Animals

Subjects were adult males from two distinct sublines derived from the Sardinian preferring (sP) rats—the msP and Scr:sP rats. The breeding program of sP rats began in 1981 (University of Cagliari, Italy). Beginning with the 13th generation of sP rats, a subset of rats was provided to a separate program for selective breeding at the University of Camerino, Italy. After 20 generations of selective breeding for high alcohol preference, this distinct subline was renamed msP (27). The msP rats in the present study (n=25 for the strain comparison and n=79 for the home cage consumption study) were from the 58th–59th generations of selective breeding. A rederivation of the msP line, to generate sublines homozygous for the G allele (GG) and homozygous for the A allele (AA), was performed as described in Ref. (35). Rats from these sublines (n=9 GG, n=8 AA) were used for analysis of food intake when no alcohol was available. All behavioral studies, which involved only msP rats, were performed at the University of Camerino (Camerino, Italy). While these studies only include male subjects, it should be noted that sex differences in CRF₁ signaling have been reported in the locus coeruleus (37) that might increase female responsiveness to conditions of altered CRF₁ expression, although sex differences were not observed following germline Crhr1 deletion (38).

In the 32nd generation of selective breeding of sP rats according to high alcohol preference at the University of Cagliari, a subset of sP rats were provided by Dr. G. L. Gessa to The Scripps Research Institute and maintained as a colony without further selective breeding. Rats from this colony were designated as Scr:sP (http://rgd.mcw.edu/rgdweb/report/strain/main.html?id=2302666). The Scr:sP subline, like the sP line from which they were derived (39), show high spontaneous home cage alcohol intake (5–7g/kg daily) and preference (>80% vs. water) and display high levels of anxiety-like behavior (9, 40). The Scr:sP rats in the present study (n=20) were from the 23rd–24th generations of intra-line breeding at the Committee on the Neurobiology of Addictive Disorders of The Scripps Research Institute (La Jolla, CA, USA).

For the strain comparison, adult male outbred Wistar rats (n=21) obtained from Charles River (Raleigh, NC, USA) were used as controls. In addition, adult male Indiana P rats (41), generously provided by Dr. L. Lumeng (Indiana University School of Medicine, Indianapolis, IN, USA), also were subjects (n=23) at The Scripps Research Institute. P rats have been bred at Indiana University School of Medicine since original establishment of the line by selective breeding from a Wistar rat colony at the Walter Reed Army Hospital (the Walter Reed Wistar rat) (42). P rats demonstrate high alcohol intake in the home cage (>5g/kg/day) (43), yielding pharmacologically relevant blood alcohol levels (44).

Rats were housed under a 12-h reversed light cycle under constant temperature (20–22°C) and humidity (45–60%). Rats were housed three per cage in standard, wire-topped plastic cages containing sawdust bedding, with food (Camerino: 4RF18, Mucedola, Settimo Milanese, Italy; Scripps: LM-485 Diet 7012, Harlan, Madison, WI, USA) and water available ad libitum. Rats for behavioral studies received environmental enrichment prior to experimentation, at which time rats were singly housed to allow for recording of drinking and food consumption, as detailed below. At the conclusion of experiments, rats were euthanized by carbon dioxide inhalation.

Sequencing of the Crhr1 Promoter

Crhr1 promoters were sequenced in all subjects. For the strain comparison, 6 primer pairs were designed to sequence the ~2,300 bases upstream of the Crhr1 start codon (Figure ​(Figure1A)1A) and obtained from Integrated DNA Technologies (Coralville, IA, USA). The target amplicons, named P0-P5, were selected to generate an accurate, complete sequence of the proximal promoter region using slight overlap between all adjacent amplicons (Figure ​(Figure1B).1B). Primer pairs P0 and P1 were designed to overlap with the promoter sequence in which the msP Crhr1 polymorphisms were reported previously (31), with each amplicon containing one of the two known SNPs, to ensure successful sequencing. Primer sequences are provided in Table ​Table1.1. Tail DNA was extracted and polymerase chain reaction (PCR) was carried out with all samples (~30ng DNA/well) run on a single 96-well plate for each primer pair with seven wells reserved as no template controls. Sequencing reactions were run in both the forward and reverse directions. ABI trace files for all high quality sequencing results (defined as Phred scores>20 for at least 75% of the sequence) were analyzed using Mutation Surveyor. Only mutations identified in analysis of both the forward and reverse direction sequencing reactions were considered to be real mutations. This approach for complete sequencing of the 2,300bp upstream of the Crhr1 start codon was utilized for the strain comparison study.

