Erythropoietin Ameliorates Damage to the Placenta and Fetal Liver Induced by Exposure to Lipopolysaccharide

Article Outline

• Abstract
• 1. Introduction
• 2. Methods
  • 2.1. Animal preparation
  • 2.2. Experimental protocol
  • 2.3. Necropsy
  • 2.4. Histology
  • 2.5. Analysis of placental structure
    • 2.5.1. Qualitative
    • 2.5.2. Quantitative
    • 2.5.3. Placental injury index
  • 2.6. Analysis of fetal liver injury
    • 2.6.1. Qualitative
    • 2.6.2. Quantitative
  • 2.7. Statistical analysis
• 3. Results
  • 3.1. Structural analysis of the placenta
    • 3.1.1. Qualitative
    • 3.1.2. Quantitative
    • 3.1.3. Placental injury index
  • 3.2. Structural analysis of the liver
  • 3.3. Regression analyses
• 4. Discussion
  • 4.1. Mechanisms underlying LPS-induced placental damage
  • 4.2. LPS exposure induces damage in the fetal liver
  • 4.3. rhEPO treatment ameliorates LPS-induced injury
• 5. Conclusions
• Acknowledgements
• References
• Copyright

Abstract 

Intrauterine infection and inflammation have been causally linked to preterm birth and fetal brain injury. Using an ovine model of endotoxin–induced brain injury we have recently shown that recombinant human erythropoietin (rhEPO) reduces brain injury and protects against damage to myelination in major myelinated axon tracts. Our present objective was to determine whether rhEPO is also protective of the placenta and the fetal liver, organs which could influence fetal well-being. At 107 ± 1 days of gestational age (DGA) chronically catheterized fetal sheep were randomly assigned to receive, on 3 consecutive days, either: 1) an i.v. bolus dose of lipopolysaccharide (LPS; ∼0.9 μg/kg; n = 8); 2) i.v. bolus dose of LPS, followed at 1 h by 5000  IU/kg of rhEPO (LPS + rhEPO, n = 8); 3) rhEPO (n = 3). Seven untreated fetuses served as controls (n = 7). The placenta and fetal liver were examined histologically at 116 ± 1 DGA; a placental injury index was formulated comprising measures of placental area, apoptosis, tissue injury and the size of the intervillous space. In LPS-exposed fetuses this index was greater than in control or rhEPO alone fetuses (p < 0.02). Treatment of LPS-exposed fetuses with rhEPO resulted in a reduction in the index (p < 0.05) and in the extent of liver necrosis. We conclude that rhEPO offers protection to the placenta and fetal liver in the presence of acute inflammation.

Keywords: Endotoxin, Placenta, Injury, Inflammatory response

Back to Article Outline

1. Introduction 

Intrauterine infection and inflammation have been causally linked to preterm birth and fetal brain damage [1], [2], but the mechanisms involved are not clearly identified. In order to investigate the underlying mechanisms we have established a model of lipopolysaccharide (LPS)-induced brain injury in preterm fetal sheep [3]. In this robust model repeated exposure of the fetus to endotoxin leads to profound reductions in umbilical–placental blood flow [4] and placental injury [5]. Reduced placental perfusion and placental injury likely impair fetal–maternal gas exchange, contributing to the observed fetal hypoxemia [3] and reduced cerebral oxygen delivery [4], and ultimately brain injury. Endotoxin given to the fetus by a sustained infusion [6], rather than as a bolus [3], does not cause fetal hypoxemia; however the placenta is damaged and fetal brain injury still occurs. This suggests that inflammation with or without hypoxemia can activate pathways resulting in fetal brain injury. A recent study in human preterm infants with brain injury concluded that disturbed placental circulation underlies the development of brain injury in a majority of cases, although additional factors such as the fetal inflammatory response may operate in the case of chorioamnionitis [7].

Many approaches have been trialled to treat perinatal brain injury but with the exception of hypothermia in some cases [8], [9], they have had limited success or untoward side effects. There are presently no data to indicate whether therapeutic agents primarily aimed at protecting the fetal brain also offer protection to the placenta, with the possibility of minimizing ongoing consequences to the nervous system. This is highly relevant as studies have shown an association between placental lesions and later neurological impairment in some very low birth weight infants [10]. Furthermore, it has been suggested that disturbed placental circulation underlies the development of periventricular white matter injury in cases with prenatal or peripartum brain injury [7]. We have recently shown that, in the preterm ovine fetus, the pleiotrophic cytokine erythropoietin (EPO) significantly reduces brain injury induced by repeated exposure to LPS [11]; however, it is not known whether rhEPO is also protective of the placenta or other fetal organs. EPO has a wide range of actions which might underlie its putative protective qualities including anti-inflammatory [12], anti-apoptotic [13] and anti-oxidant effects [14].

