NEW YORK (GenomeWeb) − Researchers in Toronto have developed a method to dissect and sequence DNA from structures within the nucleus of individual cells, allowing them to study the spatial arrangement of genes in great detail.
Chromosomal DNA is not randomly distributed in the nucleus − the position of genes is highly organized. Certain gene loci come together in sub-nuclear structures, called nuclear bodies, which contain proteins and regulate their expression.
Researchers have used a number of techniques − such as ChIP-seq and variations of chromosome conformation capture, or 3C – to study protein-DNA interactions in nuclear bodies and other protein complexes in cell populations.
"But what we really wanted was another method that could be used on a single-cell basis," David Bazett-Jones, a senior scientist at the Hospital for Sick Children and a professor of biochemistry at the University of Toronto, told In Sequence. Such a method, he said, would allow them to pick out individual nuclear bodies from within a nucleus and study the genes associated with it.
The method he and his colleagues came up with, published last month in the journal Small, involves dissecting DNA from the sub-nuclear structures under a scanning electron microscope followed by sequencing. It was developed in collaboration with Yu Sun, a professor of mechanical engineering at the University of Toronto.
The approach has the potential to "provide unique information about the spatial arrangement of the genome inside cells that cannot be easily acquired by any other methods," said Job Dekker, a professor of biochemistry and molecular pharmacology at the University of Massachusetts Medical School, who developed the 3C method and was not involved in the development of the new method.
Imaging approaches, he explained, require a priori knowledge of the loci that might be present in a nuclear structure, and 3C-type methods, while unbiased, provide no information about the absolute position of loci inside the nucleus. "This new approach solves both these limitations by allowing identification of genomic loci associated with observable nuclear structures," he said.
For their proof-of-concept study, the researchers focused on two types of nuclear bodies: histone locus bodies, or HLB, and promyelocytic leukemia nuclear, or PML, bodies. HLBs are known to associate with histone gene loci on chromosomes 1 and 6. Mammalian cells typically contain up to 20 PMLs but certain leukemia cells have far fewer.
The scientists first labeled the nuclear bodies of interest with an immunofluorescent tag, froze the cells, and cut them into sections about 300 nanometers thick. Under a fluorescent microscope, they then identified nuclear bodies. After transferring the slide to a scanning electron microscope, they correlated the SEM image with the fluorescence image, allowing them to hone in on the exact nuclear body they were interested in.
They then introduced a nano spatula, made of carbon-coated glass with a tip smaller than 100 nanometers, and used it to scoop up chromatin from the target area containing low femtogram amounts of DNA. After retracting the tip from the SEM, they PCR-amplified the DNA, followed by Sanger sequencing.
One challenge they had to overcome was the extracted material blowing off the tip when air was reintroduced to the SEM. To prevent that from happening, they welded the sample onto the tip using a technique called electron beam induced deposition.
Another problem was contamination with DNA from other sources. The researchers solved that by ligating linkers containing PCR priming sites to the DNA after cutting it with restriction enzymes. That way, only the DNA from the cells they studied would be amplified in the PCR reaction.
The success rate of the method was about 16 percent, but according to David Anchel, a researcher in Bazett-Jones' lab and one of the lead authors, this can be improved by decreasing the number of PCR cycles and switching to next-generation sequencing, so that not only those DNA regions that are efficiently amplified get sequenced.
Another way to improve the method is to prevent damage to the DNA from the electron beam, which makes it impossible to amplify the DNA. For example, the sample could be protected by coating it with a chemical such as glycerol, Bazett-Jones said.
In the future, the fluorescent tags could also be replaced by gold-tagged antibodies, which can be seen in the SEM, so a fluorescence microscope would no longer be required.
In addition to the SEM method, Bazett-Jones and his team are working on an entirely different approach for sequencing sub-nuclear structures that does not require a SEM or a nanomanipulator, he said. It uses a two-photon laser to create double-strand breaks in the DNA near the structure of interest, which is imaged under a fluorescence microscope. The broken DNA is then tagged with linkers, pulled into a tube, amplified, and sequenced. "It is very similar to this SEM approach, it has its own pitfalls and own advantages and disadvantages," he said, noting that they plan the method for publication in a journal in the near future.
The current method is low in throughput – it takes about a day to process 10 cells, the nano-dissection being the bottleneck. "This tool is really a hypothesis generator, as well as confirming some ideas about how nuclear bodies might function," Bazett-Jones said. It identifies genes present at a particular site that can then be further studied with population-based techniques. "With even a few genes, you suddenly have all kinds of hypotheses as to why those genes [are there], how important it is that they are there, what happens when you prevent them from going to these bodies."
In the HLBs they analyzed, the researchers found DNA from chromosomes 1 and 6 to be enriched, as expected from prior studies. For PML bodies, they identified a number of novel genomic loci that were significantly associated with them, which they further studied using fluorescence in situ hybridization.
Bazett-Jones and his team now plan to use the technique to study the function of specific PML bodies with unique characteristics, for example, those that are enriched in a certain transcription factor. "Now we can go in and say, 'Around that particular body, what sets of genes do we find that we would not find in a neighboring PML body that doesn't have that high level of that transcription factor?'" he said.
Such research could have implications for our understanding of cancer. Leukemia cells, for example, often only retain one or two PML bodies, he said, and knowing whether there is a unique set of genes they associate with could help explain the disease state of those cells.