected between TN and HS animals on the number of tertiary follicles per ovary (P = 0.20) or follicle diameter (P = 0.96; Figure 1C and D). Large, dominant follicles were apparent in all ovaries regardless of group.
Heat stress did not impact follicle size or number of follicles. During the maximal HS load for each day, the pigs undergoing HS had increased (P < 0.01) average rectal temperatures compared to pigs in TN conditions (A) due to increased environment temperature (B). HS had no effect on the number of tertiary follicles per ovary (C) nor the follicle diameter (D). This indicates that while the gilts undergoing the HS treatment had elevated body temperatures, tertiary follicles were still able to develop under HS conditions mimicking a diurnal pattern.
Heat stress increases vacuolization of the oocyte and surrounding granulosa cells
To determine the effects of HS on cellular morphology of the oocyte and the surrounding GCs, TEM was performed on ovarian sections (n = 3 per treatment). Heat stress qualitatively increased the incidence of vacuolization of the oocytes and GCs, observed as spherical white spaces, compared to TN oocytes and GCs (Figure 2A–C). The GC surrounding oocytes from HS ovaries had increased incidence of sequestered cytoplasm of smaller diameter (Figure 2B), while the oocytes from HS ovaries contained large vacuole-like structures (Figure 2C). These results indicate that HS may not alter the number of tertiary follicles, though HS does affect the cellular morphology of oocytes and surrounding GC.
Environmental heat stress increases vacuolization in follicles. Ovaries from TN (n = 3) or HS (n = 3) gilts were examined by TEM to characterize cellular morphology of follicles. (A) Representative images of follicles from either a TN or HS ovary. The follicles of gilts exposed to HS conditions had increased space between the oocyte and surrounding GC. (B) The HS GC had increased vacuole-like structures. The oocytes of gilts objected to HS also had increased vacuole-like structures. (C) The arrows denote vacuole-like structures where cytoplasm is being sequestered.
Markers of autophagy are increased in the ovary due to heat stress
Western blots using total ovarian protein were conducted to characterize the protein abundance of autophagy-induction markers BECN1, LC3B-II, and the abundance of the ATG5-ATG12 complex. Heat stress increased ovarian protein abundance of BECN1 (P = 0.001) compared to ovaries from TN gilts (Figure 3A and B). The abundance of LC3B-II was increased in HS ovaries (P = 0.003), as well as the ratio of LC3B-II to LC3B-I (P = 0.006; Figure 3B) demonstrating an increase in the induction of autophagosome formation. Cyclical HS did not affect ATG5 or ATG16 abundance in whole ovarian protein. The abundance of ATG12 in complex with ATG5 decreased in the HS ovaries compared to ovaries from animals in TN conditions (P = 0.002; Figure 3A and B). Though the reduction of ATG12 in complex with ATG5 due to HS was unexpected, the increase of both BECN1 and cleaved LC3B-II supports that HS induces components of autophagy signaling in the ovary.
Heat stress alters autophagy-related protein abundance in the ovary. Gilts either underwent 5 days of cyclical HS or TN conditions after synchronization into the follicular phase. Western blotting of whole ovary protein for each gilt (n = 12) with antibodies directed toward autophagy-related proteins (A) showed an increase in protein abundance due to HS for BECN1, LC3B-II alone, and the ratio of LC3B-II/LC3B-I (B). Asterisk denotes P < 0.05.
Immunohistochemical (IHC) staining of ovarian sections was conducted for both BECN1 and LC3B. BECN1 was detected mainly in interstitial tissues (Figure 4A–D) and the GCs of tertiary follicles (Figure 4G–J). LC3B was also detected mainly in interstitial tissues (Figure 5A–D), as well as oocytes of growing follicles (Figure 5G–J). The abundance of oocyte specific signal pertaining to either BECN1 or LC3B was quantified, and a difference between TN and HS groups was not detected (P > 0.05; data not shown). These immunostaining results indicate that autophagy-related proteins were present in ovaries from either TN or HS gilts.
Localization of BECN1 in the ovary. BECN1 was localized to the ovarian interstitial tissues (A–D) and tertiary follicles (G–J). Representative grayscale images of the fluorescein signal alone (A, C, D, G, I, K) and representative composite images of antibody-labeled BECN1 (green) and DAPI-stained chromatin (blue) (B, D, F, H, J, L) are presented. Negative controls using rabbit IgG instead of rabbit anti-BECN1 primary antibody were used (E, F, K, L), and each image was captured at the same exposure where no signal was detected in the negative controls. White arrows point out areas of detected signal. White bar represents 100 μm.
