|Home | About | Journals | Submit | Contact Us | Français|
Sir Joseph Barcroft asserted that the fetus develops under an oxygen tension comparable to that experienced on the summit of Mount Everest; the ‘Everest in utero’ hypothesis (Barcroft, 1946). The experiments he performed, and those of others, demonstrated clearly that oxygen availability is an important determinant of fetal growth. The question arises as to what happens during pregnancy when the feto‐placental unit suffers hypoxia and fetal oxygenation might be expected to be reduced further. Is the placenta able to adapt to maintain the oxygen flux to the fetus? Placental hypoxia has been implicated in the pathophysiology of various complications of pregnancy, including non‐genetic or infectious causes of growth restriction, allthough it has never been confirmed by measurements performed in vivo. By contrast, pregnancy at altitude represents an experiment of nature in which the maternal arterial oxygen tension (P aO2.) is known to be significantly lowered, dropping from 95 mmHg at sea level to ~50 mmHg at 3500–4000 m (Krampl et al. 2000; Postigo et al. 2009). This decrease is partially compensated for in terms of oxygen carriage by physiological responses, with an increase in the maternal haematocrit and haemoglobin concentration (Julian et al. 2009; Postigo et al. 2009). Nonetheless, the incidence of small‐for‐gestational age births increases in humans at high altitude, with birth weight being reduced by ~100 g for every 1000 m of elevation (Giussani et al. 2001). The effect is greatest in recent migrants of non‐indigenous origin (Moore et al. 2001; Julian et al. 2007) and is independent of other risk factors, including smoking and socioeconomic class (Giussani et al. 2001).
In comparison with the well‐established change in birth weight in high‐altitude pregnancies, the effect on the placenta is contentious. Increases, decreases or no difference in human placental weight have all been reported (Jackson et al. 1987; Reshetnikova et al. 1994; Tissot van Patot et al. 2003; Zamudio, 2003; Tissot van Patot et al. 2010). This variability in the data is probably a result of intrinsic differences among the populations studied, such as ethnic background, dietary intake and altitudinal levels, as well as the mode of delivery and processing of the placenta. Animal studies have produced equally variable results. There was no change in placental weight when guinea‐pigs and mice were exposed to 12% oxygen throughout pregnancy (Bacon et al. 1984) or from embryonic day (E)14.5 (Cuffe et al. 2014), respectively, whereas placental weight increased in rats exposed to 13% oxygen from E6.5 onwards (Richter et al. 2012).
Upon adverse conditions, the fate of cells is determined by the balance between activation of stress‐response and growth signalling pathways. Hypoxia leads to an immediate reduction in energy‐demanding cellular functions aiming to conserve resources. These functions include the synthesis of non‐essential proteins, which is achieved by a selective arrest of translation mediated principally through phosphorylation of eukaryotic initiation factor eukaryotic initiation factor 2 subunit α (eIF2α) (Koritzinsky et al. 2006; Fahling, 2009). Over the longer term, changes in protein kinase B‐mechanistic target of rapamycin (Akt‐mTOR) signalling affect metabolic, angiogenic, anti‐oxidative and other adaptive responses that promote the survival of a cell within its new environment (Manning & Cantley, 2007; Sengupta et al. 2010).
The activities of many components in the Akt‐mTOR pathway are regulated by oxidative or endoplasmic reticulum (ER) stress at multiple levels, including gene expression, protein translation and phosphorylation (Koritzinsky et al. 2006; Yung et al. 2007; Yamaguchi et al. 2008; Yung et al. 2011). Phosphorylation of residues Thr308 and Ser473 critically regulates Akt activity and downstream substrate specificity (Guertin et al. 2006; Jacinto et al. 2006), whereas the phosphorylation level of 4E binding protein 1 (4EBP‐1) is the readout of mTORC1 activity and regulates protein synthesis (Hay & Sonenberg, 2004). ER stress can attenuate Akt protein translation (Yung et al. 2007), as well as up‐regulate gene expression of 4EBP‐1. Mice with a transgenic deletion of Akt1 show a decrease in mTOR signalling, and a reduction in placental and fetal weight by 50% and 20%, respectively (Yang et al. 2003; Yung et al. 2008). Akt‐mTOR signalling is also reduced in human growth restricted placentas (Yung et al. 2008), indicating the importance of Akt‐mTOR signalling in placental growth. In a recent study, we also reported that Akt‐mTOR signalling is reduced in the human placenta from high altitude, despite no change in placental weight (Yung et al. 2012).
