Reduced oxygen supply during the pre- and perinatal period often leads to acquired neonatal brain damage. So far, there are no reliable markers available to assess the hypoxic cerebral damage and the resulting prognosis during the immediate postnatal period. Thus we aimed to determine whether the hypoxia-inducible transcription factors (HIF-1 and HIF-2) and/or their target genes in the placenta represent reliable indicators of hypoxic distress of the developing brain during systemic hypoxia at the end of gestation. To this end, pregnant mice were exposed to systemic hypoxia (inspired O2 fraction: 6%, 6 h) at gestational day 20. This hypoxic exposure significantly increased HIF-1α and HIF-2α protein levels in brain and placental tissue. Compared with normoxic controls, an increase of HIF-1α-immunoreactive neurons and HIF-2α-positive glial cells and vascular endothelial cells was observed in hypoxic cerebral cortex and hippocampus. In placenta, HIF-1α and HIF-2α were expressed in labyrinthine layer with increased staining intensity during hypoxia compared with normoxia. Significant upregulation of VEGF mRNA and protein in brain and placenta, as well as erythropoietin protein in placenta, indicated activity of the HIF system upon fetal hypoxia. Notably, hypoxia did not affect expression of the HIF target genes inducible nitric oxide synthase and GLUT-1. Taken together, at gestational day 20, systemic hypoxia led to upregulation of HIF-α in mouse brain that was temporally paralleled in placenta, implying that α-subunits of both HIF-1 and HIF-2 are indeed early markers of hypoxic distress in vivo. If our data reflect the situation in humans, analysis of the placenta will allow early identification of the hypoxic brain distress occurring near birth.
- developing brain
- mouse placenta
- vascular endothelial growth factor
- inducible nitric oxide synthase
hypoxic and ischemic complications during the perinatal period are common etiological factors of acquired neonatal brain damage associated with severe long-term neurodevelopmental disabilities (48). Beyond the broad spectrum of risk factors of acute hypoxia/ischemia at birth arising from impaired materno-/feto-placental blood flow or acute anemia (placental bleeding or uterine rupture), growing evidence arises from clinical studies that especially antenatal hypoxic events are significant causes of perinatal hypoxic brain damage, including remittent antepartum fetal metabolic acidosis and asphyxia (28), as well as chronically compromised prenatal fetal oxygen and energy availability (40). Etiologically, structural and functional placental abnormalities, intrauterine growth retardation, or maternal diseases, such as maternal hypotension, preeclampsia, or diabetes mellitus (28, 40), predestinate to acute decompensation of fetal energy and oxygen supply close to and during delivery, including global failure of oxygen delivery to the developing brain.
In this critical clinical situation, reliable predictive indicators to identify high-risk neonates during the very early postnatal period, when neuroprotective options might be efficient (48), are limited (15, 41). Based on the experimental insights into the molecular pathways that define the immediate responses of the developing brain to hypoxia per se, we have, therefore, developed a rodent model of acute antepartum systemic hypoxia at the last day of mouse gestation (GD20) to analyze early indicators of fetal brain hypoxia, which might have future diagnostic implications. At this developmental stage, maturation of mouse placenta corresponds to that of mature human placenta (49), and development of mouse brain at GD20 approximately mirrors human brain developmental stage at midgestation (11). During this period, vasculogenesis, differentiation of oligodendrocytes, neuronal development, as well as axonal and dendritic sprouting are highly active processes (11, 25).