Localization of LC3B in the ovary. LC3B was localized to the ovarian interstitial tissues (A–D) and oocytes of growing follicles (G–J). Representative grayscale images of the fluorescein signal alone (A, C, D, G, I, K) and representative composite images of antibody-labeled LC3B (green) and DAPI-stained chromatin (blue) (B, D, F, H, J, L) are presented. Negative controls using rabbit IgG instead of rabbit anti-LC3B primary antibody were used (E, F, K, L), and each image was captured at the same exposure where no signal was detected in the negative controls. White arrows point out areas of detected signal. White bar represents 100 μm.
Anti-apoptotic signaling is increased in the ovary due to heat stress
B-Cell lymphoma 2 (BCL2) family members, BCL2 and BCL2 like 1 (BCL2L1; formerly BCL-xL), can regulate both autophagy and apoptosis through interaction with BECN1 or BCL2 associated X, apoptosis regulator (BAX), respectively [30–32]. The abundance of total BCL2 was not affected by HS, but HS increased the abundance of phosphorylated BCL2 at Thr56 (Figure 6A and B). The abundance of BCL2L1 was also increased due to HS (Figure 6A and B), and BCL2L1 was detected in the interstitial tissues (Figure 7A–D) and GC and theca cells of tertiary follicles (Figure 7G–J) from both TN and HS ovaries. There was an increase in BCL2L1 abundance in prophase I arrested oocytes (Figure 8A and C) and the somatic cells of primordial follicles of ovaries from gilts subjected to HS, visualized via immunofluorescence (Figure 8B and D; P < 0.05). The phosphorylation of BCL2 at the Thr56 amino acid is known to decrease apoptotic signaling [33]. Heat stress increased the phosphorylation of BCL2 at Thr56 as well as BCL2L1 abundance in the ovary. Collectively, 5 days of cyclical HS altered apoptotic signaling in the ovary via BCL2 and BCL2L1 in the ovary (Figure 6A and B), and specifically BCL2L1 in the oocyte, as quantified through IHC (Figure 8A). Also, there was little to no detection of caspase 3 (CASP3) cleavage in either the TN or HS ovaries (data not shown), suggesting that apoptosis was not induced in the ovary at the level of HS the gilts experienced [34].
Heat stress impacts anti-apoptotic protein abundance in the ovary. Anti-apoptotic signaling is increased in the ovary due to HS. Western blotting of whole ovary protein for each gilt (n = 12) with antibodies directed toward either total BCL2, phosphor-BCL2 Thr56, or BCL2L1 (A) showed an increase in phosphorylation of BCL2 at Thr56, but not total BCL2, and in increase in the abundance of BCL2L1 (B). Immunohistochemistry detected signal for BCL2L1 in the interstitial tissues and tertiary follicles for animals that underwent either TN or HS conditions, albeit not different between treatments (C). Asterisk denotes P < 0.05. White bar represents 100 μm.
Localization of BCL2L1 in the ovary. BCL2L1 was localized to the ovarian interstitial tissues (A–D) and tertiary follicles (G–J). Representative grayscale images of the fluorescein signal alone (A, C, D, G, I, K) and representative composite images of antibody-labeled BCL2L1 (green) and DAPI-stained chromatin (blue) (B, D, F, H, J, L) are presented. Negative controls using rabbit IgG instead of rabbit anti-BCL2L1 primary antibody were used (E, F, K, L), and each image was captured at the same exposure where no signal was detected in the negative controls. White arrows point out areas of detected signal. White bar represents 100 μm.
Heat stress increases BCL2L1 abundance in the oocyte and primary follicles. BCL2L1 abundance increased in oocytes and primordial/primary follicles due to HS. Immunohistochemistry using a primary antibody against BCL2L1 and a fluorophore-linked secondary antibody was done for ovary sections from gilts that either underwent 5 days of TN or cyclical HS conditions. BCL2L1 protein abundance was significantly increased in oocytes (A and C) and in primary follicles (B and D). ImageJ was used to quantify the fluorescent signal in oocytes. Asterisk denotes P < 0.05. White bar represents 100 μm.
Discussion
Seasonal infertility refers to the reduction in reproductive competence during the summer months [35]. Though seasonal infertility is due to a multitude of factors, such as photoperiod [36], associated elevated temperatures during the summer months have a large or at least cumulative effect on reproductive competence [37,38]. Collectively, HS costs the global animal agriculture industries billions of dollars per year. There are numerous mechanisms through which HS compromises animal production [3]. With respect to reproduction specifically, HS compromises fecundity through a variety of mechanisms including the induction of altered ovarian function [2].
