Role of PABP4 in intrauterine growth restriction
Item statusRestricted Access
Embargo end date06/06/2024
In mammals, two members of the poly(A)-binding protein (PABP) family of RNAbinding proteins, PABP1 and PABP4, are widely expressed in a range of adult tissues. PABP1, the prototypical family member, has been intensively characterised as central regulator of both global and mRNA-specific translation and mRNA turnover in vitro and in cell lines, but almost nothing is known about the roles of these so-called “somatic PABPs” in mammalian physiology. Here, we aim to address this, building on unpublished data in our Pabp4-/- mice by investigating its surprising requirement for the establishment/maintenance of healthy pregnancy. Previous work in the lab has shown that maternal Pabp4 deletion results in pregnancies which exhibit intra-uterine growth restricted (IUGR) fetuses from e15.5, and an increased prevalence of late-gestational fetal death (akin to stillbirth). Stillbirth affects approximately 1 in 225 pregnancies in the UK, and IUGR is not only a major risk factor for stillbirth but also for prematurity, neonatal mortality, childhood neurodevelopmental disorders and adverse adult cardio-metabolic health. In the last few decades, little improvement in stillbirth rate have been realised, with progress in understanding the underlying basis being hampered by a lack of relevant animal models. As the pregnancy defects in Pabp4-/- mice are independent of fetal and paternal genotype, these mice present a powerful, unique model for studying maternally-derived IUGR and stillbirth, addressing this priority area of unmet clinical need. Here I characterise this phenotype further to determine if there is sexual dimorphism in fetal outcome, and subsequently focus on three aspects of pregnancy that are commonly associated with adverse pregnancy outcomes, including reduced fetal growth: (I) altered placental morphology/function, (II) maternal and fetal nutrient status, and (III) maternal haematology. Previous work had not considered effects of fetal sex, despite adverse pregnancy conditions (e.g. uterine stresses, nutrient deficiencies) often causing sexually dimorphic phenotypes in the offspring (e.g. differential survival rates). These differences are thought to arise from different growth strategies associated with male and female fetuses. Thus, to answer this important gap in our knowledge, I measured fetal growth and survival at late gestation (e18.5) in three crosses, wild-type female x wild-type male (WT cross), Pabp4 knock-out female x wild-type male (maternal cross) and wild-type female x Pabp4 knockout male (paternal cross, which controls for Pabp4 heterozygosity of fetuses). This analysis confirmed the presence of the previously observed phenotype, and upon stratification by sex, revealed that in PABP4-deficient dams both sexes are affected by IUGR and are equally likely to survive until late pregnancy. The placenta is an organ that is unique to pregnancy and plays critical role not only in the supply of nutrients and gases (and removing waste) but also in directing a series of maternal adaptions across a wide range of organs that are necessary to support different stages of pregnancy. Placental insufficiency is a term used in the clinic that describes a placenta that appears incapable of fully supporting growth and development. This often manifests in a change in fetal-placental weight ratio, and this was previously observed in pregnant Pabp4-/- dams. Prior work in Pabp4-/- mice also uncovered volumetric changes in placental morphology, where there was a reduction in maternally-derived decidua basalis and a reciprocal increase in the conceptusderived junctional zone by e18.5. The placenta is considered a plastic organ and it remains unclear in most cases whether size/morphology changes are causative or an attempt to adapt to an adverse pregnancy. To ascertain if these effects on the placenta could be direct, the expression of PABP4 within the placenta was examined. In lategestational placenta (e18.5) PABP4 was found to be expressed in all three main functional layers: the maternally-derived decidua basalis and conceptus-derived junctional zone and labyrinth. Within the decidua, PABP4 was expressed in decidualised stromal cells and in invading trophoblasts (conceptus-derived). It was also expressed within the vast majority of trophoblasts in the junctional zone. However, qualitative analysis with markers for the major cell types within the placenta, ruled out the simplest hypothesis: that the reduction in decidua basalis volume is due to loss of a specific cell-type (or -types), as neither decidualised stromal cells (Pabp4-/-), nor invading trophoblasts (Pabp4+/-) were visibly reduced. Marker analysis also did not reveal a gross