Open in a separate window

Figure 1

Alignment of amplicons used for Crhr1 promoter sequencing. Primer pairs were designed using the Crhr1 promoter sequence obtained from the University of California, Santa Cruz genome browser to generate six overlapping amplicons for complete coverage of the promoter region, based in part on previously published sequences (31). Chromosomal alignment of the overlapping Crhr1 promoter amplicons is shown relative to (A) the Crhr1 gene sequence and (B) the Crhr1 proximal promoter region (based on UCSC Genome Browser, rat assembly July 2014). While consistent with the relative lengths of amplicons produced, the visual representation is not drawn to scale.

Once it was clear which Crhr1 promoter nucleotide positions were likely to be altered in msP rats, a more targeted genotyping approach was taken for the rats involved in the behavioral studies. Because only 2 polymorphisms were observed across all rats in the strain comparison study, PCR primers were designed to target only the expected polymorphism sites, encompassing positions −2,092 and −1,835bp relative to ATG start of Crhr1. Primer extension assays were performed using the Sequenom iPlex Gold reagent kit (San Diego, CA, USA), and extension products analyzed by mass spectrometry using a Sequenom Mass Array system. Base calls scored as “low probability” by the Sequenom software were not used in the final analysis.

Home Cage Food and Alcohol Intake

To explore further the functional significance of the Crhr1 promoter genotype within the msP line, behavioral studies were performed in male msP rats (n=79, weight range during experimentation 333–536g, run in two cohorts of 32 and 47 rats, respectively). Rats were single-housed prior to the first alcohol access day to quantify home cage intake. To ascertain the relation of Crhr1 promoter genotype to home cage alcohol intake, msP rats were provided ad libitum concurrent, continuous access to 10% alcohol (0.56kcal/ml), water and food pellets (Mucedola 4RF18, 2.6kcal/g) until a stable baseline alcohol intake (5–7g/kg/day) was achieved (maximum 15% daily variation, reached in 3weeks). Alcohol was prepared as a 10% (v/v) solution in tap water, and intake of both solutions and food was measured daily at 2, 4, 8, 12, and 24h after the onset of the dark cycle. As no time-dependent differences were observed, data are presented as 7-day means of 24-h intake. All consumption measures are normalized to body weight to account for size differences in both caloric requirement and pharmacological efficacy. To assess differences in focus on alcohol as a fluid or caloric source, data are additionally presented as percent alcohol preference [100×10% (v/v) alcohol intake/total fluid intake] and as percent calories derived from alcohol (100×alcohol intake in kcal/total kcal intake), per (17, 45). To further assess the impact of alcohol availability on chow intake, an additional group of rats from a rederivation of the msP line to separately breed AA and GG homozygous rats (35) was provided with ad libitum access to food pellets (Mucedola 4RF18, 2.6kcal/g) and water in the absence of alcohol. Data are presented as 3-day means of 24-h food intake.

Crhr1 Polymorphisms and the Regulation of CRF₁ Expression

Hansson et al. originally reported 2 G-to-A substitutions in the promoter region of Crhr1 of msP rats on one or both gene loci in a majority (85%) of msP rats (31). The present study found lower frequency of the A nucleotides, occurring in 68% of msP rats in the strain survey and 48% of msP rats in the behavior cohort. These frequency differences may result from random selection of rats from the population for the various experiments, as Crhr1 promoter genotype was not determined until the conclusion of experiments, or from genetic drift in the frequency of the AA loci in the msP parent line over time. The rederived parallel AA and GG msP lines now allow for direct assessment of the behavioral impact of the promoter polymorphism (35), particularly helping to delineate the phenotypes observed in the parent msP line that may have resulted from genes other than Crhr1 cosegregating with AA status.

