Cornell molecular geneticist Paul Soloway wants to know where epigenetic marks coincide – and not just in which patients, or in which tissues, or even in which cells, but on which individual stretches of chromatin. He and co-PI Harold Craighead, an engineering physicist at Cornell, have an Epigenomics Roadmap grant that will help them find out. They’re developing a turnkey device to probe and sort fluorescently-tagged fragments on the nanoscale.
What will this device be like?
“Right now it’s a silica wafer fabricated using etching methods that are commonly used in the electronics industry, mounted on a fluorescent scope. … Essentially a nanoscale tube with electric current hooked up to it. It’s very similar to a flow cytometer in terms of the principle of operation. But instead of flowing cells we’d be flowing chromatin fragments” prepped by routine methods. “And instead of them going through a micron-scale device they would go through a nanometer scale device.
“We would like to put it in a box, with the optics integrated – to reduce the cost, make it more user-friendly, and obviate the need for a scope.”
Nanoscale Sorting: Not Your Standard Flow
“The scale is actually very important for the throughput: if we were to do this using a standard flow cytometer, our throughput would be 15,000 fold lower than it can be with a single nanoscale device.(We anticipate running many tubes in parallel, magnifying the difference that much more.) We do use routine physical principles of electric field potentials to flow materials through the device. But that’s probably where the similarity ends.
“The flow chamber itself is 250 nm x 500 nm in a tube, and the interrogation volume is .16 femtoliters (0.16×10^-10 μl). Individual molecules can be directed through the nanoscale device.
“The input of material can be far lower than is required for chromatin IP. Chromatin IP might require on the order of 10^7 cells. We could do this kind of analysis with literally 10s of cells.
“And the throughput is pretty high. We haven’t pushed the limits … but 4000 molecules per minute per channel is pretty routine, and we can pretty much image whatever labels they may have on them. And our devices can carry thousands of parallel channels (though we currently only use one at a time).”
Applications in Epigenetics Research
What will it do for epigenetics research?
“Chromatin would be stained with multiple, independent probes, each detecting its own individual epigenetic mark, rather than doing chromatin IPs with one antibody at a time.
“The idea of single molecule analysis is fundamentally different from bulk population analysis. This is essentially a single molecule method. There are great advantages to looking at things at the single molecule scale, one of them being when there are coincident marks you can unambiguously identify them.
“With chromatin IPs, you’re querying an entire population [and] you need to do sequential chromatin IPs [Re-ChIP] to validate that you really, truly have coincidence of marks. Otherwise, how can you tell whether a given nucleosome carries histone H3 tri-methylated on both lysine 9 and lysine 27?”
Is the device just for histone mods?
“The earlier work from Harold, my collaborator, had used essentially the same devices for imaging individual molecules of DNA. And they had a proof of principle paper, which showed they could take lambda DNA markers and get a histogram that looked very similar to a densitometry scan of the gel. Except rather than using nanogram quantities of DNA this used on the order of 10,000 individual DNA molecules.
“Our goal is to be able to include all epigenetic marks that are measurable as long as there’s a fluorescent probe that can be developed to detect it. So we do envision being able to detect simultaneously histone modifications as well as DNA methylation.”
Probing Epigenetic Marks
“To detect methylated DNA, we use the probe generously provided by Adrian Bird’s lab, which is an expression vector for MBD1 – methyl DNA binding protein 1. And that can bind to methylated DNA in a duplex.”
For chromatin, “we’ve selected aptamers for very specific reasons right now, but we’d like to generalize this for antibodies.
“We image the material as it’s flowing through, collecting the fluorescent signal data through appropriate detectors. Right now we are interrogating just one nanoscale channel, but we envision it’s quite reasonable to multiplex these,” using a CCD.
Then what?
Just like with flow cytometry, determining which marks are found together in a given cell type or following a given treatment might be all that’s needed.
Or populations that might have two, or three, or more epigenetic marks of interest can be sorted “based on color – in a preparative manner, so that we can then choose the chromatin fragments that have the marks of interest.
“We have some of our custom-designed circuitry, which can on-the-fly read the fluorescent signature of molecules as they’re coming through the interrogation volume, and make a decision as to how to sort them and where to sort them. The sorted material would then be shunted into a chamber, which would then be collected for downstream analysis. And we envision something like Solexa sequencing of the output — very similar to ChipSeq that is done right now.