Uteroplacental insufficiency (UPI) leads to intrauterine growth restriction (IUGR), which predisposes infants toward renal insufficiency early in life and increases the risk of kidney-related adult morbidities, such as hypertension. This compromised in utero environment has been demonstrated to impair nephrogenesis, as evidenced by a reduced nephron endowment in humans and in rats rendered IUGR by UPI. Concordantly, we have observed that IUGR rats have increased kidney p53 protein levels associated with increased apoptosis. Several factors can regulate p53 gene expression and activity, including posttranslational modifications and protein-protein interactions in the cell. Among these, two important mechanisms are 1) phosphorylation of the amino terminal serine 15 [phospho-p53 (Ser15)], which increases p53 stability and apoptotic activity, and 2) the murine double-minute (MDM2) functional circuit that limits further p53-induced apoptosis by promoting proteosomal degradation of p53. We hypothesize that UPI induces an increase in phospho-p53 (Ser15) in association with an absent MDM2 response, predisposing the kidney to increased apoptosis. To test our hypothesis, we induced IUGR through bilateral uterine artery ligation of the pregnant rat. UPI significantly increased phospho-p53 (Ser15), as well as ataxia teleangiectasia-mutated kinase/A-T-related kinase and dsDNA-activated protein kinase kinase levels, which induce phosphorylation of p53. In contrast, UPI induced no increase in kidney MDM2 mRNA and protein levels in IUGR pups. We conclude that among multiple mechanisms that affect nephrogenesis, UPI induces an increase in p53 phosphorylation without a corresponding increase in MDM2 expression, and we speculate that this response may contribute to the increased apoptosis previously described in the IUGR kidney.
- ataxia teleangiectasia-mutated kinase
- A-T-related kinase
- dsDNA-activated protein kinase kinase
intrauterine growth restriction (IUGR) caused by uteroplacental insufficiency (UPI) is a morbidity associated with several common complications of pregnancy, such as preeclampsia and maternal diabetes (57). Epidemiological studies have demonstrated that IUGR neonates experience increased morbidity and mortality rates, including impaired renal function, and are at increased risk to develop adult morbidities, such as hypertension (12, 27, 51, 53). In chronic fetal hypoxia, such as seen in UPI, peripheral blood flow and blood flow to the kidneys is reduced to maintain brain, heart, and adrenal perfusion (39). This is known as the brain-sparing effect and leads to a condition termed “asymmetrical growth restriction.”
Interestingly, both human and animal studies have shown that IUGR results in smaller kidneys with decreased nephron number (2, 8, 49, 51). UPI affects multiple components of the fetal milieu, so a single mechanism responsible for abnormal kidney development is unlikely. For instance, nephrogenesis involves rapid remodeling of structures, which requires massive apoptosis, making this is an important and active mechanism of normal human and rodent fetal kidney development (24). Our laboratory has previously demonstrated that rats rendered IUGR by bilateral uterine artery ligation also suffer reduced nephron number, in association with increased apoptosis (41). A key regulator of apoptosis is p53, which acts as both an active component of the apoptosis cascade and a transcription factor (19, 46). In response to stress, hypoxia, and damaged DNA, p53 accumulates in the cell nucleus and is activated as a transcription factor (4). Activation of p53 initiates or inhibits the expression of numerous genes that mediate cell cycle arrest or induce apoptosis (5). Many of the initial insults that characterize our model of UPI and IUGR, including moderate hypoxia, acidosis, hypoglycemia, and decreased levels of key growth factors, such as IGF-I, can potentially increase p53 expression in different organs, including the kidney (9, 18, 44, 52). Indeed, in our IUGR rat animal model, kidney p53 protein levels are increased at birth. Furthermore, the expression of apoptosis-related molecules Bcl-2 and Bax are also affected in the kidney of these IUGR pups. Bcl-2 and Bax contribute to the signaling pathway that activates caspase-3, which is necessary for the chromatin condensation and DNA cleavage that characterize apoptosis. Through its function as a transcription factor, p53 negatively regulates Bcl-2 expression (antiapoptotic protein) and positively regulates Bax (proapoptotic protein) expression (48, 59).
