Genomewide Identification of Protein B Locations Using Chromatin Immunoprecipitation Coupled
Abstract Interactions between cis-acting elements and proteins play a key role in transcriptional regulation of all known organisms . 
To better understand these interactions , researchers developed a method that couples chromatin immunoprecipitation with microarrays ( also known as ChIP-chip ) , which is capable of providing a whole-genome map of protein-DNA interactions . 
This versatile and high-throughput strategy is initiated by formaldehyde-mediated cross-linking of DNA and proteins , followed by cell lysis , DNA fragmentation , and immu-nopurification . 
The immunoprecipitated DNA fragments are then purified from the proteins by reverse-cross-linking followed by amplification , labeling , and hybridization to a whole-genome tiling microarray against a reference sample . 
The enriched signals obtained from the microarray then are normalized by the reference sample and used to generate the whole-genome map of protein-DNA interactions . 
The protocol described here has been used for discovering the genomewide distribution of RNA polymerase and several transcription factors of Escherichia coli . 
Keywords ChIP-chip , chromatin immunoprecipitation , microarray , RNA polymer-ase , transcription factor , transcription factor binding 
1 Introduction
In the postgenomic era , systematic and high-throughput technologies allow us to enumerate biological components on a large scale . 
As one of the approaches , chromatin immunoprecipitation coupled with microarrays ( ChIP-chip ) has been used to explore the genomewide interactions between proteins and cis-acting elements , such as a comprehensive identification of transcriptional regulatory regions of the human genome [ 1 ] and in other organisms [ 2 -- 8 ] . 
Maps of genomewide protein-DNA interactions are essential in understanding many fundamental biological components , such as the logic of regulatory networks , chromosome structure , DNA replication , DNA repair , heat shock response , and metabolism [ 9 -- 14 ] . 
Due to a dramatic improvement in microarray technologies , it is possible to create high-resolution maps of genomewide protein-DNA interactions [ 1 , 3 ] . 
Although many variations of the ChIP-chip protocol exist , the essential steps begin with an in vivo fixation of protein-DNA complexes mediated by formaldehyde . 
The cells are then lysed and the DNA fragmented to a desired size using sonication . 
The protein bound DNA is then enriched through immunoprecipitation by the specific antibody against the protein of interest ( or against epitope-tag flanked with the target protein ) , then purified of the protein through a heatmediated reversal of the cross-links . 
The DNA then is purified , amplified , and hybridized to a microarray [ 15 -- 20 ] . 
Several features make the ChIP-chip protocol difficult for all applications . 
First , not every available antibody is efficiently applicable to the immunoprecipitation of the cross-linked protein-DNA complexes [ 17 ] . 
This limitation probably is due to antibody specificity and an affinity against the target protein and epitope masking by formaldehyde-mediated cross-linking . 
To address this limitation , epitope-tagging methods have been developed to use in ChIP-chip of yeast and E. coli [ 4 , 21 ] . 
A second limitation is amplifying the ChIP DNA , which often is required to obtain sufficient amounts of DNA for labeling and hybridization without introduction of bias . 
Generally , two PCR-based methods have been widely used to amplify the ChIP DNA . 
The first method uses a degenerate oligo that randomly anneals to DNA [ 22 ] , while the other uses ligationmediated PCR ( LM-PCR ) , and achieves roughly 100 - to 1,000-fold amplification [ 2 ] . 
Finally , current microarray platforms are fairly expensive , but these costs are beginning to decrease significantly as new technologies are developed and competition among suppliers increases . 
This chapter describes the ChIP-chip protocol used for mapping the RNA polymerase binding in Escherichia coli along with several transcription factors . 
2 Materials
2.1 Cell Culture and Cross-Linking
1 . 
10X M9 salt stock ( M9 minimal medium ) : Dissolve 60 g Na HPO , 30 g 2 4 KH PO , 5 g NaCl , and 10 g NH Cl in dH O ; adjust to 1 L final volume ; and 2 4 4 2 sterilize by autoclaving for 20 min at 15 psi ( 1.05 kg/cm2 ) on a liquid cycle . 
