| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Departments of Obstetrics & Gynecology,
b Preventive Medicine and Community Health,
c Pathology, The University of Texas Medical Branch, Galveston, Texas 77555
d Division of Endocrinology, Robert Wood Johnson School of Medicine, New Brunswick, New Jersey 08903-0019
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
developmental biology, embryo, placenta, pregnancy
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
The adrenomedullin gene and its protein product are highly conserved across species, including human, rat, and porcine [4]. Adrenomedullin has a unique six amino acid residue ring structure and C-terminal amidation similar to CGRP and amylin [5]. The six-membered amino acid ring structure (amino acid 16–21) connected by one disulfide bond (between Cys16 and Cys21) was found to be responsible for the vasodilator activity in human (h) AM [5, 6]. Champion et al. [6] demonstrated that hAM (15–22) possessed vasodepressor activity similar to that of hAM (1–52) in the systemic arterial pressure of the rat. However, hAM (22–52) lacked vasodepressor activity when injected intravenously in doses up to 300 nmol/kg [6, 7]. Administration of synthetic hAM antagonist, composed of amino acids 22–52, blocked the decreased peripheral vascular resistance associated with hAM (1–52) therapy [5, 6]. Human AM (22–52) is thus a specific antagonist to the vasodilator activity of hAM (1–52). It is not known whether a naturally occurring form of hAM (22–52) antagonist exists.
With regard to human pregnancy, Macri et al. [8] identified AM peptide in second trimester human amniotic fluid by RIA and AM peptide and RNA in fetal membranes by immunocytochemical analysis, Western blot analysis, reverse transcriptase-polymerase chain reaction (RT-PCR), and in situ RT-PCR [8]. Di Iorio et al. [9] identified AM peptide in maternal plasma and amniotic fluid by RIA. Di Iorio et al. [10] compared human pregnancies at 26 wk gestational age that were complicated by intrauterine growth restriction (IUGR). In IUGR pregnancies, mean immunoreactive AM values for umbilical cord plasma (63.7 ± 34.2 pg/ml) were double those in control pregnancies (38.1 ± 14.8 pg/ml), suggesting a compensatory role for AM [10].
The human AM precursor, preproAM, a 185 amino acid residue peptide, is processed to proAM, a 164 amino acid residue peptide, and finally to AM, a 52 amino acid residue peptide. Proadrenomedullin N-terminal 20 peptide (PAMP) is processed from preproAM and has vasodilator properties independent from AM [1]. Caron and Smithies [11] recently reported fetal death at midgestation in adrenomedullin gene (Adm) Adm-/Adm- mice. The cause of death appeared to be severe nonimmune fetal hydrops associated with cardiovascular defects (overdeveloped ventricular trabeculae and underdeveloped vascular smooth muscle in the aorta and large arteries). Hay and Smith [12] summarized discussion from a recent meeting held in Japan. They reported that deletion of the entire Adm gene in the germ line, the homozygous knockout (Adm-PAMP-/-), produced embryos with little vascularization in the yolk sac and thin umbilical cords. Also, the few fully formed vessels in the placenta of the knockout embryos were revealed to be "leaky" on angiography. These particular knockout mice [12], however, could be rescued when AM was administered by osmotic minipump. Heterozygotes (Adm-PAMP+/-) were fully viable but expressed only half the tissue concentration of AM present in wild-type mice [12].
Comparing the model described by Caron and Smithies [11] with that described by Hay and Smith [12] suggests that preproAM and PAMP are present both in the embryo and pregnant mouse in the Caron and Smithies [11] model. In contrast, in the Hay and Smith [12] model, preproAM and PAMP should be present in the heterozygote pregnant mouse, but PAMP is absent in the embryo. Caron and Smithies [11] did not examine the placentas associated with these embryos.
Taken together, the spatial and temporal relationship of AM to gestational tissues suggests that AM is a vital endogenous vasodilator peptide responsible for fetal development and maintenance of normal placental function during gestation. Therapeutic interventions that can favorably alter the course of fetal growth restriction could prove invaluable and may assist in the design of appropriate therapeutic interventions to reverse aberrant fetal and placental growth and development in human pregnancy. We hypothesized that AM antagonist will decrease placental and fetal weight, increase maternal systolic BP, and increase fetal reabsorptions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
Timed pregnant rats (250–300 g body weight) were purchased from Harlan Sprague Dawley (Houston, TX). The animal-care facilities at the University of Texas Medical Branch (UTMB), Galveston, TX, received groups of pregnant rats on Day 11 of gestation (a positive sperm smear was Day 1). All animals were given free access to food and water. The Animal Care and Use Committee at UTMB approved all of the procedures for our study. Six pregnant rats were used in each experimental group.