Our primary objective in the present study was to determine whether administration of recombinant human EPO (rhEPO) reduces the structural damage that is induced in the placenta by acute exposure to LPS; this model was chosen to replicate an acute fetal inflammatory event. EPO and its receptor (EPOR) are expressed in the ovine placenta from about 100 days of gestation (DGA) [15]. In addition we have investigated the effects of rhEPO on the fetal liver following exposure to LPS. The fetal liver is directly exposed to various antigens, pathogens and toxins reaching the fetus via the placenta; however relatively little is known about the fetal liver's capacity to cope with prenatal insults such as infection or inflammation. Furthermore, the fetal liver is the likely site of at least a proportion of EPO production in the fetus [16]; consequently any injury to the liver during intrauterine life could compromise its ability to provide ongoing endogenous protection to the fetus.

Back to Article Outline

2. Methods 

2.1. Animal preparation 

All animal procedures were approved by the Monash University Animal Ethics Committee. Date-mated pregnant Merino x Border-Leicester ewes (n = 23) underwent aseptic surgery at 102 ± 1 days after mating (term is ∼147 days) for the chronic implantation of catheters into a fetal femoral artery (for blood sampling and recording arterial pressure) and vein (for drug infusions). Ampicillin sodium (1 g i.m.) was administered to ewes for 3 days after surgery. To monitor fetal well-being, fetal arterial blood (0.3 ml) was sampled on the day after surgery and then on alternate days until necropsy.

2.2. Experimental protocol 

At 107 ± 1 DGA fetuses were randomly assigned to receive, intravenously, either LPS (∼0.9 μg/kg estimated fetal weight; Escherichia coli, 055:B55, Sigma Chemical, St. Louis, MO, USA), or an equivalent volume of saline. One hour later rhEPO (5000 IU/kg fetal body weight, i.v., Epoetin alfa, Janssen-Cilag, Australia) or an equivalent volume of saline was administered. This protocol was repeated for three consecutive days. Study groups were: [1] controls (n = 7) comprising 4 saline-infused fetuses and 3 non-catheterised fetuses; [2] saline followed by rhEPO (rhEPO, n = 3); [3] LPS followed by saline (LPS, n = 8); and [4] LPS followed by rhEPO (LPS + rhEPO; n = 8). Fetal arterial blood samples (0.3 ml) were collected 5 min before, and then hourly for 6 h after, administration of LPS or saline (on each study day); in each sample we measured arterial oxygen saturation (SaO₂), partial pressures of oxygen and carbon dioxide (PaO₂, PaCO₂), pH (pHa), haematocrit (Hct) and haemoglobin (Hb) concentration (ABL 520, Radiometer, Copenhagen, Denmark), and blood glucose and lactate concentrations (YSI 230 STAT; YSI Inc., Yellow Springs, OH, USA). Fetal arterial pressure was recorded continuously (PowerLab8/30, ADInstruments Pty Ltd, Australia) via the arterial catheter connected to a pressure transducer; the digitized data were subsequently analysed over 2 min periods at 1 h intervals to obtain systolic, diastolic and mean arterial pressures (MAP). Fetal heart rate (HR) was derived from the arterial pressure waveform. We measured the baseline values of each physiological variable (i.e. values immediately prior to administration of LPS, saline or rhEPO) and the maximal change in each variable during the first 6 h after these administrations.

2.3. Necropsy 

The ewe and fetus were euthanized with an overdose of sodium pentobarbitone (130 mg/kg, i.v.) administered to the ewe at 116 ± 1 DGA. All placentomes were weighed, following which 6 randomly selected placentomes were immersion-fixed in 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB; pH 7.4). Fetal body weight was measured and the liver removed, weighed and immersion-fixed in 4% PFA in 0.1 M PB. Blocks of tissue (2 mm thick) were collected from the central region of placentomes and the inferior medial lobe of the fetal liver and embedded in paraffin; the presence of gross tissue injury, visualised as white necrotic tissue, was noted during dissection.

2.4. Histology 

For each animal, transverse sections (5 μm) of placenta and fetal liver were stained with haematoxylin and eosin (H&E) for qualitative and quantitative analysis (see below). Sections were also prepared for immunohistochemistry and reacted for the following antibodies: rabbit anti-mouse caspase-3 antibody (1:500; R&D Systems, NSW Australia) to identify apoptotic cells (placenta only) and mouse anti-human Ki-67 antibody (1:100; clone Mib-1; Dako Denmark) to identify proliferating cells using previously established methods including antigen retrieval [17]. Sections were then incubated in appropriate species-specific biotinylated secondary antibodies. The avidin–biotin–peroxidase complex (Vector Laboratories, Burlingame, CA, USA) was used to visualise the immunoreactive product; prior to coverslipping, sections were counterstained with 0.1% thionin. Two sections from each of four randomly selected placentomes per animal were stained with von Kossa's stain to identify calcium deposits, indicative of long-term tissue injury, and counterstained with H&E [18].

2.5. Analysis of placental structure 

H&E-stained sections were examined for evidence of overt injury including tissue necrosis, oedema and enlarged intervillous spaces.