However, the mechanisms underlying the change in p53 protein levels in the IUGR kidney are not fully understood. In the present study, we focus on the signaling pathways and molecular mechanisms underlying the increase in p53 levels and apoptosis observed in the IUGR rat kidney. Several factors can regulate p53 gene expression and activity, including posttranslational modifications and protein-protein interactions in the cell. Among these factors, two important mechanisms regulate p53 protein levels and activity. The first one includes posttranslational modification, such as phosphorylation of the amino terminal serine 15 [phospho-p53 (Ser15)], which increases p53 stability and apoptotic activity (4). Several kinases have been identified that detect cellular stress and initiate signaling pathways through phosphorylation of p53 at Ser15. These include the ataxia teleangiectasia mutated (ATM) kinase, A-T-related (ATR) kinase, and dsDNA-activated protein kinase (DNA-PK) (4, 40).
The second mechanism that controls p53 levels and activity involves murine double-minute (MDM2), which functions to limit further p53-induced apoptosis (47). MDM2, a p53-specific E3 ubiquitin ligase, is the principal cellular antagonist of p53, blocking the interaction of p53 with transcriptional coactivators and promoting p53 proteosomal degradation (34, 54). p53 and MDM2 form one feedback loop termed “the p53 functional circuit” in which p53 positively regulates MDM2 by activating the gene mdm2 transcription, and MDM2 negatively regulates p53 by promoting p53 ubiquitination and degradation (21). Posttranslational modification of the MDM2 protein by phosphorylation at serine 166 [phospho-MDM2 (Ser166)] increases MDM2 function by increasing nuclear entry (60). MDM2 enters the nucleus to form the “p53-MDM2 complex” that then shuttles to the cytoplasm where p53 degradation takes place. The goal of this autoregulatory negative feedback loop is to maintain low levels of p53 in the absence of stress and to limit the severity of the p53 response to cellular stress.
On the basis of the fact that UPI predisposes to increased kidney apoptosis both in humans and animals, we hypothesized that IUGR induces an increase in p53 phosphorylation at Ser15 through elevated ATM, ATR, and DNA-PK kinases. Furthermore, we hypothesized that the p53-MDM2 functional circuit fails to respond to the UPI insult in the immature kidney, leading to an absent MDM2 response, and therefore, increased apoptosis.
To prove our hypothesis, bilateral uterine artery ligation (i.e., IUGR) and sham surgery (sham) were performed on day 19 of gestation in Sprague-Dawley rats (term = 21.5 days). Similar to the human, UPI results in offspring with low birth weight and asymmetrical IUGR (50). In this well-established model of asymmetrical growth restriction, IUGR pups are 20–25% lighter than the sham-operated control animals at birth [IUGR (4.00 ± 0.25) vs. sham (5.25 ± 0.22), P < 0.05], with a normal distribution of birth weights within and among litters, and no difference in litter size between control and IUGR groups (5, 25, 37).
Using this IUGR model, we quantified the effects of UPI on fetal kidney phospho-p53 (Ser15), ATM, ATR, and DNA-PK protein levels and determined localization by immunohistochemistry. To establish the MDM2 response to increased p53 expression and apoptosis in the IUGR kidney, we measured MDM2 mRNA and total protein levels, as well as [phospho-MDM2 (Ser166)] levels.
MATERIALS AND METHODS
All procedures were approved by the University of Utah Animal Care Committee and are in accordance with the American Physiological Society's guiding principles (3). These surgical methods have been previously described (41, 50). On day 19 of gestation, the maternal rats (Sprague-Dawley) were anesthetized with intraperitoneal xylazine (8 mg/kg) and ketamine (40 mg/kg), and both inferior uterine arteries were ligated (IUGR; n = 6 litters). Sham surgery was performed in control animals who underwent identical anesthetic and surgical procedures except for the uterine artery ligation (sham) (n = 6 litters). Rats recovered within a few hours and had ad libitum access to food and water. At term (21.5 days gestation), pups were delivered by caesarian section, weighed, and decapitated. Kidneys were quickly harvested and frozen in liquid nitrogen or placed in 10% formalin. Day 0 pups were genotyped using PCR for the spermatogenic gene (Sby) from the Y chromosome to ensure an equal distribution of each sex for each methodology (forward primer: 5′-ACTGTTCAAGCAGTCAGCCG; reverse primer: 5′-CTCCATGAACTTGGGGTC) (33). We randomly selected one pup from each litter for all studies without knowledge of size or position within the uterine horn (n = 6 IUGR pups and n = 6 sham pups).