Among early adaptive responses to hypoxia, hypoxia-inducible transcription factors HIF-1 and HIF-2 have been characterized as the main regulators of oxygen-dependent gene transcription that modulate oxygen and metabolic supply upon hypoxic exposure. Knockout mouse experiments emphasized the crucial role of HIFs in fetal tissues for physiological vasculo- and embryogenesis, as well as central nervous system development (24, 42). HIF is a heterodimer composed of the HIF-α and HIF-β [the latter, also termed ARNT (aryl hydrocarbon receptor nuclear translocator)] subunits, both belonging to the Per-ARNT-Sim family of basic helix-loop-helix transcription factors (13). Under normoxic conditions, HIF-α subunit is rapidly degraded by the ubiquitin-proteasome pathway mediated by specific prolyl residues that are hydroxylated by HIF-prolyl hydroxylases (prolyl hydroxylation domain protein). Reduced activity of the prolyl hydroxylation domains under hypoxia initiates stabilization of the HIF-α subunit, activation of nuclear translocation, and heterodimerization with the β-subunit and binding to hypoxia response elements of enhancers and promoters of specific target genes. As a result, numerous HIF target genes modify oxygen and energy supply, e.g., by activation of glucose utilization (e.g., GLUT-1), vasoproliferative and vasoactive effects [e.g., vascular endothelial growth factor (VEGF); inducible NO synthase (iNOS)], and cell survival [e.g., erythropoietin (EPO)] (13). The most widely expressed HIF α-subunit is HIF-1α. HIF-2α (endothelial Per-ARNT-Sim protein-1) is assumed to differ from HIF-1α, e.g., in target gene specificity and tissue-specific functions (13, 50).
In human placenta, elevated levels of both HIFs were observed under conditions of chronic hypoxia, e.g., in preeclampsia (34), fetal growth restriction (9), and gestations at high altitude (51). Augmented placental expression of HIF target genes, such as VEGF (38), are related to vasoactive adaptation to chronically compromised utero-placental blood flow and oxygen delivery in intrauterine growth retardation and preeclampsia. Furthermore, HIFs are of particular interest in hypoxic developing brain injury, because of their cytoprotective properties and induction of essential vasoactive and metabolic adaptive mechanisms (1). Different experimental neonatal models, such as hypoxic preconditioning (1, 26), focal hypoxic/ischemic (45), and global nonischemic hypoxic brain injury (27), demonstrated upregulation of cerebral HIF-1 protein (26), VEGF (27), erythropoietin (14), and iNOS gene expression (45) involved in hypoxic stress-induced molecular responses of the developing brain. Interestingly, in vivo studies in humans suggested HIF-1-regulated genes as early indicators of perinatal brain complications in preterm (15) and term newborns (10). From previous human in vivo studies on placentas derived from acute birth asphyxia, we observed a significant upregulation of VEGF and adrenomedullin levels in relation to the degree of neonatal central nervous system complications (43, 44).
However, little data are available on the regulation of the HIF system during late-gestational systemic hypoxia. Thus present investigations of endogenous HIF regulation in fetal mice during acute systemic hypoxia represent an approach to understand the impact of systemic hypoxia on placental and cerebral HIF kinetics and target gene expression, and to evaluate the hypothesis that placental HIF responses to hypoxia correlate with those of developing brain, indicating cerebral hypoxic stress. To this end, we compared hypoxia-induced changes of HIFs and HIF-regulated target genes in developing mouse brain and placenta upon acute systemic hypoxia at the end of gestation (GD20). Furthermore, we analyzed the pattern of spatial and cellular expression of HIFs in mouse brain and placenta in response to severe acute systemic hypoxia.
Pregnant C57BL/6 wild-type mice were exposed to systemic hypoxia at the last day of mouse gestation (GD20). Dams (n = 5) were kept at continuous hypoxia with an inspired O2 fraction (FiO2) of 6% for 6 h (gas mixture: 6% O2, 94% N2; Hypoxic Workstation INVIVO2 1000, Biotrace International). To enable adjustment to the hypoxic environment, O2 deprivation was done gradually by decreasing the FiO2 from 21% to 6% in 2% O2 steps every 10 min. Controls (n = 5) were kept in the INVIVO2 chamber under room air. After the incubation period, neonatal brains (hypoxia group, n = 16; controls, n = 16) and corresponding placentas (hypoxia group, n = 16; controls, n = 16) were immediately dissected (without reoxygenation of the pups), frozen in liquid nitrogen, and stored at −80°C until protein (n = 3 per group) and mRNA extraction (n = 10 per group). For immunohistochemical studies, brains and placentas (n = 3 per group) were embedded in 4% paraformaldehyde. Coronal sections at the levels of the dorsal hippocampus were examined to analyze cerebral cortex and hippocampus known to show selective vulnerability to hypoxia at early development (25). Note that mouse brain development at GD20 approximately corresponds to that of the human brain at midgestation (11). Pregnant mice were provided with food and water ad libitum. Animal experiments were performed according to protocols approved by the Kantonales Veterinäramt Zurich.
RNA isolation and RT-PCR.
