Perinatal iron and copper deficiencies alter neonatal rat circulating and brain thyroid hormone concentrations

Abstract

Copper (Cu), iron (Fe), and iodine/thyroid hormone (TH) deficiencies lead to similar defects in late brain development, suggesting that these micronutrient deficiencies share a common mechanism contributing to the observed derangements. Previous studies in rodents (postweanling and adult) and humans (adolescent and adult) indicate that Cu and Fe deficiencies affect the hypothalamic-pituitary-thyroid axis, leading to altered TH status. Importantly, however, relationships between Fe and Cu deficiencies and thyroidal status have not been assessed in the most vulnerable population, the developing fetus/neonate. We hypothesized that Cu and Fe deficiencies reduce circulating and brain TH levels during development, contributing to the defects in brain development associated with these deficiencies. To test this hypothesis, pregnant rat dams were rendered Cu deficient (CuD), FeD, or TH deficient from early gestation through weaning. Serum thyroxine (T(4)) and triiodothyronine (T(3)), and brain T(3) levels, were subsequently measured in postnatal d 12 (P12) pups. Cu deficiency reduced serum total T(3) by 48%, serum total T(4) by 21%, and whole-brain T(3) by 10% at P12. Fe deficiency reduced serum total T(3) by 43%, serum total T(4) by 67%, and whole-brain T(3) by 25% at P12. Brain mRNA analysis revealed that expression of several TH-responsive genes were altered in CuD or FeD neonates, suggesting that reduced TH concentrations were sensed by the FeD and CuD neonatal brain. These results indicate that at least some of the brain defects associated with neonatal Fe and Cu deficiencies are mediated through reductions in circulating and brain TH levels.

Figure 1

Cu and Fe content of P12 pup brains and livers. A, Liver Cu content. B, Liver Fe content. C, Brain Cu content. D, Brain Fe content. Cu and Fe content was measured in livers and half brains from P12 male pups (n = 5 for control, FeD, and PTU; n = 4 for CuD; for liver Fe measurement n = 4 for PTU) by flame AAS. Data are presented as the mean ± sem. Bars with unlike letters indicate a statistical difference (P < 0.05) by one-way ANOVA and Tukey’s or Scheffé’s multiple comparison test.

Figure 2

Perinatal Cu and Fe deficiencies lead to altered serum and brain TH levels. A, Serum total T₄. B, Serum total T₃. Serum was harvested from P12 male pups (n = 10 for control, FeD, and PTU; n = 8 for CuD). C, Brain total T₃. Half brains were harvested from P12 male pups (n = 5 for control, FeD, and PTU; n = 4 for CuD). Data are presented as the mean ± sem. Bars with unlike letters indicate a statistical difference (P < 0.05) by one-way ANOVA and Tukey’s or Scheffé’s multiple comparison test. Similar results were obtained from a second independent study (data not shown).

Figure 3

Perinatal Cu or Fe deficiencies alter P12 brain proteins. A, Brain CCO activity. Half brains were harvested from P12 male pups (n = 5 for control, FeD, and PTU; n = 4 for CuD). B, Brain CCO mRNA analysis. Half brains were harvested from P12 male pups (n = 5 for control, FeD, and PTU; n = 4 for CuD). Quantitative RT-PCR was performed for Cox IV. C, Brain COX IV, CCS, and TfR protein analysis. Western blot analyses were performed on brain homogenates for the indicated proteins. LDH protein levels were measured as a loading control for both the 15 and 8% gels. The LDH blot for the 15% gel is shown. A similar result was obtained for LDH with the 8% gel. D, Quantification of densities of COX IV, CCS, and TfR immunoblot bands in C relative to the control mean. Data are presented as the mean ± sem. Bars with unlike letters indicate a statistical difference (P < 0.05) by one-way ANOVA and Tukey’s or Scheffé’s multiple comparison test.

Figure 4

The effect of perinatal Fe and Cu deficiency on TH-responsive brain gene expression. Half brains were harvested from P12 male pups (n = 5 for control, FeD, and PTU; n = 4 for CuD), total RNA was extracted, and cDNA was synthesized from 2 μg of RNA. Quantitative RT-PCR was performed for several TH-responsive genes. A, Dio2. B, Pvalb. C, Mbp. D, Mobp. E, Bdnf IV. F, Hr. G, Egr1. H, Rc3. I, Fth1; qRT-PCR was also performed on one housekeeping gene. J, Gapdh. Relative mRNA levels are calculated relative to an internal control sample. Data are presented as the mean ± sem. Bars with unlike letters indicate a statistical difference (P < 0.05) by one-way ANOVA and Tukey’s or Scheffé’s multiple comparison test.