Native Chromatin Immunoprecipitation from Brain Tissue Using Magnetic Beads

Epigenetics is a field that has grown enormously in the last decade. Numerous biochemical pathways and pathophysiology of many diseases, have been associated with epigenetic changes taking place in cell nucleus [1,2]. Two major epigenetic modifications that influence gene expression directly are DNA methylation and post-translational histone modifications. DNA methylation is a rather stable modification that is used by the cell primarily to delineate repressive chromatin. On the other hand, histone modifications are very dynamic, and are used to regulate gene expression in response to environmental stimuli. Therefore, histone modifications are regarded as a link between the genome and the environment, hence an increasing effort is put into investigating these modifications.


Introduction
Epigenetics is a field that has grown enormously in the last decade. Numerous biochemical pathways and pathophysiology of many diseases, have been associated with epigenetic changes taking place in cell nucleus [1,2]. Two major epigenetic modifications that influence gene expression directly are DNA methylation and post-translational histone modifications. DNA methylation is a rather stable modification that is used by the cell primarily to delineate repressive chromatin. On the other hand, histone modifications are very dynamic, and are used to regulate gene expression in response to environmental stimuli. Therefore, histone modifications are regarded as a link between the genome and the environment, hence an increasing effort is put into investigating these modifications.
Chromatin immunoprecipitation (ChIP) is a method that allows investigation of association between different proteins and DNA in the chromatin. Two main types of ChIP exist: ChIP based on chemical cross-linking (XChIP) and native ChIP (NChIP) [3]. XChIP protocol is widely used for all tissue types and all proteins that interact with DNA (structural, transcription factors, polymerases etc). These proteins must be cross-linked to the DNA prior to immunoprecipitation, usually by formaldehyde [4]. Such cross-linking decreases yield because it can cause epitope alterations, thereby making the immunoprecipitation inefficient. Furthermore, XChIP uses sonication as a way of chromatin fragmentation, which results in a wide variety of fragment sizes, thereby decreasing the resolution and reproducibility of the assay [5].
NChIP on the other hand, is only applicable to histone proteins. This approach takes advantage of the natural configuration of DNA wrapped tightly around core histones, making it possible to investigate histone modifications and DNA associations in their native form, without chemical cross-linking. NChIP uses micrococcal nuclease (MNase) digestion as a way of preparing the chromatin for immunoprecipitation, a method pioneered by Hebbes [6]. MNase is an endonuclease, unique in its ability to cut double stranded DNA in the linker region between nucleosomes, whereas it only causes single strand breaks in the nucleosome core region, where DNA is tightly wrapped around histones.
The subfield of epigenetics dealing with neurological disorders is developing very fast. Immunoprecipitation from brain tissue is a necessary tool for investigation of multiple histone modifications, but the literature that deals with NChIP from brain tissue is rather scarce. Therefore, we present a complete NChIP protocol for whole brain tissue, which is routinely carried out in our laboratory. This protocol pioneers the use of magnetic beads for immunoprecipitation, as opposed to very popular sepharose beads. We have successfully applied the protocol to both hippocampus and prefrontal cortex of adult rats. Furthermore, several antibodies have been implemented in the procedure, e.g. antiacetylated histone H4 (anti-H4ac), anti-trimethylated lysine 9 on histone H3 (anti-H3K9me3), anti-trimethylated lysine 4 on histone H3 (H3K4me3), all commercially available and previously proven to be applicable to ChIP (Abcam, Cambridge, UK).

Tissue, special materials and conditions
Adult male Sprague-Dawley rats (270-290 g, Charles River Laboratories, Hamburg, Germany) were used in all experiments. To achieve sufficient chromatin yield, between 30 mg and 70 mg prefrontal cortex tissue (or a whole hippocampus) was used. All solutions were ice-cold and all centrifugations were performed at 4°C. Low protein binding tubes were used (Sarstedt, Germany) in order to minimize protein interaction with the surface of the eppendorf tubes.

1.
The frozen tissue is placed in 10 mL EB1 and disrupted 3x 10 seconds by T10 Ultra-Turrax homogenizer (IKA, Staufen, Germany) using 8 mm pistil and by increasing speed each time. The time and speed of the tissue homogenizer should be optimized to the applied tissue. We used 10 seconds on each of three consecutively increasing speed levels (starting from 3 on the homogenizer's scale) to be appropriate for both tissue types. After homogenization the sample must be spun down immediately for 20 min at 3000 g.

2.
Discard the supernatant and re-suspend the pellet in 1 mL EB2. Centrifuge the samples for 10 min at 12,000 g.

3.
Discard the supernatant and re-suspend the pellet in 1 mL Digestion Buffer.

1.
To ensure the presence of intact nuclei, DAPI staining should be performed at this stage.

3.
Centrifuge the sample at 12,000 g for 5 min and re-suspend the pellet in 100 µL MilliQ water.

4.
Analyze the samples by a fluorescent microscope with an appropriate filter set (DAPI excitation= 364 nm, emission=454 nm).