DNase I treated total RNA (Ambion, Austin TX) was extracted from 30–100 mg of day 0 IUGR and sham rat pup kidneys using the NucleoSpin RNA and Virus Purification Kit (BD Biosciences Clontech Palo Alto, CA). Total RNA was quantified using the PharmaSpec-1700 UV absorbance spectrophotometer (Shimadzu Scientific Instruments, Torrance, CA). RNA integrity was confirmed by gel electrophoresis.
Kidney mRNA levels of MDM2 were measured at day 0 with the real-time RT-PCR method as previously described (28). cDNA was synthesized using random hexamers and SUPERSCRIPT III RT (Life Technologies, Gaithersburg, MD) from 2.0 μg of DNase treated total RNA as described above. Primers and probes for MDM2 and GAPDH were designed using Primer Express Software (Applied BioSystems, Foster CA). The sequences for MDM2 were forward: 5′-tctgtgtctaccgagggtgct; reverse: 5′-tttggtctaaccagagtctcttgttc; probe: 5′-caggcacctcacagattccagcgtc. Target probes were labeled with fluorescent reporter dye carboxyfluorescein. Reporter dye emission is detected by an automated sequence detector combined with ABI Prism 7700 Sequence Detection System software (Applied BioSystems). An algorithm normalizes the reporter signal (Rn) to a passive reference and multiples 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. Rat GAPDH was used as an internal control to correct for differences in cDNA loading. Before use of GAPDH as a control, parallel serial dilution between control and IUGR cDNA are quantified to assess the validity of GAPDH as an internal control (41). Relative quantification of PCR products are then based on value differences between the target and GAPDH control using the comparative CT method (30). Cycle parameters were 50°C × 2 min, 95°C × 10 min, and then 40 cycles of 95°C × 15 s and 60°C × 60 s. Each sample was run in quadruplicate.
Immunoblotting and antibodies.
Whole kidneys were obtained from day 0 IUGR and sham rat pups. Total protein was isolated by homogenizing 30–100 mg of tissue in RIPA buffer (150 mM NaCl, 50 mM Tris pH 7.4, 1 mM EDTA, 0.25% Na-deoxycholate, 1% Igepal CA-630) with EDTA protease inhibitor (400 μl; Roche, Mannheim, Germany), centrifuged at 10,000 g for 15 min at 4°C. The supernatants were collected and stored at −80°C until use. Enriched nuclear protein was isolated for DNA-PK quantification by first pulverizing the tissue (50–100 mg) under liquid nitrogen, followed by homogenization using a Dounce homogenizer. Cytosolic supernatant was separated by centrifugation. The nuclei were resuspended in buffer and incubated at 4°C. Postincubation, the supernatant containing enriched nuclear protein was separated by centrifugation. Nuclear protein and total protein concentration was determined using the bicinchoninic acid protein assay kit method (Pierce, Rockford, IL). Total protein (20–50 μg), nuclear enriched protein (20–50 μg), Hepg2 or Hela cell lines (positive controls), and molecular weight markers were loaded and separated by XT Criterion gels (Bio-Rad Laboratories, Hercules, CA) at 200 V for 60 min. After gel electrophoresis, the proteins were transferred to PVDF membranes (Millipore, Billerica, MA) at 4°C for 1 h at 100 V or 4°C overnight at 35 V. Posttransfer, the membranes were blocked in either 5% milk TBS-T or 5% BSA TBS-T for 1 h and then washed multiple times in 1× TBS-T. After blocking, bound proteins were exposed to antibodies against phospho-p53 (Ser15) (Cell Signaling Technology, Beverly, MA), 1:1,000 in 5% BSA; MDM2 (Santa Cruz Biotechnology, Santa Cruz, CA) 1:200 in 5% milk; phospho-MDM2 (Ser166) (Cell Signaling Technology) 1:500 in 5% BSA; ATM (Santa Cruz Biotechnology) 1:500 in 5% milk; ATR (Santa Cruz Biotechnology) 1:100 in 5% milk; DNA-PK (Abcam, Cambridge, MA) 1:100 in 5% milk; and GAPDH (Abcam) 1:2,000 in 5% milk. Blots were incubated overnight at 4°C or at room temperature for 1 h. After multiple wash steps in TBS-T, membranes were probed with appropriate secondary antibody: horseradish peroxidase-conjugated anti-rabbit IgG antibody (Cell Signaling Technology), anti-mouse (Cell Signaling Technology), or bovine anti-goat (Santa Cruz Biotechnology) for 1 h at room temperature. After multiple wash steps in TBS-T, antibody signals were detected with Western Lighting enhanced chemiluminescence (PerkinElmer Life Sciences, Boston, MA) and quantified using a Image Station 2000R (Eastman Kodak/SIS, Rochester, NY). GAPDH signal was used to normalize samples.