Here we utilize an in vivo HS model to characterize the effect of HS on the ovary. This model was based on previous studies [13,39]. As in previous studies, gilts expressed HS-associated behavioral symptoms: lethargy, decreased feed-intake, and increased respiration rate (data not shown). The gilts appeared to acclimate to HS by day 5, as observed through rectal temperatures. This was expected due to previous work characterizing the response to HS in pigs [40,41], and it has been suggested that physiological acclimation to HS can itself be harmful to the animal [42,43]. Follicular development appeared to occur normally under HS, yet qualitatively HS increased the amount of unhealthy primary follicles, visualized via TEM. The data presented here suggest that HS can increase cellular stress within the ovary, and that this induces autophagy.
Autophagy is a mediator of the cellular stress response and has previously been shown to occur in the ovary [44–46]. Autophagy induction can be manifested in response to several environmental stressors such as cigarette smoke, which increases ovarian autophagosome formation in mice [45]. Heat stress compromises pig oocyte maturation and competency in in vitro studies [14,47,48], although little is known regarding the stress response within the follicle during in vivo exposure to HS.
The 5-day in vivo exposure to cyclical HS during the follicular phase increased vacuole-like structures in the oocyte and GC. The formation of areas of cytoplasmic sequestration was present in both the oocyte and surrounding GC in ovaries from heat-stressed gilts providing visual evidence of autophagosome formation. Interestingly, these vacuole-like structures appeared larger in the oocyte compared to the surrounding GC, which had areas of sequestered cytoplasm more similar to the size of mammalian autophagosomes (0.5–1.5 μm, [49]). The majority of autophagosome visualization in follicular cells has been in GC [45,50,51], and whether autophagy actually occurs in the oocyte is poorly understood. Oocytes have been classified as having autophagosome-like features [52,53], but autophagosome formation has not definitively been demonstrated in oocytes. Autophagy also appears to be necessary for oocyte maintenance, as the loss of BECN1 or ATG7 results in premature loss of oocytes [54]. Therefore, HS causes morphological changes in ovarian cells and qualitatively increased autophagosome-like features.
Mammalian oocytes have previously been shown to undergo necrosis in ovarian culture [53], developing a morphology similar to neural cells undergoing necrosis [55,56]. Autophagy is indispensable for neural cells, which do not divide after differentiation [57]. We speculate that oocytes could be utilizing autophagy in the same manner as neural cells, as neural cells and oocytes both must maintain longevity through mechanisms other than cell division. Beclin 1 complexes are central to membrane trafficking and control of autophagy in neural cells [58]. It is likely that the large areas of sequestered ooplasm seen here arise via BECN1 complexes and lead to the degradation of protein aggregates or damaged organelles.
In addition to the detection of autophagosome-like structures in the oocyte and GC, HS increased the abundance of proteins associated with autophagy induction in the pig ovary. The abundance of BECN1 and the ratio of LC3B-II to LC3B-I was increased, in comparison with ovaries from TN gilts. Whole ovary abundance of ATG5 in complex with ATG12 was decreased by HS, which was unexpected as the ATG12-ATG5 complex is part of the ubiquitin-like conjugation pathway for autophagosome formation [59]. This result could potentially be due to the difference in the dynamics between the two ubiquitin-like conjugation pathways that extend the autophagosomal membrane [23]. Additionally, this study demonstrated substantial formation of autophagosome-like structures at the single time point of ovary collection, which suggests that ATG12–ATG5 complex formation may not have been detected at its most abundant level.
Both BECN1 and LC3B were localized to the ovarian interstitial cells and follicular cells in immunostained sectioned ovarian tissue. The presence of these proteins in ovarian cells suggests that autophagy could be utilized in response to HS. The IHC signal from antibody detected BECN1 and LC3B was not quantified due to the high degree of different cell types in the ovary and signal variability within a single ovarian section. Even within the same cell type in the same ovarian tissue section, there appeared to be spatial differences in signal intensity for both BECN1 and LC3B. Though localization of BECN1 and LC3B to a specific cell type was consistent between different gilts, signal intensity appeared to be variable.
The fact that differences in BECN1 and LC3B abundance between TN and HS treatments were detected through western blotting but differences were not seen through IHC could potentially be due to changes in circulating immune cells in the ovary due to HS. In the mouse, immune cell function has been linked to estrous cyclicity [60], and feed restriction in rats reduces the number of macrophages surrounding preovulatory follicles [61]. This study does not explicitly characterize the effect of HS on autophagy in circulating immune cells.
The biological importance of autophagosome formation in the ovary in response to environmental stress hinges on the role of the autophagosome after HS-induced formation. One potential feature of autophagosome formation in the ovary could be for the purpose of mitophagy, or the selective engulfment of damaged mitochondria by autophagosomes and the subsequent degradation via lysosomes [62]. Mitophagy is of interest, as oocytes of developing follicles greatly increase the mitochondrial population [63,64], which is subsequently parsed out during holoblastic cleavage of developing embryos prior to zygotic genome activation [65–67].