Both elevated and reduced oxidative stress have been reported in placentas from high altitude (Jefferson et al. 2004; Zamudio et al. 2007; Yung et al. 2012). Oxidative stress and ER stress are closely interlinked and are increased in human placentas from cases of fetal growth restriction (Burton & Yung, 2011). ER stress activates the unfolded protein response (UPR), a cellular homeostatic response to stimuli/stresses aimed at restoration of normal ER function (Walter & Ron, 2011). It comprises three evolutionarily conserved signalling pathways, including protein kinase RNA‐like endoplasmic reticulum kinase (PERK)/eIF2α, activating transcription factor 6 and inositol requiring endoribonuclease 1 (IRE1)/X‐box binding protein 1 (XBP‐1). These pathways are activated sequentially in response to ER stress, resulting in the initiation of a series of adaptive responses, including attenuation of non‐essential protein translation, an increase of ER folding capacity and facilitation of degradation of misfolded proteins.
To investigate the effects of chronic hypoxia on placental development, pregnant mice were housed from day 0.5 of gestation inside an hypoxic chamber with a normobaric atmospheric oxygen concentration of either 16% or 13%, which is equivalent to the hypobaric oxygen concentration at ~2100 m and 3600 m, respectively. Placentas were assessed morphologically, as well as at the molecular level for adaptive responses.
All chemicals were purchased from either Sigma‐Aldrich (Poole, UK) or Fisher Scientific UK Ltd (Loughborough, UK), except where statedotherwise. Antibodies against phospho‐Akt (Ser473), phospho‐Akt (Thr308), Akt, phospho‐4EBP‐1 (Thr37/46), 4EBP‐1, phospho‐eIF2α (Ser51), eIF2α, phospho‐p38 MAPK (Thr180/Tyr182), p38 kinase, phospho‐AMP‐activating kinase(AMPK)α(Thr172), AMPKα, phospho‐heat shock protein (HSP)27 (Ser82) and HSP27 were obtained from Cell Signalling Technology (NEB, Hitchin, UK). Antibodies against XBP‐1, GRP94 and HSP60 were obtained from Abcam (Cambridge, UK), and HSP90 and HSP70 were obtained from Enzo Life Science (Exeter, UK). Anti‐GRP78 was from Transduction Laboratories (BD Biosciences, Oxford, UK) and anti‐β‐actin was obtained from Sigma‐Aldrich. OxPhos Complex Kit for mitochondrial electron transport chain (ETC) subunits was obtained from Invitrogen (Renfrew, UK).
All experimental procedures were carried out in accordance with the UKHome Office Animals (Scientific Procedures) Act 1986 as described by Drummond (2009), which mandates ethical review. C57BL/6 mice were bred and housed in conventional cages under standard conditions (12 : 12 hlight/dark cycle at 23 ± 1°C and 60% humidity) with free access to food and water. For the experiments, virgin females (8–12 weeks of age) were paired with fertile males (minimum 10 weeks of age). The presence of a copulatory plug was considered as day 0.5 of pregnancy (term ~20 days). The pregnant mice were randomly allocated to normoxic (20% O2), mild or more severe hypoxic (16% O2 or 13% O2) groups. The value of 13% O2 was chosen because previous studies in the rat showed that this is the lowest level that does not induce any change in food intake (Richter et al. 2012). Females assigned to the hypoxic groups were placed inside an hypoxic chamber, which combined a PVC isolator (PFI Plastics Ltd, Keynes, UK) with a nitrogen generator (N2MID60; Dominick Hunter Ltd, Warwick, UK), with an oxygen concentration setting of either 16% or 13%. The normoxic group was housed in the same room. Daily food and water intake in the groups were measured. Animals were killed by cervical dislocation at E18.5. Uteri were removed and immediately immersed in ice‐cold PBS. The number of pups and resorptions in each horn was counted. The placentas and fetuses were weighed, and the placentas were cut equally into two portions; one half was snap‐frozen in liquid nitrogen for western blotting analysis and the other was fixed in ice‐cold 4% paraformaldehyde in PBS for subsequent embedding in paraffin wax for stereology and immunohistochemistry. Tail tips were collected for sex determination by PCR.
