ChIP-Chip Protocols & Methods

• Chromatin Immunoprecipitation Protocol
• ChIP-Chip Data Analysis Methods

Chromatin Immunoprecipitation Protocol

Cell harvest and crosslinking complexes in vivo:

1. Grow cells to confluence in TC flask. Plate 1.552x10⁷ for BMM in T-150.
2. Stim cells
3. Save approximately 1ml supe for ELISA, store at -20°C.
4. Wash cells with 20mL room temperature PBS (Cellgro).
5. Add 20mL of PBS to cells, and then add 540uL of formaldehyde stock solution (37% Fisher #F79-500). Final concentration is 1%. Incubate for exactly 5 min at room temperature, occasionally mixing flasks.
6. Add 1.7mL of 1.5M glycine to stop the crosslinking reaction (final concentration is 0.125 M).
7. Incubate 5 min at room temperature, occasionally mixing flasks.
8. Wash 3 x 1.5 min with 20mL ice-cold PBS.
9. Add 20mL of ice cold PBS and scrape cells. Rinse with an additional 10mL PBS. Add to 50mL conical tube and centrifuge at 1500 rpm at 4°C for 5 min. Remove supernatant. Store cell pellets at -80°C.

Cell lysis and chromatin sonication:

10. Resuspend cell pellet in 750ul RIPA plus 7.5ul protease inhibitors for probe sonicator (500uL of RIPA buffer plus 5uL protease inhibitors for UCD-200). (Calbiochem Inhibitor cocktail 1 #539131, 100x stock).
11. Leave cells on ice for 30 min with occasional vortexing.
12. Disrupt cells by drawing through a 30g needle 3X, put into 1.5ml tube (or 15 ml conical BD Falcon #352095 if using UCD-200).
13. Leave cells on ice for an additional 30 min.
14. Save 25ul of pre-sonicated sample, do not centrifuge.
15. Sonicate: Wear Hearing Protection!

1. Ultrasonic Processor 130W with 3mm micro-tip: (Use in BSL2 Hood) 30% output setting (continuous) for 4 x 1 min for RAWS and 5 x 1 min for BMM. Output watts = 3. Immerse sample probe ~ 1 cm into tube (same approximate level as where the tube begins to taper). If the probe is at the wrong level, sample foaming will result and sonication efficiency drops drastically. Place tubes in ice/water for at least 1 min between each sonication. Rinse probe between samples by sonicating in dH₂O or rinsing with 70% EtOH. Spin tubes on high (13,000 rpm) for 5 min.
2. UCD-200: Put EtOH washed adaptors on tubes. Sonicate on high setting for 10 cycles of 30 sec on/30 sec off. Fill empty slots with balance tubes. Transfer sample to 1.7ml tube and spin tubes on high (13,000 rpm) for 5 min.

16. Transfer supernatant to new tube (store chromatin samples at -80°C).
17. Save 25ul of sonicated sample to check sonication.

Check Sonication:

• Use 25ul of each pre-sonicated and post-sonicated samples (steps 14 and 17) and q.s. to 250ul with RIPA Buffer.
• Reverse the Crosslinks:
• Add 15ul of 4M NaCl to each sample.
• Incubate the samples for > 5 hr at 65°C.
• Vortex samples, then to each tube add:

10uL 1 M Tris, pH 8.0

5uL 0.5 M EDTA, pH 8.0

2.5uL Proteinase K (600 U/mL Fermentas #EO0491) (or 2ul of 900U/ml)

• Incubate >1 h at 42°C. Remove phenol:chloroform:isoamyl alcohol from fridge and let it warm to room temperature.
• In the hood: Add an equal volume (282.5ul) of phenol:chloroform:isoamyl alcohol (25:24:1 Fisher #BP1752-100). Draw from the lower phase.
• Vortex tubes and spin tubes at 4000 rpm for 5 min.
• Remove aqueous (upper) phase.
• To aqueous phase (~260 uL) add:

1^st: 2 uL pellet paint (Novagen #60049)

2^nd: 26 uL 3M sodium acetate pH 5.2; Mix sample briefly.

3^rd: 520 uL 100% ethanol; Vortex sample.

