Everyone knows that 2D gels are a difficult and labor-intensive way to separate proteins. The gels are fragile and hard to handle, a spot can contain more than one protein, and with all the liquid handling steps required for spot picking and digestion — even when automated — proteins are inevitably lost to tube walls. The problem with trying to get rid of 2D gels, however, is that they are ensconced in the labs of many proteomics researchers, and that for better or for worse an image of separated proteins is an intuitive way to understand the data.
But is this really a problem? Denis Hochstrasser’s group at the University of Geneva Hospital, in collaboration with the proteomics team at Applied Biosystems, think not. Rather than fight the 2D gel establishment by trying to introduce new approaches, based on liquid chromatography, for example, the two groups have worked diligently over the last few years to develop a concept that embraces the 2D gel platform — as well as the even more established tool of Western blot analysis — to create a generic tool that could eliminate many of the deficiencies of 2D gels. They call the tool the “molecular scanner,” and Applied Biosystems exclusively licensed the technology three years ago.
Electroblots
The molecular scanner, according to Hochstrasser, is a technique for safely moving the proteins held in a 1D or 2D gel to a sturdier PVDF membrane via an electroblotting transfer step that simultaneously digests the proteins while maintaining their spatial resolution. Once deposed on the surface of the PVDF membrane, a properly configured MALDI-TOF mass spectrometer can scan the surface of the membrane, producing a spatially resolved dataset of protein identifications based on mass spectra. With the advent of MALDI TOF/TOF and MALDI Q-TOF mass spectrometry, the amino acid sequence of individual peptides could even be determined.
The idea isn’t altogether new — Hochstrasser first dreamt it up in the late ‘80s — but recent technological advancements in mass spectrometry and a methodical effort to develop the necessary chemistries have brought the idea closer to the fore. ABI researchers won’t predict when a product might hit the market — it seems likely a commercially available system is at least a year away from production — but when it does it’s likely to receive an enthusiastic response. Given the number of proteomics labs already using 2D gels, “if many of them started using [the molecular scanner], from an instrumentation manufacturing point of view, that’s an attractive market,” said Ruedi Aebersold of the Institute for Systems Biology in Seattle.
Potentially Practical
Of course running any sort of gel carries with it inherent limitations, and Aebersold, as a proponent of LC-based approaches to protein separation, is quick to point these out. Running prefractionated narrow gradient gels is unlikely to be considered a high-throughput method for separating low-abundance proteins, for example, and automating gel handling and analysis doesn’t come cheap. Nevertheless, Hochstrasser and his partners at ABI say the advances they’ve made in developing the molecular scanner represent significant advantages over the traditional approach to 2D gel analysis, and perhaps over liquid chromatography approaches as well.
On a conceptual level, scanning the surface of a membrane embedded with proteins represents a departure from the sequential approaches necessary when analyzing spot-picked proteins, or proteins eluted from a chromatograph, Hochstrasser said. In practical terms, analyzing proteins in parallel should result in substantially higher-throughput experiments. Given a MALDI instrument with sufficiently high laser repetition, a mass spectrometer could complete a preliminary peptide mass fingerprint analysis of a typical 1D gel in an hour, he added.
Scientists at ABI seconded this assertion. A MALDI mass spectrometer outfitted with a 3 Hz laser used to take several days to scan a single lane of a 1D gel, but with recent advances in MALDI technology, that same scan might take under an hour, said Dave Hicks, the director of proteomics application marketing at ABI’s proteomics facility in Framingham, Mass. “It was Denis’ invention combined with his discussions with Marvin Vestal before the introduction of the TOF/TOF, that said ‘Geez, this is going to become more than just academically interesting, this could potentially become practical in the [coming years], if that’s the direction instrumentation development is going to take,’” Hicks said.
Vella, a vice president for science and technology at ABI’s Framingham facility, added that his group has developed a cocktail of reagents that improves upon the basic concept of Western blotting for transferring the proteins out of the polyacrylamide gel. “It’s a cocktail that will allow us to electroblot all the proteins, including high molecular weight proteins, [so] that we can achieve a fairly good recovery and convert them to peptides,” he said.
Returning to the Same Spot
Hochstrasser pointed out other potential advantages, in addition to the common argument that protein spot-picking and delivery to a MALDI target causes significant amounts of protein to adhere to the liquid-handling apparatus. The molecular scanner, he said, avoids the assumption that each spot contains only one protein, because the resolution of the experiment is limited by the size of the laser, not the size of the spot cutter. Furthermore, in contrast to LC-based protein analysis techniques that typically rely on electrospray ionization, with MALDI a researcher can typically return to the same protein spot to perform additional experiments.
In fact, Hochstrasser’s Swiss collaborators on the molecular scanner project, including Ron Appel, who leads the proteome informatics group at the Swiss Institute for Bioinformatics, are working with ABI to create data visualization and storage tools that would enable researchers to build mass spectra-based image maps. This “staining by mass spectrometry” approach would allow scientists to view a preliminary peptide mass fingerprinting scan collected by a MALDI-TOF, determine where the data for protein identification were insufficient, and return to that spot using a MALDI-TOF/TOF or Q-TOF mass spectrometer to obtain the full amino acid sequence.
Technical Troublespots
But the Geneva and ABI teams still have much to resolve. Routinely scanning entire gels would generate an immense amount of mass spectrometry data, and “it’s by no means trivial to process these gel replicates in a mass spectrometer,” said Aebersold. “If you think of shooting every 100 microns, and you want to image a 20 cm by 20 cm area, and you need to shoot every spot 10 times to get a sufficient number of ions, that’s a lot of laser shots,” he said.
In addition, Aebersold pointed out that the molecular scanner as it is currently conceived lacks the ability to make quantitative measurements of protein expression. “Mass spectrometry in general is not good for quantitation,” he said. However, he added that the molecular scanner could in theory be combined with quantitative techniques, such as the ICAT reagents, which rely on stable isotope tags. ABI has expressed interest in the idea, he added.
Hicks and Vella at ABI agreed that their group must still tackle the challenge of developing software for automating the data collection and analysis, a process which could involve modifying how their AB & AB/MDS Sciex mass spectrometers are configured and operated. There are other technical issues as well, including developing a technique for easily mounting the membrane onto the MALDI target, and automatically determining accurate molecular weights following the electroblotting transfer. In addition to working with Hochstrasser''''''''''''''''s group to test the apparatus, ABI also plans to seek out additional sites to help define the capabilities of the molecular scanner and validate with various samples and workflows.
But the remaining challenges do nothing to dampen Hochstrasser’s enthusiasm for the ultimate usefulness of the instrument. After all, it is his brainchild: “I wish I had [it] for a year already,” he said. “I have so many people coming to ask our lab here to sequence and analyze proteins, [and] I would love to tell them, ‘Look, buy a 1D gel, buy the trypsin membrane, buy the collecting membrane, run the experiment, and when you have the collecting membrane come to us. We spray it with the matrix, we stick it in the MALDI-TOF, and we give you back your image with the proteins identified. It’s a dream!”
— JSM