Iodine Deficiency Exacerbates Thyroidal and Neurological Effects of Developmental Perchlorate Exposure in the Neonatal and Adult Rat

Abstract

Thyroid hormones (THs) require iodine for biosynthesis and play critical roles in brain development. Perchlorate is an environmental contaminant that reduces serum THs by blocking the uptake of iodine from the blood to the thyroid gland. Using a pregnant rodent model, we examined the impact of maternal exposure to perchlorate under conditions of dietary iodine deficiency (ID) on the brain and behavior of offspring. We observed modest reductions in thyroxine (T4) in the serum of dams and no effect on T4 in pup serum in response to maternal exposure to 300 ppm of perchlorate in the drinking water. Likewise, serum T4 was reduced in ID dams, but, as with perchlorate, no effects were evident in the pup. However, when ID was coupled with perchlorate, reductions in pup serum THs and transcriptional alterations in the thyroid gland and pup brain were detected. These observations were accompanied by reductions in the number of cortical inhibitory interneurons containing the calcium-binding protein parvalbumin (Pvalb). Alterations in Pvalb expression in the neonatal brain were associated with deficits in the prepulse inhibition of acoustic startle in adult male offspring and enhanced fear conditioning in females. These findings support and extend structural defects in the brain previously reported in this model. Further, they underscore the critical need to consider additional non-chemical stressors in the determination of hazards and risks posed by environmental contaminants that affect the thyroid system.

Keywords: brain; development; iodine deficiency; neurotoxicity; perchlorate; thyroid hormone.

Figure 1

Body weights. (A) Mean (+/−SE) body weight did not differ across treatment conditions in pregnant and lactating dams (n = 10–13 dams/treatment). (B) Treatment-related body weight increases were seen in both male (B) and female (C) pups from the Con-ClO₄ relative to the Con-Con group at these young ages. * Dunnett’s t-test, significantly different from Con-Con using α = 0.05. B-Birth.

Figure 2

Serum hormones before and after pregnancy. (A) Mean (+/−SE) serum T4 and (B) T3 in control and ID female rats prior to breeding. (C) Dam serum T4 was reduced in all treatment groups relative to Con-Con in late gestation and at weaning of pups on PN21. (D) Dam T3 did not differ among groups in late gestation but significant differences in T3 were present on PN21. * Dunnett’s t-test, significantly different from Con-Con using α = 0.05; # Dunnett’s t-test, significantly different from Con-ClO₄ using α = 0.05. Sample size range 8–16/treatment group at pre-breeding and 12–15/treatment group at GD20 and PN21.

Figure 3

Pup serum hormones. Mean (+/−SE) serum T4 (A) and T3 (B) in pups and TSH (C) in pups and dams. Reductions in serum T4 were present at all ages and were limited to the ID-ClO₄ relative to the Con-Con group. Reductions in serum T3 were also evident in this treatment group on PN2 and PN6 but were not significant on PN14. (C) Pup serum TSH was increased in the ID-CLO₄ group relative to the Con-Con group at all ages tested and in dams on PN21. On PN0, TSH was significantly elevated above that of the Con-Con group in both perchlorate-exposed groups. * Dunnett’s t-test, significantly different from Con-Con using α = 0.05. Sample size for T3 and T4 ranged from 7 to 14 litters on PN0 and 10 to 15 at older ages. Sample size for TSH was 5–12 litters on PN0 and 9–15 at older ages and in dams.

Figure 4

Thyroid glands. (A) Mean (+/−SE) thyroid gland weight increased significantly in dams and pups. (B) Nis was upregulated in the thyroid glands of pups at all ages, with the most consistent increases across age seen in the ID-ClO₄ group. (C–E) Thyroid gland transcripts were differentially affected by treatment and age, with a prominent pattern of downregulation in both perchlorate-exposed groups. (F) Gene expression changes in dams on PN21 were limited to the downregulation of Pax8 in both perchlorate-exposed groups and of TshR in all treatment groups relative to the Con-Con group. * Dunnett’s t-test, significantly different from Con-Con using α = 0. 05. Dotted line indicates FC = 1. Sample sizes ranged from 4 to 7/treatment on PN0 and 6 to 8 for older-aged pups and dams.