Western blotting analysis of protein expression and kinase phosphorylation was preformed as described previously (Yung et al. 2007). In brief, placental tissue lysates were prepared using MagNA Lyser Instrument (Roche Diagnostics, Lewes, UK) with Lysing Martrix D (MP Biomedicals, Carlsbad, CA, USA) and lysis buffer containing 20 mm Tris (pH 7.5), 150 mm NaCl, 1 mm, EDTA, 1 mm EGTA, 1% Triton X‐100, 2.5 mm sodium pyrophosphate, 1 mmβ‐glycerolphosphate, 1 mm Na3VO4 and complete mini proteases inhibitor cocktail (Roche Diagnostics). Protein concentration of the tissue lysate was determined using a Bicinchoninic acid kit (Sigma‐Aldrich). Equivalent amounts of protein were resolved by SDS‐PAGE, blotted onto nitrocellulose (0.2 μm) and analyzed by enhanced chemiluminescence (Amersham Biosciences, Little Chalfont, UK) using X‐OMAT Autoradiographic film (Sigma‐Aldrich). β‐actin or Ponceau S staining was used to normalize protein loading. Films were scanned using a flat‐bed scanner (HP G4050;Hewlett Packard, Palo Alto, CA, USA) and the intensities of the bands representing phospho‐ and total kinase forms were determined from two or three different exposures (within the linear detection range) using Image J (NIH, Bethesda, MD, USA).
Sections (7 μm) were dewaxed and rehydrated through xylene: twice for 5 min, and then 3 min each in the ethanol series: 100% ethanol, 90%, 70% and 50%. Slides were placed in PBS for 5 min and rinsed with water, stained in haematoxylin for 8 min, differentiated in acid‐alcohol, rinsed under tap water for 10 min, counterstained in eosin for 30 s and rinsed in water. They were rehydrated through the graded ethanol series as described above, but in reverse order, for 1 min each, placed in xylene twice for 5 min and mounted using DPX.
The analysis was conducted as described previously (Coan et al. 2004), except that the placental slides were scanned using a Nanozoomer (Hamamatsu Phonics UK, Ltd, Welwyn Garden City, UK) and analysed using newCAST software (Visiopharm Inc., Hørsholm, Denmark). In brief, six placentas from each group were randomly selected. Half of each placenta was exhaustively serially sectioned, and every 20th section was stained with haematoxylin and eosin. Placental volume was calculated using the Cavalieri technique, and the volume fraction of the different placental compartments was estimated by point counting.
The sex of the pups was determined using PCR approach as described by Clapcote & Roder (2005). In brief, PCR was carried out to amplify exons 9 and 10 of Jarid1c and Jarid1d on the X and Y chromosomes, respectively, using forward primer 5'‐CTGAAGCTTTTGGCTTTGAG‐3' and reverse primer 5'‐CCACTGCCAAATTCTTTGG‐3' with the profile: 94°C for 5 min, followed by 35 cycles of 94°C for 20 s, 54°C for 1 min and 72°C for 40 s, then followed by 72°C for 10 min in a GeneTouch thermal cycler (Alpha Laboratories Ltd, Eastleigh, UK). PCR products, which are of different lengths (331 bp and 302 bp from the X and Y chromosomes, respectively), were then resolved in 2% (w/v) agarose gel and the presence of X and Y chromosomes was determined.
All statistical analysis was performed in Prism, version 6.0 (GraphPad Software Inc., San Diego, CA, USA). Differences were tested using either the two‐tailed Student's t test or, when appropriate, the non‐parametric Mann–Whitney U test. P < 0.05 was considered statistically significant. For multiple comparisons, differences were tested using two‐way ANOVA, followed Bonferroni's multiple comparisons test.
