Severe uteroplacental insufficiency causes cerebral apoptosis in the fetus. Moderate uteroplacental insufficiency causes intrauterine growth retardation (IUGR) and increases the risk of postnatal neurological morbidity. In the rat, uteroplacental insufficiency and IUGR affect cerebral gene expression of Bcl-2 and predispose the newborn IUGR rat toward cerebral apoptosis when challenged with perinatal hypoxia. Expression of Bcl-2, as well as the proapoptotic protein Bax, is regulated by p53. p53 also induces MDM2 transcription, which functions to limit further p53-induced apoptosis. The predisposition of the IUGR fetus toward cerebral apoptosis suggests that the p53-MDM2 “functional” circuit may be perturbed in the newborn IUGR rat brain. We hypothesized that MDM2 cerebral expression does not increase in response to increased p53 expression or increased levels of phospho-p53 (Ser15), an activated form of p53. To prove this hypothesis, we induced IUGR through bilateral uterine ligation of the pregnant rat. Uteroplacental insufficiency significantly increased p53 mRNA, total p53 protein, and phospho-p53 (Ser15) protein levels in the brain at term. Increased expression of phospho-p53 (Ser15) and terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling-positive cells were localized to the CA1 region of the hippocampus, the subcortical and periventricular white matter, and the amygdala of the IUGR rat brain. In contrast, uteroplacental insufficiency decreased cerebral MDM2 mRNA and phospho-MDM2 (Ser166) protein levels in the IUGR rat pups. We conclude that the cerebral MDM2 response to increased p53 expression is not present in the newborn IUGR rat pup, and we speculate that this contributes to the predisposition of the IUGR fetus toward perinatal and long-term neurodevelopmental morbidities.
- white matter
- intrauterine growth retardation
uteroplacental insufficiency causes intrauterine growth retardation (IUGR) and increases the risk of perinatal and long-term neurological morbidities (35, 57, 79, 92, 94). These morbidities range from abnormal behavioral assessments to cerebral palsy and often occur without evidence of gross neurological damage. Most IUGR infants who demonstrate abnormal neurodevelopmental outcomes do not display overt neurological symptomatology in the perinatal period. Apoptosis is a potentially “silent” molecular mechanism through which a selective loss of cerebral cells could occur. Moreover, we previously demonstrated (54) that rat pups rendered IUGR through uteroplacental insufficiency are predisposed toward increased apoptosis after suffering a perinatal insult.
Apoptosis is an important and active mechanism of normal human and rodent fetal brain development (11, 67). A key regulator of apoptosis is p53, which acts as both a transcription factor and an active component of the apoptosis cascade (33, 39, 75). p53 plays a pivotal role in the cellular response to stress, in part based on posttranslational phosphorylation at multiple serine sites within its NH2- and COOH-terminal ends (65, 80). Posttranslational modifications such as phosphorylation of the COOH-terminal Ser390 (Ser392 in human, Ser389 in mouse) enhance p53 DNA binding, whereas phosphorylation of the NH2-terminal Ser15 (Ser15 in human, Ser18 in mouse) increases p53 stability and apoptotic activity (37, 41, 76).
Stability of the p53 gene product is normally regulated by MDM2, which blocks the interaction of p53 with transcriptional coactivators and promotes proteasome degradation of p53 (4, 32, 91). p53 binding to an internal promoter activates MDM2 transcription (7, 40). This negative-feedback loop, termed the “p53 functional circuit,” serves to tightly control p53 activity (39). Because of the predisposition toward apoptosis in both IUGR humans and rats, we hypothesized that uteroplacental insufficiency disrupts this circuit in the IUGR fetal rat brain.
To test this hypothesis, we induced IUGR with a rat model of uteroplacental insufficiency. Uteroplacental insufficiency is associated with the hypertensive disorders of pregnancy, which are the most common medical complications of pregnancy and lead to serious perinatal mortality and morbidity (95). The human IUGR fetus endures in utero hypoxia, acidosis, hypoglycemia, altered levels of growth factors, and hypoinsulinemia (20–22). Similarly in the rat, when induced via bilateral uterine artery ligation of the pregnant dam, IUGR induces an identical response in the late-gestation fetus (68, 70, 85). In this well-characterized and widely published rat model of asymmetric growth retardation, IUGR pups are 20–25% lighter than sham-operated control animals, and birth weights are normally distributed within and among litters (9, 45, 46, 48–50, 52, 69, 70, 78). Using this IUGR model, we quantified the effects of uteroplacental insufficiency on fetal cerebral p53 mRNA and protein levels, as well as phospho-p53 (Ser15) and phospho-p53 (Ser390) protein levels. MDM2 mRNA and protein levels were measured to determine the IUGR response to p53 gene expression and posttranslational modifications. Immunohistochemistry was then used to localize p53 expression.
