Turnover of histones and histone variants in postnatal rat brain: effects of alcohol exposure

Animal procedures

Adult Sprague Dawley rats were purchased from Charles River (Wilmington, MA) and maintained in a controlled environment with a 12-h light/dark cycle at a constant temperature (22 °C). Adult female Sprague-Dawley rats were bred at 2 months of age in our vivarium. At postnatal day 2, rat pups of both sexes were either fed a milk formula containing 11.34% of ethanol yielding a daily ethanol dose of 2.5 g/kg (alcohol-fed, AF) (Bio-Serv, Frenchtown, NJ) or an isocaloric liquid control diet (pair-fed, PF) (Bio-Serv) in which the alcohol calories were replaced by maltose-dextrin. Feedings were conducted at 1000 and 1200 h from postnatal day 2 until postnatal day 6 (PD2-PD6). In a preliminary study, we validated the use of PF as control group by comparing the biological data (BrdU incorporation and γH2A as DNA damage marker) of PF group with ad libitum fed group (AD). The data in Additional file 1: Figure S2 and Additional file 2: Figure S3 shows that the PF and AD data are similar (γH2AX foci/mm, AD-265 ± 46, PF-327 ± 58, n = 4–6; p > 0.05; BrdU⁺ cells/ mm², AD-1970 ± 173, PF-1775 ± 74, n = 4; p > 0.05). Hence, we only used PF group as a control group for alcohol treatment. For isotope tracer labeling and proteomic studies, on day 1 of feeding, rat pups (PF and AF) received an intraperitoneal bolus injection of ²H₂O as normal saline and were then fed a milk diet that was enriched with 5% ²H₂O to ensure continuous precursor labeling [30, 33]. Note that the use of the stable isotope (²H) labeled water represents an adaptation of studies that previously relied on radioactive (³H) labeled water [34]. In some experiments, pups from each experimental group were subcutaneously injected with the free radical spin trap α-phenyl-N-tert-butyl nitrone (PBN) at 100 mg/kg before the morning feeding to investigate the role of ROS production in the effects of PAE on histone turnover. After each feeding, the pups were immediately returned to the litter. In some pups, blood alcohol concentration (BAC) was measured in trunk blood 1 h after the last ethanol feeding. Feeding 11.34% v/v ethanol increases BAC level to ~ 0.2 g/dl, which is considered to be a moderate BAC. Feeding trials ran for a total of 5 days after which animals were euthanized and samples were quick-frozen. Note that samples were also collected from a subgroup of control animals, which were fed ad lib, and not given any tracer; samples from this group guide the analyses and background corrections. The need for such baseline samples is well known in tracer studies; since stable isotopes are naturally occurring, they generate a background signal that needs to be subtracted from experimental samples [35]. Animal care was performed in accordance with institutional guidelines and complied with National Institutes of Health policy.

Acid extraction of histone

Brains were collected on PD6 and mediobasal hypothalamus (MBH) and frontal cortex was dissected. The frontal cortex was defined as the tissue extending from the frontal pole caudally to the bregma 3 mm in the pups (from figure 47 to figure 50) [36]. The hypothalamus was sectioned from the posterior portion of the optic chiasm until the end of the mediobasal hypothalamus, from figure 55 to figure 59, according to Ramavhandra and Subrananian [36]. Histones were acid extracted as previously described with some modifications [37, 38]. Briefly, dissected brain tissues were homogenized on ice in Triton Extraction Buffer (TEB: PBS containing 0.5% Triton X 100 (v/v), protease inhibitors cocktail, and 5 mM sodium butyrate) at 100 mg/ml using a tight-fitting glass Dounce homogenizer by applying 20 strokes and kept on ice for 5 to 10 min. Samples were then centrifuged at 2000 rpm for 10 min at 4 °C. The supernatants were discarded, and the nuclear pellets are washed in half the volume of TEB and centrifuged as above. Histones are then acid extracted in 200 μl of 0.4 N H₂SO₄ and incubated overnight at 4 °C [36], then centrifuged at 14.000 g for 15 min at 4 °C and supernatant transferred to a clean 5 ml tube. Two milliliters of cold acetone is then added and incubated at − 20 °C overnight to precipitate histones. Histones are then collected by centrifugation at 14,000 g for 15 min, washed again with acetone then air dried, and resuspended in sterile deionized water. Protein concentration was measured using Bradford assay, and samples were stored at − 20 °C until further analysis.