Mice were subjected to normobaric hypoxia at 16% and 13% O2, equivalent to the oxygen concentration at ~2100 m and ~3600 m, respectively, for the whole of gestation. Food intake was reduced under 13% oxygen (P < 0.001), although there was no change in water intake (P = 0.932) at either level compared to the normoxic (20.8%) controls (Fig. (Fig.11 A and B). Litter size (live fetuses) was significantly reduced at 13% O2 (6.2 ± 0.5; number of litters = 10) in comparison with normoxia (7.6 ± 0.5; number of litters = 12) and 16% O2 (8.3 ± 0.3; number of litters = 4)(Fig. 1 C). There was a comparable higher incidence of embryonic loss (resorption and embryonic death) at 13% O2 (1.1 ± 0.4; n = 10) compared to normoxia (0.4 ± 0.2; n = 12) (Fig. (Fig.11 D)
As expected, exposure of pregnant animals to chronic hypoxia reduced fetal growth, with a reduction of 11.7% in fetal weight at E18.5 after 13% O2 compared to normoxia (Fig. (Fig.2).2). Interestingly, placental weight was increased by 10%. No changes in fetal and placental weight were found between 16% O2 and normoxia (Fig. (Fig.2)2) and, consequently, we focused on 13% O2 for the remainder of the study.
Next, the factors that might account for the heavier placenta under 13% O2 were investigated. First, we examined any alteration in placental structure using a stereological approach. The isolated murine placenta can be divided into four distinctive regions based on their origin or functions, including the maternal decidua layer, junctional zone (endocrine region), labyrinthine zone (nutrient and gaseous exchange region) and chorionic plate (Georgiades et al. 2002). The macroscopic structure of placentas under normoxia and 13% O2 are shown in Fig. 3 A. In the hypoxic placentas, we observed large cavities in both the junctional and labyrinthine zones, which, at high magnification, were confirmed to be blood vessels because they were lined by endothelial‐like cells and contained erythrocytes. They are either maternal arterial channels supplying or venous vessels draining the placenta, depending on their location (Georgiades et al. 2002). Therefore, the terms ‘maternal arterial blood space’ and ‘maternal venous blood space’ were included in the stereological study.
A total of six placentas from three different litters of each group were used for stereology. No significant increase in the volumes of the major compartments of the placentas was found (Fig. (Fig.33 B). However, a 91% (equivalent to 3.65 mm3) increase in the maternal blood spaces in the 13% O2 placentas was observed (Fig. (Fig.33 B); the median maternal arterial and venous blood space volumes were increased by 132% and 54% respectively, compared to the normoxic controls. The volumes of placentas under normoxia and 13% O2 were 92.7 mm3 and 102.4 mm3, respectively. Hence, the 3.65 mm3 increase in maternal blood space cannot account fully for the 10% difference in the placental weight observed.
Accordingly, we aimed to obtain evidence of enhanced placental growth under hypoxia by investigating the central Akt‐mTOR growth regulatory pathway. In comparison with normoxic placentas, there was an increase in the activity of the Akt‐mTOR pathway as shown by significant elevation of phosphorylation of Akt at both Thr308 and Ser 473, as well as of 4EBP‐1 at Thr37/46 (Fig. (Fig.44 A and B).
These results suggest that the heavier hypoxic placentas were the result of both increased placental growth and enlargement of the maternal blood vessels supplying and draining the placenta.
Both oxidative and low‐grade ER stress have been reported in human placentas from high‐altitude pregnancies (Yung et al. 2012). Therefore, we determined whether the same stresses are also observed in the hypoxic murine placentas. Surprisingly, no increases in the major oxidative stress markers, including P‐p38 kinase, HSP70 and HSP90, were observed (Fig. (Fig.55 A and B), and phosphorylation of the small HSP27 was even reduced > 50% (Fig. (Fig.55 B). Furthermore, immunohistochemical staining for lipid peroxidation with antibodies directed against 4‐hydroxynonenal revealed no difference between placentas from normoxia or 13% O2 (Fig. (Fig.55 C). These results suggest that, when analysed as a pooled group, the hypoxic murine placentas did not suffer from oxidative stress.
In comparison with oxidative stress, the UPR pathways appear to be more sensitive to hypoxia (Yung et al. 2012). Increases of ~100% and >70% were observed for P‐eIF2α/eIF2α and XBP‐1 respectively, whereas other markers, including GRP78 and GRP94, remained unchanged. These results indicate activation of PERK and IRE1α but not the activating transcription factor 6 pathway (Fig. (Fig.66 A and B). This indicates that hypoxic murine placentas suffer mild ER stress.