Total RNA was extracted using RNAzol-B isolation kit (WAK-Chemie Medical, Bad Homburg, Germany), according to the manufacturer's instructions. RT-PCR was performed as described previously (43). Commercial reagents (TaqMan PCR Reagent Kit, Perkin-Elmer) and conditions were applied according to the manufacturer's protocol. The PCR reaction was performed in an ABI 7500 real-time PCR thermocycler (Applied Biosystems). All reactions were performed in duplicate using β-actin and porphobilinogen deaminase (PBGD) as endogenous controls. The following primers and TaqMan probes based on published reports were used: β-actin, forward: 5′-ATGCTCCCCGGGCTGTAT-3′; reverse: 5′-TCACCCACATAGGAGTCCTTCTG-3′; TaqMan probe: 5′(FAM)-ATCACACCCTGGTGCCTAGGGCG-(TAMRA)-3′; HIF-1α, forward: 5′-AGACAGACAAAGCTCATCCAAGG-3′; reverse: 5′-GCGAAGCTATTGTCTTTGGGTTTAA-3′; TaqMan probe: 5′(FAM)-CTGCCACTTTGAATCAAAGAAATACTGTTCCTGAG-(TAMRA)-3′; iNOS, forward: 5′-CAGCTGGGCTGTACAAACCTT-3′; reverse: 5′-CATTGGAAGTGAAGCGTTTCG-3′; TaqMan probe: 5′(FAM)-CGGGCAGCCTGTGAGACCTTTGA-(TAMRA)-3′; PBGD, forward: 5′-ACAAGATTCTTGATACTGCACTCTCTAAG-3′; reverse: 5′-CCTTCAGGGAGTGAACAACCA-3′; TaqMan probe: 5′(FAM)-TCTAGCTCCTTGGTAAACAGGCTCTTCTCTCCA-(TAMRA)-3′; VEGF; forward: 5′-GCACTGGACCCTGGCTTTACT-3′; reverse: 5′-ACTTGATCACTTCATGGGACTTCTG-3′; TaqMan probe: 5′(FAM)-CCATGCCCAGTGGTCCCAGGCTG-(TAMRA)-3′.
Real-time PCR analysis for GLUT-1 and EPO has been performed using mouse-specific TaqMan-based gene expression assays (Applied Biosystems, Foster City, CA; EPO, catalog no. Mm00433126_m1, Glut-1, catalog no. Mm00441473_m1).
Western blot analysis.
For Western blot analysis, separate nuclear and cytosolic protein lysates were performed. Whole placentas and brains were homogenized in lysis buffer [10 mmol/l HEPES, 10 mmol/l KCl, 1.5 mmol/l MgCl2, 1 mmol/l EDTA, 1 mmol/l dithiothreitol, 1 mmol/l phenylmethylsulfonyl fluoride (PMSF), and 1 μg/ml protease inhibitors (PI) in distilled water]. Samples were centrifuged at 4°C for 15 min, the supernatant (cytosolic extracts) was separated and stored at −80°C. The resulting pellet (nucleus) was resuspended in buffer (25% glycerol, 20 mmol/l HEPES, 20 mmol/l KCl, 1.5 mmol/l MgCl2, 1 mmol/l EDTA, 1 mmol/l dithiothreitol, 1 mmol/l PMSF, and 1 μg/ml PI in distilled water), and then high-salt buffer (20 mmol/l HEPES, 1 mmol/l EDTA, 1.5 mmol/l MgCl2, 25% glycerol, 1.2 mol/l KCl, 1 mmol/l dithiothreitol, 1 mmol/l PMSF, and 1 μg/ml PI in distilled water) was added. Thereafter, samples were centrifuged for 10 min at 4°C, and the supernatants (nuclear extracts) were kept at −80°C. The protein content was analyzed by commercial protein assay (BCA protein assay, Pierce Biological, Rockford, IL). Protein (40 μg) was run on a 7.5% SDS-polyacrylamide gel, transferred to a nitrocellulose membrane, blocked in 5% nonfat milk-Tris-buffered saline for 90 min at room temperature, and incubated overnight at 4°C with the polyclonal rabbit anti-HIF-1α antibody (Novus Biologicals, Littleton, CO; 1:1,000) or the polyclonal rabbit anti-HIF-2α antibody (Novus Biologicals, Littleton, CO; 1:500). After washing, blots were incubated with a horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature, followed by chemiluminescent detection. Western blotting was quantified using MCID (Imaging Research) imaging system to measure the densities of protein signals on radiographic films. β-Actin was used as a loading control.