Preparation of histone samples for proteomics analysis

Extracted histones were prepared for proteomic analyses as described by [39], with some modifications. Briefly, histones were first derivatized with propionic acid to block unmodified lysines. Briefly, 30 μg of acid-extracted histones were dried down in a SpeedVac, then were dissolved in 15 μl 0.1 M ammonium bicarbonate solution and 10 μl of deionized water. The pH of the samples was adjusted to 8 using concentrated ammonium hydroxide solution and reacted with 20 μl of freshly prepared propionic anhydride/2-propanol solution (1:3 v/v). The solution was immediately neutralized with 6 μl of concentrated ammonium hydroxide (12 N). If needed, the pH of the solution was adjusted to 8 using additional concentrated ammonium hydroxide. Samples were incubated at 37 °C for 15 min. To remove solvent, the samples were evaporated for 30 min in a SpeedVac. To perform an additional round of propionylation, samples were dissolved in 20 μl of deionized water; pH of the samples were adjusted to 8 using concentrated ammonium hydroxide and propionylated with anhydride/2-propanol solution (1:3 v/v). Dried samples were then dissolved in 50 μl of 0.1 M ammonium bicarbonate; the pH was adjusted to 8 using concentrated ammonium hydroxide as needed. Histones were digested with 15 μl of 100 ng/mL sequencing grade Tripsin (Promega, Madison, WI) at 37 °C for 6 h. To stop the digestion, 5 μl of acetic acid was added to each sample and then concentrated to < 10 μl using a SpeedVac. After adjusting the pH to 8, propionylation was repeated twice. The volume of each sample was adjusted to 20 μl using 5% trifluoroacetic acid, and 10 μl aliquot was purified and desalted using a C18 Ziptip (Millipore, Darmstadt, Germany). Eluted samples were dried in a SpeedVac and reconstituted in 30 μl of 1% acetic acid for LC-MS analysis.

LC-MS/MS analysis of proteolytic peptides

Proteomic analysis was carried out by nano-flow LC-MS/MS with a Thermo Acclaim PepMap RSLC C18 nano-column (150 mm × 75 μm, 2 μm, 100 Å) using mobile phase A and B (1% formic acid in water and 0.1% formic acid in acetonitrile, respectively). Tandem mass spectrometry was performed on a Thermo LTQ Oribtrap Elite mass spectrometer (Thermo Scientific, Bremen, Germany) in positive ion mode. Each full scan MS at a resolution of 60,000 (at 400 m/z) was followed by 15 collision induced dissociation (CID) MS/MS (or MS2) scans. Dynamic exclusion was triggered by three consecutive selections within 30 s and last for 90 s.

To quantify 2H incorporation into histones, we manually integrated the areas under each mass isotopomer of tryptic peptides identified in protein database searches. Peptides that cannot be assigned to unique proteins are excluded. The mass isotopic distribution for all selected peptides are quantified as a function of time; the fractional synthesis rate (FSR) is then calculated from the shift in the distribution profile [30, 31].

GC-MS analysis of plasma water labeling

The ²H-labeling of plasma water was determined using GC-MS following the exchange of label with acetone [40]; under alkaline conditions, the ²H that is present in water will exchange with hydrogen that is bound to acetone, the enrichment of acetone is then quantified. Briefly, samples are prepared by incubating 5 μL of plasma or known standards in a 2 ml glass screw-top GC vial at room temperature for 4 h with 2 μL 10 N NaOH (Fisher Scientific, Maltham, MA) and 5 μL of acetone (Sigma, St. Louis, MO). The GC-MS is programmed to inject 5 μL of headspace gas from the GC vial in a splitless mode using an isothermal run (Agilent 5973 MS coupled to a 6890 GC oven fitted with an Agilent DB-5MS column), the mass spectrometer is operated in selected ion monitoring (electron impact ionization, m/z 58 and 59, 10 ms dwell time per ion).