Store at room temperature . 
2 . 
10X glucose stock : Dissolve 20 g glucose in dH O , adjust to 1 L final volume , 2 and sterilize by passing the solution through a 0.22 - µm filter . 
Store at room temperature . 
3 . 
500X MgSO stock ( 1 M ) : Dissolve 22.85 g MgSO .6 H O in dH O , adjust the 4 4 2 2 volume to 0.1 L , and sterilize by autoclaving for 20 min at 15 psi ( 1.05 kg/cm2 ) on a liquid cycle . 
Store at room temperature 
4 . 
1000X CaCl stock ( 0.1 M ) : Dissolve 1.47 g CaCl .2 H O in dH O , make up the 2 2 2 2 volume to 0.1 L , and sterilize by autoclaving for 20 minutes at 15 psi ( 1.05 kg / cm2 ) on a liquid cycle . 
Store at room temperature . 
5 . 
100X trace element solution : Dissolve 16.67 g FeCl .6 H O , 0.18 g ZnSO .6 H O , 3 2 4 2 0.12 g CuCl .2 H O , 0.12 g MnSO . 
H O , 0.18 g CoCl .6 H O , 0.12 g Na MoO , 2 2 4 2 2 2 2 4 and 22.25 g Na EDTA .2 H O in dH O ; adjust to 1 L final volume ; and sterilize by 2 2 2 passing the solution through a 0.22 - µm filter . 
Store at room temperature . 
6 . 
37 % formaldehyde solution ( Fisher Scientific , F79-500 ) . 
Store in the chemical hood . 
7 . 
2.5 M glycine solution : Dissolve 187.68 g glycine in dH O , make up the final 2 volume to 1 L , and sterilize by passing the solution through a 0.22 - µm filter . 
Store at room temperature . 
8 . 
Tris-buffered saline ( Sigma , T5912 ) . 
Store at 4 °C . 
2.2 Cell Lysis, Preparation of Chromatin Complexes, and Immunoprecipitation
1 . 
Lysis buffer : 10 mM Tris-HCl , pH 7.5 , 100 mM NaCl , and 1 mM EDTA . 
Store at 4 °C . 
2 . 
Lysozyme ( Epicentre , R1810M ) . 
Follow storage instructions provided by supplier . 
3 . 
Protease inhibitor cocktail ( Sigma , P8465 ) : Dissolve at 200 mg/mL in DMSO and dilute in four volumes of dH O. Make fresh prior to use and keep cold on ice . 
2 4 . 
IP buffer : 100 mM Tris-HCl , pH 7.5 , 200 mM NaCl , 2 mM EDTA , and 2 % Triton X-100 . 
Store at 4 °C . 
5 . 
Misonix 3000 sonicator equipped with microtip . 
6 . 
Antibody : Anti-RNAP mouse antibody ( Neoclone , W0001 , W0002 ) , anti-myc mouse antibody ( Santa Cruz Biotechnology , sc-40 ) , Mouse IgG ( Upstate , 12 -- 371 ) . 
Follow storage instructions provided by supplier . 
7 . 
Bead washing buffer : Dissolve BSA ( Sigma , A7906 ) at 5 mg/mL in PBS . 
Make fresh prior to use and keep cold on ice . 
8 . 
Dynabeads : Pan mouse IgG ( Invitrogen , 112.05 ) . 
Follow storage instructions provided by supplier . 
3 . 
IP washing buffer 3 ( W3 buffer ) : 10 mM Tris-HCl , pH 8.0 , 250 mM LiCl ( Sigma , L7026 ) , 1 mM EDTA , and 1 % Triton X-100 . 
Store at 4 °C . 
4 . 
TE buffer : 10 mM Tris-HCl , pH 8.0 , and 1 mM EDTA . 
Store at 4 °C . 
5 . 
IP Elution buffer : 10 mM Tris-HCl , pH 8.0 , 1 mM EDTA , and 1 % SDS . 