Experimental Protocol
The mean body weight of the rats on Day 14 (study initiation) was 276.3 ± 7.7 g. Osmotic minipumps (Alza, Palo Alto, CA) with varying pumping rates and duration of infusion were inserted on specified days of gestation. The types of minipumps used included model 2ML2 with a pumping rate of 5 µl/h and a 14-day infusion period, model 2ML1 with a pumping rate of 10 µl/h and a 7-day infusion period, model 1003D with a pumping rate of 1 µl/h and a 3-day infusion period, or model 2001D with a pumping rate of 8 µl/h and a 1-day infusion period. An initial time-course study was performed. Pumps were inserted on Day 14 of gestation to continuously deliver AM antagonist (250 µg/rat/day) or vehicle control. Gestational Day 14 was chosen to correspond with the onset of rapid placental growth in normal rat gestation. The dose of AM was chosen based on preliminary work with AM (C.Y. and S.J.W.), published experience with CGRP [13], and the theoretic difference in relative potency between the two similar vasodilator peptides.
The dose of AM antagonist was 10x the dose of AM used in the prior studies. Pregnant rats were killed on Gestational Days 15, 16, 17, 18, 20, or 22. Based on the pumping rate and duration of infusion, we adjusted the drug concentration in the pumps to provide the specified daily dose of the drug. In one group of pregnant rats, we began continuous infusion of AM (25 µg/rat/day) on Day 14 of gestation; animals were killed on each of Days 15, 16, 17, and 18 to obtain placental and fetal weight and placental histology.
Following the time-course study, we modified the doses of AM antagonist. The osmotic minipumps were inserted on Day 14 of gestation; the rats were killed on Gestational Day 22 prior to labor. Based on our initial time-course studies, we adjusted the dose of AM antagonist to 125 µg/rat/day, 250 µg/rat/day, 500 µg/rat/day, 750 µg/rat/day, or vehicle control.
In all cases, the drugs were placed in the minipumps and the osmotic minipumps were inserted s.c. into the dorsum of the pregnant rats while animals were under anesthesia. Anesthesia consisted of a combination of ketamine (45 mg/kg body weight; Fort Dodge Laboratories, Fort Dodge, IA) and xylazine (5 mg/kg body weight; Burns Veterinary Supply, New York, NY). One of the authors (S.J.W.) synthesized hAM (1–52) and hAM antagonist (22–52) by use of solid-phase t-butoxycarbonyl chemistry; the hAM or hAM antagonist was then purified and characterized by mass spectrometry, amino acid analysis, and sequencing. The hAM and hAM antagonist peptides were dissolved in sterile saline solution.
In the initial studies performed by Kitamura et al. [1], human AM (1–52) was infused into male Wistar rats, resulting in decreases in rat cAMP levels. This suggests that there is sufficient homology between human and rat AM and its receptors to administer the human form of this peptide to the rat.
We killed the rats with CO2. Each gestational sac was dissected intact; we recorded the gestational sac weights, placental weights, pup weights, litter size, and pup mortality for each group. The fluid weight for each sac was calculated by subtracting placental and fetal weight from sac weight. The pups, placentas, and membranes were inspected grossly for change in color, shape, or obvious birth defects.
Systolic Blood Pressure
Systolic blood pressure (BP) was measured daily in all rats using a pneumatic tail cuff device (Narco-Biosystems, Houston, TX) from Day 14 of gestation until study termination. We obtained BP values from three consecutive measurements; those values were averaged and analyzed as the mean systolic BP of a given rat for each day. Systolic BPs were included in the analysis regardless of the viability of the pups.
Histopathology
For this portion of the study, drug infusion was begun on Gestational Day 14 with AM antagonist 250 µg/rat/day, AM 25 µg/rat/day, or vehicle control. The animals were killed on each of Gestational Days 15, 16, 17, and 18. Rat placentas were obtained from each treated animal for histologic evaluation. The placentas were fixed in Bouin solution and a 2-mm perpendicular section through the central portion of the placenta was processed, embedded in paraffin, and cut in 6-µm sections. Sections were stained with hematoxylin and eosin (H&E). A technician (blinded to the treatment group) coded the placental blocks using a random numbers table. Each placenta was graded independently by the principal investigator (A.G.W.) and senior pathologist (M.R.G.), who also were blinded to the respective group assignments. Any discrepancies in scores were resolved by joint review. The code was not broken until all scores were finalized. Necrosis was defined as evidence of cell death as manifested by severe cell swelling or cell rupture, denaturation and coagulation of cytoplasmic proteins, and breakdown of cell organelles [14]. We evaluated the following components: maternal decidual necrosis, necrosis in the placental labyrinth, quality and quantity of vessels in the placental labyrinth, and the presence or absence of necrotic fetal remains. In grading necrosis, 0 = none, 1 = 0–10% necrosis, 2 = 10–50% necrosis, and 3 = >50% necrosis. In grading quality of vessels, 0 = no vessels present, 1 = vessels present but deficient in development, 2 = vessels normal, and 3 = vessels congested.
