Jeffrey Perkel has been a scientific writer and editor since 2000. He holds a PhD in Cell and Molecular Biology from the University of Pennsylvania, and did postdoctoral work at the University of Pennsylvania and at Harvard Medical School.
In the world of gene regulation, perhaps no topic is hotter than epigenetics. And when it comes to epigenetics, perhaps no technique is more ubiquitous that ChIP.
ChIP, or chromatin immunoprecipitation, uses an antibody to a chromatin-associated protein, such as a histone or transcription factor, to identify sequences associated with that protein or protein modification. It’s a way to map epigenetic changes across the genome. “ChIPping” with antibodies to histone H3 trimethylated on lysine-4 (i.e., H3K4me3) tends to highlight active genes, for instance, while antibodies to H3K27me3 demarcate genes that are silent.
ChIP is just one of a family of related techniques. Researchers can use RIP (RNA immunoprecipitation) to identify RNAs associated with specific RNA-binding proteins and MeDIP (methyl-DNA immunoprecipitation) to pull down methylated chromatin sequences.
In theory, these methods are simple. In practice, they are anything but. “In general, the ChIP process is very complex and can be quite lengthy,” says Barbara Kaboord, research and development manager for protein biology at Thermo Fisher Scientific. Here we review three key variables to bear in mind as you set about incorporating these methods in your own lab.
Make good chromatin
As with most lab methods, nucleic-acid IP operates on the GIGO principle: “Garbage in, garbage out.” In this case, the input is chromatin, and if your starting material is of poor quality—well, you know the rest.
“Getting chromatin is easy. Getting ChIP-able chromatin is a different story,” says John Rosenfeld, group leader for epigenetics research and development at EMD Millipore.
In particular, chromatin must be broken down into pieces a few hundred base pairs long, typically using sonication. But sonication is highly variable. Do it too long or too harshly, and you risk denaturing your protein, dissociating it from the nucleic acid or destroying the epitope your antibody recognizes. The process requires considerable optimization, and the sheer profusion of options can be daunting, says Rosenfeld. “With a given cell line and sonicator there literally are an infinite number of protocols that can work.”
The alternative is enzymatic fragmentation, an approach that is undeniably simpler and gentler, and also more reproducible, says Chris Fry, ChIP development group leader at Cell Signaling Technologies (CST).
CST’s SimpleChIP® and SimpleChIP Plus Enzymatic Chromatin IP Kits rely on micrococcal nuclease, and according to Fry, they make it easy for users to dial in on the desired fragment size. “We have used it on more than 20 different cell lines and tissues, and you always get the same fragment length,” he says, as long as users stick to a constant ratio of enzyme to cell number.
Thermo Fisher Scientific’s Pierce Magnetic ChIP Kit also uses micrococcal nuclease but includes an optional sonication step, as well, says Barbara Kaboord, research and development manager for protein biology. In this case, Kaboord explains, sonication is not for fragmentation so much as sensitivity, as it helps release more chromatin from the nucleus. New users, says Kaboord, “might be more comfortable with enzymatic manipulation; it’s a bit more straightforward and easier to optimize.”
That said, the enzymatic approach does have drawbacks, says Life Technologies scientist Loni Pickle. For one thing, nucleases sometimes exhibit sequence biases, thereby potentially skewing results. Enzymatic digestion also is not always compatible with formaldehyde crosslinking, a key step in ChIP. Finally, she says, “it’s a little harder to optimize fragment size.”
As a result, Pickle says, researchers using digestion “have to do more work to show enzymatic digestion isn’t affecting … results, whereas sonication is so well accepted.”
If you’re going to try sonication, Rosenfeld says there are four key variables to consider: cell density, power and the time and number of cycles. “Make sure you’re not optimizing all of them [at once],” he says.
Validate your antibodies
The other key ingredient of a successful nucleic-acid IP experiment is the antibody you use for the pull-down. Nucleic-acid IP is far more demanding of antibodies than, say, Western blotting, says Karin Abarca Heidemann, director of research and development at Rockland Immunochemicals. Antibodies in Western blots recognize a denatured protein, for instance. But “ChIP is an application where you need to make sure the antibody recognizes the tertiary structure of the protein or an exposed epitope.”
Your best bet is to get ChIP (or RIP or MeDIP)-validated antibodies. Rockland, for instance, offers a line of validated antibodies for epigenetics applications called Epi-Plus™. According to Abarca Heidemann, Epi-Plus antibodies are validated across multiple assays, including Western blotting, immunohistochemistry, immunofluorescence, immunoprecipitation and ChIP—a process that can take the company three to four months per antibody.
At CST, antibodies are put through the proverbial ringer, including high- and low-expressing cells, RNA interference, tagged expression constructs and treatments that are known to change protein expression, localization or modification, says Fry. The company even uses peptide arrays to assess specificity of their histone modification antibodies. “You shouldn’t settle for a Western blot,” he says. “The more assays you have showing specificity, the more confidence you can have that it will pick up the right target.”