Phosphorylation of eIF2α suppresses the translation of proteins in the mitochondrial ETC (Colleoni et al. 2013) and the level of some subunits was found to be reduced in human placentas at high altitude (Colleoni et al. 2013). Therefore, we investigated whether the high
P‐eIF2α in the hypoxic murine placentas was associated with changes in the mitochondrial ETC. HSP60 is a mitochondrial chaperone responsible for the transport and refolding of proteins from the cytoplasm into the mitochondrial matrix. It was increased significantly by 35% under hypoxia (Fig. (Fig.77 A and B), suggesting the probable existence of mitochondrial stress. Indeed, the level of some mitochondrial ETC complex subunits was reduced by >60%, including complex V (ATP5a), complex IV (MTCO1), complex III (UQCRC 2) and complex II (SDHB) (Fig. (Fig.77 A and B). Surprisingly, despite these changes, there was no indication of an energy crisis in the hypoxic placentas because phosphorylation of the energy sensing kinase, AMPKα, showed no difference. However, we did observe that the maternal glucose concentration at E18.5 was reduced significantly by 29% [7.6 ± 0.8 mmol (n = 8) to 5.5 ± 0.3 mmol (n = 9) P = 0.011] (Fig. (Fig.8).8). These results suggest that hypoxic placentas might carry out more anaerobic respiration to compensate for any reduction of ATP generation resulting from the defective mitochondrial ETC.
There is growing evidence that sexual dimorphism plays a crucial role in determining the outcome of pregnancies affected by stress (Gabory et al. 2013). Therefore, we reanalyzed the data on fetal and placental weights using two‐way ANOVA to take the sex of the offspring into account. In the normoxic group, there were 27 male and 32 female pups and, in the hypoxic group, there were 30 male and 34 female pups (Table 1). Both oxygen concentration and sex of the offspring had significant effects on placental and fetal weights, as well as placental efficiency (P < 0.005 in all cases). There was a significant interaction between the two effects for placental weight (P < 0.05) but not for the other parameters. Placental weight was greatest in males under hypoxic conditions (Fig. (Fig.99 A). These effects could not be explained by differences in litter size because this was constant (P > 0.05) across the groups (Table 1). Placental efficiency, as expressed by the ratio between fetal and placental weight, did not show a significant difference between females and males under hypoxia, suggesting that the heavier placentas in males mitigated the growth restriction induced by hypoxia.
Additionally, we re‐examined the levels of total protein and, where appropriate, phosphorylation of various markers of oxidative and ER stress, the mitochondrial ETC, and kinases in the AMPK and Akt‐mTOR pathways. Most of the markers did not show a difference between the sexes, although HSP70 and HSP27 appeared to be consistently increased in females under hypoxia, indicating potential exposure to oxidative stress (Fig. (Fig.99 B). These results suggest that female placentas may be more susceptible to hypoxia‐induced stresses than male placentas, possibly rendering them less able to respond effectively and resulting in more severe fetal growth restriction. Larger studies enabling statistical analyses are required to confirm this point.
Normobaric housing of pregnant mice at 13% oxygen for their entire gestation induced ~12% reduction in fetal weight. This is greater than the 7% reduction reported for mice housed under 12% from E14.5 onwards (Cuffe et al. 2014). Interestingly, the detrimental effect on fetal growth induced was sex‐specific, with female fetuses being affected to a greater extent than male fetuses. The birth weight of human males is known to be heavier than that of females (Misra et al. 2009), and this effect is magnified by altitude. In a study of pregnancies of European ancestry, male birth weight was ~1.8% heavier than female birth weight at low altitude (430 m), whereas this difference widened to ~3.7% at high altitude (3600 m) (Soria et al. 2013). This finding suggests the involvement of sex differences in hypoxia‐induced growth restriction, which is consistent with our observations.
By contrast, placental weight was increased. This change was probably not induced by the small drop in food intake because we have previously reported a reduction in placental weight in undernourished mice (Coan et al. 2010). More probably, it is a result of the chronic hypoxia. Cuffe et al. (2014) reported no change in placental weight in mice after exposure during mid‐ to late‐gestation, although their data indicate a similar trend to a reduction in placental efficiency. In rats, placental weight was significantly increased after exposure to 13% oxygen from day 6 (Richter et al. 2012), whereas a greater placental cross‐sectional area was reported under 11% oxygen (Rosario et al. 2008). Further analysis indicated that sex‐specific changes in the placenta may account for the male–female differences in the fetal weight that we observed. The placentas of male fetuses were heavier, and demonstrated less evidence of oxidative stress than their female counterparts, and so may have been able to compensate for the reduced maternal P aO2. Nonetheless, placental efficiency did not show a significant difference between male and females. Maintaining growth under adverse conditions may be a risky approach by the male fetus that could result in higher embryonic lethality.