Store at room temperature . 
3 . 
Protease K solution , 20 mg/mL ( Invitrogen , 25530-049 ) . 
4 . 
3 M sodium acetate , pH 5.2 . 
5 . 
Qiagen PCR Purification Kit ( Qiagen , 28106 ) . 
6 . 
2X SYBR green PCR master mix ( Qiagen , 204145 ) . 
Follow storage instructions provided by supplier . 
7 . 
iCycler real-time PCR detection system ( Bio-Rad , CA ) . 
8 . 
Sequenase ( USB , 70775Z ) , 5X Sequenase buffer ( supplied with Sequenase ) , and Sequenase dilution buffer ( supplied with Sequenase ) . 
9 . 
0.1 M DTT ( USB , 70726 ) . 
10 . 
0.5 mg/mL BSA ( diluted from 10 mg/mL stock from NEB , B9001S ) . 
11 . 
Rand 9-Ns primer : 5 ′ TGGAAATCCGAGTGAGTNNNNNNNNN 3 ′ . 
12 . 
Rand univ primer : 5 ′ TGGAAATCCGAGTGAGT 3 ′ . 
13 . 
dNTP mix ( Takara , 4030 ) . 
14 . 
pfu turbo polymerase ( Stratagene , 600135 ) and 10X pfu buffer ( supplied with polymerase ) . 
1 . 
Cy3-labeled nine-mers ( TriLink Biotechnologies , N46-0001-50 ) . 
2 . 
Cy5-labeled nine-mers ( TriLink Biotechnologies , N46-0002-50 ) . 
3 . 
Random nine-mer buffer : 125 mM Tris-HCl , pH 8.0 , 12.5 mM MgCl , and 2 0.175 % β-mercaptoethanol . 
4 . 
50X dNTP mix solution : 10 mM Tris-HCl , pH 8.0 , 1 mM EDTA , 10 mM dNTP . 
5 . 
100 U Klenow fragment ( NEB , M0212M ) . 
6 . 
MAUI hybridization unit ( BioMicro Systems , Utah ) . 
7 . 
NimbleGen custom microarrays ( Design ID : 1881 , Escherichia coli whole-genome tiling array consisting of 371,034 oligonucleotides spaced 25 bp apart across the whole genome ) . 
8 . 
NimbleGen Array Reuse Kit 40 ( NimbleGen , KIT001-2 ) . 
9 . 
Axon scanner , model 4000B . 
10 . 
Cy3 CPK6 50-mer ( IDT , Custom oligo synthesis ) 
13 . 
0.1 M DTT . 
14 . 
Wash I : 250 mL ddH O , 2.5 mL 20X SSC , 5 mL 10 % SDS , and 250 µL 0.1 M 2 DTT . 
15 . 
Wash II : 250 mL ddH O , 2.5 mL 20X SSC , and 250 µL 0.1 M DTT . 
2 16 . 
Wash III : 250 mL ddH O , 625 µL 20X SSC , and 250 µL 0.1 M DTT . 
2 
2.6 Normalization and Peak Identification
SignalMap ( www.nimblegen.com ) , Matlab ver 7.0.4 with bioinformatics toolbox ( www.mathworks.com ) , Microsoft Excel ( www.microsoft.com ) , Mpeak ( the complete program is available from http://www.stat.ucla.edu/~zmdl/mpeak/ ) . 
3 Methods
3.1 Cell Culture and Cross-Linking
1 . 
Add 2.8 mL of 37 % formaldehyde solution directly to each 100mL culture that contains the number of cells used for a ChIP-chip experiment ( see Note 1 ) . 
Continue to incubate with gentle shaking for 20min at room temperature ( see Note 2 ) . 
2 . 
Add 5 mL of 2.5 M glycine solution directly to each 100 mL sample followed by incubation for 5 min at room temperature . 
3 . 
Centrifuge at 4,700 g for 5 min at 4 °C and pour off the supernatant . 