• Let stand 5 min at room temperature.
• Spin at 13,000 rpm for 10 minutes.
• Remove supernatant and air dry the pellet.
• Resuspend in 30 uL of molecular grade H₂O (Store samples at -20°C).
• Treat with Rnase A:
• 4 ul chromatin prep
• 1 ul 100mM Tris-HCl (pH 7.5)
• 4 ul 25mM MgCl₂
• 1 ul Rnase A (1/200)
• Incubate for 30 min in a 37°C waterbath and hold on ice.
• Run samples on 2% agarose gel with a 100bp DNA ladder.

Protein Concentration: Perform BioRad DC Protein assay on chromatin samples to determine protein concentration. Use Microplate assay protocol.

*note: The BSA used for protein assay is different than the BSA used for different steps in the IP and amplification.

• Prepare Reagent A': For every 1.0ml of Reagent A needed add 20ul of Reagent S.
• In a mixing plate prepare BSA standard dilutions in duplicate (Pierce #23209 stock concentration = 2mg/ml). The dilution of RIPA in water depends on the dilution of samples:

1.5mg/ml = 75ul BSA + 25 ul H₂O (with RIPA)

1.0mg/ml = 67ul 1.5mg/ml dilution + 33 ul H₂0 (with RIPA)

0.5mg/ml = 50ul 1.0mg/ml dilution + 50 ul H₂0 (with RIPA)

0.25mg/ml = 50ul 0.5mg/ml dilution + 50 ul H₂0 (with RIPA)

0.125mg/ml = 50ul 0.25mg/ml dilution + 50 ul H₂0 (with RIPA)

0.0625mg/ml = 50ul 0.125mg/ml dilution + 50 ul H₂0 (with RIPA)

0.0313mg/ml = 50ul 0.065mg/ml dilution + 50 ul H₂0 (with RIPA)

0.0 = 50ul H₂0 (with RIPA)

• In mixing plate prepare sample dilutions in duplicate in 20ul total volume:

Mouse BMM cells dilution = 1/5

Mouse Mast cells dilution = 1/10

RAW cells dilution = 1/20

• Pipet 5ul of standards and samples into a clean 96-well flat bottom plate.
• Add 25ul of Reagent A' (A+S) to each sample.
• Add 200ul of Reagent B to each sample. Pipet to mix. If bubbles form use a P1000 to gently blow air into the well to pop bubbles.
• Incubate 15 min at room temperature. Samples will be stable for about 1 hour.
• Read Absorbance at 750 nm using pathlength correction.

Sample preclearing and immunoprecipitation: Dilute samples (1ml final volume) to 0.5mg/ml for BMM.

For ChIP-chip: make an additional 60ul of 0.5mg/ml for BMM to use as real input chromatin controls: final volume = 1.06ml. Pull out 60ul and qs to 250ul with RIPA and set aside to reverse the crosslinks in step 29. Use the remaining 1ml for the next step:

18. Sample preclearing: Add 30uL of Salmon sperm DNA/protein-A agarose beads (Upstate #16-157), use large bore pipet tips. Incubate for 15 min with end over end mixing at 4°C.
19. Spin tubes at 7500 rpm for 2 min.
20. Transfer supernatant to new tube, discard beads.
21. Add antibody (concentration of antibody must be empirically determined) and 10ul BSA in 10X TBS (#37520 Pierce) to supernatant.
22. Incubate overnight at 4°C with end over end mixing; put on parafilm to prevent leaking or evaporation.
23. Add 70uL of Salmon sperm DNA/protein-A agarose beads. Incubate 30 min at 4°C with end over end mixing.
24. Spin tubes at 3000 rpm for 3 min.
25. Carefully remove the supernatant (try not to disturb the beads) Keep the beads!!!!!
26. Wash the beads for 5 min at RT with end over end mixing with 1mL of the following buffers. Spin tubes at 3000 rpm for 3 min after every wash to pellet the beads.

1. 3x RIPA
2. 3x RIPA + NaCl
3. 2x LiCl
4. 2x TE (for final TE wash use 30g needle to thoroughly remove all wash from beads before extraction)

27. Extract immunoprecipitates by adding 100uL of 1% SDS in TE to the beads. Mix and incubate the samples at 65°C for 15 min (in water bath). Spin tubes at 3000 rpm for 3 min, transfer eluate to fresh tube and wash beads with a further 150uL of 1% SDS in TE as above (incubate for 15 min at 65°C in water bath). Respin and add to first eluate. Save the eluates... the beads are no longer attached to what we want.