Data analysis and calculation of protein kinetics

As noted above, the first step in the LC/MS analyses is to identify the peptide species. This is done from the accurate mass determination and the fragmentation spectrum. Examples of histone peptide spectra are shown in Additional file 3: Figure S1, and a summary of the histone peptides identified and used for the determination of histone turnover is listed in Table 1. In the case of the peptide vTIAQGGVLPNIQAVLLPkkTESHHk, which is derived from hypothalamic histone H2A, we observed a signal at 1001.57 a.m.u. with a charge state of 3+; this reflects propionyl modifications from the chemical derivatization which are observed at the N-terminal, K119, K120, and K126 residues. The second step in the LC/MS analyses is to determine if the mass isotopomer distribution in a spectrum is shifted from the respective control samples. For example, all proteins have a natural isotopic distribution; they exist as a mixture of slightly different molecular weights. This results from the fact that the various elemental building blocks are not made of homogenous pools, e.g., carbon is present in the body as both ¹²C and ¹³C, hydrogen is present as both ¹H and ²H, etc. A standard approach to better visualize the isotopic distribution of a target analyte involves normalizing the abundance of the heavy isotopes to what is referred to as the “M0” or “monoisotopic” peak, i.e., the analyte which contains only ¹²C, ¹H, ¹⁴N, ¹⁶O, etc. Therefore, in the case of peptide vTIAQGGVLPNIQAVLLPkkTESHHk, we divide each signal by that of the M0 signal (an example of this is shown in Fig. 1).

Isotopic distributions are determined after correction for natural background labeling.

Western blot analysis of total histones

Total histone levels were determined by western blot analysis. Whole brain tissue samples were processed for acid extraction of histones and quantification using the Bradford Assay (Bio-Rad Laboratories). Eight to ten micrograms of total histones were run on a 15% SDS PAGE and transferred to PVDF membrane. Membranes were blocked with 5% nonfat dry milk in TBST at room temperature for 2 h then incubated with different primary antibodies for histone proteins at 4 °C overnight. The primary antibodies used were rabbit anti-H2A (cat# 2578, 1:1000, Cell Signaling) and rabbit anti H3 (cat# ab1791,1:1000), rabbit anti-H3.3 (cat# ab62642, 1:1000) and rabbit anti-H2Az (cat# ab4174, 1:1000) from Abcam. We also used rabbit anti-H4 (cat# 07-108, 1/1000, Millipore) and mouse anti-β-actin (cat# CP01-1EA,1:5000, Calbiochem). Membranes were then washed in TBS-T and incubated with the appropriate secondary antibodies conjugated to a horseradish peroxidase followed by chemilumenescence detection using ECL reagents (Thermo Fisher Scientific).

Immunofluorescence staining for γH2Ax

Twenty millimeters of frozen sections from PF and AF, prefrontal cortex (bregma 4.7 to 3.2 mm), and hypothalamic arcuate nucleus (ARC) (bregma − 1.8 to −4.3) were used for these studies. Sections were fixed with 4% paraformaldehyde and stained overnight at 4 °C with mouse anti-γH2Ax (EMD Millipore, Billerica, MA; 1/250) antibodies. The secondary antibody used was Alexa Fluor 594 donkey anti-mouse (Abcam, Cambridge, MA; 1/500). Sections were mounted with a mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI) to counterstain nuclei (Invitrogen, Waltham, MA). γH2Ax foci were quantified using ITCN (Image-based Tool for Counting Nuclei for ImageJ, NIH, Bethesda).

Detection of 8-hydroxy 2 deoxyguanosine (8-OHdG)

A competitive ELISA for 8-OHdG was performed using a commercial 8-OHdG ELISA kit (Abcam, Cambridge, MA), following the manufacturer’s instructions. The DNA was purified from frontal cortex and hypothalamus using the DNeasy Blood and Tissue Purification kit (Qiagin, Germantown, MD). DNA quantity and purity were determined using NanoDrop-1000 version 3.7 (Thermo Scientific, Maltham, MA). Enzymatic digestion was performed using nuclease P1 (pH 5.3; Sigma, St. Louis, MO) at 50 °C for 1 h and treated with alkaline phosphatase (pH 8.5; New England BioLabs, Ipswich, MA) at 37 °C for 30 min. Samples were boiled for 10 min and placed on ice for 5 min. DNA hydrolysates were analyzed by ELISA, following the manufacturer’s instructions, and OD was measured at 450 nm using a plate reader (Thermo Scientific, Maltham, MA), and the level of 8-OHdG was determined for each sample from a standard curve.