In von Kossa-stained sections, the total cross-sectional areas of the placentome, haemophagous zone, fetal/maternal interdigitating tissues, fetal tissue and areas of tissue injury (von Kossa positive) were measured on a digitising tablet (Sigma Scan Pro Plus v. 4.0, Jandal Scientific, CA); areas of injury were expressed as percentages of the total cross-sectional area, as described previously [5]. In H&E-stained sections from each of two placentomes per animal, the density (cells/mm²) of apoptotic bodies (confirmed by comparing with activated caspase 3-stained sections) was assessed in at least six randomly chosen fields, excluding regions of damaged tissue. In these sections, the density (cells/mm²) of neutrophils, monocytes and lymphocytes was determined at 10 sites across the placentome; classification of cells was based on morphology. For each fetus, H&E-stained sections were scored for the degree of intervillous spaces: 0 = no prominent intervillous spaces, 1 = presence of prominent intervillous spaces, 2 = enlarged intervillous spaces, 3 = exaggerated intervillous spaces and associated distortion of the fetal villi. For each fetus, the density of Ki67-immunoreactive (IR) cells was determined in 20 randomly selected fields; counts included both fetal and maternal compartments as localisation was not always clear. For each variable, a mean value was calculated for each animal and then a group mean determined.

To assess the overall extent of structural alterations to the placenta we formulated a placental injury index. This index included placentome cross-sectional area (mm²), the presence of gross tissue injury, the percentage of placenta occupied by calcium deposits (positive von Kossa staining), the size of intervillous spaces (scoring system) and the density of apoptotic cells (cells/mm²). For each animal, data for each variable was ranked from the most affected (26 points) to least affected (1 point); for identical values the same ranking score was given to each fetus. The placental injury index for each animal was determined by adding the scores from each category; a group mean was then calculated.

2.6. Analysis of fetal liver injury 

H&E-stained sections were examined for evidence of overt injury including tissue necrosis, haemorrhage and dilated blood vessels.

In at least 2 sections per fetus the densities (cells/mm²) of Ki67-IR cells was determined in 20 randomly selected fields; a mean value was calculated for each animal and a group mean determined.

2.7. Statistical analysis 

Differences between groups were analysed by two-way repeated ANOVA (factors: day and treatment; physiological data) or one-way ANOVA (histological data); or Kruskal–Wallis test (placental injury index scores); post-hoc analysis corrected for multiple comparisons was used where appropriate. For all histological analyses the examiner was blinded to the identity of the fetus, with the code not being disclosed until analysis was complete. Differences were considered to be significant at p < 0.05. Results are expressed as mean of means ± SEM. Linear regression analysis was performed, in GraphPad Prism (v 4.02), to correlate the mean of the maximum change in SaO₂ and MAP over the three experimental days with the placental injury index; a) across all groups and b) within each group (control, LPS, LPS + rhEPO, rhEPO).

Back to Article Outline

3. Results 

Physiological and neuropathological data have been reported previously [11]. In brief, administration of saline or rhEPO alone had no significant effect on any measured physiological or neuropathological variable. Exposure to LPS alone, resulted in fetal hypoxemia and hypotension (both p < 0.05) and significant brain injury; treatment of LPS-exposed fetuses with rhEPO did not significantly alter the physiological effects of LPS but reduced the degree of brain injury [11].

Organ weights. At necropsy there were no differences between groups in fetal body weight, or in brain, liver or placental weights when expressed either as absolute weights or relative to body weight (Table 1). There was no difference in the number of placentomes between groups (Table 2).

Table 1. Fetal organ and placental weights.

Values represent mean ± SEM. No significant differences between groups. rhEPO: recombinant human erythropoietin; LPS: lipopolysaccharide.

Table 2. Placental parameters.

Values represent mean ± SEM; *p < 0.05 compared to controls; **p < 0.02 compared to controls and rhEPO alone; #p < 0.05 compared to LPS + rhEPO. rhEPO: recombinant human erythropoietin; LPS: lipopolysaccharide.

3.1. Structural analysis of the placenta 

In von Kossa-stained sections counterstained with H&E, placentomes from control (Fig. 1A, D, G) and rhEPO alone fetuses showed no evidence of overt injury. At high magnification (Fig. 1G), the interdigitated fetal and maternal tissues could be easily distinguished; fetal tissues were lined with cuboidal trophoblastic cells and giant binucleate cells while maternal tissues were lined with squamous epithelium and dark staining syncytium. In fetuses exposed to LPS alone (Fig. 1B, E, H) but not LPS + rhEPO (Fig. 1C, F, I), degenerating fetal and maternal tissues were observed.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1 

  Transverse sections of placenta from control (A, D, G), LPS alone (B, E, H) and LPS + rhEPO (C, F, I) fetal sheep at 116 days' of gestation stained with von Kossa's stain for calcium deposits and counterstained with haematoxylin and eosin. In control fetuses there was no signs of overt tissue injury (A,D,G) with the fetal and maternal tissues being closely associated (G). Positive von Kossa staining, indicative of tissue injury (black precipitate), was present in fetuses exposed to LPS alone (B, E). This was associated with oedema in the fetal villus tissue (E, asterisks) and increased intervillous spaces (E, H arrow heads) and apoptosis in fetal tissues (H, arrows). Less damage occurred in LPS-exposed fetuses treated with rhEPO (C, F, I). Ki67-IR (arrows) in control (J), LPS alone (K) and LPS + rhEPO (L) illustrating a higher density in LPS + rhEPO fetuses. M, maternal tissue; F, fetal tissue; F&M, interdigitated fetal and maternal tissue. Scale bars: A-C = 500 μm; D–F = 100 μm; G-L = 20 μm.