For immunohistochemical analysis, 3-μm-thick sections of paraformaldehyde-embedded tissues were used (3 brains and corresponding placentas per group). After heat-induced epitope retrieval, washing (Tris-buffered saline/Tween 20, 0.05%), and blocking with normal goat serum, sections were incubated with the monoclonal mouse anti-HIF-1α (α67, Novus Biologicals, Littleton, CO; 1:100) or the polyclonal rabbit anti-HIF-2α antibody (Novus Biologicals; 1:400) overnight at 4°C. Negative controls were performed by omitting the primary antibody. After washing, sections were incubated with a biotinylated secondary antibody for 60 min. Antibody detection was amplified by a commercial catalyzed signal amplification kit (VECTASTAIN Elite ABC Kit; Linaris, Wertheim, Germany), based on the streptavidin-biotin-peroxidase reaction, according to manufacturer's instructions. Reactions were visualized with diaminobenzidine (Sigma, St. Louis, MO). To define the identity of cerebral HIF-1α or HIF-2α immunopositive cells, brain sections were costained with antibodies against NeuN (neuronal nuclei), a routinely used neuronal marker, or GFAP (glial fibrillary acidic protein), which is routinely used to identify glial cells, especially astrocytes. Incubations of monoclonal mouse anti-NeuN (Chemicon International, Hampshire, UK; 1:200) and polyclonal rabbit anti-GFAP (Sigma, St. Louis, MO; 1:100) were performed at room temperature for 2 h, followed by Alexa Fluor 488-conjugated secondary antibody (Invitrogen, Eugene, OR). During all incubations, sections were stored in a humidified chamber.
Tissue VEGF and EPO concentrations were measured by specific ELISA (Quantikine, mouse VEFG Immunoassay, mouse Erythropoietin Immunoassay; R&D Systems Europe, Abingdon, UK), according to the manufacturer's recommendations. The mean detection limits for VEGF and EPO were 3.0 and 18.0 pg/ml, respectively. All samples were assayed in duplicate. Protein was determined by the BCA protein assay reagent (Pierce). Results are given as picograms per milligram of protein.
Data are expressed as means ± SE. Statistical significance was determined by one-way ANOVA and two-way ANOVA for repeated measurements (P value < 0.05).
To investigate the impact of severe systemic oxygen deprivation on hypoxia-induced HIF responses in the developing brain of mouse fetuses and corresponding placenta shortly before birth, we exposed pregnant mice on their last day of gestation (GD20) to systemic hypoxia of 6% O2 for 6 h. Examination of hematoxylin-eosin staining of the fetal brains and corresponding placentas showed no morphological differences between normoxic and hypoxic tissues (data not shown).
Stabilization of HIFs in the fetal mouse brain and placenta of pups derived from normoxic pregnant mice.
HIF-1α and HIF-2α proteins were detectable in all developing brains and placentas kept at normoxic conditions (Fig. 1, A and D). This observation suggests constitutive stabilization of HIF-1α protein during intrauterine brain development that reflects a physiological hypoxic environment.
Enhanced HIF-1α protein stabilization in developing mouse brain and placenta during severe, acute, systemic hypoxia.
Hypoxia did not change HIF-1α mRNA expression in developing brains (HIF-1α/PBGD mRNA: 1.073 ± 0.056, n = 10, normoxia vs. 1.297 ± 0.081, n = 10, hypoxia) and placentas (0.483 ± 0.070, n = 10, normoxia vs. 0.755 ± 0.109, n = 10, hypoxia). However, hypoxia strongly induced HIF-1α protein stabilization in both developing brain and placenta compared with normoxia (Fig. 1, A and B). This increase was significant, as shown by densitometric analysis (Fig. 1B). HIF-1α signals were mainly found in nuclear rather than in cytoplasmic protein fraction of brain (Fig. 1C) and placental tissues (data not shown).
