Maternal cocaine administration in mice alters DNA methylation and gene expression in hippocampal neurons of neonatal and prepubertal offspring

Abstract

Previous studies documented significant behavioral changes in the offspring of cocaine-exposed mothers. We now explore the hypothesis that maternal cocaine exposure could alter the fetal epigenetic machinery sufficiently to cause lasting neurochemical and functional changes in the offspring. Pregnant CD1 mice were administered either saline or 20 mg/kg cocaine twice daily on gestational days 8-19. Male pups from each of ten litters of the cocaine and control groups were analyzed at 3 (P3) or 30 (P30) days postnatum. Global DNA methylation, methylated DNA immunoprecipitation followed by CGI(2) microarray profiling and bisulfite sequencing, as well as quantitative real-time RT-PCR gene expression analysis, were evaluated in hippocampal pyramidal neurons excised by laser capture microdissection. Following maternal cocaine exposure, global DNA methylation was significantly decreased at P3 and increased at P30. Among the 492 CGIs whose methylation was significantly altered by cocaine at P3, 34% were hypermethylated while 66% were hypomethylated. Several of these CGIs contained promoter regions for genes implicated in crucial cellular functions. Endogenous expression of selected genes linked to the abnormally methylated CGIs was correspondingly decreased or increased by as much as 4-19-fold. By P30, some of the cocaine-associated effects at P3 endured, reversed to opposite directions, or disappeared. Further, additional sets of abnormally methylated targets emerged at P30 that were not observed at P3. Taken together, these observations indicate that maternal cocaine exposure during the second and third trimesters of gestation could produce potentially profound structural and functional modifications in the epigenomic programs of neonatal and prepubertal mice.

Figure 1. Evaluation of the structure and total DNA content of hippocampal pyramidal tissues in P3 progeny of saline-treated and cocaine-treated mice.

(a) Typical micrograph of HistoGene-stained coronal section through the hippocampus of a P3 Saline pup. (b) Micrograph of a comparable section from an age-matched offspring of a cocaine-treated mother. (c) Relative amounts of total DNA content between the offspring of saline-treated and cocaine-treated mothers at P3 and P30. Horizontal bars show mean difference in DNA content between Cocaine and Saline groups at each age, while vertical bars show the SEM (n  = 10/group/age). For either age group, the inter-sample ranges of coefficient of error (CE) values are provided at the bottom of the chart. C1, C2, C3, hippocampal regions; DG, dentate gyrus. Note the absence of detectable histological differences between micrographs a and b.

Figure 2. Histology of hippocampal sections before and after laser microdissection.

Photomicrograph of a typical HistoGene-stained coronal section through the hippocampus of a 3-day-old mouse (a), and the same section following excision of the hippocampal pyramidal layer by microdissection on a Leica ASLMD laser capture microdissection system (b). HIPpl, hippocampal pyramidal layer; DG, dentate gyrus; SUB, subiculum. Note the precision of the cut.

Figure 3. Relationship between the pixel density of MeDIP-generated signals and the number of methylated CpGs present in test DNA fragments of similar size (449, 446, 440, 445, and 449 bp) but containing different numbers of methylated CpGs (0, 11, 25, 36, or 45, respectively).

Each data point represents the mean±SEM values for 3 clones. The linear regression line was fitted using GraphPad Prism software. Note the linearity of the plotted relationship, indicating that the MeDIP assay can quantitatively detect the number of methylated CpGs within a DNA fragment.

Figure 4. Global DNA methylation and DNMT expression levels in the hippocampal pyramidal layer of P3 offspring from saline-injected or cocaine-treated dams.

(a) Global levels of cytosine methylation in DNA samples of control and cocaine groups each expressed as percentage relative to the methylation signal in Epigentek's methylated DNA standard. (b) Expression levels of DNA methyltransferases DNMT1, DNMT3a, and DNMT3b each calculated as a ratio relative to GAPDH as internal standard. Similar results were obtained with ß-actin used as internal standard. Each column represents mean±SEM of 10 samples. *p<0.05 compared to the saline-control group.