Wash each pellet three times with one volume of ice-cold TBS ( see Note 3 ) and resuspend the cell pellet in the TBS remaining after decanting the supernatant . 
Transfer the sample to a new 1.5-mL tube and centrifuge at the maximum speed ( 15,800 g ) for 1 min . 
4 . 
Remove all supernatant using a pipette and store the cell pellet at − 80 °C until use . 
3.2 Cell Lysis, Preparation of Chromatin Complexes and Immunoprecipitation
1 . 
Completely resuspend the cell pellet in 0.5 mL of lysis buffer . 
Add 40 µL of protease inhibitor cocktail and 0.5 µL of lysozyme solution . 
Incubate the sample for 30 min at 37 °C on a rocker , and then add 0.5 mL of IP buffer and 40 µL of protease inhibitor cocktail . 
Continue to incubate on ice until the lysate is cleared ( see Note 4 ) 
2 . 
Shear the lysate by sonicating for four 20 s pulse with a Misonix microtip sonicator at output setting 2 ( see Note 5 ) . 
To avoid overheating the sample , keep the sample on ice at least 1 min between cycles . 
Centrifuge at 15,800 g for 10 min at 4 °C to clarify the chromatin solution . 
Take 10 µL of the chromatin solution to use as `` total DNA ( tDNA ) '' sample and store at − 20 °C for further use . 
3 . 
Split the chromatin solution into two 0.5-mL aliquots . 
Add 1 µg of specific antibody to one aliquot , and 1 µg of mouse IgG to the other ( see Notes 6 and 7 ) . 
Incubate samples overnight at 4 °C on a rocker . 
1 . 
For each sample , wash 50 µL of Dynabeads Pan mouse IgG beads three times with 1 mL of bead washing buffer . 
Add the incubated samples ( `` with specific antibody [ iDNA ] '' and `` with mock antibody [ mDNA ] '' ) to the washed magnetic beads ( see Note 8 ) . 
Continue incubation for at least 6 h at 4 °C on a rocker at 8 rpm . 
2 . 
Collect the magnetic beads using an MPC magnet and remove the supernatant by aspiration . 
If needed , save the supernatant for the `` unbound fraction sample . '' 
Sequentially , wash the beads twice with 1 mL of W1 buffer , once with 1 mL of W2 buffer , once with 1 mL of W3 buffer , and once with 1 mL of TE buffer ( see Notes 9 and 10 ) . 
3 . 
Resuspend the beads in 200 µL of IP elution buffer and add 190 µL of that solution to the tDNA sample . 
Continue to incubate overnight at 65 °C to reverse cross-links . 
3.4 Purification, qPCR and Amplification of DNA
1 . 
Pull down the magnetic beads using an MPC magnet and transfer 200 µL of supernatant to a new tube . 
Add 200 µL of TE buffer and 8 µL of RNaseA solution ( 10 mg/mL ) to each sample . 
Continue to incubate for 2 h at 37 °C . 
Add 4 µL of protease K solution ( 20 mg/mL ) to each sample and continue to incubate for 2 h at 55 °C . 
Purify DNA using the Qiagen PCR Purification Kit and elute with 50 µL of ddH O ( see Note 11 ) . 
At this point , qPCR can be done using the iDNA , 2 mDNA , and tDNA to confirm the enrichment fold required to run a microarray ( see Note 12 ) . 
2 . 
Set up the round A reaction mix on ice as described in Table 9.1 ( see Note 13 ) . 
3 . 
In a PCR tube , mix 7 µL of the iDNA or mDNA , 2 µL of 5X Sequenase buffer , and 1 µL of 40 µM Rand 9-Ns primer . 
Cycle to 94 °C for 2 min then cool to 10 °C . 
Add 5.05 µL of round A mix . 
Ramp up from 10 °C to 37 °C over 8 min , hold at 37 °C for 8 min , heat to 94 °C for 2 min , then cool to 10 °C . 