In addition to the reduced frequency of the dual polymorphism in the current msP cohort, slight differences in the locations of both SNPs were identified in the genetic analysis, with one located 1,835bp (rather than 1,836bp) upstream of the receptor start codon, and the second SNP at 2,092bp (rather than 2,097bp) upstream. This discrepancy in position may reflect modification of the genome sequence since the original identification of the polymorphisms. Both loci lie well outside the proximal core promoter region (47), which could explain the apparent lack of involvement of the variant alleles in basal amygdala Crhr1 expression (31, 32). However, in the rederived msP lines, Crhr1 expression in the BNST was higher in AA than in GG or Wistar controls, suggesting that the polymorphism can affect Crhr1 gene expression in a brain region-specific manner, despite the distance from the core promoter region. Distal DNA regions as far away as 100kb can alter transcription via various mechanisms, resulting in enhancement, repression or insulation of transcription factor activity (47–50). Such effects can occur via looping of the intervening DNA sequence that apposes the distal upstream sequence to the proximal promoter (49, 51), as well as by regulation of DNA availability through methylation (50). Despite these possible mechanisms by which distal Crhr1 promoter regions may regulate its expression, at present the data are largely lacking for genotype differences in basal Crhr1 levels. Nonetheless, distal regions of the Crhr1 promoter may provide a site for experience-based modulation of Crhr1 transcription that would confer the differential sensitivity to CRF₁ antagonism and stress-induced reinstatement observed in AA msP rats (35).

One mechanism by which stress may alter CRF₁ function, via altered receptor expression, involves changes in distal promoter methylation status. The Crhr1 promoter region lies within a CpG island, and behavioral experiences that alter anxiety-like behavior have been shown to increase methylation on one CpG site located 1,348 bases upstream of the Crhr1 translation start in the amygdala (52). Recently, stress-related changes in gene methylation status and expression of another receptor (Oprl1) were shown in msP rats using a stress paradigm that increased msP rats’ alcohol consumption (53), indicating that stressors that exacerbate alcohol intake in msPs might accomplish this via altered methylation, and thereby expression, of specific gene targets. Since AA msP rats are anticipated to be more susceptible to stress-induced elevations in drinking based on yohimbine exacerbation of operant alcohol self-administration (35), stress-related increases in Crhr1 promoter methylation might differentially impact Crhr1 gene expression in AA vs. GG msP rats. Together these data highlight the possible involvement of distal regions of the Crhr1 promoter as putative sites of stress-related regulation of CRF₁ expression, and point to the identification of the molecular mechanisms by which the variant A alleles may confer increased sensitivity to stress and CRF₁ modulation of alcohol self-administration as an important focus to uncover new therapeutic avenues for alcohol dependence.

CRF₁ Regulation of Alcohol Intake

Increased activation of the CRF system is evident in the progression to substance dependence, including for alcohol [reviewed in Ref. (54)]. Pharmacological and genetic studies specifically implicate activation of extended amygdala CRF₁ receptors in mediating the excessive alcohol self-administration resulting from chronic intermittent exposure (7, 8) or access to alcohol sufficient to produce excessive (10) or binge-like (11, 15–17) drinking. Before CRF₁ antagonists were found to reduce alcohol self-administration in nondependent msP rats (31), recruitment of the CRF system had been conceptualized, even within alcohol-preferring rat lines (9, 36), as a mechanism of transition to alcohol dependence that resulted from chronic intermittent alcohol exposure. Indeed, chronic intermittent exposure to alcohol in Wistar rats leads to alcohol dependence and increased alcohol intake in association with increased extracellular CRF levels during withdrawal (6), depletion of amygdala CRF tissue content (3, 55), elevated CRF₁ expression in the basolateral and medial amygdala (56), and potentiation of GABA and glutamate release in the central amygdala (57, 58). msP rats also express elevated CRF₁, even under alcohol-naïve conditions (31), and show elevated function of the GABA and glutamate systems in the central amygdala (46, 59, 60). Binge-like DID alcohol consumption in C57BL/6J mice also altered CRF levels and electrophysiological properties of the central amygdala (15), but in contrast to alcohol dependence and msP findings, CRF immunoreactivity increased, presynaptic GABA release was unchanged, and CRF lost the ability to enhance evoked GABAergic currents after DID. Thus, CRF₁ antagonists’ ability to similarly reduce alcohol consumption in multiple models of heightened alcohol intake may arise from differential adaptation at the neuronal level based on alcohol access parameters utilized and stimuli producing heightened alcohol intake (7, 8, 11, 12, 16, 34, 61–63). The regulation of alcohol intake by CRF₁ encompasses its activation by both CRF and urocortin-1, which has been shown to specifically regulate escalating, but not controlled, alcohol intake (45). How urocortin-1 impacts alcohol intake in preferring rat lines, including msP rats, and whether this might be a site of neuronal adaptation yielding high alcohol intake remain to be determined, although greater density of urocortin-1 fibers across multiple preferring, vs. non-preferring, rat lines (64) suggest likely contribution of urocortin-1 to the high drinking phenotype. Surprisingly, while the msP (27), P (41), and Scr:sP (9, 40, 65) strains of alcohol-preferring rats each display high levels of alcohol intake, only msP rats are sensitive to CRF₁ antagonists in the absence of experimentally induced alcohol dependence (9, 27, 35, 36). The alcohol dependent-like neuronal activity profiles displayed in the central amygdala of msP rats (46, 59, 60), which may contribute to the rats’ increased anxiety-like phenotype (32, 46), are of particular interest for further examination as a basis for msP rats’ high alcohol preference.