This simultaneous upregulation of HIF-1α in developing brains and placentas during intrauterine hypoxia prompted us to investigate spatial and cellular localization of HIF-1α by immunohistochemical analysis. Normoxic brains at GD20 (Fig. 2, A–C) revealed HIF-1α-positive immunoreactivity in cerebral cortex throughout cortical layers II-VI and in hippocampus, whereas hypoxia significantly augmented the cytoplasmic and nuclear signal intensity of HIF-1α-immunoreactive cells in cerebral cortex and hippocampus (Fig. 2, D–F). Interestingly, whereas in normoxic cortex positive staining was stronger in the cytoplasm of cells located in the upper layers (Fig. 2B), in hypoxic cortex HIF-1α immunoreactive cells were detectable in the upper and lower layers that revealed strong nuclear staining (Fig. 2E). By double-immunolabeling, we demonstrated a colocalization of HIF-1α and NeuN (Fig. 2, G and H). Furthermore, positive signals were found in several glial cells and vascular endothelium of the hypoxic fetal brains (Fig. 2, I and J).
In parallel, tissue sections of normoxic placentas (Fig. 2, L and M) showed positive immunoreactivity for HIF-1α in syncytiotrophoblasts of the labyrinthine layer, spongiotrophoblasts, and vascular endothelium. In line with the Western blot data (Fig. 1), the intensity of HIF-1α staining increased in hypoxic placentas (Fig. 2, N and O). Whereas the pattern of immunoreactive cell types did not alter, hypoxic exposure enhanced nuclear signals.
Enhanced HIF-2α protein stabilization in developing mouse brain and placenta during severe, acute, systemic intrauterine hypoxia.
HIF-2α protein was detectable by Western blot analysis in normoxic brains and placentas (Fig. 1D). Hypoxia induced a strong elevation of HIF-2α protein levels in both brain and placenta (Fig. 1D). Spatial and cellular distribution of HIF-2α is presented in Fig. 3, which shows representative immunohistochemical stainings of cerebral and placental sections. In developing brains, HIF-2α-positive cells were weakly present in normoxic cerebral cortex and hippocampus (Fig. 3, A and B). Systemic hypoxia strongly augmented the number of immunoreactive cells in cerebral cortex and hippocampus (Fig. 3, C and D). Protein accumulation was present in glial cells, as shown by colabeling with GFAP (Fig. 3, E and F) and vascular endothelial cells.
Normoxic placentas (Fig. 3, G and H) showed HIF-2α-positive immunoreactivity in spongiotrophoblasts and vascular endothelium. In hypoxic placentas, HIF-2α-positive cells were found in similar distribution, however, with stronger intensity of nuclear staining than in normoxic tissues (Fig. 3, I and J).
Differential expression of HIF target genes in developing mouse brain and placenta in response to systemic hypoxia.
Exposure of pregnant mice at GD20 to hypoxia significantly increased VEGF mRNA expression in developing brains (Fig. 4A, P < 0.01) and, in parallel, in placentas (P < 0.05). At the protein level, VEGF upregulation was found in both tissues by ELISA (Fig. 4B). Notably, VEGF protein induction by hypoxia was more prominent in the brain (P < 0.01) than in placenta (not significant).
Unexpectedly, there were no significant effects of intrauterine hypoxia at GD20 on cerebral and placental EPO mRNA expression (Table 1). However, at the protein level, EPO concentrations were significantly increased in hypoxic placentas compared with normoxic tissues (Fig. 4C). Expression of GLUT-1 and iNOS mRNA was not regulated by short-term hypoxia, even though these genes represent classical transcriptionally regulated HIF targets (Table 1).
The present mouse model of severe acute systemic hypoxia in the absence of concomitant experimental ischemia at the last day of gestation (GD20) assessed concomitant hypoxia-induced changes of HIFs in developing brain and placenta. Our main goal was to establish that HIFs and their target genes are placental indicators of hypoxic distress in the developing brain during acute antepartum hypoxia. The degree of hypoxia used in our experiments (FiO2 6%, 6 h) is comparable to the severity of hypoxia in altitudes of 9,000–10,000 m above sea level (22) and is tolerated by adult mice (39). Of note, this exposure is more severe than that so far used in rodent models of hypoxic preconditioning (1, 2). Although there are well-established rodent models of postnatal hypoxic-ischemic brain damage (30, 47) mostly performed in the 7-day-old rat, there are, in fact, very few studies on true intrauterine hypoxia (20). As susceptibility to hypoxic injury changes with developmental age (25) and severity of hypoxia (3), comparison of these different studies is very limited.