5.
The nuclei isolated from the rat brain can be seen as individual ovals (Figure 1a) or in groups of several (Figure 1b), but all nuclei should have a nice round (oval) shape and look intact.

Micrococcal nuclease digestion
Before the MNase digestion, the amount of chromatin present in the sample should be assessed by spectrophotometry in the presence of 0.1% SDS. Subsequently, dilute the chromatin to a concentration of 0.5 mg/mL with Digestion Buffer (see above).

1.
50 U of MNase (Thermofischer Scientific, USA) per 0.5 mg chromatin should digest the sample in 4 min, at 37°C. It is optimal to use a thermomixer set to 600 rpm to avoid sedimentation of the nuclei.

2.
Stop the digestion by adding EDTA to a final concentration of 5 mM and place on ice.

3.
In order to avoid over-or under digestion of the chromatin, a time-course optimization should be performed on the desired tissue, followed by gel electrophoresis on 1.2% agarose in the presence of 0.1% SDS. Ethidium Bromide staining should be performed after the electrophoresis, as SDS binds to ethidium bromide.

4.
Centrifuge the sample for 5min at 11,600 g and transfer the supernatant to a new tube -this is the first supernatant fraction (S1).

5.
Resuspend the pellet in 0.5 mL Resuspension Buffer and transfer the solution to a dialysis tube (10 kDa pore width). Dialyse the sample in 4 L Resuspension Buffer overnight at 4°C on magnetic stirrer.

6.
Transfer the sample to an eppendorf tube and centrifuge for 10 min at 2000 g. Retain the supernatant -this is the second supernatant fraction (S2).

7.
Visualization of the two supernatant fractions by 1.2% agarose gel electrophoresis should be performed to assess the degree of digestion (Figure 2). The measurement of DNA quantity should be performed using spectrophotometer. Typical yield from hippocampus was approx. (2.5 mg/mL) and from PFC approx. (0.5 mg/mL). Repeat this step and combine the two eluates afterwards, obtaining 500 µL final eluate.
12. Mix 0.1 mL input fraction (step 4) with 0.4 mL Elution Buffer and incubate continuously (total of 30 min.) alongside the samples as in step 6.

DNA analysis by qPCR
After Proteinase K treatment the samples contain DNA fragments ready for further processing. We have used several DNA purification kits, e.g. MinElute PCR Purification (Qiagen, Germany) or ChIP DNA Clean & Concentrator (Zymo Research, USA), with satisfactory yield for many qPCR investigations.

Results
We have run all reactions initially in triplicates, and after confirming the reproducibility we continued running them in duplicates. In order to check the specificity of the NChIP on the rat brain tissue, we chose an active gene (actin (Actb), AC_000080.1 ) associated with acetylation on histone H4, and a gene which is not expressed in the brain (gamma globin (Hbe1), NC_005100.3), thus not associated with acetylated histone H4 to the same degree. The results indicate a very high degree of histone H4 acetylation on the actin gene promoter as compared to the globin gene ( Figure 3). We have calculated 12-fold enrichment in the histone H4 acetylation of the actin gene compared to globin gene using the following equation: fold enrichment= 2^(Ct input actin-Ct H4Ac actin)/2^(Ct input globin-Ct H4Ac globin), Ct input actin=31.16, Ct H4Ac actin=32.09, Ct input globin= 32.71, Ct H4Ac globin=37.32.
Comparing the relative enrichment of the pull-down between XChIP and NChIP methodologies, using the same antibodies, reveal both high specificity and sensitivity of NChIP superior to XChIP. Abcam as an antibody manufacturer uses XChIP with sepharose beads to validate their products as ChIP-grade. However, less enrichment relative to input is achieved by their approach, compared to the present protocol. For the activating H3K4me3 modification, their best enrichment is ~27% of input, whereas we precipitate 40-88% of input with this antibody (Figure 4a) (Abcam) [7]. An even clearer picture is true for the repressive modification H3K9me2. Using this antibody, Abcam precipitates ~4.5% of input, whereas we enrich by 25-98% of input (Abcam) [8]. Furthermore, Figure 4 clearly demonstrates the specificity of the NChIP protocol, since active and inactive genes are more/less associated with active and repressive modifications, respectively. The minute SEM-values of Figure 4 validates the reproducibility of the NChIP protocol, as these results stem from 6 different animals.
which is described for the first time in this paper, increases the immunoprecipitation efficiency, therefore the amount of tissue used in the experiment can be lowered. This is an important issue, as the amount of tissue is often a limiting factor. The final result indicates a 12-fold increased association of H4-Ac with the active gene compared to an inactive gene, thereby confirming high specificity of the protocol.