There were no differences (p > 0.05) between any groups in the mean cross-sectional area of placentomes nor in the proportion of the cross-sectional area occupied by the haemophageous zone, fetal tissue or the interdigitated fetal/maternal tissues (Table 2). In von Kossa-stained sections from control and rhEPO fetuses, there were a few, very minor areas of positive staining; in contrast, areas containing calcium deposits were present in all LPS-exposed fetuses (Table 2); due to the variability in the response to LPS the effect did not reach significance. Calcium deposits were primarily located in fetal tissues in regions of overt degeneration (Fig. 1B, E), but were also present in regions where such damage was not evident. In the most affected LPS fetus, the damaged area exceeded 12% of the total fetal and maternal tissues. In general the most severe damage occurred towards the capsular surface, likely due to the narrowing of blood vessels in this region and thus a potentially greater degree of tissue hypoxemia. In placental tissues associated with larger calcium deposits the mesodermal core in the fetal compartment appeared oedematous (Fig. 1E). Throughout the remainder of the placentome there was an apparent reduction, relative to controls, in the volume of the villous vascular tree resulting in prominent intervillous spaces (Fig. 1H). In LPS + rhEPO fetuses (Fig. 1C, F, I) the area of von Kossa-stained tissue was not different from control levels (p > 0.05; Table 2). Oedema and enlarged intervillous spaces were not prominent features in placentomes from LPS + rhEPO animals (Fig. 1I). Overall, there was no difference between groups in the intervillous space score (Table 2) although, as indicated above, there were regions of significant expansion of this space in the most affected LPS fetuses.

In control fetuses, apoptotic cells were present mainly in the fetal compartment, at low density suggesting a constant turnover of cells, most likely trophoblasts (Table 2). In LPS-exposed fetuses, apoptotic cells were concentrated in regions of overt injury but were also present throughout the placentome (Table 2, Fig. 1H). There was a non-significant trend for a reduction in the density of apoptotic cells in rhEPO-treated LPS-exposed fetuses compared to LPS alone; there was no significant difference between densities in LPS + rhEPO and control fetuses (Table 2, p > 0.05). Ki67-IR cells were present across the placenta; it appeared that proliferation occurred mainly in the fetal compartment. Across both compartments cell densities were not different between control (Fig. 1J), LPS alone (Fig. 1K) and rhEPO alone but were greater in LPS + rhEPO-treated fetuses (Fig. 1L) compared to controls (Table 2). There were no differences (p > 0.05) between groups in the density of neutrophils, lymphocytes or monocytes at the time point examined (data not shown).

In LPS-exposed fetuses, the placental injury index was greater than in control fetuses (p < 0.02), rhEPO alone fetuses (p < 0.02), and LPS + rhEPO-treated fetuses (p < 0.05, Table 2). The index was not significantly different between control, rhEPO and LPS + rhEPO-treated fetuses (p > 0.05).

3.2. Structural analysis of the liver 

Control fetuses exhibited a typical hepatic structure comprising hepatocytes, sinusoids, portal triads and central veins (Fig. 2A, B). In 38% of fetuses that received LPS alone, regions of necrosis were evident, predominantly bordering the central vein (Fig. 2D, E). Inflammatory cells including neutrophils, granulocytes and eosinophils were observed around the periphery of necrotic lesions. Such necrosis was distributed irregularly across liver sections. Dilatation of sinusoids was evident across the liver in the majority of fetuses. In contrast, necrotic lesions and associated inflammatory cells were not seen in the liver of any LPS + rhEPO fetuses (Fig. 2G, H) and sinusoids were not as dilated. In all fetuses, extramedullary hematopoiesis was evident, comprised of erythroid islands and associated cells assumed to be myeloid precursors. There was a trend for an increase in cell proliferation in this tissue but also amongst hepatocytes in fetuses exposed to LPS, with or without rhEPO treatment (control 356 ± 129, LPS 945 ± 144, LPS + rhEPO 1227 ± 374 cells/mm²). This is illustrated by comparing control (Fig. 2C) with LPS (Fig. 2F) and LPS + rhEPO (Fig. 2I) fetuses.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 2 

  Transverse sections of the liver stained with haematoxylin and eosin (A, B, D, E, G, H) or Ki-67-IR for cell proliferation (C, F, I) in control (A–C), LPS alone (D–F) and LPS + rhEPO (G–I) fetuses. In control fetuses (A, B) plates of hepatocytes were arranged around the central vein (asterisks). In fetuses exposed to LPS alone, tissue necrosis (D, arrow heads) was present usually distributed around the central vein (D, E); necrosis was not seen in LPS + rhEPO fetuses (G, H). LPS exposure results in increased cell proliferation (F, brown cells) compared to control (C); cell proliferation was further enhanced in fetuses treated with rhEPO following LPS (I). Scale bars: A, D, G = 500 μm; B, E, H = 20 μm; C, F, I = 50 μm.