Statistical Analysis
Placenta, pup, sac, and fluid weight (grams) and maternal systolic BP (mm Hg) are expressed as mean ± SEM, compared using one-way ANOVA for individual days and compared using repeated measures of ANOVA for changes across all days. Pup mortality and fetal reabsorption were compared using the Fisher exact or chi-square test with Yates correction. These analyses were computed using Epi Info 6 (CDC 1997, available at: http://www.cdc.gov/epiinfo/index.htm); P < 0.05 was considered statistically significant [15].
The categorical outcomes of decidual necrosis, villus necrosis, and vessel formation were analyzed using the general estimating equations approach with an exchangeable correlation structure using the SAS procedure GENMOD. Each litter was considered a cluster of correlated data. Group differences were evaluated, but the data were too sparse to evaluate differences between days [15].
The repeated measures of BP over time were analyzed using the SAS procedure MIXED with an exchangeable covariance structure for the factors of treatment group, day, and the treatment group x day interaction. Data from each litter was analyzed as correlated data. Least squares means are reported. Measurement of maternal systolic BP was not corrected for fetal viability status; all treated mothers were included within the analysis [15].
The outcomes of placental, sac, fluid, and fetal weights (and ratios) were analyzed using the SAS procedure MIXED for the factors of treatment group, day, and the treatment group x day interaction. Data from the same litters were considered correlated clusters. An exchangeable covariance structure was used to account for correlated data arising from each litter. Pairwise comparisons between groups and days were made using the least squares means option. Data for some treatment groups was collected only at Day 22. For those analyses, treatment was the only factor in the model [15].
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
Time course effects from Gestational Day 14 to Day 22 This portion of the study evaluated the fetoplacental growth effects of AM antagonist infusion initiated on Gestational Day 14 and continued through Gestational Day 22 as compared with vehicle control. The detrimental effects of AM antagonist infusion appeared to manifest in both the placenta and fetus after 24 h of continuous drug exposure as determined by significantly decreased placental and fetal growth during the time-course experiment (Fig. 1). The placental and fetal weights were measured only for the viable pups remaining in each treatment group (Fig. 1, A and B).
|
Amniotic fluid weight and total sac weight were evaluated between Gestational Days 15 and 20 only; sacs with reabsorptions present were excluded from this analysis. There was no change (P = 0.95) in daily amniotic fluid weight (grams) by treatment group (AM antagonist vs. control) despite significantly decreased placental (P < 0.001) and fetal weight (P < 0.001) in AM antagonist-exposed pups and placentas (Fig. 1c). The ratio of fluid weight/fetal weight in the AM antagonist group was significantly greater (P < 0.001) than the control group at each day sampled. There were a total of 27 fetal reabsorptions in AM antagonist-exposed pups (n = 511) compared with 4 in control pups (n = 266) (P = 0.018, chi-square with Yates correction). This calculation included all AM antagonist treatment groups between Gestational Day 16 and Day 22.
Parallel mixed model ANOVAs were performed on the outcome measures of fetal and placental weights. For both placental and fetal weights, a full model with main effects for day, treatment group, and group x day interaction revealed highly significant day (P < 0.001), treatment (P < 0.001), and group x day interaction effects. Interpretation of the interaction in both models is that the relationship among treatment groups changes from day to day. At each gestational day sampled, placenta and pups exposed to AM antagonist are significantly decreased in weight vs. controls.
The AM-only treatment group did not show any difference in fetal weight compared with controls on Gestational Days 15, 16, 17, and 18. A significant increase in placental weight, however, was observed on Gestational Day 18 (AM only, 0.60 ± 0.02 g, vs. control, 0.47 ± 0.02 g, P < 0.001).
