Long-term modulations in the vertebral transcriptome of adolescent-stage rats exposed to binge alcohol

Abstract

Aims: Dangerous alcohol consumption practices are common in adolescents, yet little is known about their consequences on attainment of peak bone mass and long-term skeletal integrity. We previously demonstrated that binge alcohol-exposed adolescent rats showed site-specific reductions in accruement of bone mineral density and bone strength, which were incompletely recovered following prolonged alcohol abstinence. Currently, we analysed the vertebral transcriptome of adolescent rats following alcohol treatment and abstinence to identify long-term molecular changes in the lumbar spine.

Methods: Sixty male adolescent Sprague-Dawley rats were assigned to one of six treatment groups receiving binge alcohol (3 g/kg) or saline i.p., 3 consecutive days (acute binge), 4 consecutive weekly (3-day) binge cycles (chronic binge) or 4 weekly binge cycles followed by a 30-day abstinence period (chronic binge with abstinence). Following treatment, lumbar vertebrae were assayed for global transcriptional changes using gene array technology.

Results: Analysis of the adolescent rat vertebral transcriptome identified clusters of binge alcohol-sensitive genes displaying differential expression patterns starting before bone damage was seen and persisting after alcohol treatment was discontinued. Functional grouping of these gene clusters identified candidate cellular pathways affected following acute and chronic binge treatment, as well as pathways remaining modulated following abstinence.

Conclusions: These results demonstrate that binge alcohol exposure can produce disruptions of normal bone gene expression patterns in the adolescent rat that persist well beyond the period of active intoxication. This data may have relevance to peak bone mass attainment and future risk of skeletal disease in adolescents engaging in repeated binge-drinking episodes.

Fig. 1

Heat map: depicting expression of genes significantly altered by acute binge alcohol. Normalized gene expression is shown by colour according to the scale for genes that were P < 0.05 and at least 1.5-fold different in alcohol samples when compared to controls (permutative Wilcoxon–Mann–Whitney non-parametric analysis with Benjamini and Hochberg false discovery rate multiple testing correction). Samples and genes were both clustered in hierarchical fashion according to Euclidean distance metrics and the average linkage distance rule. Outlier samples were removed from heat map.

Fig. 2

Heat map: depicting expression of genes significantly altered by chronic binge alcohol. Normalized gene expression is shown by colour according to the scale for genes that were P < 0.05 and at least 1.5-fold different in alcohol samples when compared to controls (permutative Wilcoxon–Mann–Whitney non-parametric analysis with Benjamini and Hochberg false discovery rate multiple testing correction). Samples and genes were both clustered in hierarchical fashion according to Euclidean distance metrics and the average linkage distance rule.

Fig. 3

Pathways significantly affected by acute binge alcohol treatment. PANTHER was performed on a list of 300 genes significantly changed by acute alcohol. The resulting pathways were deemed significant by Chi-squared test with Bonferroni’s multiple testing correction. Data are displayed as pathways with significant over-representation of genes from within each differentially expressed list, compared to a reference list. Gene numbers represented in each pathway are given in parentheses as (Observed/Expected).

Fig. 4

Pathways significantly affected by chronic binge alcohol treatment. PANTHER was performed on a list of 180 genes significantly changed by chronic alcohol. The resulting pathways were deemed significant by Chi-squared test with Bonferroni’s multiple testing correction. Data are displayed as pathways with significant over-representation of genes from within each differentially expressed list, compared to a reference list. Gene numbers represented in each pathway are given in parentheses as (Observed/Expected).

Fig. 5

Heat map: depicting expression of genes significantly altered by chronic alcohol followed by abstinence. Normalized gene expression is shown by colour according to the scale for genes that were P < 0.05 and at least 1.5-fold different in alcohol samples when compared to controls (permutative Wilcoxon–Mann–Whitney non-parametric analysis). Samples and genes were both clustered in hierarchical fashion according to Euclidean distance metrics and the average linkage distance rule. Outlier samples were removed from heat map.

Fig. 6

Pathways significantly affected after chronic binge alcohol treatment followed by abstinence. PANTHER was performed on a list of 512 genes significantly changed following alcohol abstinence. The resulting pathways were deemed significant by Chi-squared test with Bonferroni’s multiple testing correction. Data are displayed as pathways with significant over-representation of genes from within each differentially expressed list, compared to a reference list. Gene numbers represented in each pathway are given in parentheses as (Observed/Expected).

Fig. 7

Venn diagram analysis depicting overlap between differentially expressed gene lists. Acute and chronic binge alcohol and alcohol abstinence gene lists were used. Numbers in each circle represent total gene numbers in each differential expression list. Numbers in regions where circles overlap represent the overlap between the representative gene lists.

Fig. 8

Validation of microarray data for select genes by qRT-PCR. Period homolog 1 (Per1), period homolog 2 (Per2), aryl hydrocarbon receptor nuclear translocator (Arntl), nuclear receptor subfamily 1 group D1 (Nr1d1), lymphoid enhancer factor 1 (LEF1), bone morphogenetic factor 2 (Bmp-2) and osteocalcin (Bglap). For each gene, both the microarray and PCR data are shown relative to the control group. n = 5/group. Data are shown as the mean ± SEM. *P < 0.05 for qRT-PCR data when compared to control by one-way ANOVA. #P < 0.05 for microarray data when compared to control by Wilcoxon–Mann–Whitney non-parametric analysis followed by Benjamini and Hochberg multiple testing correction.