3.3. Regression analyses 

There was no correlation between SaO₂ and MAP and the placental injury index across all groups (SaO₂, r² = 0.0197, p = 0.54; MAP, r² = 0.00008, p = 0.97) or within each group (p > 0.05).

Back to Article Outline

4. Discussion 

Using an ovine model of intrauterine inflammation induced by fetal LPS exposure, we have previously shown that repeated, acute LPS exposure leads to fetal physiological perturbations as well as injury to the fetal brain and placenta [3], [5]. More recently we have shown that rhEPO has protective effects on LPS-induced fetal brain injury [11]. We now show that treatment with rhEPO also protects the placenta and fetal liver against injury induced by LPS exposure. In the placenta, we found that there was a significantly greater placental injury index in LPS-exposed fetuses compared with controls: when LPS-exposed fetuses were treated with rhEPO, the placental injury score was significantly lower than in the LPS-treated fetus and not different from values in controls. We acknowledge the limitations in using such an index including assigning equal weighting to all parameters; at this stage it is not possible to determine which pathologies might have the greatest adverse influence on the fetus. In the fetal liver we showed that LPS-induced perivascular necrosis was reduced by rhEPO treatment. This protection is achieved by rhEPO without ameliorating the physiological effects of acute LPS exposure, most notably fetal hypoxemia and hypotension, and therefore is likely to occur at a cellular level. This is the first study in a long-gestation species to demonstrate that an agent which is protective to the fetal brain [11], [19], [20] has beneficial effects for other organs acutely exposed to LPS.

4.1. Mechanisms underlying LPS-induced placental damage 

The morphological alterations that were observed in the placenta after LPS exposure are similar to those that we have previously observed using the same protocol of repeated acute LPS administration to the fetal circulation [5]. In the present study we have also demonstrated an increase in apoptosis, as has been described for other species [21]. We did not observe an increase in cell proliferation as reported in an ovine study of endotoxin exposure [22]; we note that their study examined a shorter time frame than the present study. The mechanisms underlying LPS-induced placental injury are not yet understood and are likely to be multifactorial but could involve tissue hypoxia or local release of cytokines as part of an inflammatory cascade. Any proposed mechanism must take into account that both fetal and maternal tissues were affected, albeit the majority of injury occurred in the fetal compartment. In humans the LPS receptor, CD14, is expressed on fetal membranes [23] and placental trophoblastic cells [24]; Toll-like receptor 4 (TLR4), the main mediator of immune responses to LPS is also expressed on placental trophoblasts [25]. We have previously demonstrated that the cascade of events induced by activation of this pathway in the ovine placenta involves upregulation of nuclear factor κ-B [26], a transcription factor for proinflammatory cytokines. LPS exposure of human placental explants has been shown to increase cytokine production [27]; examination of ovine placental cytokines was not included in our protocol.

Placental injury could also be caused by fetal physiological responses to LPS exposure. We have previously established that repeated exposure of the preterm fetus to LPS results in systemic hypoxemia, hypotension, elevated plasma proinflammatory cytokines [3] and a marked reduction in umbilical-placental blood flow [4]. Reduced placental flow could be due, at least in part, to increased vascular permeability and placental oedema, potentially resulting in an increase in umbilical-placental vascular resistance [28] and an increased separation between maternal and fetal circulations. Evidence for placental oedema was found in the present study in the swollen fetal tissues observed on histological examination. In some cases we also observed an apparent decrease in the volume of the vascular tree within the placenta, resulting in more prominent intervillous spaces which would be deleterious to placental exchange function. Together with data from our previous studies, we consider it likely that fetal LPS exposure results in increased vascular permeability and oedema in the placenta, decreased perfusion and tissue hypoxia.

4.2. LPS exposure induces damage in the fetal liver 

Damage to the fetal liver, in vivo, by exposure to LPS alone has not been reported previously although fetal hepatic damage resulting from galactosamine/LPS [29], perinatal asphyxia in humans [30] and chronic placental insufficiency in sheep [31] has been reported. We found a distinctive pattern of necrosis with injury to hepatocytes predominantly occurring around the central hepatic vein; this is the region most distant from oxygenated arterial blood and presumably particularly vulnerable to insults. Suggested mechanisms for hepatic injury following LPS exposure include TNF-alpha release by the liver [29] and the production of peroxynitrite, a powerful oxidant [21]. Dilatation of hepatic sinusoids, such as we observed with LPS exposure, is considered to be a primary event leading to hepatic injury [32] although the causation and functional impact of the dilatation are not understood. In our model, dilatation could result from inflammation of endothelial cells lining the sinusoids [33] and/or the effects of free radicals such as nitric oxide produced during the systemic inflammatory response [21]. In LPS-exposed livers, compared to controls, there appeared to be an increase in Ki-67-IR cells. This is in accord with a recent autopsy study of neonates with chorioamnionitis; increased erythropoiesis was associated with increased myelopoiesis, thought to represent an active fetal response to intrauterine infection [34].