Figure 5. Genomic characteristics of abnormally methylated CGIs in the hippocampal pyramidal layer of P3 cocaine-treatment pups.

Information associating abnormally methylated CGIs with various genomic elements was obtained by BLAT analysis of the affected CGIs as determined from the MeDIP/CGI array profiling experiments. The data is limited to the genomic elements that are present in at least 5 CGIs. The number of CGIs associated with a particular genomic element is represented by horizontal bars across from the element's name. The length of the bar is proportional to the number of CGIs; this number is also given at the end of the row. The number of CGIs with sequences matching more than one position in the genome is provided in parenthesis. The tendency for CGIs to be associated with more than one genomic element is reflected in the partial overlap of the bars on most rows. Note that overall there are more hypomethylated CGIs than hypermethylated CGIs.

Figure 6. CpG methylograms of selected abnormally methylated promoters in P3 mice.

From the MeDIP/CGI array-generated methylation profiles, ten cocaine-induced abnormally methylated CGIs known to be associated with gene promoter regions were selected and cloned, followed by bisulfite sequencing to identify the relative locations of methylated CpGs in the gene promoter sequence. For each chart, the name of the promoter is shown on the left vertical while the horizontal numbers mark the CpGs on the sequence; the CpGs are numbered from the 5'end of the clone. The vertical numbers on the right (1–5) indicate data rows for samples from each of 5 animals from the control (upper, DRN) and cocaine (lower, COC) groups. For each sample, three clones were sequenced, hence the three sub-rows available to each sample for methylation mapping. For each sub-row representing a clone, white squares indicate unmethylated CpGs while black-filled squares indicate methylated CpGs. If all three clones were consistent, the filled boxes would appear for each sub-row for that sample. If all 5 samples (mice) were consistent, then the filled boxes would extend vertically across all 5 rows. The relative density of filled boxes between the control (upper 5) samples and the cocaine (lower 5) samples provides a visual feedback on the relative levels of CpG methylation in these DNA regions. For example, clones of DYRK3 or COQ7 promoters contained lower proportions of methylated CpGs in the cocaine-treatment pups than in the saline-control offspring, whereas the converse applies in the case of GPR73 or MAPK1 clones.

Figure 7. Effects of maternal cocaine exposure on global DNA methylation and DNMT gene expression in 30-day-old pups.

(a) Relative levels of global DNA methylation in hippocampal pyramidal tissues of P30 pups. Each bar represents the mean±SEM (n = 10). *p<0.05 compared to saline-control group. (b) Expression levels of DNA methyltransferases relative to GAPDH as internal standard. *p<0.05 compared to the respective saline-control group.

Figure 8. Effects of maternal cocaine exposure on CGI DNA methylation in P30 compared to P3 pups.

(a) Cocaine-induced hypermethylated and hypomethylated CGIs in P30 compared to P3 pups. Shown are the numbers of CGIs with altered methylation levels in cocaine-treated animals relative to saline-controls at each age. For P30 data, the number of CGIs with altered methylation levels that did not show a change at P3 (new) are indicated by a cross-hatched segment of the bar. Also indicated by differential shading are CGIs that were significantly altered at both P3 and P30, retaining or reversing their direction of change at P30 compared to P3 (retained direction; changed direction). (b) Hypermethylated and hypomethylated promoter-associated CGIs at P3 and P30. For both ages, CGIs associated with promoters of neural tissue-specific genes (neural genes) are represented by white bars, while CGIs associated with promoters of non-neural tissue-specific genes (housekeeping genes) are represented by black bars. In addition, for P30, the number of CGIs with altered methylation state undetectable at P3 (new) and the number of CGIs corresponding to the CGIs affected at P3, which either retained or reversed the direction of change (changed direction/retained direction), are indicated as in a. (c) Abnormally methylated repetitive element-associated CGIs at P3 and P30.