All procedures were approved by the University of Utah Chancellor's Animal Research Committee and are in accordance with the American Physiological Society's guiding principles (3). These surgical methods were described previously (43, 46–55, 72, 78, 82). On day 19 of gestation, the pregnant rats were anesthetized with intraperitoneal xylazine (8 mg/kg) and ketamine (40 mg/kg), and both uterine arteries were ligated (IUGR group; n = 6 litters). Term gestation in the rat is 21.5 days. Sham surgery was performed on control animals that underwent identical anesthetic and surgical procedures except for the uterine artery ligation (control group; n = 6 litters). Day 0 pups were delivered by cesarean section (n = 6 litters each for control and IUGR) at term, 2.5 days after the bilateral uterine artery ligation. Brains were quickly removed and either processed for histological preparation or frozen.
Cloning of rat MDM2 cDNA.
Total RNA was isolated from Sprague-Dawley rat brain. cDNA was synthesized with random primers and the Superscript III first-strand synthesis system for RT-PCR (Invitrogen, Carlsbad, CA). The cDNA was then amplified with mouse MDM2-specific primers: forward 5′ atgtgcaataccaacatg and reverse 5′ actaaatttctgtagatcat. A 323-bp PCR product was cloned into pCR2.1, sequenced, and found to be 96% homologous to the mouse sequence.
Total RNA was extracted from day 0 brains with a RNeasy Mini Kit (Qiagen, Valencia, CA), treated with DNase3 (Ambion, Austin, TX), and quantified with ultraviolet absorbance. RNA integrity was confirmed by gel electrophoresis.
Brain mRNA levels of MDM2 and p53 were measured by real-time RT-PCR as previously described (58). Probe and primers were designed with Primer Express software (Applied Biosystems, Foster City, CA). Probes were labeled with N-(3-fluoranthyl)maleimide reporter dye and 6-carboxytetramethylrhodamine quencher dye. The sequences for MDM2 were forward 5′ tctgtgtctaccgagggtgct, reverse 5′ tttggtctaaccagagtctcttgttc, and probe 5′ caggcacctcacagattccagcgtc. The primers and probe for p53 were forward 5′ cccaccatgagcgttgct, reverse 5′ ccacccggataagatgttgg, and probe 5′ aggagcgacgaccaggccgtcaccatca. cDNA was synthesized from 2 μg of DNase-treated total RNA as described above. Reporter dye emission was detected by an automated sequence detector combined with ABI Prism 7700 Sequence Detection System software (Applied Biosystem). An algorithm normalizes the reporter signal (Rn) to a passive reference and multiplies the standard deviation of the background Rn in the first cycles by a default factor of 10 to determine threshold cycle (CT). CT has a linear relation with the logarithm of the initial template copy number (34). Rat GAPDH was used as an internal control to correct for differences in cDNA loading. Relative quantification of PCR products are then based on value differences between the target and GAPDH control with the comparative CT method (64). Cycle parameters were 50°C for 2 min, 95°C for 10 min, and then 40 cycles of 95°C for 15 s and 60°C for 1 min. Samples were run in triplicate.
Immunoblotting and antibodies.
Whole brains were homogenized in ice-cold lysis buffer (150 mM NaCl, 50 mM Tris pH 7.4, 1 mM EDTA, 0.25% Na-deoxycholate, 1% Igepal CA-630) with complete mini EDTA protein inhibitors (Roche, Mannheim, Germany) and 0.1 M PMSF and centrifuged at 10,000 g at 4°C for 15 min. The supernatants were collected and stored at −80°C until use. Protein concentrations were determined by the bicinchoninic acid method (Pierce, Rockford, IL). Proteins were separated by 10% SDS-PAGE ready gels (Bio-Red) and transferred to nitrocellulose membranes in standard transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol). After blocking the membranes with 5% milk in Tris-buffered saline (TBS) for 1 h, bound proteins were exposed to antibodies against p53, phospho-p53 (Ser15) (Cell Signaling Technology), phospho-p53 (Ser390) (ABR Affinity BioReagents), human MDM2 (SMP14; Santa Cruz Biotechnology), and phospho-MDM2 (Ser166) (Cell Signaling Technology) overnight at 4°C. After extensive washing in TBS with 0.1% Tween 20 (TBST), a 1/2,000 dilution of horse anti-mouse or goat anti-rabbit horseradish peroxidase secondary antibody (Cell Signaling Technology) was applied and incubated for 1 h at room temperature. After extensive washing in TBST, blots were detected with Western Lightning enhanced chemiluminescence (PerkinElmer Life Sciences, Boston, MA) with Biomax film (Amersham, Little Chalfont, UK) and quantified by densitometry or by quantification on a Kodak Image Station 2000R (Eastman Kodak/SIS, Rochester, NY).