4.3. rhEPO treatment ameliorates LPS-induced injury 

The mechanisms underlying the beneficial effects of rhEPO treatment on the structure of the placenta and fetal liver in the LPS-exposed fetus, like the neuroprotective effects of rhEPO, are unclear and still being elucidated. In our studies, rhEPO treatment did not alter LPS-induced systemic hypotension and hypoxemia in the fetus. Across all fetuses, either with or without rhEPO, there was no correlation between the mean changes in SaO₂ or MAP with the extent of the placental injury; neither was there a correlation when only LPS + rhEPO fetuses were analysed. Therefore it appears that, in the injury induced by LPS exposure, rhEPO is not acting by ameliorating systemic fetal hypoxemia or hypotension but may be acting by limiting the injurious effects of inflammation in the placenta and fetal liver at the cellular level.

In the placenta, rhEPO treatment appeared to reduce oedema and preserve the fetal villous vascular tree; both of these effects may occur as a result of rhEPO protecting against an LPS-induced increase in vascular permeability. Although the influence of rhEPO on the placental vasculature has not been examined specifically, rhEPO has been shown to stimulate cell proliferation and angiogenesis in endothelial cell cultures [35]. Furthermore, it is known that rhEPO protects junctional proteins in the blood–brain barrier, preventing leakage of plasma proteins into the brain [36]; it is conceivable, therefore, that rhEPO has a protective effect on the integrity of the placental vasculature. In the liver, amelioration of damage might also relate to the protection of endothelial cells as swelling of sinusoids was reduced. Based on our observations, it is conceivable that rhEPO treatment causes a further increase in the number of Ki-67-IR cells compared to LPS-exposed fetuses; some of these cells were located in the erythroid islands but many were amongst hepatocytes, possibly representing a response directed at regenerating hepatic tissue in the face of an inflammatory insult. There was only a modest increase in the hematocrit in both groups.

It is also possible that rhEPO might be exerting protective effects by up-regulating anti-apoptotic genes [13] or inhibiting the activation of apoptotic genes, as has been reported for other tissues; in accordance with this we observed a trend for a decrease in the density of apoptotic cells in the placenta after rhEPO treatment and a general preservation of the integrity of the tissue. In agreement with our findings, it has recently been reported that EPO attenuates LPS-induced splenic and thymic apoptosis in rodents [37]. The reduction of apoptosis in lymphoid tissue during sepsis is a vital aspect in maintaining an appropriate immune response. As in the liver, rhEPO treatment caused an increase in the proliferative response in the placenta; similarly, it might represent a protective mechanism against the inflammatory insult.

Back to Article Outline

5. Conclusions 

Together with our previous study on the fetal brain [11], we demonstrate in a long-gestation species that rhEPO offers protection to the placenta and fetal liver, increasing its potential as an anti-inflammatory therapy. Recently, a randomised control study of a small group of very preterm human infants has shown that there were no significant adverse effects with early high dose rhEPO treatment for brain injury paving the way for a larger trial [38], [39]. Our present study provides support for these clinical trials.

Back to Article Outline

Acknowledgements 

We thank Lisa Cardamone for collection of the placentas, Amy Shields for assistance with preparation of the manuscript and aspects of the analysis and Drs Prithi Bathal and Theonia Boyd for histopathology advice. The study was supported by the National Health and Medical Research Council of Australia; this body had no involvement in the collection, analysis and interpretation of the data or writing of the manuscript.