“The biggest pain point [in ChIP] is probably making sure you have a good antibody,” says Rosenfeld, whose company, EMD Millipore, offers a line of 90 or so ChIP-validated antibodies called ChIPAb+™, plus a smaller number of RIPAb+™ reagents for RIP.
For methylated DNA enrichment, users can use antibodies targeting 5-methylcytosine (6-mC) or 5-hydroxymethylcytosine. (Active Motif’s MeDIP and hMeDIP kits operate on this principle, for instance.) Alternatively, users can enrich methylated sequences using methyl-binding domain proteins (MBDs), a technique sometimes called MIRA (Methylated CpG Island Recovery Assay). Active Motif’s MethylCollector™ Ultra kit uses such proteins to enrich methylated sequences.
According to Kyle Hondorp, product manager at Active Motif, the difference between antibodies and MBPs is that anti-5-mC antibodies recognize methyl marks in the context of single-stranded DNA, while MBDs interact with double-stranded molecules—a difference that can complicate some downstream processing steps.
Researchers studying RNA-protein interactions also have several options. Traditional RIP is essentially ChIP without the crosslinking step, and a number of companies offer antibodies and kits for this purpose, including EMD Millipore. There’s also iCLIP, a variant of RIP that maps interactions on the RNA sequence itself.
To study chromatin-associated RNAs, there’s Active Motif’s RNA ChIP-IT® kit, basically a ChIP kit optimized to recover chromatin-associated RNAs. EMD Millipore has two similar kits in the pipeline (one with crosslinking to identify more transient interactions, and one without), which should be available by the end of Q1, says Rosenfeld.
Alternatively, for researchers interested in RNA-protein interactions from the RNA perspective, there are techniques such as ChIRP, CHART and RAP, all of which use a biotinylated probe to pull out complementary RNAs and whatever associated proteins are along for the ride. The Pierce Magnetic RNA-Protein Pull-Down Kit is a variant of this approach. According to Kaboord, this kit enables researchers to attach a specific RNA to a bead and use it to fish for interacting proteins, which they can then identify by Western blotting or mass spectrometry.
Consider cell number
ChIP traditionally has been performed using cultured cells, and sample inputs of 1 million to 10 million cells per reaction were typical. Today, researchers increasingly are applying the technique to more precious samples, such as tumor and tissue biopsies, stem cell populations, archived formalin-fixed paraffin-embedded (FFPE) samples or cells enriched by fluorescence-activated cell sorting (FACS) or laser-capture microdissection. In such cases, a million cells can be hard to come by.
The actual number of cells you need depends on the target you’re looking at, says Kaboord. Histone modifications are relatively abundant, and researchers ChIPping such targets typically can get away with at least 10-fold less starting material. Transcription factors, though, generally are less abundant. In that case, she says, the typical input is about a million cells, but, “if you have good signal-to-noise, you can probably cut back in the future.”
“A million cells in general is not that hard to achieve, certainly in tissue culture, but it is a daunting number when you realize that that is a per-reaction requirement,” says Rosenfeld.
One reason for that high cell number, Rosenfeld explains, is the method’s high background. Increasingly, though, vendors have been optimizing their kits to handle smaller input amounts, gaining better recoveries by (among other things) optimizing buffers and replacing porous agarose beads with magnetic ones.
EMD Millipore gained a big boost in sensitivity for its Magna ChIP™ HiSens kit by reducing detergent in buffer, siliconizing tubes to prevent material from sticking to the sides and also incorporating a simple “tube switch step” before the final elution, Rosenfeld says. It now requires only about 100,000 cells for transcription factors and as few as 10,000 for histone modifications.
Enzymatic fragmentation also boosts recovery relative to sonication, says Fry, simply by virtue of being gentler. CST’s SimpleChIP Plus Enzymatic Chromatin IP Kits kits call for about 4 million cells – enough to get about three IPs from a single 15-cm dish of adherent cells, he says. But “with histones, we can get away with 10-fold less and still get great signal.”
These optimized methods also are typically shorter than they once were, and more high-throughput. According to Pickle, ChIP protocols used to take days. Now, Life Technologies’ MAGnify™ ChIP kit protocol takes just five hours, and it can easily be extended to large sample numbers.
Still, for all their optimizations, nucleic-acid IP methods look largely the same as they always have, enriching antibody-chromatin complexes on beads. One company, though, has developed an alternative approach.
Chromatrap®, from Porvair Sciences, a subsidiary of filtration company Porvair plc, is a protein A/G-coupled BioVyon filter that captures antibody complexes, no beads required. According to company CEO Ben Stocks, the method takes just five hours from start to finish, is compatible with automation and yields as much as 25 times more material, all while requiring less sample on the front end. Both spin column and microplate-based formats are available.
For researchers looking for an ever-greater return on ever-shrinking samples, such optimizations could be just what they need to make chromatin immunoprecipitation a reality.