BrdU incorporation assay in vivo

In vivo bromodeoxyuridine (BrdU) incorporation was performed to analyze cell proliferation in the brain of rat pups. BrdU (Sigma, St. Louis, MO) (50 mg/kg) was subcutaneously administrated everyday during ethanol feeding from PD4 to PD6 before sacrifice of pups. The brains were then collected and snap frozen in dry ice. Frozen brains from PF and AF from prefrontal cortex (bregma 4.7 to 3.2 mm) and hypothalamic arcuate nucleus (ARC) (bregma − 1.8 to − 4.3) were sectioned at a 20 μm thickness using a Cryostat (Leica CM1900; Leica Microsystems, Buffalo Grove, IL), then fixed with 4% paraformaldehyde then treated with 2 N HCl solution for 30 min at 37 °C, followed by a 10 min incubation with boric acid buffer to neutralize the acid, then washed in PBS. After permeabilization and blocking, sections were incubated with anti-BrdU monoclonal antibody (Sigma, St. Louis, MO; 1:500) overnight at 4 °C, and then with the secondary antibody Alexa Fluor 488 donkey anti-mouse (Abcam, Cambridge, MA; 1:500) for 1 h at room temperature. Sections were mounted with a mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI) to counterstain nuclei (Invitrogen, Waltham, MA). BrdU positive cells were counted using ITCN (Image-based Tool for Counting Nuclei for ImageJ, NIH, Bethesda).

RNA extraction and real-time polymerase chain reaction

Statistical analysis

Statistical analysis of data was performed using Graph Pad Prism software version 5.0 (LA Jolla, CA). Data were analyzed using unpaired student’s t test or one way ANOVA with Newman-Keuls post-hoc tests where appropriate. All results are presented as standard error of the mean (SEM). P < 0.05 was considered as significant.

PAE altered the turnover rates of canonical histones and histone variants

To explore the effects of alcohol exposure on histone turnover in postnatal rat brain, we administered a milk diet containing 11.34% of ethanol (2.5 g/kg per day, alcohol-fed, AF) or an isocaloric liquid control diet (pair-fed, PF) to 2-day-old Sprague Dawley rat pups for 5 days. During the feeding, ²H₂O was added to the milk diet for continuous labeling of newly made proteins. Samples were then collected for determining the protein kinetics via LC/MS.

As expected, there is a change in the relative abundance of the isotopic peaks in animals following 5 days of chronic dosing with the ²H₂O tracer as compared to the spectra that is observed in control animals, which did not receive any ²H₂O; there is a marked difference between the AF and PF groups (Fig. 1). Similar data sets were collected for the various histones that are discussed herein. The fractional synthesis rates (FSRs) are determined from the total change in isotopic labeling of the peptide product as outlined in the “Methods” section. It is important to note that the overall logic which is used here, including the data that are acquired, is virtually identical to that described by Zee et al. [41] with the exception of the shift in isotopic distribution. For example, since Zee et al. [41] administered a more highly substituted precursor (i.e., ¹³C₆ ¹⁵N₂-lysine) in cultured HeLa cells, they observed a shift in a region of the mass spectra where there is less natural background [40]. Efforts by our group have focused on the use of ²H₂O for reasons noted earlier in regard to the ease of administration in chronic long-term in vivo studies. We also observe a change in the molecular weight (as shown by the shift in isotope distribution in Fig. 1); however, this occurs in a region of the mass spectra where there is more natural background labeling (this is corrected as part of the calculations).

To convert the data contained in Fig. 1 to kinetic rates, e.g., FSR and half-life, one follows a few basic calculations that are outlined in the “Methods” section. The first step involves a calculation of the total enrichment, i.e., subtract the background distribution observed in control animals. The second step is to sum the excess labeling and normalize that against the water labeling; this requires an additional constant factor which is derived from knowledge of the amino acid sequence and the expected equilibration of ²H from body water with the respective amino acids [30, 31]. These calculations yield an estimate of the FSR.