Back to Article Outline

References 

1. Dammann O, Leviton A. Infection remote from the brain, neonatal white matter damage, and cerebral palsy in the preterm infant. Semin Pediatr Neurol. 1998;5:190–201
   • View In Article
   • Abstract
   • Full-Text PDF (1266 KB)
   • CrossRef
2. Dammann O, Kuban KC, Leviton A. Perinatal infection, fetal inflammatory response, white matter damage, and cognitive limitations in children born preterm. Ment Retard Dev Disabil Res Rev. 2002;8:46–50
   • View In Article
   • MEDLINE
   • CrossRef
3. Duncan JR, Cock ML, Scheerlinck JP, Westcott KT, McLean C, Harding R, et al. White matter injury after repeated endotoxin exposure in the preterm ovine fetus. Pediatr Res. 2002;52:941–949
   • View In Article
   • MEDLINE
   • CrossRef
4. Dalitz P, Harding R, Rees SM, Cock ML. Prolonged reductions in placental blood flow and cerebral oxygen delivery in preterm fetal sheep exposed to endotoxin: possible factors in white matter injury after acute infection. J Soc Gynecol Investig. 2003;10:283–290
   • View In Article
   • MEDLINE
   • CrossRef
5. Duncan JR, Cock ML, Rees S, Harding R. The effects of repeated endotoxin exposure on placental structure in sheep. Placenta. 2003;24:786–789
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (451 KB)
   • CrossRef
6. Duncan JR, Cock ML, Suzuki K, Scheerlinck JP, Harding R, Rees SM. Chronic endotoxin exposure causes brain injury in the ovine fetus in the absence of hypoxemia. J Soc Gynecol Investig. 2006;13:87–96
   • View In Article
   • CrossRef
7. Kumazaki K, Nakayama M, Sumida Y, Ozono K, Mushiake S, Suehara N, et al. Placental features in preterm infants with periventricular leukomalacia. Pediatrics. 2002;109:650–655
   • View In Article
   • CrossRef
8. Gunn AJ, Wyatt JS, Whitelaw A, Barks J, Azzopardi D, Ballard R, et al. Therapeutic hypothermia changes the prognostic value of clinical evaluation of neonatal encephalopathy. J Pediatr. 2008;152:55–58
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (186 KB)
   • CrossRef
9. Gluckman PD, Wyatt JS, Azzopardi D, Ballard R, Edwards AD, Ferriero DM, et al. Selective head cooling with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised trial. Lancet. 2005;365:663–670
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (120 KB)
   • CrossRef
10. Redline RW, Wilson-Costello D, Borawski E, Fanaroff AA, Hack M. Placental lesions associated with neurologic impairment and cerebral palsy in very low-birth-weight infants. Arch Pathol Lab Med. 1998;122:1091–1098
    • View In Article
    • MEDLINE
11. Rees S, Hale N, De Matteo R, Cardamone L, Tolcos M, Probyn M, et al., Erythropoietin is neuroprotective in a preterm ovine model of endotoxin-induced brain damage. J Neuropathol Exp Neurol, in press.
    • View In Article
12. Gorio A, Gokmen N, Erbayraktar S, Yilmaz O, Madaschi L, Cichetti C, et al. Recombinant human erythropoietin counteracts secondary injury and markedly enhances neurological recovery from experimental spinal cord trauma. Proc Natl Acad Sci U S A. 2002;99:9450–9455
    • View In Article
    • MEDLINE
    • CrossRef
13. Siren AL, Fratelli M, Brines M, Goemans C, Casagrande S, Lewczuk P, et al. Erythropoietin prevents neuronal apoptosis after cerebral ischemia and metabolic stress. Proc Natl Acad Sci U S A. 2001;98:4044–4049
    • View In Article
    • MEDLINE
    • CrossRef
14. Genc S, Kuralay F, Genc K, Akhisaroglu M, Fadiloglu S, Yorukoglu K, et al. Erythropoietin exerts neuroprotection in 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-treated C57/BL mice via increasing nitric oxide production. Neurosci Lett. 2001;298:139–141
    • View In Article
    • MEDLINE
    • CrossRef
15. Kim MJ, Bogic L, Cheung CY, Brace RA. Placental expression of erythropoietin mRNA, protein and receptor in sheep. Placenta. 2001;22:484–489
    • View In Article
    • Abstract
    • Full-Text PDF (2598 KB)
    • CrossRef
16. Davis LE, Widness JA, Brace RA. Renal and placental secretion of erythropoietin during anemia or hypoxia in the ovine fetus. Am J Obstet Gynecol. 2003;189:1764–1770
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (145 KB)
    • CrossRef
17. Rees S, Loeliger M, Munro K, Shields A, Dalitz P, Dieni S, et al. Cerebellar development in a baboon model of preterm delivery: impact of specific ventilatory regimes. J Neuropathol Exp Neurol. 2009;68:605–615
    • View In Article
    • CrossRef
18. Rungby J, Kassem M, Eriksen EF, Danscher G. The von Kossa reaction for calcium deposits: silver lactate staining increases sensitivity and reduces background. Histochem J. 1993;25:446–451
    • View In Article
    • MEDLINE
    • CrossRef
19. Demers EJ, McPherson RJ, Juul SE. Erythropoietin protects dopaminergic neurons and improves neurobehavioral outcomes in juvenile rats after neonatal hypoxia-ischemia. Pediatr Res. 2005;58:297–301
    • View In Article