Add 1.2 µL mixture of 0.9 µL of Sequenase dilution buffer and 0.3 µL of Sequenase enzyme 
Ramp up from 10 °C to 37 °C over 8 min , hold at 37 °C for 8 min , then cool to 4 °C . 
Dilute the samples by addition of 45 µL of ddH O. 2 4 . 
Set up the round B reaction mix on ice as described in Table 9.2 . 
Transfer 15 µL of the diluted template into a new PCR tube and add 85 µL of the round B reaction mix to each tube . 
Prepare four tubes per sample to achieve enough DNA for microarray hybridization . 
Cycle to 94 °C for 30 s , 40 °C for 30 s , 50 °C for 30 s , and 72 °C for 2 min ( 25 cycles ) ( see Note 14 ) . 
Purify the amplified DNA using a Qiagen PCR Purification Kit and elute with 120 µL EB buffer supplied with the kit . 
Use one purification column per two reactions and combine two elutions in a new tube . 
The total volume per sample should be 240 µL . 
Add 24 µL of ice-cold 3 M sodium acetate ( pH 5.2 ) and 700 µL of ethanol . 
Continue to incubate overnight at − 20 °C . 
3.5 Labeling, Hybridization, and Scanning
1 . 
Centrifuge at 37,000 g at 4 °C for 30 min and wash the pellet with cold 80 % ethanol . 
Dry the pellet and dissolve the pellet in 9 µL ( iDNA ) and 7 µL ( mDNA ) of ddH O , respectively . 
Dilute 1 µL of the sample in 99 µL of EB buffer and 2 measure the DNA quantity and quality using a spectrophotometer . 
The DNA yields range from 5 to 10 µg and A should be between 1.8 and 2.0 . 
260/28 
2 . 
Dilute Cy3 and Cy5 dye-labeled nine-mers to 1 OD in 42 µL random nine-mer buffer . 
Aliquot to 40 µL individual reaction volumes in 0.2 mL thin-walled PCR tubes . 
Add 1 µg of iDNA and mDNA to the Cy5 and Cy3 tubes , respectively , and bring the final volume to 80 µL using ddH O. Denature the samples in a thermo-2 cycler at 98 °C for 10 min and then quickly chill in an ice-water bath . 
Add 20 µL of 50X dNTP mix solution and mix well by pipetting at least 10 times . 
3 . 
Incubate at 37 °C for 2 h in a thermocycler ( light sensitive ) and then stop the reaction by the addition of 10 µL of 0.5 M EDTA . 
Transfer the reaction to a 1.5-mL tube . 
Precipitate the labeled samples by adding 11.5 µL of 5 M NaCl and 110 µL of isopropanol to each tube . 
Vortex and incubate for 10 min at room temperature in the dark . 
4 . 
Centrifuge at 37,000 g for 10 min and remove the supernatant with a pipette . 
Rinse the pellet with 500 µL of 80 % ice-cold ethanol and centrifuge again at 37,000 g for 2 min . 
Remove the supernatant with a pipette and dry the pellet for 5 min in a SpeedVac using low heat and protection from light . 
Rehydrate the dried pellets in 25 µL of ddH O. Vortex for 30 s and quick spin to collect con-2 tents at bottom of tube . 
Measure the A in each sample to determine DNA 260 concentration . 
Typical yields range from 10 to 30 µg per reaction . 
5 . 
Set MAUI hybridization unit to 42 °C and allow time for the temperature to stabilize . 
Combine 13 µg of the both iDNA ( Cy5 ) and mDNA ( Cy3 ) into a single 1.5 mL tube ( see Note 15 ) . 
Dry the combined contents in a SpeedVac on low heat . 
Resuspend the sample in 10.9 µL of ddH O and vortex to com-2 pletely dissolve the sample . 
Spin down the tube briefly to collect the contents in the bottom . 
6 . 
Using the NimbleGen Array Reuse Kit , add 19.5 µL of 2X hybridization buffer , 7.8 µL of hybridization component A , 0.4 µL of Cy3 CPK6 50-mer oligo , and 0.4 µL of Cy5 CPK6 50-mer oligo to each sample ( see Note 16 ) . 