Day 0 IUGR and sham rat kidney sections were deparaffinized and rehydrated in a graded series of ethanol and water. The slides were then incubated in a preheated, high pH, antigen retrieval buffer (DakoCytomation, Carpinteria, CA), in a Coplin jar in a 95°C water bath and then washed with water. Sections were incubated in a 3% H2O2 solution for 15 min at room temperature (20–22°C) to quench endogenous peroxidase activity. After being rinsed with tap water, sections were washed in 0.1% PBS (Sigma, St. Louis, MO) for 5 min. Next, the slides sections were blocked for endogenous advidin and biotin by incubating with advidin and biotin solutions (Vector Laboratories, Burlingame, CA) for 15 min at room temperature and then washed in PBS three times. Sections were then blocked using a blocking buffer (0.5% casein in PBS; Sigma) at room temperature for 30–60 min. Sham and IUGR sections were then treated with phospho-p53 (Ser15) rabbit polyclonal antibody (Abcam) 1:100 in PBS, ATM rabbit polyclonal antibody (Abcam) 1:50 in PBS, ATR (N-19) goat polyclonal antibody (Santa Cruz Biotechnology) 1:200 in PBS, and DNA-PK mouse monoclonal antibody (Abcam) 1:50 in PBS. Antibody specificity was tested by treating negative control sections with appropriate secondary antibodies (Vector Laboratories) 1:200 in PBS, normal serum (rabbit Ig2/FITC DakoCytomation) 1:200 in PBS, and advidin and biotin blocking buffer (Vector Laboratories) straight. Positive control sections were treated with anti-pan cytokeratin (Sigma-Aldrich, St. Louis, MO) 1:1:000 in PBS. Each slide was incubated with the primary antibody overnight at 4°C in a humidified chamber. The next day, sections were washed in PBS and were exposed to the appropriate biotinylated secondary antibody in PBS containing 10% Tween-20 for 1 h at room temperature (Vector Laboratories) and washed in PBS. Following exposure to a tyramide signal amplification Biotin System (PerkinElmer), slides were washed in PBS for 15 min, stained with diaminobenzidine (Sigma), counterstained with hematoxylin (Sigma), dehydrated in a graded series of ethanol, and coverslipped with Cytoseal 60 (Stephens Scientific, Kalamazoo, MI).
All data presented are expressed as means ± SE. Western blotting data were analyzed using the unpaired Student's t-test. Real- time RT-PCR results were analyzed using one-way ANOVA, and when this was significant, the Fisher least significant difference post hoc test was computed. A two-tailed probability value of less than 0.05 was considered statistically significant.
UPI induced no increase in kidney MDM2 mRNA and protein levels in IUGR pups.
Day 0 mRNA levels of MDM2 were measured using real-time RT-PCR. Day 0 protein levels of total MDM2 and phospho-MDM2 (Ser166) were quantified using Western blot analysis. GAPDH was used as an internal control. Results are reported as IUGR mean percent of control ±SE (n = 6 litters). UPI did not affect MDM2 mRNA expression and protein levels in IUGR pups: MDM2 mRNA was 91 ± 4% of sham values (P = 0.16), total MDM2 protein was 100 ± 7% of sham values (P = 0.7), and phospho-MDM2 (Ser166) protein was 82 ± 11% of sham values (P = 0.18) (Fig. 1).
UPI increases ATM, ATR, and DNA-PK protein levels, as well as phosphorylation of p53 at Ser15.