Figure 5

Dam and pup perchlorate in serum and milk. (A) Mean (±SEM) serum perchlorate in dams during late gestation and at the end of lactation and (B) pups over the first two weeks of life. (C) Perchlorate was present in the milk bands of pups on PN2 and at higher concentrations than in the serum at the same age. Sample sizes per treatment group were 6 for serum and 5–7 for milk band.

Figure 6

Thyroid hormone action in brain. Mean (±SE) relative expression of TH⁻ responsive genes in the brains of rat pups on PN14. A predominant pattern of downregulation was evident, largely limited to the ID-ClO₄ treatment condition. Relative expression was significantly lower in several transcripts in the cortex (A), hippocampus (B), and cerebellum (C). * Dunnett’s t-test, significantly different from Con-Con using α = 0.05. Sample size varied from n = 7–12/treatment condition for cortex and hippocampus and n = 4–6 for cerebellum.

Figure 7

Immunoreactivity for Pvalb⁺ neurons in pup neocortex. Fewer Pvalb⁺ cells were present in three cortical regions of ID-ClO₄-exposed pups. (A) Representative Pvalb immuno reactivity images from one animal from each treatment condition for (A) cingulate, CC; (B) somatosensory, SS; and (C) auditory, Aux cortex. (B) Regions in the anterior and posterior cortex selected for cell counts are indicated by dotted ovals on images on the far left. High-magnification images of Pvalb+ cells are shown on the far right for the area indicated by the bounding box in the sections of the middle panel. (D) Group mean (+/−SE) cell counts in each treatment condition are presented as mean per litter, collapsed across sex. Mean contrasts supported a significant reduction in Pvalb⁺ cell numbers that was restricted to the ID-ClO₄ group in all three cortical regions. Sample size n = 10–12 litters/treatment condition. * Dunnett’s t-test, significantly different from Con-Con using α = 0.05.

Figure 8

Immunostaining for the GABA synthesis enzyme, GAD-67, was conducted on a series of sections from Con-Con and ID-ClO₄ (n = 5/treatment condition). No difference was detected in GAD-67 immunostaining in the (A) cingulate or (B) somatosensory cortex based on qualitative assessments by 4 independent observers blind to the treatment condition.

Figure 9

Trace fear conditioning was altered in females (left panels) but not males (right panels). Trace fear conditioning was altered in female (left panels) but not male (right panels) offspring. (A) Training Phase: Mean activity counts in 15-s bins during trace fear acquisition training did not differ among treatment conditions. All groups learned equally well. (B) Context Learning: In a test of conditioning to context 24 h after training, the suppression of activity relative to training baseline was seen in females (left) of the ID-ClO₄ group, while male activity levels (right) were unchanged relative to Con-Con. (C) Cue Learning: A similar but less robust pattern was seen in cue learning in females (E). No significant difference in cue learning in females or cue or context learning was observed in males (D,F). Litter was used as a random variable and sample sizes were 17, 18, 26, and 19 for female offspring and 15, 20, 12, and 17 for male offspring in groups Con-Con, ID-Con, Con-ClO₄, and ID-ClO₄, respectively. * Contrast relative to Con-Con significant using α = 0.05, Bonferroni-adjusted for multiple comparisons.

Figure 10

Acoustic startle response (ASR) and prepulse inhibition (PPI). Mean (+/−SE) baseline ASR was increased in (A) adult female offspring from the Con-ClO₄ treatment condition and (B) male offspring from both the Con-ClO₄ and the ID-ClO₄ conditions. (C) Prepulse inhibition of the ASR was increased as a function of prepulse intensity, but no treatment-related differences were evident in females. (D) Adult male offspring of the ID-ClO₄ group exhibited a diminished suppression of the ASR at the 6, 10, and 18 dB levels. Litter was used as a random variable in the statistical model and sample sizes varied from 15 to 26/sex/treatment group and 10 to 13 litters/treatment group. * Contrast relative to Con-Con significant using α = 0.05, Bonferroni-adjusted for multiple comparisons.