Mix the tube briefly , and then spin down to collect the contents in the bottom and place at 95 °C for 5 min . 
7 . 
Immediately transfer the tube to the MAUI 42 °C sample block and hold at this temperature until ready for sample loading ( see Note 17 ) . 
Place the MAUI mixer SL hybridization chamber on the array using the provided assembly/disassembly jig and carefully follow MAUI setup instructions . 
Use the braying tool to remove all air bubbles from the adhesive gasket around the outside of the hybridization chamber . 
8 . 
Load the sample using the pipette supplied with the MAUI station , following manufacturer 's instructions . 
During loading , a small amount ( 3 -- 7 µL ) of the sample may flow out of the outlet port . 
Confirm that there are no bubbles in the chamber . 
9 . 
Place the loaded array into one of the four MAUI bays and let it equilibrate for 30 s. Wipe off any sample leakage at the ports and adhere MAUI stickers to both ports . 
Close the bay clamp and select mix mode B. Hold down the mix button to start mixing . 
Confirm that the mixing is in progress before closing the cover . 
Hybridize the sample overnight 
10 . 
Remove chip from MAUI hybridization station , load it back into the MAUI assembly/disassembly jig , and immerse in the shallow 250 mL Wash I ( see Note 18 ) . 
While the chip is submerged , carefully peel off the mixer . 
Gently agitate the chip in Wash I for 10 -- 15 s ( see Note 19 ) . 
11 . 
Transfer the slide into a slide rack in the second dish of Wash I and incubate 2 min with agitation . 
Transfer to Wash II and incubate 1 min with agitation . 
Rock the dish to move the wash over the tops of the arrays . 
Transfer to Wash III and incubate for 15 s with agitation . 
12 . 
Remove the array and spin dry in an array-drying unit for 1 min . 
Store the dried array in a dark desiccator and proceed immediately the scanning of the arrays . 
3.6 Data Normalization and Peak Identification
1 . 
Conceptually , the normalization approach using sum of intensity of each channel is the simplest way , whose assumption is that the total DNA used is same for both channels . 
Calculate the sum of each Cy5 and Cy3 channel and the ratio between the total intensity of Cy5 and that of Cy3 . 
Multiply the ratio ( N ) to each data point ( G , R ) and then calculate log ratio ( R ) of each point ( Eq . 
9.1 -- Eq . 
k k 2 9.3 ) . 
Log ratio can be used for the peak identification step ( see Note 20 ) . 
2 Narray ∑ Rk N = k = 1 Narray ∑ Gk k = 1 
2 . 
The binding sites should appear in the data as runs of consecutive points with enhanced amplitude shown in Fig. 9.1 . 
Several peak identification methods have been developed using an error model to compute p-values for single array probes [ 2 ] , a sliding window approach with Gaussian error function [ 23 ] , double-regression model [ 1 ] , a percentile approach [ 24 ] , a hierarchical empirical Bayes model [ 25 ] , a joint-binding deconvolution method [ 26 ] , a tiled model-based analysis of tiling-arrays method [ 27 ] , and a variance stabilization approach [ 28 ] . 
Of those methods , we used the double-regression method to identify the protein binding sites across genome 
4 Notes
1 . 
The protocol describes the growth of E. coli MG1655 in minimal media to the mid exponential phase . 
Because the protein-DNA interactions are very sensitive to the physiological state of the cell , it is very important to control the growth conditions as tightly as possible . 
Generally , each immunoprecipitation requires 5 × 107 ∼ 1 × 108 cells ( approximately , OD 0.4 ~ 1.0 ) . 
To find out all the potential 600 promoters of E. coli , the cells were treated with rifampicin for 30 min [ 3 ] . 
2 . 
The cross-linking time should be empirically optimized to each protein-DNA-antibody combination . 