Reversal of crosslinks and final DNA purification:

28. Add 15uL of 4M NaCl to each sample including input controls (Tv=265uL).
29. Incubate the samples for > 5 hr at 65°C.
30. Vortex samples, then to each tube add:

10uL 1M Tris, pH 8.0

5uL 0.5M EDTA, pH 8.0

2.0uL Proteinase K (900 U/mL Fermentas #EO0491)

31. Incubate >1 h at 42°C. Remove phenol:chloroform:isoamyl alcohol from fridge and let it warm to room temperature.
32. In the hood: Add an equal volume (282.5ul) of phenol:chloroform:isoamyl alcohol (25:24:1 Fisher #BP1752-100). Draw from the lower phase.
33. Vortex tubes and spin tubes at 4000 rpm for 5 min.
34. Remove 240ul aqueous (upper) phase.
35. In the hood: Add 240ul of Chloroform:isomyl alcohol (24:1) to aqueous phase.
36. Vortex tubes and spin tubes at 4000rpm for 5 min.
37. Remove 220ul of aqueous phase.
38. To aqueous phase add:

1^st: 2uL pellet paint (Novagen #60049) or 5ul glycogen (Fermentas)

2^nd: 22uL 3M sodium acetate pH 5.2; Mix sample briefly.

3^rd: 4400uL 100% EtOH; Vortex sample.

39. Let stand 5 min at room temperature.
40. Spin at 13,000 rpm for 10 minutes.
41. Discard the supernatant and wash pellet with 440uL 70% EtOH.
42. Remove supernatant and air dry the pellet.
43. Resuspend DNA in 30uL of molecular grade water and store at -20°C.

Solutions: All solutions are made with Millipore water and should be steri-flipped or autoclaved prior to use. Always wear gloves and a lab coat and rinse all glassware with Millipore water prior to use.

1.5M glycine: 75.1 MW x final volume (in Liters) x 1.5M = grams of glycine

3M Sodium Acetate pH 5.2: Dissolve 408.3g of sodium acetate-3H₂0 in 800 ml of H₂0.

Adjust pH to 5.2 with glacial acetic acid. Bring up to 1L. Autoclave.

10% SDS: Wear a mask!! Dissolve 10g of electrophoresis-grade SDS in 75mL H₂0.

Heat to 68°C and stir to dissolve. Adjust pH to 7.2 with HCl if necessary. Final

volume is 0.1L. Do not autoclave!

RIPA Buffer - final volume is 50mL

10 mM Tris-HCl, pH 8.0 500 uL of 1.0 M stock

140 mM NaCl 1750 uL of 4M stock

1% Triton X-100 500 uL

0.1% SDS 500 uL of 10% stock

1% Deoxycholic acid sodium salt 0.5 g

TE Buffer - final volume is 50 mL

10 mM Tris-HCl, pH 8.0 500 uL of 1.0 M stock

1 mM EDTA, pH 8.0 100 uL of 0.5 M stock

1xRIPA Buffer with 1M NaCl - final volume is 100mL

0.1% SDS 1 mL of 10% stock

0.1% sodium deoxycholate 0.1 g

1% Triton X-100 1 mL

1 mM EDTA 200uL of 0.5 M stock

0.5 mM EGTA 100uL of 0.5 M stock

10 mM Tris-HCl, pH 8.0 1 mL of 1.0 M stock

140 mM NaCl 3.5mL of 4M

NaCl 5.844 g

LiCl Buffer - final volume is 100mL

0.25 M LiCl 1.06 g

1% IgePal CA-630 (same as Nonidet P-40) 1 mL

1% sodium deoxycholate 1 g

1 mM EDTA 200uL of .5 M stock

10 mM Tris-HCl pH 8.0 1 mL of 1 M stock

(2xRIPA with NaCl and LiCl buffers are from V.A. Spencer et.al. METHODS 31 (2003) 67-75.)

Primer A2 HPLC Purified PCR (adapted from Affymetrix protocol):

• Use 30% (10µl) of ChIP or input DNA for initial round of amplification.