change in junctional zone cell types that could account for its increased volume. As an alternative way of understanding what happens to the placenta in PABP4-deficient dams, transcriptomic analysis of e15.5 placentas was undertaken. IUGR is present by this time, which is prior to the majority of fetal death. The analysis revealed dysregulations in a subset of mRNAs in heterozygous placentas, and additional changes in placentas from Pabp4-/- dams. An absence of a global effect on mRNA stability, is potentially consistent with work in cell lines suggesting that PABP4 may have a more mRNA-specific role. Indeed, changes in a small number of mRNAs were present, but these failed validation when tested on an independent cohort of placentas. This lack of effect could suggest that PABP4 does not play an important role in determining mRNA stability in this tissue, or may more simply reflect PABP4 still being present in many cell-types within the placenta, as conceptus derived cells will be heterozygous. Nutrient availability is key for fetal growth. Prior work using glucose tracers indicated that this nutrient was normally transported across the placenta of fetuses from Pabp4-/- dams. However, the maternal availability of glucose in Pabp4-/- dams, fetal glucose and glycogen or placental glycogen levels had not been established. Levels of other important nutrients such as lipids were also unknown. Analysis of the fetal and maternal nutrient status revealed dysregulated fetal glucose and glycogen in fetuses from Pabp4-/- dams in the absence of maternal dysregulation. These effects were not seen in the paternal cross so were not the result of fetal heterozygosity. One explanation for reduced fetal glucose/glycogen despite apparently normal availability and transport capacity, is that glucose is being shunted into a different metabolic pathway to make up for a deficit in another nutrient. One important nutrient class is lipids (e.g. free fatty acids, triglycerides, cholesterol). In healthy fetuses, triglycerides are metabolised to provide the fatty acids and glycerol needed for essential processes such as membrane growth (via phospholipid synthesis) and cholesterol is essential for membrane fluidity and hormone synthesis. However, when glucose is not readily available lipids can substitute as energy source. Due to the interrelatedness of glucose and lipids levels, triglyceride and cholesterol levels were examined in e18.5 fetuses and placentas when IUGR and hypoglycaemia are well established. Surprisingly, fetal triglyceride levels were unaffected by either fetal heterozygosity or maternal genotype, despite fetal hypoglycaemia in Pabp4-/- dams. Equally, placental and maternal triglycerides were normal. Total cholesterol was normal in Pabp4-/- dams, but increased in their pups and placentas. The precise mechanism by which cholesterol levels are increased remains to be elucidated but its maternal genotype-dependent nature suggests a maternal or placental, rather than fetal, origin. Lastly, as maternal anaemia is associated with IUGR (up to 40 % of women at risk of anaemia during pregnancy), and PABP4 is implicated in erythrocyte maturation and haemoglobin synthesis in cell lines, a haematopoietic analysis was undertaken. Both non-pregnant and pregnant females were analysed to determine if any changes were pre-existing or pregnancy-specific. Whilst not the focus, no changes in white blood cell (WBC) populations were detected, which is informative as immune cells play a critical role in pregnancy. Interestingly, prior to pregnancy, PABP4-deficient females exhibited microcytic erythrocytes. Surprisingly, rather than being exacerbated by pregnancy, this phenotype had resolved by e18.5, however it is currently unclear at what point during pregnancy this happens. Conditional Pabp4 knock-out experiments determined that the microcytic red blood cell (RBC) phenotype is not intrinsic to haematopoietic cells, and further revealed that loss of PABP4 in haematopoietic cell lineages did not recapitulate the IUGR phenotype, suggesting that PABP4 function in WBCs or RBCs is not required for healthy pregnancy. In summary, I present data suggesting that the basis of maternal Pabp4-/- genotypedependent IUGR may be complex, as fetuses in Pabp4-/- have volumetric changes in their placentas that may be causative or compensatory, fetuses display multiple nutrient dysregulations, and Pabp4-/- adult females have a microcytic red blood cell phenotype that is ameliorated by late pregnancy. Future work will be required to address their relevance for the reproductive phenotype and delineate the dysregulated pathways. This knowledge would allow for the development of predictive markers for adverse pregnancy outcomes or future therapeutic avenues for IUGR and stillbirth.