Immunohistochemistry analysis for phospho-p53 Ser15 localization.
Coronal sections of formalin-fixed brains were deparaffinized and rehydrated in a graded series of ethanol and distilled H2O (dH2O) with a final wash in PBS. Immunohistochemistry was performed as previously described (2). The slides were then subjected to an antigen retrieval procedure in which the slides were put in a slide tray with high-pH target retrieval solution (DakoCytomation, Carpinteria, CA) in a microwave oven and heated at high power for 165 s and at low power for 8 min, after which they were cooled to room temperature and washed with tap water. Sections were then incubated in a 3% H2O2 solution for 30 min at room temperature (20–22°C) to quench endogenous peroxidase activity. After being rinsed with tap water, sections were washed in PBS for 10 min and incubated in a blocking buffer (2% normal goat serum, 2% bovine serum albumin, 0.8% Triton X-100, 0.2% nonfat dry milk in PBS) at room temperature for 1 h. Sections were treated with phospho-p53 (Ser15) rabbit polyclonal antibody at 1:100 dilution in blocking buffer overnight at 4°C in a humidified chamber. The next day, sections were washed in PBS containing 0.2% Tween 20 (PBST) three times and were exposed to biotinylated goat anti-rabbit immunoglobulins for 1 h. After exposure to a TSA Biotin System (PerkinElmer Life Sciences), slides were washed in PBST for 15 min, stained with diaminobenzidine (DAB; Sigma, St. Louis, MO), counterstained with hematoxylin (Sigma), dehydrated, and coverslipped with Cytoseal 60 (Stephens Scientific, Kalamazoo, MI).
Terminal deoxynucleotidyl transferase dUTP nick-end label assay.
Coronal sections of brain were deparaffinized and rehydrated in a graded series of ethanol. Slides were then rinsed in dH2O, washed in PBS, and incubated for 15 min at room temperature with 20 μg/ml of proteinase K (Sigma-Aldrich). Slides were rinsed in dH2O, washed twice in PBS, and then incubated in 2% H2O2, to quench endogenous peroxidase activity. After a wash in dH2O, sections were incubated with terminal deoxynucleotidyl transferase (TdT) and biotinylated dUTP (Trevigen, Gaithersburg, MD) for 60 min at 37°C in a humidified chamber. The reaction was stopped with a stop/wash buffer, followed by a wash in PBS and incubation in PBS with avidin DH and biotinylated horseradish peroxidase H (Vectastain Elite ABC kit, Vector Laboratories, Burlingame, CA) for 20 min. After two PBS washes, the chromagen DAB (0.07 mg/ml) (Sigma) and urea H2O2 (1.6 mg/ml) in dH2O were added for 8 min, turning labeled DNA fragments brown. Sections were washed in dH2O, counterstained with methyl green, dipped in dH2O, 85%-95%-100% ethanol, Hemo-De (Fisher Scientific, Pittsburgh, PA), and coverslipped. Positive control sections were made by exposing sections to a DNase I solution before the TdT step, and negative control sections were made by substituting dH2O for TdT. Both controls were included each time sections were stained.
All data presented are expressed as mean ± SE percentage of control. ANOVA (Fisher's protected least significant difference) and the Student's unpaired t-test determined statistical significance. We accepted P < 0.05 for statistical significance.
Cerebral p53 gene expression.
mRNA and protein levels of p53 were measured with real-time RT-PCR and Western blotting, respectively. GAPDH was used as an internal control. Uteroplacental insufficiency and subsequent IUGR increased cerebral p53 mRNA and protein levels to 152 ± 18% and 138 ± 13% of control values, respectively (P < 0.05) (Fig. 1).
Cerebral phospho-p53 (Ser15 and Ser390) levels.
Protein levels of phospho-p53 (Ser15 and Ser390) were quantified by Western blotting with GAPDH as an internal control. Uteroplacental insufficiency and subsequent IUGR significantly increased cerebral phosphor-p53 (Ser15) levels to 175 ± 6% of control values (P < 0.01; Fig. 1). In contrast, uteroplacental insufficiency did not significantly affect phosphor-p53 (Ser390) levels in IUGR brains (80 ± 16%; Fig. 1).
Phosphor-p53 (Ser15) immunohistochemistry and TdT-mediated dUTP nick-end labeling.
