Bioinformatics analyses reveal age-specific neuroimmune modulation as a target for treatment of high ethanol drinking

Abstract

Background: Use of in silico bioinformatics analyses has led to important leads in the complex nature of alcoholism at the genomic, epigenomic, and proteomic level, but has not previously been successfully translated to the development of effective pharmacotherapies. In this study, a bioinformatics approach led to the discovery of neuroimmune pathways as an age-specific druggable target. Minocycline, a neuroimmune modulator, reduced high ethanol (EtOH) drinking in adult, but not adolescent, mice as predicted a priori.

Methods: Age and sex-divergent effects in alcohol consumption were quantified in FVB/NJ × C57BL/6J F1 mice given access to 20% alcohol using a 4 h/d, 4-day drinking-in-dark (DID) paradigm. In silico bioinformatics pathway overrepresentation analysis for age-specific effects of alcohol in brain was performed using gene expression data collected in control and DID-treated, adolescent and adult, male mice. Minocycline (50 mg/kg i.p., once daily) or saline alone was tested for an effect on EtOH intake in the F1 and C57BL/6J (B6) mice across both age and gender groups. Effects of minocycline on the pharmacokinetic properties of alcohol were evaluated by comparing the rates of EtOH elimination between the saline- and minocycline-treated F1 and B6 mice.

Results: Age and gender differences in DID consumption were identified. Only males showed a clear developmental increase difference in drinking over time. In silico analyses revealed neuroimmune-related pathways as significantly overrepresented in adult, but not in adolescent, male mice. As predicted, minocycline treatment reduced drinking in adult, but not adolescent, mice. The age effect was present for both genders, and in both the F1 and B6 mice. Minocycline had no effect on the pharmacokinetic elimination of EtOH.

Conclusions: Our results are a proof of concept that bioinformatics analysis of brain gene expression can lead to the generation of new hypotheses and a positive translational outcome for individualized pharmacotherapeutic treatment of high alcohol consumption.

Keywords: Alcoholism; Bioinformatics; Drinking-in-Dark; FVB/NJ × C57BL/6J F1 Mouse; Medications Development; Minocycline.

Figure 1

Age influenced male, but not female, DID alcohol consumption. Ethanol consumption is shown for F1 mice over four days using the DID paradigm for A. males and B. females. The bars represent average ethanol intake in g/kg/day over the four days for adolescent (P30, light textured; n = 18) and adult (P70, dark textured; n = 18) mice as mean ± S.E.M. Two-way ANOVA revealed a significant overall difference in drinking between the adolescent and adult male mice (F (1,34) = 8.139, p < 0.05) but not in the female mice (F (1, 13) = 1.854, p = ns; n = 7 (P30), n = 8 (P70)). Tukey's post-hoc analysis revealed a significant day effect between adolescent and adult male mice for days 3 and 4 (*p < 0.05).

Figure 2

Pathway analysis of alcohol-induced gene expression revealed significant over-representation of immune pathways predominantly in adult mice, with little or no significance in the adolescents. Various biological pathways that directly or indirectly have a role in immune response signaling are shown, with the false discovery rate (FDR) protected significance (q < 0.05) value for both age groups stated for each pathway. Three unbiased WebGestalt (Zhang et al., 2005) analyses were used to determine pathway differences in age- and alcohol-related brain transcriptome changes (AgeC = comparison between adolescent and adult controls, P30E = adolescent control compared to adolescent ethanol drinking, P70E = adult control compared to adult ethanol drinking). The cartoon was generated based on potential mechanisms, as determined in silico, that may influence binge alcohol consumption. Arrows indicate direction of response with initial Tlr4 action leading to either MyD88 dependent or independent downstream changes. It is unknown how alcohol directly acts on neuroimmune function, yet genetic and functional studies have shown a role for LPS, Cd14 and Tlr4. Therefore, a boxed “?” is used to depict possible action. Mechanisms known to occur in the cytoplasm and nucleus are shown in their respective compartments divided by gray bars. However, although action is likely to predominantly occur via microglia, cell specificity remains to be determined in future studies. Abbreviations are: Ccl2/MIP, chemokine (C-C motif) ligand 2 / macrophage inflammatory protein; Cd14, monocyte differentiation antigen Cd14; IL-1B, interleukin 1 beta; IL6, interleukin 6; Jak-STAT, Janus kinase - signal transducer and activator of transcription; LPS, lipopolysaccharide; MAPK, mitogen-activated protein kinase; MyD88, myeloid differentiation primary response 88; NF-κB, nuclear factor of kappa light polypeptide gene enhancer in B cells; Tlr4, Toll-like receptor 4.

Figure 3

Minocycline administration reduced DID drinking in adult (P70), but not adolescent (P30) F1 mice. The bars represent average ethanol consumption in g/kg over the 4 days of measurement as mean ± S.E.M. for adolescent (P30) and adult (P70), saline (open bars) or minocycline (filled bars) treated male or female mice. Minocycline administration did not affect ethanol consumption in adolescent male (F (1,13) = 1.810, p = ns) or female (F (1,15) = 2.458, p = ns) mice, but caused a significant reduction in adult DID drinking in male (F (1,16) = 15.06, p < 0.05) and female (F (1, 19) = 31.30, p < 0.001) mice, compared to saline treated mice as determined using 3-way ANOVA. (*p < 0.05, ***p < 0.001, ns = not significant).

Figure 4

Minocycline administration did not affect blood ethanol concentrations (BEC) in adolescent and adult male and female DID treated F1 mice. The bars represent average BEC in mg/ml measured at the end of DID paradigm as mean ± S.E.M. for saline (open bars) or minocycline (filled bars) treated male and female mice. A 3-way ANOVA revealed no significant effect of minocycline on BEC in either the adolescent or adult F1 mice (F (1, 48) = 0.3622, p = ns), compared to saline treated mice. (ns = not significant).

Figure 5

Minocycline administration did not affect ethanol elimination in either A. male and B. female adult (P70) F1 mice. Average BEC at 120, 150 and 180 min post-gavage are represented as mean ± S.E.M. for the saline (triangles) and minocycline (squares) treated mice. Lines represent linear regression analysis for estimation of elimination rates. Minocycline treatment did not alter the elimination in either the male or female mice (F (1, 24) = 0.4929, p = ns), compared to saline treated mice. (ns = not significant).

Figure 6

Minocycline administration reduced DID drinking in adult (P70), but not adolescent (P30) C57BL/6J mice. The bars represent average ethanol consumption in g/kg over the 4 days of measurement as mean ± S.E.M. for saline (open bars) or minocycline treated (filled bars) male and female mice. Minocycline administration did not affect ethanol consumption in adolescent male (F (1, 12) = 1.639, p = ns) and female (F (1, 14) = 1.426, p = ns) mice, but caused a significant reduction in adult male (F (1, 40) = 12.93, p = ns) and female (F (1, 20) = 1.580, p < 0.05), compared to saline treated mice. (*p < 0.05, **p < 0.01, ***p < 0.001, ns = not significant).

Figure 7

Minocycline administration did not affect ethanol elimination in either A. male or B. female adult (P70) B6 mice. Average BEC at 165, 210, 255, and 300 min post-gavage are represented as mean ± SEM for the saline (triangles) and minocycline (squares) treated mice. Lines represent linear regression analysis for estimation of elimination rates. Minocycline treatment did not alter the rate of ethanol elimination in either the male or female mice (F (1, 24) = 0.0615, p = ns), compared to saline treated mice. (ns = not significant).