    • MEDLINE
    • CrossRef
20. Sola A, Wen TC, Hamrick SE, Ferriero DM. Potential for protection and repair following injury to the developing brain: a role for erythropoietin?. Pediatr Res. 2005;57:110R–117R
    • View In Article
    • MEDLINE
21. Asagiri K, Nakatsuka M, Konishi H, Noguchi S, Takata M, Habara T, et al. Involvement of peroxynitrite in LPS-induced apoptosis of trophoblasts. J Obstet Gynaecol Res. 2003;29:49–55
    • View In Article
22. Garnier Y, Kadyrov M, Gantert M, Einig A, Rath W, Huppertz B. Proliferative responses in the placenta after endotoxin exposure in preterm fetal sheep. Eur J Obstet Gynecol Reprod Biol. 2008;138:152–157
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (394 KB)
    • CrossRef
23. Martin RW, Scott JR, Stika CS, Hill WC, Fortunato SJ. Inflammatory cytokine (interleukins 1, 6, and 8 and tumor necrosis factor-alpha) release from cultured human fetal membranes in response to endotoxic lipopolysaccharide mirrors amniotic fluid concentrations - discussion. Am J Obstet Gynecol. 1996;174:1861–1862
    • View In Article
24. Okada T, Matsuzaki N, Sawai K, Nobunaga T, Shimoya K, Suzuki K, et al. Chorioamnionitis reduces placental endocrine functions: the role of bacterial lipopolysaccharide and superoxide anion. J Endocrinol. 1997;155:401–410
    • View In Article
    • MEDLINE
    • CrossRef
25. Kumazaki K, Nakayama M, Yanagihara I, Suehara N, Wada Y. Immunohistochemical distribution of toll-like receptor 4 in term and preterm human placentas from normal and complicated pregnancy including chorioamnionitis. Hum Pathol. 2004;35:47–54
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (795 KB)
    • CrossRef
26. Briscoe T, Duncan J, Cock M, Choo J, Rice G, Harding R, et al. Activation of NF-kappaB transcription factor in the preterm ovine brain and placenta after acute LPS exposure. J Neurosci Res. 2006;83:567–574
    • View In Article
    • MEDLINE
    • CrossRef
27. Malek A, Sager R, Schneider H. Effect of hypoxia, oxidative stress and lipopolysaccharides on the release of prostaglandins and cytokines from human term placental explants. Placenta. 2001;22(Suppl. A):S45–S50
    • View In Article
    • Abstract
    • Full-Text PDF (122 KB)
    • CrossRef
28. Garnier Y, Coumans A, Berger R, Jensen A, Hasaart TH. Endotoxemia severely affects circulation during normoxia and asphyxia in immature fetal sheep. J Soc Gynecol Investig. 2001;8:134–142
    • View In Article
29. Josephs MD, Bahjat FR, Fukuzuka K, Ksontini R, Solorzano CC, Edwards CK, et al. Lipopolysaccharide and D-galactosamine-induced hepatic injury is mediated by TNF-alpha and not by Fas ligand. Am J Physiol Regul Integr Comp Physiol. 2000;278:RR201–R1196
    • View In Article
30. Tarcan A, Tiker F, Guvenir H, Gurakan B. Hepatic involvement in perinatal asphyxia. J Matern Fetal Neonatal Med. 2007;20:407–410
    • View In Article
    • CrossRef
31. Cheung CY, Bogic L, Gagnon R, Harding R, Brace RA. Morphologic alterations in ovine placenta and fetal liver following induced severe placental insufficiency. J Soc Gynecol Investig. 2004;11:521–528
    • View In Article
    • MEDLINE
    • CrossRef
32. DeLeve LD. Hepatic microvasculature in liver injury. Semin Liver Dis. 2007;27:390–400
    • View In Article
    • CrossRef
33. Pober JS, Sessa WC. Evolving functions of endothelial cells in inflammation. Nat Rev Immunol. 2007;7:803–815
    • View In Article
34. Miranda RN, Omurtag K, Castellani WJ, Casas L, Quintanilla NM, Kaabipour E. Myelopoiesis in the liver of stillborns with evidence of intrauterine infection. Arch Pathol Lab Med. 2006;130:1786–1791
    • View In Article
35. Ribatti D, Presta M, Vacca A, Ria R, Giuliani R, Dell'Era P, et al. Human erythropoietin induces a pro-angiogenic phenotype in cultured endothelial cells and stimulates neovascularization in vivo. Blood. 1999;93:2627–2636
    • View In Article
    • MEDLINE
36. Martinez-Estrada OM, Rodriguez-Millan E, Gonzalez-de Vicente E, Reina M, Vilaro S, Fabre M. Erythropoietin protects the in vitro blood-brain barrier against VEGF-induced permeability. Eur J Neurosci. 2003;18:2538–2544
    • View In Article
    • MEDLINE
    • CrossRef
37. Koroglu TF, Yilmaz O, Ozer E, Baskin H, Gokmen N, Kumral A, et al. Erythropoietin attenuates lipopolysaccharide-induced splenic and thymic apoptosis in rats. Physiol Res. 2006;55:309–316
    • View In Article
    • MEDLINE
38. Fauchere JC, Dame C, Vonthein R, Koller B, Arri S, Wolf M, et al. An approach to using recombinant erythropoietin for neuroprotection in very preterm infants. Pediatrics. 2008;122:375–382
    • View In Article
    • CrossRef
39. Juul SE, McPherson RJ, Bauer LA, Ledbetter KJ, Gleason CA, Mayock DE. A phase I/II trial of high-dose erythropoietin in extremely low birth weight infants: pharmacokinetics and safety. Pediatrics. 2008;122:383–391
    • View In Article
    • CrossRef