Oxygen crosses the placenta by simple diffusion, and the principal driving force is the concentration gradient between the two circulations. A maternal–fetal gradient of ~10 mmHg is considered to be necessary for placental exchange (Browne et al. 2015) and may have influenced the evolution of the placental mammals (Falkowski et al.
2005). Oxygen crosses cell membranes freely, and so placental exchange is considered to be flow‐limited rather than diffusion‐limited (Wilkening & Meschia, 1983). A high blood flow on either side of the placenta will refresh and deplete the supply and recipient pools, respectively, ensuring that the maximum gradient is maintained. This effect will become even more important in the high‐altitude situation where the gradient is inevitably lower as a result of reduced maternal P aO2. Indigenous populations demonstrate higher uteroplacental blood flows at high altitude than recent migrants, and this is assumed to explain their higher birth weights (Browne et al. 2015). Hence, the enlarged maternal arterial blood spaces supplying (and venous spaces draining) the placenta seen in our hypoxic animals may represent a compensatory mechanism. They are consistent with the concept that structural changes are more successful on the maternal rather than the fetal side of the high‐altitude placenta (Mayhew, 1991). Trophoblast invasion and remodelling of the uterine spiral arteries may mediate this effect because these are enhanced in rats maintained under hypoxia during pregnancy, leading to greater vessel diameters (Rosario et al. 2008). The critical period for exposure was found to be E8.5–9.5, and so this effect will not have operated in the study of Cuffe et al. (2014), which only involved exposure to hypoxia during mid‐ to late‐gestation. This could account for the fact that no increase in placental weight was observed in their study.
The changes in maternal blood volume could not fully account for the increase in placental weight. Activation of Akt‐mTOR signalling suggests that there may also have been enhanced placental growth under hypoxia because this pathway plays a crucial role in the regulation of placental size. Akt‐mTOR signalling is down‐regulated in placentas from growth restricted pregnancies (Yung et al. 2008) and elevated in pregnancies of obese women with large babies (Jansson et al. 2013). Therefore, the increased Akt‐mTOR signalling in the hypoxic placentas is consistent with our finding of a heavier placenta, although the difference in volume generated may have been too small to be detected stereologically because of the small sample size. This result was contradictory to our previous finding of down‐regulation of Akt‐mTOR signalling in human placentas from high altitude, although, in that study, there was a concomitant reduction in growth of the placental villous tree (Yung et al. 2012). Activation of Akt signalling by hypoxia has been reported in other cell types (Alvarez‐Tejado et al. 2001), although the mechanisms that are responsible remain unclear. Activation of de novo synthesis of RNAs and/or proteins is required because the global protein and RNA synthesis inhibitors, cyclohexamide and actinomycin D, both suppressed hypoxia‐induced Akt phosphorylation (Alvarez‐Tejado et al. 2001). Sex differences in the response of the human placenta to adverse stimuli are increasingly being recognized. Differences in gene expression associated with high‐level functions, such as protein synthesis, hormone secretion and growth, have been reported (Clifton, 2010; Osei‐Kumah et al. 2011; Buckberry et al. 2014). Further analyses are required to determine whether they might account for the differences observed in the present study.