• Set first round:
• DNA 10µl
• 5X sequenase buffer 4µl (Sequence USB#70775Y)
• Primer A2 (200 µM) 4µl (GTTTCCCAGTCACGGTCNNNNNNNNN) HPLC Purified

• Cycle conditions:
• 94°C 4 min
• Snap freeze (on ice)
• 10°C hold - add 1^st mix; place snap frozen samples back in cycler
• 10°C 5 min
• ramp to 37°C over 9 min (1°C/20sec)
• 37°C hold for 8 min
• Repeat the following 3 times:
• 94°C 4 min
• Snap freeze (on ice)
• 10°C hold - add 1.0µl diluted Sequenase; place snap frozen samples back in cycler
• 10°C 5 min
• ramp to 37°C over 9 min (1°C/20sec)
• 37°C hold for 8 min
• Hold 4°C

• During the first 10°C hold, add:
• 1^st mix:
• 0.2µl 10mg/ml BSA
• 1µl 0.1M DTT
• 0.5µl 25mM dNTPs
• 1.0µl diluted Sequenase (1/10 from 13 U/µl stock, use dilution buffer)

• Upon completion of Primer A2, purify PCR products using NucleoFast 96well PCR plates.
• Transfer samples to plate
• Apply 15 mm Hg pressure for 10-15 min. Once well appears dry, apply vacuum for an additional 30 sec.
• Add 70µl of dH₂0 to each well.
• Cover wells with plate sealer and vortex for 5 min at Speed 1 or 2.
• Pipet sample up and down minimum of 10x.
• Transfer sample to new tube.

• Primer B2 PCR:

• PCR:
• 20µl of Primer A2 DNA from above
• 10µl 10X PCR Buffer (plus Mg++) (Roche 11146165001)
• 3µl 25mM MgCl₂
• 1.5µl 25mM dNTPs + dUTP
• 4µl 100µM Primer B2 (GTTTCCCAGTCACGGTC)
• 10µl 1U/µl Taq (Roche)
• 49.25ul dH₂O

• Cycle conditions:
• 95°C 30 sec
• 45°C 30 sec
• 55°C 30 sec
• 72°C 1 min
• Repeat for total of 15 cycles
• 95°C 30 sec
• 45°C 30 sec
• 55°C 30 sec
• 72°C 1 min (+5 sec every round)
• Repeat for total of 15 cycles
• Hold 4°C

• Check amplified DNA on 2% agarose gel (3µl sample + 2µl dH2O + 1µl dye).
• Purify Primer B PCR products using NucleoFast 96well PCR plates - elute in 50µl.
• Quantitate with NanoDrop:
• Blank with 2ul dH₂O
• Measure absorbance of 1ul of sample (A₂₆₀ should be between 0.1 and 1.0 to ensure concentrations are accurate.)
• PCR Samples with specific primers used to PCR IP and run on 2% agarose gel.
• If necessary: speed vac 7.5ug of each sample down to <39.45µl at 45°C.

Fragment Samples:

• 4.8µl 10X cDNA Fragmentation buffer
• 1.5µl UDG 10U/µl
• 2.25µl APE1 100U/ µl
• < 39.45µl Amplified DNA (7.5 µg)
• Qs to 48µl with dH₂O

• Use a thermal cycler and incubate samples at 37°C for 1 hour, 93°C for 2 min, 4°C for at least 2 minutes. Hold samples on ice. Have samples Bio-Analyzed to determine quality of fragmented product.

TdT Labeling:

• DNA from fragmentation 45µl
• 5X TdT Buffer 12µl
• 5mM GeneChip DNA Labeling Reagent 1µl
• Terminal Deoxytransferase 2µl

• Incubate at 37°C for 1 hr, 70°C for 10 min, hold at 4°C (must be at 4°C for a minimum of 2 minutes).

• Check labeling:

• Prepare Neutravidin solution of 2mg/ml in PBS.
• Mix samples:
• 1^st: 2µl sample + 3µl dH₂0
• 2^nd: 2µl sample + 5µl Neutravidin Solution
• Incubate 5 minutes.
• Add 6x loading dye to each sample (1.0µl and 1.4µl)
• Load Sample in 4-20% TBE gel, with 1X TBE Buffer. Use two 50bp ladder lanes, one on each side of samples.
• Run the gel at 150 volts until the front dye almost reaches the bottom (approx 1 hr).
• While gel is running, prepare 50mL of a 1X SYBR Gold solution in TBE for staining the gel. For fresh solution a gel should be stained for approximately 10 minutes.