Our high-resolution accurate mass LC-MS analyses allowed us to identify several histone-derived peptides as well as measure the incorporation of ²H. First, we measured ²H incorporation in the newly made replication-dependent canonical histones H2A, H2B, H3, and H4. These histones have long half-lives, and their synthesis is tightly linked to DNA replication [42, 43]. Figure 2 clearly shows that H2A followed by H3 has the highest FSR, reflecting faster turnover rates of these histones in both the brain regions studied, the frontal cortex (F(3, 10) = 71.37, p < 0.0001, Fig. 2a) and the hypothalamus (F(3, 10) = 124.8, p < 0.0001, Fig. 2b). Figure 2 also clearly shows that PAE significantly decreased the FSR of H2A, H2B, H3, and H4, reflecting a decrease in turnover rates and therefore an increase of the half-lives of these histones in the frontal cortex and the hypothalamus (Fig. 2, Table 3 and Additional file 4: Table S1).

	Half-life in frontal cortex (d)		Half-life in Hypothalamus (d)
Histone	PF	AF	PF	AF
H2Aun	5.9 ± 0.3	9.7 ± 0.8**	7.2 ± 0.4	13.1 ± 1.3**
H2AK120meK126ac	6.9 ± 1.1	11.7 ± 3.6	6.8 ± 0.5	10.7 ± 2.3**
H2Bun	27.4 ± 8.9	41.5 ± 15.2	20.5 ± 0.2	65.5 ± 12.1***
H2BK9ac	14.2 ± 4.4	12.8 ± 5.8	9.4 ± 1.5##	20.7 ± 2.8** #
H3un	11.0 ± 0.9	14.6 ± 1.1**	14.3 ± 0.9	20.5 ± 1.0**
H3K1ac	7.5 ± 0.4##	8.7 ± 0.4##	8.9 ± 0.2#	17.5 ± 5.9*
H3K9	10.0 ± 0.6	20.2 ± 2.9	19.8 ± 3.0	33.2 ± 1.7
H3K9me2	25.2 ± 6.7	21.4 ± 1.3	18.2 ± 2.8	22.3 ± 3.1
H3K9me3	23.1 ± 5.8	46.1 ± 15.8	38.7 ± 5.7a	69.8 ± 4***
H3K9ac	6.4 ± 0.4#	12.1 ± 3.3	10.5 ± 2.0#	19.9 ± 6.7
H4un	16.2 ± 1.7	29.8 ± 6.3**	28.4 ± 3.4	40.3 ± 1.1*
H4K5acK9acK13ac	8.2 ± 0.7##	8.5 ± 0.1###	11.5 ± 2.9#	14.8 ± 2.6#
H1	10.0 ± 1.3	20.4 ± 5.9***	12.3 ± 1.2	19.8 ± 1.2*
H1.1	3.9 ± 0.7	7.9 ± 0.7*	5.7 ± 0.7	10.8 ± 2.2*
H1.4	12.7 ± 2.8	13.9 ± 1.5	11.1 ± 0.6	18.9 ± 0.9***
H1.5	11.3 ± 1.8	20.2 ± 8.2	10.1 ± 0.6	14.5 ± 0.3*
H3.3	4.2 ± 0.6	7.8 ± 0.5**	9.3 ± 2.0	11.3 ± 3.5
H2Ax	12.1 ± 1.6	14.3 ± 1.5	15.1 ± 5.0	39.6 ± 9.1*
H2Az	5.1 ± 0.9	8.6 ± 2.0**	6.0 ± 0.5	8.2 ± 0.5*

Half-lives are calculated from FSR data as described in the “Methods” section. d day, Un unmodified, Ac: acetylated, me2 di-methylation, me3 tri-methylation. Data are shown as mean ± SE, n = 4 per group, *p < 0.05, **p < 0.01, ***p < 0.001, AF vs PF, #p < 0.05, ##p < 0.01, ###p < 0.001 Nac vs Ac

We also measured the turnover rates of newly labeled replication-independent histone variants (i.e., H3.3, H2Az, and H2Ax). The data shown in Fig. 2c, d and Additional file 4: Table S1 indicates that the histone variant H3.3 has a faster turnover rate in the frontal cortex (Fig. 2c) than in the hypothalamus (Fig. 2d). PAE dramatically decreased H3.3 turnover rate in the frontal cortex with no significant change in the hypothalamus. The turnover rate of H2Az was decreased in both brain regions; however, the turnover rate of H2Ax was not affected in the frontal cortex but significantly decreased in the hypothalamus. These findings indicate that PAE differentially affects the turnover rates of replication-independent histone variants and that differences exist between both brain regions studied.