Day 0 protein levels of ATM, ATR, DNA-PK kinases, and of phospho-p53 (Ser15) were quantified using Western blot analysis. GAPDH was used as an internal control. Results are reported as IUGR mean percent of control ±SE. In other words, if IUGR was the same as sham, it would be 100%. UPI significantly increased IUGR kidney protein levels of ATM, ATR, and DNA-PK kinases to 224 ± 38% (P < 0.01), 160 ± 20% (P < 0.05), and 168 ± 10% (P < 0.01) of sham values, respectively (n = 6 litters). Concomitantly, UPI significantly increased levels of kidney phospho-p53 (Ser15) in IUGR pups (121 ± 7% of sham, P < 0.05) (Fig. 2).
Immunohistochemistry for localization of phospho-p53 Ser15, ATM, ATR, DNA-PK kinases.
Localization of phospho-p53 (Ser15), ATM, ATR, DNA-PK kinases was mainly associated with cortical distribution. Following UPI, phospho-p53 (Ser15), ATM, ATR, and DNA-PK expression appeared greater in the IUGR kidney compared with the control, similar to the Western blot analysis findings reported above. Phospho p53 (Ser15) immunolabeling was associated exclusively with cell nuclei observed both in the interstitial mesenchymal cells as well as tubules, whereas all three kinases had positive immunoreactivity associated in both nuclei and cytoplasm of proximal and distal tubules of the cortex (Fig. 3).
UPI affects 3–10% of pregnancies in the Western Society and is associated with increased risk to develop short-term morbidities, such as renal insufficiency, and long-term morbidities including cardiovascular disease and hypertension (6, 7, 16). UPI leads to an abnormal intrauterine environment exposing both the human and rat IUGR fetus to stressors, such as acidosis and hypoxia (13, 37, 38). This compromised in utero environment has been demonstrated to impair nephrogenesis, as evidenced by a reduced nephron endowment in humans and experimental IUGR animals (31, 32, 49).
Other animal models of IUGR in the rat, such as UPI, induced by placement of a silver clip around the aorta below the renal arteries (2), by partial ligation of one uterine horn (31), and by maternal undernutrition during late gestation (58), have also shown a significant decrease in nephron number and the development of hypertension in the offspring. Similarly, antenatal administration of dexamethasone in rats and ewes has been associated with a reduction in the number of nephrons and in glomerular filtration rate in the offspring, as well as adult onset hypertension (10, 14, 56). However, although these studies confirm a clear link between IUGR prenatal overexposure to glucocorticoids and reduced nephron number, none have assessed the molecular mechanisms underlying abnormal renal development.
Nephrogenesis depends heavily on apoptosis and therefore can be affected by altered apoptotic homeostasis (24, 29). For example, previous studies have shown that chronic or moderate insults upregulate renal p53 mRNA and induce apoptosis (11, 36). Similarly, our group has previously described that rats rendered IUGR by bilateral uterine ligation have a predisposition toward increased kidney apoptosis leading to a 25% decrease in nephron number (41). In this study, UPI induced renal p53 genomic DNA hypomethylation in association with increased p53 and Bax mRNA levels, as well as decreased Bcl-2 mRNA levels. These novel findings suggest a molecular mechanism at a genomic level through which IUGR induces fetal renal apoptosis and a resultant permanent loss in glomeruli.
In the present study, we further investigate the molecular mechanisms and signaling pathways underlying the increase in p53 protein levels and apoptosis observed in the IUGR rat kidney. Several factors can regulate p53 protein levels and activity, including modifications and protein-protein interactions in the cell. In the present study, UPI significantly increased kidney phospho-p53 (Ser15), as well as ATM, ATR, and DNA-PK kinase levels. By immunohistochemistry analysis, we established that kidney phospho-p53 (Ser15), and the kinases expression appeared greater in the outermost part of the cortex, which corresponds to the nephrogenic zone. This region of the kidney is likely more vulnerable during compromised renal development. Importantly, the increase in phospho-p53 (Ser15) in the IUGR rat kidney seems to affect both the mesenchymal cells and the tubular structures in the nephron, indicating that the increase in apoptosis is most likely a global phenomenon, affecting both already-formed nephrons, as well as potential mesenchymal cells.
