Long-term effects of iron deficiency during pregnancy on offspring blood pressure have been demonstrated previously in several strains of rat [10, 12, 17, 18]. The long-term outcomes for the developing fetus are similar across these studies, and whilst it is assumed that these programmed attributes arise in response to identical maternal adaptations to the nutritional insult across all strains, this has never been directly assessed. This knowledge is required to improve understanding of the mechanisms of nutritional programming, which are currently thought to involve modification of fetal gene expression and resetting of organ structure, cellular function and endocrine regulation . Previous work in the RHL strain identified alterations in iron homeostasis in response to a low iron diet during pregnancy [17, 18]. This present study aimed to extend these findings to the more widely used Wistar strain of rat, and to examine any differences between the strains in terms of iron homeostasis and response to a low iron diet.
This study clearly demonstrates that the maternal and fetal livers are very sensitive to the feeding of a low iron diet prior to and during gestation. Additionally, whilst the total iron content of the placenta was unaffected by diet, a reduction in placental non-haem iron was observed in response to a low iron diet in both strains. These data are in good agreement with previous work with the RHL strain [7, 17]. The reduction in maternal liver content in response to dietary restriction was associated with changes in the expression of key regulators of iron transport. Expression of the transferrin receptor (TfR) mRNA was up-regulated in the maternal liver by more than two-fold in both strains of rat. This is interpreted as a compensatory response to increase iron uptake into the liver cells and has previously been shown to be reversed by iron supplementation at selected points during pregnancy in the RHL . Maternal hepatic hepcidin mRNA expression was vastly down-regulated in response to the feeding of a low iron diet in both strains of rat, almost to the point of being undetectable. Decreased hepcidin expression has previously been observed  and can be reversed by supplementation of iron during the second half of rat pregnancy . As hepcidin prevents iron efflux [22, 23], down-regulation may therefore contribute to meeting the greater requirements for iron during pregnancy.
Maternal ceruloplasmin mRNA expression was found to be unaffected by the feeding of a low iron diet, suggesting that transcriptional regulation is not necessarily related to the level of iron efflux from the cell. There has previously been contradictory evidence published regarding the sensitivity of ceruloplasmin to iron status [24–26]. However, in an in vivo rat model, there was no effect of iron status on plasma ceruloplasmin concentration or activity or hepatic mRNA expression . Instead, it has been suggested that oxidase activity could be altered indirectly via changes in redox state . Similarly to the maternal liver, TfR protein expression was increased in the fetal livers in response to prenatal dietary iron restriction, but only in the RHL strain. No change in expression of ferroportin was observed in either strain. However, the reduction in hepcidin mRNA expression in the fetal liver in response dietary iron restriction in both strains may modulate ferroportin internalisation via ubiquitination [28, 29].
Placental expression of the transferrin receptor was increased in response to dietary iron restriction in both strains of rat, as previously observed in the RHL . This may act to enhance iron uptake into the placenta and optimise transfer to the fetus. There was no associated increase in ferroportin protein expression, but its responsiveness to dietary iron status may instead be related to the observed reduction in fetal hepcidin mRNA expression modulating the internalisation of placental ferroportin [17, 22, 28]. In both strains of rat, placental DMT1+IRE mRNA expression was increased in response to dietary iron restriction, which would act to allow release of iron from intracellular vesicles. This is in agreement with previous work with the RHL , and suggests that adaptive mechanisms optimise the availability of transporters that allow iron to cross the placenta. However, it must be noted that work with a murine DMT1 knock-out model indicated that this transporter does not play an essential role in materno-fetal iron transfer and that another transporter must be responsible for iron uptake by fetal-derived cells .
Interestingly, there was considerable difference in liver iron content between the strains, with Wistar dams having approximately twice the quantity of iron present in their livers in comparison to RHL rats. This suggests greater promotion or efficiency of absorption and storage of iron in the maternal liver. Wistar dams also exhibited higher placental total and haem iron content in comparison to RHLs, which was associated with a greater placental weight. The opposite was observed in the fetal liver, where total iron content was significantly higher in fetuses of the RHL strain. This direct comparison of iron status between two different strains of rat highlights the care which needs to be taken when comparing findings from different experimental models and suggests that programmed phenomena, such as hypertension, do not necessarily arise from similar maternal adaptations to nutritional insults across strains or species. Whilst our research has identified common developmental processes and pathways which are affected across different strains and dietary insults during pregnancy, there may be other differences in response which impact on the progression of the programmed phenomenon . Importantly, the liver iron content of the iron restricted Wistar rat was actually higher than the liver iron content of the control RHL rat, which demonstrates the inherent difficulty in labelling treatment groups as ‘deficient’ in comparison to an internal control in the absence of clinical thresholds.
The data suggests greater maternal to fetal iron transfer in the RHL rats, despite a reduction in placental weight in response to low iron diet in this strain only. This is consistent with the increased expression of placental iron transporters in this strain. Placental ferroportin protein and DMT1+IRE expression were significantly higher in rats of the RHL strain compared to the Wistar strain, which may account for a greater transfer of iron to the fetus at the expense of maternal liver iron stores. RHL rats also tended to have lower maternal liver TfR expression, although this marginally failed to reach statistical significance. Hepcidin mRNA expression was shown to be significantly lower in RHL fetal livers compared to Wistars, which may permit more ferroportin to be present at the membrane to facilitate increased transport of iron to other tissues. Unfortunately direct comparisons of the fetal liver iron transporters could not be made between strains, as expression of the normalising protein (tubulin) differed between strains, perhaps indicating structural (cytoskeletal) differences between liver cells of the two strains of rat. However, the lack of response of TfR protein expression to dietary iron restriction in Wistar fetal livers may explain the lower fetal liver iron content in this strain.
It has been previously shown that 72% of iron in the fetal liver is derived from maternal iron stores, and that the remaining 28% is from maternal dietary absorption . We therefore went on to consider whether the differences in maternal iron status between the strains could be due to differences in absorption of iron from the diet in the duodenum. Iron absorption is known to be enhanced when the body is iron deficient , as indicated by increased expression of DMT1 in the duodenum . However, no significant differences in duodenal DMT1+IRE, ferroportin or hephaestin expression were observed between the two strains of rat. The differences in maternal liver iron content cannot, therefore, be explained through differences in the expression of intestinal iron transporters, although a differential response to pregnancy cannot be excluded at this point. It is concluded that the differences in maternal liver iron content observed between the two strains are most likely a result of differences in the efficiency of placental transfer of iron to the fetus, as indicated by the greater fetal hepatic iron content and increased placental expression of iron export molecules observed in the RHL strain. Despite this, birth weight was reduced in response to iron restriction in the RHL strain only, indicating that the adaptive responses protected fetal iron status but not birth weight.