In addition, we investigated whether these changes in histone turnover are associated with changes in histone abundance; western-blot analysis was used to determine the quantity of total histones in the whole brain of PF and AF animals. Figure 2e clearly shows that PAE had no effect on the abundance of H2A, H3, H4, H3.3 and H2Az histone proteins, suggesting that the total quantity of histones is not affected by PAE and that alcohol’s effect is most likely altering the turnover and therefore the newly made histone pool.

PAE altered the turnover rates of linker histone H1 and its variants

As their name indicates, linker histones bind to the linker DNA located between two adjacent nucleosomes and provide an additional stabilizing effect on chromatin. We next explored the turnover rates of the linker histone H1 and its variants, H1.1, H1.4, and H1.5 (Fig. 3 and Additional file 4: Table S1). We found that the linker histone H1 and its variant H1.1 had a higher FSR in the frontal cortex than in the hypothalamus. Regardless of the brain region, H1.1 histone had the highest turnover than the other linker histones (F(3, 11) = 15.02, p < 0.0003 in the frontal cortex, Fig. 3a; F(3, 10) = 12.31, p < 0.001 in the hypothalamus and 3B). In addition, PAE dramatically decreased the FSR and therefore the turnover rates of H1 and H1.1 in the frontal cortex with no significant changes in H1.4 and H1.5 turnover rates (Fig. 3a). In the hypothalamus, however, the rates of all linker histones were significantly decreased, although to a lesser extent than in the frontal cortex (Fig. 3b).

PAE altered the turnover rates of modified histones

To explore the rates of histone turnover as function of their PTMs, we identified acetylated and methylated peptides of the above aforementioned replication-dependent canonical histones. We determined the relative FSR of the newly labeled modified histone peptides, which then was contrasted to the FSR of the unmodified peptides (Figs. 4, 5, and 6 and Additional file 4: Table S1). Peptide sequences of modified and unmodified H2A, H2B, H3, and H4 analyzed in this study are shown in Table 1.

Figure 4 shows data that were generated from the frontal cortex and the hypothalamus examining the turnover of acetylated peptides identified as H2B, H3, and H4 histones. As reported by others [41, 44], our method confirms the findings that acetylated histones have a significantly faster turnover than the unmodified histones. We found that PAE significantly decreased the turnover rates of unmodified H2B, H3, and H4 in the frontal cortex and the hypothalamus, but had no effect on the acetylated forms in the frontal cortex (Fig. 4a, c, and e) while it decreased the turnover rates of acetylated H2B, H3, and H4 in the hypothalamus (Fig. 4b, d, and f). We were the first to identify a new H2A peptide that was double modified with a methyl group on lysine 120 and an acetyl group on lysine 126 (i.e., H2AK120meK126ac). H2AK120meK126ac has not been reported previously. We calculated the relative FSR of the newly labeled double modified H2A and compared its turnover rate to the unmodified H2A peptide. Interestingly, the turnover rate of H2AK120meK126ac was not significantly different from the unmodified peptide in both brain regions (Fig. 5 and Additional file 4: Table S1). PAE in a similar manner significantly decreased the turnover rates of unmodified as well as doubly modified H2A (Fig. 5 and Additional file 4: Table S1). Furthermore, the analysis of di- and tri-methylated histone H3 revealed that progressively methylated histones have slower turnover rates than the unmodified histones. Figure 6 and Additional file 4: Table S1 show that H3K9me3 had a slower turnover rate than H3K9me2 and H3K9un in the hypothalamus (Fig. 6b). Interestingly, PAE decreased the turnover rates of H3K9un as well as H3K9me3 and had no significant effect on H3K9me2. Histone H3 was shown to be either methylated or acetylated on lysine K9. In this same set of samples, we were able to simultaneously identify an acetylated H3 on lysine 9 (i.e., H3K9ac). Consistent with the findings in Fig. 5, H3K9ac had a faster turnover than the unmodified peptide and the methylated peptides (H3K9me2 and H3K9me3), and in a similar manner was decreased by PAE in both brain regions studied (F(3, 8) = 13.00, p < 0.0019 in the frontal cortex, Fig. 6a; F(3, 10) = 10.99, p < 0.0017 in the hypothalamus, Fig 6b).