The main findings of the present study were, first, that systemic hypoxia at the very end of pregnancy (GD20) induced a simultaneous upregulation of the major HIFs, HIF-1 and HIF-2, in developing mouse brain and placenta, implying that placental HIFs represent markers of fetal cerebral hypoxic distress. Second, acute systemic hypoxia differentially affected transcriptional activity of specific HIF target genes, indicating time- and cell-specific sensitivity to hypoxia in placenta and the newborn brain.
Enhanced stabilization of HIFs in brain and placenta during systemic hypoxia.
In developing brains at GD20, constitutive expression of HIF-1α during normoxic conditions was observed. This was anticipated at this early stage of brain maturation, reflecting physiological hypoxic environment (26). Systemic hypoxia strongly induced nuclear accumulation of HIF-1α in cerebral cortex and hippocampus. Thus the regional distribution of cerebral hypoxic stress responses described in our study is in line with the observations on effects of early postnatal hypoxia in rodents demonstrating selective vulnerability and apoptotic neuronal cell loss in rat cerebral cortex, hippocampus, and subventricular zone (18, 30). In terms of cellular distribution, increase of HIF-1α levels and HIF-1 DNA binding activity in response to hypoxia has been shown in vitro in neurons (5) and astrocytes (31), as well as in vivo in neonatal rodent brain (26). Our results provide convincing evidence that cerebral HIF-1 response to systemic hypoxia in vivo involves different cell types at this early stage of brain maturation. In agreement with Chavez et al. (5), accumulation appeared most prominently in hypoxic neurons and, to a lesser extent, in glial cells, as well as in vascular endothelial cells. In contrast to studies on postnatal day 6 rat brains exposed to hypoxic preconditioning (8% O2; 3 h) (2), we observed cerebral HIF-2α induction by systemic hypoxia in very immature brains, indicating differential sensitivity in relation to developmental stage and degree of hypoxia. In analogy to in vitro observations (5), cellular distribution of HIF-2α expression was different compared with HIF-1α. This indicates that HIFs reflect complementary systems to modulate hypoxic stress responses, e.g., at the level of target gene activation, as suggested from erythropoietin regulation (50).
Consistent with physiological hypoxia during early development (23, 24), HIFs are strongly expressed in first trimester human placentas and decrease with gestational age (33). Our data on HIF-1α and HIF-2α protein accumulation in normoxic mouse placenta demonstrate constitutive activity of the HIF system at the end of gestation that extends observations on very early developmental stages of mouse placenta (36). We found HIF-1α as well as HIF-2α protein strongly accumulated in the hypoxic placenta. Functional activity of HIFs is confirmed by concomitant increase of VEGF and EPO levels compared with normoxic controls. In agreement with human observations (33), protein analyses showed an overlapping cellular localization of HIFs in labyrinthine layer. In contrast, at early stages of mouse development (embryonic days 7.5–11.5) (36), systemic hypoxia (6–7% O2, 4 h) induced spatially different expression of HIFs. Beyond different experimental conditions, distinct cellular sensitivity to changes in oxygen tissue partial pressure during development (23, 33), as well as several growth factors and inflammatory cytokines (6, 52), could explain the different observations. The fact that placental HIFs are readily upregulated by hypoxia supports complementary functions of both transcription factors.
Our results imply that placental adaptation mechanisms at GD20 do not prevent the onset of fetal hypoxia. This is in contrast to observations at midgestation (36). We show for the first time that upregulation of HIFs shows a temporally parallel dynamic in developing brain and placenta during severe intrauterine hypoxia as a consequence of placental and fetal hypoxia. This is of special interest as considerable age- and organ-specific differences in temporal dynamics of HIF-1α stabilization and activation are known from adult and neonatal rodents exposed to hypoxia (2, 39).
Differential regulation of HIF target genes in brain and placenta during systemic hypoxia.