Dose response effect from Gestational Day 14 to Day 22 Four different doses of AM antagonist were infused continuously from Gestational Day 14 to Day 22. The specific doses of AM antagonist were 125 µg/rat/day, 250 µg/rat/day, 500 µg/rat/day, or 750 µg/rat/day, respectively. We assessed fetal weight and morbidity and placental weight in all of these AM antagonist doses on Gestational Day 22 only. The fetal weights using all four doses of AM antagonist were significantly lower than controls (P < 0.001). There was, however, no significant difference in fetal weight among the doses used. Despite differing doses of AM antagonist, the rates of fetal mortality and reabsorption did not vary significantly. Similarly, placental weights were significantly lower than controls with all four doses of AM antagonist (P < 0.001). Again, there was no significant difference in placental weight between all four AM antagonist doses. The decreased placental and fetal weights in the AM antagonist-treated groups did not exhibit a dose-dependent change.
On gross examination of the pups, the fetal abdomens appeared to be more prominent from Gestational Day 18 to Day 20 and seemed to be distended with fluid, an effect corresponding to length of exposure to AM antagonist.
Effects of Adrenomedullin Antagonist on Placental Pathology
In this portion of the study, we examined rat placentas obtained from AM, AM antagonist, and vehicle control mothers; animals were killed on each of Days 15, 16, 17, and 18. Only those placentas without gross evidence of fetal reabsorptions were evaluated. In the treated rats, histopathologic changes were more prominent in the fetal compartment of the placenta rather than in the maternal compartment.
Within the placenta, maternal vascular spaces were identified by lack of endothelial cell lining, and fetal vessels were identified by endothelial-lined vascular spaces containing varying proportions of nucleated red blood cells [16]. The AM antagonist-treated rats exhibited decreased fetal vascular development (grade 0 or 1) in the labyrinth and increased necrosis (grade 2 or 3) in the decidua and labyrinth on Day 18 of gestation (Figs. 2 and 3). Because fetal vessels in the AM-treated groups were similar to vessels in the control group, these two groups were combined for statistical analysis. The analysis of fetal vessels included evaluations of a total of 57 placentas in the AM antagonist group and a total of 60 placentas in the control and AM groups. Animals treated with AM antagonist only vs. control and AM demonstrated significantly deficient vessel formation (grade 0 or 1) in the labyrinth (P = 0.04). The finding of deficient and absent vascular development in the AM antagonist group suggests a role for AM in angiogenesis. Because there was no apparent deficit in vessel formation until 96 h of exposure to AM antagonist, we speculate that vessels originally formed and later regressed by an as yet unknown mechanism.
|
|
In addition to the microscopic changes described above, we noted gross necrotic change (green discoloration) around the edges of the placentas (decidua and membranes) at approximately 72 h after drug exposure. Beginning with Day 18, AM antagonist-exposed fetoplacental units appeared to have developed polyhydramnios and fetal ascites.
Effects of Adrenomedullin Antagonist on Systolic Blood Pressure
We measured systolic BP in all mothers, including those with pup mortality. Drug infusions were started on Gestational Day 14, and we compared three treatment groups: high-dose AM antagonist (500 µg/rat/day and 750 µg/rat/day), low-dose AM antagonist (125 µg/rat/day and 250 µg/rat/day), and vehicle control (Fig. 4). Blood pressures were similar between the two different doses in the low-dose group and between the two different doses in the high-dose group. Both AM antagonist groups (high dose vs. low dose) had decreased systolic BP between Gestational Days 15 and 16. Beginning on Gestational Day 17, systolic BP in the low-dose group was similar to control. The high-dose group showed elevated systolic BP over the control group from Gestational Day 17 to Day 20.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
We used the time-course and dose-response studies to assess the complete spectrum of AM antagonist effects. In every case of AM antagonist exposure, deleterious effects to the fetus and placenta were observed. Fetal mortality rates were greater than controls at each time point and dose of AM antagonist examined. Fetal weights in exposed pups were lower at each time point and AM antagonist dose than the control weights; a similar effect of decreasing weight was noted in the placentas exposed to AM antagonist. The total sac weight increased in controls from Gestational Day 16 to Day 20; sac and fluid weight were not examined on Day 22. In contrast, amniotic fluid weight was unchanged between control and AM antagonist groups from Gestational Day 15 to Day 20. Amniotic fluid weight increased in both groups with advancing gestational age. The deleterious effects of AM antagonist on fetoplacental growth was apparent with 125 µg/rat/day, the lowest dose used in this study. Future studies using lower doses are required to assess a true dose-response effect.