Since these types of studies generally assume that the kinetics under investigation are described by a single compartment model, and reflect a first-order process, then one can convert the FSR data into protein half-lives using the standard equation half-life = ln 2/FSR [94]. Table 3 presents a comparative summary of the calculated half-lives of the different histone peptides identified in this study.

PAE induces oxidative DNA damage

Several studies including ours have shown that alcohol exposure induces ROS production and results in oxidative DNA damage. To investigate whether oxidative DNA damage is involved in the changes of histone turnover observed in response to PAE, we next measured the phosphorylation of histone H2Ax on serine 139, referred to as γH2Ax, an early marker of DNA damage. Figure 7 indicates that PAE significantly increased γH2Ax foci formation in the frontal cortex (Fig. 7a) and the hypothalamus (Fig. 7b) of AF rat pups compared to PF. To confirm these findings, we also measured the formation of 8-hydroxy 2 deoxyguanosine (8-OHdG), a known marker for DNA oxidative damage. Figure 7 shows that PAE significantly increased the formation of 8-OHdG in the frontal cortex and the hypothalamus (Fig. 7c) of AF animals compared to PF. To determine the role of free radicals in the PAE on oxidative DNA damage, we used the spin trap agent PBN. PBN treatment efficiently reduced the formation of 8-OHdG in brain tissue from AF-PBN animals compared to AF animals (Fig. 7d).

The effect of the free radical spin trap PBN on histone turnover

We next investigated the effects of the spin trapping agent PBN on histone turnover. In this study, a subset of rat pups from each experimental group (PF and AF) was subjected to a subcutaneous injection of 100 mg/Kg body weight of PBN once a day. Histones were extracted from prefrontal cortex, digested and analyzed using LC-MS/MS. As shown in Fig. 8, PBN treatment did not inhibit the effects of PAE on H3, H2Az, or H2Ax turnover. However, PBN administration significantly increased H3.3 and H4 turnover rates in AF-PBN animals compared to AF alone. These findings suggest that the free radical spin trap PBN efficiently prevented the effects of PAE on the turnover of H4 and H3.3 histones, but not on H3 and H2Az, indicating a differential effect of this reagent on histones in the brain.

PAE changed cell proliferation and the expression of cell cycle and histone synthesis genes

The PAE-mediated decrease in turnover rates of the replication-dependent canonical histones prompted us to further explore some of the possible underlying molecular mechanisms involved. Since the bulk of canonical histones are synthesized during the S-phase of the cell cycle, we investigated the effects of PAE on cell proliferation using BrdU incorporation and fluorescence microscopy. The number of BrdU-positive cells was counted, and as shown in Fig. 9, PAE significantly decreased the number of proliferating cells in frontal cortex (Fig. 9a, c) and hypothalamus (Fig. 9b, c) of AF animals compared to PF. However, the decrease in cell proliferation was more pronounced in the frontal cortex compared to the hypothalamic region (p < 0.0003) (Fig. 9).

It is well established that the synthesis of canonical histones is tightly coupled to DNA synthesis during the S-phase of the cell cycle. To elucidate some of the molecular mechanisms underlying the decrease in cell proliferation, we next investigated the effects of PAE on the expression of cell cycle genes. PAE significantly decreased the gene expression of cyclin E in frontal cortex (Fig. 10a) and had no significant effect in the hypothalamus (Fig. 10b) of AF rat pups compared to PF. In addition, PAE had no significant effect on cdk2 gene expression (Fig. 10a, b). While PAE dramatically decreased NPAT gene expression in the frontal cortex and the hypothalamus of AF animals, there was no significant change in the expression of SLBP (Fig. 10c, d). Next, we examined the impact of these changes on histone mRNA expression. Figure 10 indicates that PAE had no significant effect on histone mRNA levels in frontal cortex (Fig. 10e) or hypothalamus (Fig. 10f) of AF rat pups compared to PF.