In addition, our findings indicate that female, but not male, placentas might suffer from low‐grade oxidative stress under hypoxia because they express higher levels of HSP70 and HSP27. Unlike the stress kinase p38, which reacts rapidly in response to stress, HSPs are usually the second wave for long‐term protection against stress‐induced denaturation of proteins (Feder & Hofmann, 1999). The mechanisms underlying the greater susceptibility of a female placenta to hypoxia‐induced oxidative stress are unknown. However, we speculate that they may involve sex‐specific differences in placental cytokine profile. Cytokines such as tumour necrosis factor (TNF)α can induce the production of reactive oxygen species, resulting in oxidative stress. In pregnancies complicated by asthma, the female placenta expresses high levels of mRNAs encoding pro‐inflammatory cytokines, including TNFα, interleukin (IL)‐1β, IL‐6, IL‐5 and IL‐8, whereas there is no significant change in the male placenta compared to corresponding controls (Scott et al. 2009). Additionally, maternal concentrations of TNFα, IL‐6 and IL‐8 are increased significantly during the third trimester of pregnancy at high altitude (Coussons‐Read et al. 2002), indicating that hypoxia potentially promotes placental cytokine production. Furthermore, analysis of sexual dimorphism in zebrafish liver enzymes reported higher transcription of the gene for glutathione peroxidase 1 in male compared to female organs (Zheng et al. 2013). Therefore, the high pro‐oxidant production with low anti‐oxidant defence is consistent with the oxidative stress observed in female placentas.
Although there were differences in the oxidative stress detected in the male and female hypoxic placentas, the degree of ER stress was equivalent. Activation of the different UPR pathways is closely linked to the severity of ER stress (Yung et al. 2008). In the human placenta at high altitude, we only detected activation of PERK/eIF2α arm, whereas, in the hypoxic murine placentas, IRE1/XBP‐1 signalling was also increased, indicating a potentially higher degree of ER stress. However, hypoxia can induce different severities of ER stress in different placental cell types (Yung et al. 2012) and so these data should be interpreted with caution. ER stress and activation of the UPR are usually related to protein synthesis inhibition through phosphorylation of eIF2α, although this inhibition is selective. mRNAs containing small upstream open reading frames within their 5′‐untranslated regions or internal ribosome entry sites sequences are translated independent of eIF2α regulation (Lu et al. 2004). Translation initiation is also regulated by eukaryotic initiation factor 4 (eIF4) family members (Gingras et al. 1999), which in turn are regulated by 4EBP‐1. 4EBP‐1 binds to eIF4E, preventing its interaction with eIF4 family members during the formation of the translation initiation complex. Phosphorylation of 4EBP‐1, which is mediated by mTORC1, blocks this interaction (Mothe‐Satney et al. 2000). Therefore, the increased mTOR activity in hypoxic murine placentas indicates that eIF2α independent protein translation is probably ongoing, thereby increasing placental growth.
Consistent with an arrest of translation, we did observe a remarkable reduction in protein subunits of the mitochondrial ETC complexes. We have previously reported that increased phosphorylation of eIF2α induced by salubrinal, a specific phosphatase inhibitor for p‐eIF2α, suppresses ETC complex subunits (Colleoni et al. 2013). Additionally, we also demonstrated down‐regulation of the subunits in high‐altitude placentas (Colleoni et al. 2013). Reduction of ETC complex II activity probably reduced the number of electrons feeding into the ETC, thereby preventing electron leakage and production of reactive oxygen species. This effect could help to explain why no change in oxidative stress was observed in the male placentas, and only low‐grade stress in the female ones. Despite the reduction in ETC complexes, activity of the energy sensing kinase, AMPKα, did not change, indicating no crisis in cellular energy level. However, we did observe a reduction of maternal glucose concentration in the hypoxic animals, indicating that there may have been a general shift to anaerobic respiration as the major energy source.
Taken together, our results show that the murine placenta undergoes fine‐tuning mediated by growth signalling and stress‐response pathways to ensure optimal fetal survival and growth in response to chronic hypoxia. The female appears to follow a conservative approach, generating a smaller placenta and fetus to ensure survival of the offspring, whereas the male appears to be more aggressive with increased placental growth to extract more resources from the mother and maintain maximal fetal growth. The latter is a more risky strategy because there may be more fetal and neonatal deaths, although it is consistent with what is reported in humans (Eriksson et al. 2010). This sexual dimorphism in placental responses towards adverse conditions may be beneficial with respect to the continuity of the species by natural selection because small populations of males are sufficient to mate with larger numbers of females.
The authors declare that they have no competing interests.
GJB, DSC‐J and HWY designed the study. HM, JHWV and HWY performed the experiments. GJB, DSC‐J and HWY analysed the data and wrote the paper. All authors approved the final version.
This study was supported by a grant from the Wellcome Trust (084804/2/08/Z).