The physiologic data obtained is strengthened by the histopathologic changes seen in the AM antagonist group. We postulate that the necrosis in the labyrinth is secondary to the deficient vascular development in the placenta and decreased utero/placental/fetal oxygenation. Decidual necrosis may occur through a deficiency in the maternal vascular circulation. These results are similar to the deficient vascular development described in the Adm gene knockout models [11, 12]. Although no change in fetal weight was noted with exogenous AM treatment alone on Gestational Day 18 compared with control, our time-course study of AM alone did indicate increased placental weight. Therefore, we chose to study the AM-only treated placentas for histopathologic change. It is possible that the vascular congestion observed in the AM-only group is responsible in part for the increased placental weight identified on Gestational Day 18.
As reviewed previously, the dose-response study suggests that as little as 125 µg/rat/day of AM antagonist will cause decreased fetoplacental growth, yet increased doses of AM antagonist are required for the elevation of systolic BP. The only difference observed in systolic BP between treatment groups occurred with the high-dose AM antagonist treatment, 500 µg/rat/day and 750 µg/rat/day. This systolic BP difference was not statistically significant and occurred between Gestational Days 17 and 20 only. The systolic BP elevation may represent generalized vasoconstriction associated with potentially decreased uteroplacental flow in association with AM antagonist treatment. This effect suggests alternative systemic mechanisms for the action of AM in rat gestation rather than simply the result of deficient placental vasculogenesis.
Adrenomedullin is a vasodilator that exerts its effects through dual pathways in each of two vascular cell types. In endothelial cells, the action of AM appears to involve the generation of NO and its diffusion to the adjacent smooth muscle cells [17]. In the vascular smooth muscle cells, AM stimulates cAMP [18–20]. Endothelial cells are present in the developing fetal vessels in the placenta; endothelial and smooth muscle cells are present in the fetal and maternal vessels. These cells in the placenta can contribute to local production of AM and have a beneficial vascular effect. Exogenous infusion of AM antagonist could potentially antagonize locally produced AM in the placenta as well as circulating AM.
The observations in Adm gene deletion mice, i.e., poor vascular development in both models [11, 12] and fetal hydrops as described by Caron and Smithies [11], further support the importance of AM in early vascular development. In both knockout models, fetal AM appears essential for fetal development. It appears that fetal preproAM and PAMP are also necessary for proper placental development. Either endogenous maternal PAMP [12] may prolong survival of the embryo but not allow fetal survival or exogenous maternal administration of AM [12] may "rescue" the deficient embryo and placenta. This is consistent with our finding of increased fetal demise and decreased placental and fetal weight when AM antagonist is administered, whereas total lack of fetal AM is lethal. It appears that maternal level of AM is also imperative for normal fetal survival and growth.
Few embryonic lethal genes that directly disrupt the fetoplacental unit have been identified [21]. A common theme linking these embryonic lethal genes, including Adm, is their critical role in angiogenesis and vasculogenesis. For example, vegf disruption results in fetal death at Gestational Day 11–12, secondary to impaired angiogenesis. Deletion of one of the vegf receptor genes, Flt-1, causes embryonic lethality at midsomite stage [21].
In summary, the antagonism of AM function during rat gestation results in diminished placental and fetal growth and impaired placental vasculogenesis.
|
NOTE ADDED IN PROOF |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
|
FOOTNOTES |
|---|
1 This work was supported in part by NIH HL58144 and WRHR HD01269. 
2 Correspondence: Andrea G. Witlin, The University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-1066. FAX: 409 747 1669; agwitlin{at}utmb.edu
Accepted: April 24, 2002.
Received: December 3, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Chauhan, U. Yallampalli, L. Reed, and C. Yallampalli Adrenomedullin 2 Antagonist Infusion to Rats During Midgestation Causes Fetoplacental Growth Restriction Through Apoptosis Biol Reprod, December 1, 2006; 75(6): 940 - 947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Al-Ghafra, N.M. Gude, S.P. Brennecke, and R.G. King Increased adrenomedullin protein content and mRNA expression in human fetal membranes but not placental tissue in pre-eclampsia Mol. Hum. Reprod., March 1, 2006; 12(3): 181 - 186. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Penchalaneni, S. J. Wimalawansa, and C. Yallampalli Adrenomedullin Antagonist Treatment During Early Gestation in Rats Causes Fetoplacental Growth Restriction Through Apoptosis Biol Reprod, November 1, 2004; 71(5): 1475 - 1483. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
—, low-dose AM antagonist (125 µg/rat/day + 250 µg/rat/day); —
—, high-dose AM antagonist (500 µg/rat/day and 750 µg/rat/day). Osmotic minipumps were inserted Day 14; the mother rats were killed each day on Gestational Days 15–18, 20, and 22. Trend toward increased systolic blood pressure on Days 17–20 with the use of high-dose AM antagonist