We found that the cross-linking time ( 20 min ) and for-maldehyde concentration ( 1 % ) described in this protocol are generally applicable to RNAP ( β and β ′ subunits ) and several transcription factors of E. coli . 
However , theoretically insufficient cross-linking would result in the ina-bility to capture protein-DNA complexes . 
On the other hand , overcross-linking can make giant protein-DNA complexes and cause the epitope-masking problem . 
Formaldehyde should be used with appropriate safety measures , such as protective gloves , glasses , and clothing and adequate ventilation . 
Formaldehyde waste should be disposed of according to regulations for hazardous waste . 
3 . 
PBS can be used to wash cell pellets instead of TBS . 
4 . 
Cell lysis condition varies depending on the target cell type . 
For E. coli , the lysis method using lysozyme works very well . 
If the lysozyme is not available , alternative methods , such as French press or glass-bead-based lysis , can be used . 
5 . 
The sonication step should be optimized to achieve the optimum fragmentation of DNA ( 300 -- 1,000 bp ) . 
One way to optimize the sonication step is to take 20 µL of the chromatin solution from each sonication cycle and determine the average size of the fragmented DNA in the chromatin solution . 
Add 80 µL of IP elution buffer to each 20 µL of chromatin solution and continue to incubate at least 6 h ( or overnight ) at 65 °C to reverse cross-link the chromatin complexes . 
Add 100 µL of TE and 4 µL of RNaseA solution ( 10 mg/mL ) and continue to incubate for 2 h at 37 °C . 
Add 2 µL of protease K solution ( 20 mg/mL ) and continue to incubate for 2 h at 55 °C . 
Purify the DNA with a Qiagen PCR Purification Kit and elute using 30 µL of EB buffer supplied with the kit . 
The average size of DNA then is analyzed on a 2 % agarose gel ( Fig. 9.2 ) . 
Under - and overfragmentation result in a loss of resolution of binding events and more noise in the microarray analysis , respectively ( 16 ) . 
6 . 
Not every available antibody is efficiently applicable to the ChIP-chip , as mentioned in introduction . 
Unfortunately , routine assay methods such as Western blotting can not be used to test whether an antibody would be suitable for ChIP-chip . 
To determine whether an antibody is suitable for ChIP-chip experiments requires an actual ChIP assays followed by qPCRs of several known-binding sites . 
To address the limited use of antibodies , epitope-tagging methods have been developed for ChIP-chip of yeast and E. coli without alteration in function of target proteins [ 4 , 21 ] . 
The main advantages of epitope-tagging are the availability of a universal antibody and the ability to insert multiple copies of the epitope , which increases the immunoprecipitation yield [ 21 ] 
7 . 
When epitope-tagged proteins are being used , wild-type cells can be used as a control . 
In this case , add the same antibody against the epitope used to both the epitope-tagged sample and the wild-type sample [ 21 ] . 
8 . 
Alternatively , the antibodies can be conjugated to the Dynabeads Pan mouse IgG beads prior to the immunoprecipitation . 
Wash 50 µL of the magnetic beads three times , using the bead washing buffer , and resuspend the washed beads in 250 µL of bead washing buffer . 
Add 10 µg of antibody to the magnetic beads and incubate overnight at 4 °C on a rocker . 
Wash the beads three times in 1 mL of bead washing buffer and collect the beads using a MPC magnet prior to use . 
Then , add 0.5 mL of the sheared chromatin solution to the antibody preconjugated magnetic beads . 
Continue to incubate overnight at 4 °C on a rocker . 
9 . 
Add 1 mL of the washing buffer to each tube and gently resuspend beads . 
This can be done by removing the tubes from the MPC magnet and rotating the tubes for 1 min with rocker . 
Collect the magnetic beads using an MPC magnet and remove the supernatant by aspiration . 
Normally , to remove chromatin complexes that are nonspecifically bound to the antibody or magnetic beads , intensive washing steps are needed . 
Do not minimize this step . 
10 . 
Because not every protein-antibody interaction can bear the high salt and detergent conditions , we recommend washing the beads four times using only W1 buffer and once using TE buffer . 