ChIP-Chip Data Analysis Methods

Two different 25-mer Affymetrix GeneChip® oligonucleotide microarrays were used for the ChIP-chip analysis. One was a custom array, which densely tiled 20 kb upstream and 20 kb downstream (and selectively, the coding regions) of genes chosen based on differential expression in preliminary microarray studies using murine RAW 264.7 cells stimulated for 60 minutes by LPS, Pam3CSK4, or Pam2CSK4. The other was the Affymetrix mouse promoter array. Hybridization and scanning of the custom tiling array was carried out using standard protocols and reagents from Affymetrix.

Analysis of the ChIP-chip data was carried out using Model-based analysis of tiling arrays (MAT), after first mapping the microarray probes to the NCBI Build 37 mouse genome assembly (mm9, July 2007), and filtering out probes that mapped to repeat regions. Probes mapping to multiple locations were excluded from all analysis. Retained probe sequences were then checked for potential overlaps with regions of repeat sequences. Tables of repeats identified by RepeatMasker and those termed simple repeats were downloaded from the UCSC mouse genome annotation browser for the July 2007 assembly. Probe sequences overlapping with a repeat region by at least a single nucleotide were also excluded from all analysis. The filtered probe mappings were then used to create custom BPMAP files for use with MAT.

MAT was run in a hierarchical manner utilizing both "multiple replicates of ChIP-chips and controls" and "single replicates of ChIP-chips/multiple replicates of controls" scenarios to efficiently identify enriched segments that were consistent across biological replicates. For most antibodies and conditions, three biological replicate experiments were performed. To illustrate our approach, take as an example analysis of the binding of transcription factor PU.1 in bone-marrow derived macrophages stimulated with LPS for 4hrs. Three independent ChIP-chip biological replicate experiments investigating PU.1 binding in macrophages stimulated with LPS for four hours were performed, which, after ChIP, DNA fragmentation, labeling, hyrbridization, etc, produced three CEL files. These three CEL files were analyzed together using MAT in the "multiple replicates of ChIP-chips and controls" scenario. For controls, we used our database of negative control CEL files for ChIP experiments using normal rabbit serum (NRS) in place of an antibody. For the custom tiling array, we have 30 NRS ChIP-chip CEL files, for the Affymetrix promoter array, we have 10 NRS ChIP-chip CEL files. MAT analysis parameters were set to their default values (intensity analysis: BandWidth = 300, MaxGap = 300, MinProbe = 10, Tvalue = 0), and segments with P-values less than 0.2 were retained for downstream processing. Running the three PU.1 ChIP-chips in comparison to the NRS ChIP-chips using MAT generates an output file (.bed.xls) of enriched segments and their associated MAT scores. In our experience, these enriched segments obtained from a "replicate combined" type of analysis generally constitute a superset of segments that would be identified from running the replicates individually with MAT. Thus, a segment being enriched in this analysis does not imply that the segment is consistently enriched across all replicates.

To address the issue of replicate consistency, we also run each of the individual PU.1 CEL files in comparison to the negative control set of ChIP-chips, giving rise to three .bed.xls files with enriched segments for each individual replicate. We also run each of the individual NRS CEL files in comparison to the complete set of negative control NRS CEL files. We then go back to the original .bed.xls file from the replicate combined analysis and find the MAT scores for the segments in the individual replicates that overlap. We then take these MAT scores from the mapping and perform a simple one-sided t-test to identify segments that have statistically significantly higher MAT scores for PU.1 than the negative control NRS ChIP-chips. Given the possibility for outlier ChIP-chips, we also perform the analysis using all possible subset pairs of ChIP-chips for the antibody, and retain the enrichment statistics for the best combination. In this manner, we identify both "high confidence" (enriched segments consistent across all three replicates) and "moderate confidence" (enriched segments consistent across at least two biological replicates) segments. Empirically defined thresholds for reporting segments are: MAT > 1.4 (for replicate combined .bed.xls file from which candidate enriched segments are selected), p < 0.05 (for t-test comparing segments across replicates), and ΔMAT (difference between average of PU.1 MAT scores and NRS MAT scores) > 0.8. Figures illustrating our overall approach are shown below.

Click to see a larger image.

Click to see a larger image.