Immunochemistry of whole brain sections was used to localize expression of phospho-p53 (Ser15). Phospho-p53 (Ser15) expression appeared greater in the IUGR group compared with the control group (Fig. 2) in hippocampus, amygdala, and white matter, similar to the Western blot findings presented above. Similarly, DNA fragmentation associated with apoptosis and detected by the TdT-mediated dUTP nick-end labeling (TUNEL) assay was also localized to hippocampus, amygdala, and white matter, with the appearance of fewer TUNEL-positive cells in the controls compared with the IUGR group (Fig. 3).
MDM2 partial cDNA clone.
The rat MDM2 partial cDNA cloned here was 96% (E value 3e−147) homologous to the mouse and corresponded to base pairs 254–576 of the murine sequence (accession no. NM010786.2) (Fig. 4A). The rat MDM2 partial sequence cloned in this study also exactly matched the GenBank entry “Rattus norvegicus similar to mdm2 gene product” (accession no. XM235169.2). This sequence was then used to design a probe and primers for real-time PCR to measure MDM2 mRNA levels.
Cerebral MDM2 gene expression.
Uteroplacental insufficiency significantly decreased cerebral MDM2 mRNA to 64 ± 8% of control values (P < 0.05), as measured by real-time RT-PCR (Fig. 4B). IUGR similarly decreased phospho-MDM2 (Ser166) protein levels to 66 ± 4% of control values (P < 0.01), whereas total MDM2 protein levels were not significantly affected.
IUGR complicates up to 6% of all pregnancies (25). Both humans and rats rendered IUGR by uteroplacental insufficiency are vulnerable to perinatal hypoxia and neurodevelopmental morbidity (54, 86, 89, 90). In IUGR rats, increased cerebral apoptosis characterizes this vulnerability (54). Apoptosis contributes to the subsequent pathogenesis of cerebral ischemia in juvenile and adult animals, and the relative contribution of apoptosis toward cerebral cell death increases with the immaturity of the animals studied (17, 73, 77, 88, 96). This is particularly true for the CA1 region of the hippocampus (28). Adult animals respond to various stresses with the increased cerebral expression of p53 and a subsequent MDM2 response. For example, traumatic brain injury induces apoptosis in association with increased expression of p53 and MDM2 over the course of 1–2 days in adult male Sprague-Dawley rats (56). Similarly, adult male Sprague-Dawley rats respond to chronic administration of morphine by increasing MDM2 mRNA levels in the hippocampus, amygdala, and cortex; moreover, middle cerebral artery occlusion also induces MDM2 expression in spontaneously hypertensive male Sprague-Dawley rats (38, 83). The findings of the present study demonstrate that fetal IUGR rat brain fails to respond to the transient insult of uteroplacental insufficiency and increased p53 expression by increasing MDM2 gene expression. This dysfunctional p53-MDM2 negative-feedback loop suggests a molecular mechanism for the predisposition of immature animals toward cerebral apoptosis vs. older animals and the subsequent neurological morbidity.
The initial step in the p53-MDM2 negative-feedback loop or “functional circuit” is activation of MDM2 transcription by binding of the p53 protein to an internal MDM2 promoter (P2), located near the 3′ end of intron 1 (6, 40). RNA from the MDM2 P2 promoter is more efficiently translated than RNA from the P1 promoter (44). Posttranslation modification of the MDM2 protein by phosphorylation at Ser166 increases MDM2 function by increasing nuclear entry (63, 97). The nuclear MDM2 protein functions as a E3 ubiquitin ligase that binds to the NH2 terminus of p53, which inhibits p53 transcription factor activity and ubiquinates the COOH terminus of p53. Ubiquination of p53 subsequently leads to export from the nucleus and degradation of p53 (24, 31, 66). This process is mitigated by p53 phosphorylation on Ser15, which decreases binding affinity between p53 and MDM2 (74, 76, 93).
Phosphorylation of Ser15 on p53 appears to initiate stress-induced activation of p53 (19). IUGR-induced phosphorylation at Ser15, but not Ser390, supports this previous observation and suggests that phosphorylation of Ser15 on p53 is a specific response to the intrauterine milieu associated with uteroplacental insufficiency. The present study is among the first to document a specific posttranslational modification to p53 to a perinatal insult.