Human studies support the involvement of CRF/CRF₁ in alcohol dependence (23, 66), despite the fact that clinical trials for CRF₁ antagonists have been unsuccessful to date (23). Elevated CRF levels were seen in the cerebrospinal fluid of recently withdrawn alcoholics (67). Similar to the Crhr1 promoter variants seen in the msP line, human CRHR1 SNPs have been associated with facets of alcohol dependence. Variant CRHR1 haplotypes in adolescents predicted future problem drinking patterns, including binge drinking and lifetime prevalence of intoxication and alcohol dependence (68). CRHR1 SNPs also predicted higher levels of alcohol consumption in already dependent individuals (68) and associated with reduced P300 levels, which are observed in alcoholics (69). CRHR1 SNPs predicted greater future drinking in relation to recent stress history (70), implicating CRF₁ systems in stress-induced drinking, consistent with observations in msP rats (31, 35), C57BL/6J mice (12) and Crhr1 knockout mice (71). Genetic associations of CRF signaling to alcohol phenotypes in humans is not restricted to CRHR1, because several studies link genetic variants in CRHBP, which encodes the CRF-binding protein, to features of alcohol dependence. Polymorphisms in CRHBP, which is hypothesized to modulate the amount of CRF or urocortin-1 available to interact with its receptors, have been related to decreased EEG alpha wave power (72), a feature observed in alcoholics (72, 73), and are more prevalent in alcoholics with comorbid anxiety disorders (74). CRHBP polymorphisms have also been related to severity of stress-induced alcohol craving (75) and have been hypothesized to impact anxiety or drinking in alcohol-dependent individuals (76). Individual CRHBP and CRHR1 SNPs have been shown to coordinately predict alcohol use disorder comorbidity in a panel of schizophrenic patients (77). Importantly, Ribbe et al. (77) demonstrated elevated levels of CRHR1 mRNA relative to CRHBP mRNA in blood from individuals carrying the dual polymorphism, suggesting that CRF₁ activation by CRF or urocortin-1 predominates over CRF/CRF-BP or urocortin-1/CRF-BP interactions in individuals with high predisposition to develop alcohol use disorders. Additional gene variants of CRHR1 and CRHBP have been associated with various mood and anxiety disorders as well as alcohol use disorders [reviewed in Ref. (23)], implicating the CRF₁ system in various mental illnesses characterized by emotional dysregulation. These human findings resonate with the present data in suggesting a key role for CRF₁ signaling in some forms of alcoholism, perhaps particularly those with stress history or comorbid anxiety.

The authors thank S. Whitney Barnes for technical assistance, Michael Arends for proofreading the manuscript, and Dr. Marisa Roberto for financial assistance. The authors also thank Dr. Lawrence Lumeng for providing the P rats to The Scripps Research Institute and Dr. Nicholas Gilpin for providing samples from the P rats for genotyping. This is publication number 21809 from The Scripps Research Institute.