Evidence for the importance of autophagy is that murine models with loss of ATG7 function, critical for both the cleavage of LC3 and the formation of the ATG5-ATG12 complex, result in the accumulation of defective mitochondria in multiple tissue types [25,68–70]. The observed increase of BECN1 in this study is congruent with others demonstrating that mitochondrial damage is associated with increased puncta of BECN1 in oocytes [71].
This study investigated a single time point after 5 days of HS thereby making a detailed characterization of autophagic flux not possible. However, the results suggest that HS increased autophagosome formation in the oocyte and surrounding follicular cells, providing impetus to investigate specific upstream regulators of autophagy, such as protein kinase B subunit 1 (AKT1) and mechanistic target of rapamycin (mTOR; formerly mammalian target of rapamycin) [72,73], in future studies. It has also been previously shown that HS increases the phosphorylation AKT1 in the ovary [13], which could potentially inhibit mTOR, thereby inducing autophagy [74]. Understanding these mechanisms, which are important for further elucidating what cellular structures are being targeted by autophagosomes, could further lead to the development of strategies to mitigate the effects of HS on reproduction.
One such feasible approach to mitigate HS-induced deleterious effects on livestock reproduction is to manipulate autophagy through diet. This is feasible because affecting autophagy through dietary supplementation has been shown to decrease memory loss [75], increase immunity to viral infections [76], and alleviate fatty liver disease [77]. As such, it seems plausible that increased autophagy through dietary supplementation could increase livestock animal reproductive efficiency during abiotic or cellular stress. Speculatively, because autophagy is a mechanism by which the oocyte mitigates environmental stress, the addition of autophagy inducers in in vitro maturation media could increase oocyte quality in techniques such as in vitro fertilization and embryo transfer. There is also a myriad of chemicals that increase autophagy, which could be valuable to include in assisted reproductive techniques.
Autophagy and apoptosis are regulated in tight coordination, partly through BCL2 family member proteins, which are important regulators of apoptosis during mammalian ovary development [78–82]. The dual role of BCL2 family members to regulate both autophagy and apoptosis is mediated by the ability of BCL2 and BCL2L1 to prevent apoptosis by inhibiting the formation of mitochondrial pores that release cytochrome C [83], while BCL2 and BCL2L1 can also interact with BECN1 to regulate autophagy [31,32].
Phosphorylation status of BCL2 mediates its regulatory capacity, as phosphorylation at the threonine 56 (T56) position is associated with anti-apoptotic signaling, and the deletion of T56 eliminated its ability to inhibit mitochondrial depolarization [33]. Our study demonstrates increased ovarian T56 phosphorylation of BCL2 due to HS. We also observed a fivefold increase in the abundance of BCL2L1 in the HS oocyte compared to the ovaries of TN gilts, suggesting that anti-apoptotic signaling is increased in the ovary due to HS concomitant with markers of autophagy induction.
The abundance of BCL2L1 was increased in ovaries from gilts subjected to HS, specifically in the oocytes, compared to TN. As BCL2L1 regulates both autophagy and apoptosis through its interaction with BECN1 or BAX, respectively [84], not only does this shed some light on how autophagy could be regulated in the oocyte, but also provides rationale to further characterize autophagy specifically in the oocyte undergoing HS. Our group has previously shown that HS not only reduced the maturation rate of oocytes, but also lowered the developmental competency of oocytes that entered metaphase II [14].
Not only does HS affect the oocyte specifically, but it is also detrimental to the surrounding follicular cells. The space between the oocyte and GC appeared increased and GCs were morphologically less healthy in the follicles from ovaries of gilts that underwent HS compared to the ovaries of TN gilts. If HS affects primordial or primary follicles in the ovary, then this could decrease the amount of healthy follicles in the follicular reserve, thereby decreasing the lifetime reproductive capacity of an animal [85].
In conclusion, markers of autophagy induction are elevated in the pig ovary due to HS adding to our understanding of intraovarian signaling contributing to folliculogenesis and oocyte development during HS. Subsequent experiments are warranted to test whether the competency of produced oocytes can be increased by taking advantage of artificial induction of autophagy, either in vitro or in vivo, thereby creating mitigation strategies to maintain reproductive integrity under environmental stress.
Acknowledgment
The authors would like to acknowledge Tracey Pepper at the Iowa State University Microscopy and NanoImaging Facility for her assistance with TEM, Adrianne Kaiser for her continued technical support, as well as other members of the Ross, Keating, and Baumgard research groups at Iowa State University for their help with this project.
References
Author notes
Conference Presentation: Hale BJ, Hager C, Al-Shaibi A, Baumgard LH, Keating AF, Ross JW. 2015. Heat Stress Induces Autophagy in Pig Ovaries during Follicular Development. 48th Annual Meeting of the Society for the Study of Reproduction, San Juan, Puerto Rico June 18–22.