The stability of p53 is regulated by phosphorylation at multiple sites that promote the interaction between p53 and transcriptional coactivators, such as p300 (43). For example, in response to UV, ionizing irradiation, and cisplatin, p53 has been shown to be phosphorylated on Ser15 by either ATM or ATR kinases, as part of the stabilization process of p53 required to induce G1/S cell-cycle arrest and apoptosis (4, 40). Interestingly, the induction of apoptosis on oncogenic stress is downregulated secondary to inhibition of ATM/ATR kinases, indicating their important role in the stabilization of p53 (40). DNA-PK has also been shown to be activated in response to DNA double-stranded breaks and to phosphorylate p53 on Ser15 (26). However, its role in vivo does not seem to be as clear as the role of ATM/ATR for p53 activation (20, 55).
Our present work is the first to describe an increased phosphorylation of kidney p53 on Ser15 through elevated kinase activity of ATM/ATR and DNA-PK in response to UPI at the time of birth. However, the role and relative importance of each kinase in our animal model requires further investigation.
In contrast to increased p53 phosphorylation, UPI induced no increase in kidney MDM2 mRNA, and total and phospho-MDM2 protein levels in IUGR pups on day 0 of life. It is possible that kidney MDM2 mRNA and protein levels in the IUGR rat could be affected if we separated the cortex and medulla for the analysis. However, the day 0 rat kidney is extremely small to technically isolate cDNA and protein from different regions. Under normal circumstances, the p53 protein is kept at low levels and is short lived because of continuous degradation mediated largely by MDM2. In unstressed cells, both p53 and MDM2 shuttle between the nucleus and the cytoplasm. In the nucleus they form a complex; the RING finger domain of MDM2 ubiquinates p53, causing p53 to exit the nucleus. Once in the cytoplasm, p53 degradation mediated by MDM2 is completed (35). Posttranslational modification of the MDM2 protein by phosphorylation at Ser166 increases MDM2 stability and consequent degradation of p53 (15). However, and as part of the cellular response to stress, phosphorylation of p53 disrupts the interaction between the two proteins, interfering with both MDM2-dependent p53 nuclear export and ubiquitination of p53 (21). This complex functional circuit is essential for regulating p53 activity, but different aspects of it can be defective as observed with stress-induced modifications to MDM2. For example, our group has recently reported that the fetal IUGR rat brain fails to increase MDM2 expression in response to the transient insult of UPI and increased p53 expression (22). Similarly, hypoxia downregulates the expression of MDM2 in human papilloma virus E6-infected cells resulting in nuclear accumulation of p53 and leading to cell cycle arrest or apoptosis (1). Interestingly, in the present study, MDM2 mRNA expression was not upregulated in the presence of increased p53 nor was there an increased phosphorylation of MDM2. This dysfunctional p53-MDM2 negative feedback loop suggests that the IUGR kidney in the newborn is more vulnerable to the UPI insult due to an immature response, and represents a molecular mechanism for the predisposition toward kidney apoptosis and the subsequent morbidity.
It is necessary to be cautious when applying data from a rat model to human pathophysiology. First, we need to consider the timing and extent of UPI experienced by humans that range across a continuum, in contrast to the insult imposed on the fetal rat in this model of UPI, which is severe and specific and occurs relatively late in gestation. Second, we need to consider the timing of nephrogenesis. In the human fetus, active nephrogenesis occurs at 15–24 wk gestational age and is complete by 36 wk gestational age (23, 42). In the rat kidney, formation of nephrons continues for an additional 6–8 days after birth through induction of undifferentiated mesenchyme in the nephrogenic zone (45). Therefore, both the timing and extent of the insult, as well as the different window of kidney development in rats vs. humans are important factors to keep in mind. However, Hinchliffe et al. (17) reported a significant decrease in mean glomeruli count in both a group of IUGR stillbirths and in eight IUGR infants who died within 1 yr of birth (postnatal group), indicating a lack of postnatal compensation in nephron size.
In summary, UPI and subsequent IUGR impact multiple mechanisms that affect fetal nephrogenesis. Among these mechanisms, this study shows that the insult of UPI induces an increase in p53 phosphorylation without a corresponding increase in MDM2 expression in the kidney of IUGR rats. We speculate that the developing kidney exhibits an immature response to the UPI environment, which may contribute to the vulnerability previously described in the fetal IUGR kidney.
This research was supported by National Institute of Child Health and Human Development Grant R01-HD-41075 (to R. H. Lane) and the University of Utah Children’s Health Research Center (to M. Baserga).
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 © 2006 the American Physiological Society