If more stringent conditions are required , increase salt concentrations such as NaCl and LiCl in W2 and W3 buffer , respectively . 
At the final washing step using TE , the magnetic beads are collected on the tube wall loosely . 
Do not use aspiration but only a pipette . 
11 . 
Instead of a Qiagen PCR Purification Kit , it is possible to use the conventional purification method , consisting of a phenol : chloroform : isoamyl alcoho extraction and ethanol precipitation . 
We chose the Qiagen kit to maintain consistency among samples . 
12 . 
Examples for qPCR are shown in Fig. 9.3 . 
We used the ChIP-chip approach to study the association of RNA polymerase across the genome of E. coli under aerobic conditions . 
The ChIP DNA samples were used as a template for qPCR with primer pairs of promoter regions from pgi , cyoA , and sdhC , whose gene expressions were up-regulated under aerobic conditions . 
As a control , the promoter region of dmsA was used , as the gene is down-regulated under aerobic conditions . 
13 . 
DNA amplification consists of two steps . 
Round A involves two rounds of DNA synthesis using the immunoprecipitated DNA as template , a partially degenerate primer ( Rand 9-Ns primer ) , and T7 Sequenase . 
Round B consists of 25 -- 30 cycles of PCR using a primer ( Rand universal primer ) that anneals to the specific region of the Rand 9-Ns primer . 
14 . 
The number of cycles should be optimized prior to the PCR amplification to prevent amplification bias . 
The random amplification method does not amplify DNA linearly , so the linearity depends on the number of cycles . 
Prepare five amplification tubes and sample each tube at the 15th , 20th , 25th , 30th , and 35th cycles . 
Measure the DNA concentration and perform qPCR as described in Note 12 . 
Compare both the quantity and quality of each cycle . 
15 . 
The tubes should be protected from light during handling to prevent photob-leaching of the light-sensitive Cy dyes . 
16 . 
CPK6 50-mer oligos are include in the hybridization as controls and hybridize to alignment features on NimbleGen arrays . 
They are required for proper extraction of array data from the scanned image . 
17 . 
This procedure describes the process for hybridization of samples prepared by chromatin immunoprecipitation and amplified by random PCR on NimbleGen custom microarrays ( Design ID : 1881 , Escherichia coli whole-genome tiling array , consisting of 371,034 oligonucleotides spaced 25 bp apart across the whole genome ) . 
The use of ORF arrays has limited power of ChIP experiments , since most transcription factor binding sites are located in the intergenic region and , therefore , not included on these arrays . 
The most robust array design for ChIP-chip has contiguous tiled DNA fragments that represent the entire genome , including the noncoding regions [ 1 , 3 , 17 ] . 
18 . 
Prior to removing the array from the MAUI Hybridization Station , prepare two 250 mL dishes of Wash I , and one each for Wash II and Wash III . 
One dish for Wash I should be shallow and be wide enough to accommodate the array and mixer loaded in the MAUI assemble/disassembly jig . 
The lid from a 1-mL pipette tip box works well . 
Place the remaining three wash solutions in 300 mL Tissue-Tek slide staining dishes . 
19 . 
Peel the hybridization chamber off very slowly to prevent the slide from cracking . 
Do not let the surface of the slide dry out at any point during washing . 
20 . 
Various normalization methods can be used to normalize the tiling array data . 
Alternatively , another normalization method for the tiling array data is the use of a biweight mean [ 24 ] . 
Calculate the log of the ratio of Cy5 to Cy3 for each 2 data point and then subtract the biweight mean of this log from each data 2 point . 
Acknowledgments The protocol described here was based on previous work by many other research groups in this field . 
The pioneers in this field are Dr. Young 's group at MIT , Dr. Lieb 's group at the University of North Carolina , Dr. Grunstein 's group at Yale University , Dr. Ren 's group at UCSD , and others . 
We thank anyone whose work was not referenced in here . 
This work is supported by NIH research grant no . 
GM62791 .