Phospho-p53 (Ser15) affects multiple cellular processes that are relevant to the developing brain. Phospho-p53 (Ser15) inhibits cell cycle progression by causing arrest at the G1 phase (15). p53, which functions as a transcription factor for many genes including p21CIP/WAF and the apoptosis-related proteins Bcl-2 and Bax, requires p53 Ser15 phosphorylation for activation (8, 13, 42, 71). Moreover, mutation of Ser15 to Ala renders p53 ineffectual at both blocking cell cycle progression and initiating apoptosis, and the expression of phosphor-53 (Ser15) correlates directly with the amount of apoptosis (23, 84).
p53 also perpetuates the apoptosis cascade by direct interaction with mitochondria and activation of Bax, a proapoptotic protein (14, 62). As a result, because perinatal hypoxia increases Bax expression in the IUGR rat brain, the presence of increased p53 levels in IUGR brain potentially lowers the cerebral apoptosis threshold and thereby leaves the newborn IUGR animal particularly vulnerable to a second insult independent of transcription (54).
Many of the initial insults that characterize this model of uteroplacental insufficiency and IUGR potentially increase p53 expression, including acidosis, hypoglycemia, decreased levels of key growth factors such as IGF-I, and moderate hypoxia (26, 36, 68, 81). The component of hypoxia, which is missing from purely nutritional models of IUGR, adds significant physiological relevance to our model of IUGR because it mimics the human condition of uteroplacental insufficiency (20, 22). Hypoxia-induced neuronal apoptosis is a p53-dependent process; furthermore, hypoxia mediates the phosphorylation of p53 at Ser15 and may contribute to failure of the IUGR MDM2 response (5, 30, 98). In colorectal carcinoma cell lines infected with human papillomavirus, hypoxia decreases expression of MDM2 (1).
Hypoxia is a common characteristic of other animal models of early brain injury. An important model that has provided a great deal of insight into the pathogenesis of hypoxic-ischemic injury involves ligation of one common carotid artery followed by systemic hypoxia (87). Similar to the present study, the “Vannucci” model localizes injury to the hippocampus and white matter, as well as the cerebral cortex, when performed in either the rat or the rabbit (16, 87). Furthermore, simply exposing newborn rats to 20 min of 100% nitrogen induces neuronal loss in the CA1 region of the hippocampus in association with increased expression of p53 and TUNEL-positive staining (29). It is interesting that adult rats from this latter study were repeatedly slower than controls in various behavioral tests, demonstrating a loss of functional skills in correlation with increased central nervous system apoptosis.
Because of the practical difficulties associated with studies involving prenatal insults, few studies exist that focus on the specific cerebral molecular responses during this time period. Fetal developmental biology is unique because of the interactive physiology and biochemistry among mother, placenta, and fetus; furthermore, the fetal brain is phenotypically different from the postnatal brain, which has been exposed to the hormones and environmental stresses associated with parturition and ex utero life. In the guinea pig, reduction of placental blood flow in the second half of gestation decreases hippocampal volume, reduces the total number of CA1 pyramidal neurons, and causes ventriculomegaly (18, 59, 60). In the rat, clamping of the uterine artery on day 17 of gestation decreases fetal hippocampal weight, hippocampal DNA, and hippocampal incorporation of thymidine into DNA on day 1 of postnatal life (12). Although both of these studies provided the groundwork for the present study, neither investigated the molecular basis for their findings.
Caution is necessary, of course, when attempting to apply data from animal models to human pathophysiology. The fetal rat is physiologically immature relative to the human, and the insult imposed on the fetal rat in our model of uteroplacental insufficiency is severe and specific. In contrast, the timing and impact of uteroplacental insufficiency experienced by humans range across a continuum. It is worth noting, however, that IUGR and white matter damage are often associated findings at necropsy and that in monochorionic twins, the sibling who suffers periventricular leukomalacia (white matter injury) is more likely to be IUGR and characterized by reversed end-diastolic umbilical artery flow (27). Furthermore, apoptosis often characterizes the histopathology associated with periventricular leukomalacia in premature infants (10). Interestingly, Maneru et al. (61) found that teenagers with a history of perinatal asphyxia suffer from decreased hippocampal volume on voxel-based morphometry, as well as reduced long-term verbal memory.
In summary, this study finds that the rat fetus responds to the insult of uteroplacental insufficiency by upregulating p53 gene expression and phospho-p53 (Ser15) without a corresponding increase in MDM2 expression. These are novel findings that emphasize that immature animals in utero respond differently from more mature ex utero animals. Furthermore, the IUGR-induced changes in p53 expression localize in the CA1 region of the hippocampus and white matter, similar to what is found in other animal models and humans suffering perinatal insults.
This research was supported by National Institute of Child Health and Human Development Grants R01-HD-41075 (R. H. Lane) and National Heart, Lung, and Blood Institute Grants R01-HL-62875 (K. H. Albertine) and by the March of Dimes (R. H. Lane).
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 © 2005 the American Physiological Society