As HIF-regulated gene expression shows differences attributable to cell type and severity of hypoxia (2, 39), we examined the relationship between cerebral and placental HIF target gene expression following systemic hypoxia. In fact, levels of HIF target genes responded quite differently. There was an early upregulation in VEGF mRNA levels in both tissues, which was accompanied by unchanged EPO, GLUT-1, and iNOS mRNA expression. At the protein level, cerebral VEGF response was more prominent than placental, while that of EPO was more pronounced in the placenta than in the brain. Of note, our results indicate that severe short-term hypoxia upregulates placental and cerebral VEGF concentrations at the mRNA and protein level.
VEGF is well characterized as one of the critical factors in physiological vasculogenesis and neurogenesis and is constitutively expressed during early placental (4) and brain development (32) at normoxic conditions, as shown in this study. During systemic hypoxia, placental and cerebral VEGF expression significantly increased. In addition to its neuroprotective role in chronic postnatal hypoxia (32), our data point to VEGF induction as a readily available cerebral response to acute systemic hypoxia, indicating cerebral hypoxic stress. Of note, these data support our laboratory's previous observations of increased VEGF expression in human placentas of asphyxiated newborns developing severe hypoxic-ischemic encephalopathy (44). Extending observations on human placental VEGF response to acute hypoxia (44) and oxidative stress (7), our data suggest that acute systemic hypoxia has the potential to alter VEGF levels in both developmental brain and placenta in a temporally parallel kinetics. Differences found in induction of placental VEGF mRNA and protein levels might reflect posttranscriptional modification by inflammatory mechanisms (7), as well as inducible stress proteins such as heme oxygenase-1 (12). Additionally, in chronic placental insufficiency, hypoxia-induced overexpression of soluble Flt1 receptor is suggested to diminish biological activity of VEGF, as observed in human in vivo studies on placentas derived from intrauterine growth retardation gestations (29).
Systemic hypoxia did not change mRNA levels of iNOS in mouse placenta and brain at GD20. Although iNOS belongs to the group of classical HIF target genes, hypoxia-induced iNOS upregulation has been reported to be involved in excitotoxic and vascular responses to reoxygenation (17, 19, 45) rather than to acute hypoxia. Similarly, GLUT-1 mRNA was not affected by acute hypoxia, even though it has been shown that it is upregulated by hypoxia in cells of blood-brain barrier and astrocytes (46), as well as human placental villous explants exposed to 6 and 12 h of 1% oxygen (33). In a similar way, EPO mRNA levels were not altered by short-term hypoxia. Indeed, cerebral EPO, well known to promote neuronal survival during ischemia (14), is strongly induced during hypoxia/ischemia (5). However, in line with studies on postnatal hypoxia (8% oxygen, 3 h) (2), our data suggest that EPO and GLUT-1 are not among the genes early regulated at the mRNA level during cerebral hypoxia at early stage of brain maturation. In this context, the present results support the possibility that HIF target genes show a time-dependent and cell-specific differential regulation during systemic hypoxia, and that VEGF belongs to the group of genes involved in the early response to hypoxia in developing brain, as opposed to EPO, GLUT-1, and iNOS. This might be true for placental gene expression as well (35), except for EPO. In our study, early placental EPO protein increase indicates relevant placental tissue hypoxia and different organ-specific sensitivity of EPO to hypoxia. It seems important to consider these different dynamics and cell-specific sensitivity of HIF-regulated genes during short-term systemic hypoxia in terms of their potential use as molecular indicators of perinatal hypoxic distress (10, 15, 44).
Perspectives and significance.
Present results confirm that HIFs represent a readily available response to severe short-term hypoxia in mouse placenta and developing brain at GD20, in parallel. If our data reflect the situation in humans, analysis of the placenta might allow early identification of hypoxic brain distress occurring near birth. Of note, concerning clinical practicability, collection of placental specimens immediately after delivery and their appropriate storage after shock-refreezing are clinically and technically practicable without major preanalytic tissue degradation (43, 44).
This work was supported by grants of Fonds für Forschung und Lehre (R. Trollmann); Medical Faculty, University of Erlangen, Deutsche Forschungsgemeinschaft (R. Trollmann); and the Swiss National Science Foundation (M. Gassmann).
The authors are grateful to Jessica Braun and Bianca Saam for skilled technical assistance, to Jorge Soliz for support on the animal experiments, and to Kerstin Amann for discussion of the placental immunohistochemical results.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society