NEW YORK – While capillary electrophoresis (CE) has long played a niche role in proteomics, technology developments and shifting research trends could increase the approach's prominence.
In particular, the increasing speed of mass spectrometry instruments and the move toward analysis of extremely small samples, down even to the single-cell level, could lead researchers to more seriously consider CE as a separation technique for certain bottom-up proteomics applications.
Generally speaking, CE separates molecules based on their charge. Two forms of CE, capillary zone electrophoresis, which separates molecules based on their mass and charge, and capillary isoelectric focusing, which separates molecules based on their isoelectric points (the pH at which they have no net charge), are most commonly used for proteomics, though both techniques are far less common in proteomics than liquid chromatography (LC).
The technology has a number of potential advantages compared to LC — including greater separation speed and better sensitivity — that have made it an intriguing option for proteomics researchers, but due to certain technical challenges it has remained little used in the field.
Perhaps most notably, CE systems have small loading capacities, which have limited the amount of sample researchers can analyze when working with CE. Another issue has been the difficulty of interfacing CE systems with mass spec instruments in a way that achieves robust and reproducible movement of sample from one to the other.
These remain challenges, but CE developers have made progress addressing them in recent years.
Boston-based mass spec firm 908 Devices, which sells a chip-based CE separation device called the ZipChip, is tackling the loading capacity issue by incorporating small solid phase extraction (SPE) beds in its chips, which allows for on-chip sample preconcentration.
"You can flow your sample through the [SPE bed], pre-enrich it right next to the injection site, and then inject it right onto the [CE] chip," said Will Thompson, principal scientist at 908 Devices. He said the company had managed in this way to concentrate samples by around 100-fold.
The effort is "still very much an R&D exercise," Thompson said, adding that the company continues to evaluate the approach with collaborators. "But it's really kind of a requirement. To do CE in a proteomics space, you have to be able to do some sort of preconcentration step there, whether it is offline or online."
New York City-based CE firm CMP Scientific is likewise looking to push CE into proteomics. CEO James Xia said his company is also involved in R&D work to increase the amount of sample that CE separations can handle.
Xia said injection amounts are still a limitation for the technology, but noted that the move within proteomics toward extremely small samples could also heighten CE's relevance even despite its sample loading limits.
"I think CE is the right tool for single-cell proteomics," Xia said, suggesting that the technology's combination of speed and high sensitivity is well suited to single-cell applications. He added that workflows developed to allow for sample processing in the CE capillary could also help reduce sample loss, which is a key area of concern for single-cell proteomics.
"If you don't really have a stationary phase, and you are able to operate under conditions where things don't bind to surfaces, CE might have some fundamental advantages there," said 908 Devices' Thompson.
In fact, last month a team led by Northeastern University researcher Alexander Ivanov, who is among the leaders in implementing CE for proteomics, published a paper in Analytical Chemistry using on-capillary cell lysis for a top-down single-cell proteomic experiment. By processing the cells in the capillary, the researchers were able to eliminate sample handling steps and reduce sample loss, they wrote.
To the extent CE has made inroads into proteomics, it has largely been for top-down and intact protein work, as the technology's ability to work with non-denatured proteins as well as its ability to maintain higher separation efficiency with large molecules than LC has drawn interest from researchers in this space.
Xia said, though, that he has seen growing interest among bottom-up proteomics researchers. Last month, the company signed a comarketing agreement with Agilent Technologies to integrate its CE systems with Agilent's mass specs for applications including proteomics.
Jarrod Marto, a researcher at the Dana-Farber Cancer Institute and a member of 908 Devices' scientific advisory board, agreed, noting that he sees interest "picking up a little bit."
Like Thompson and Xia, Marto said single-cell proteomics and other research methods with limited sample sizes, such as analysis of tumor biopsies, are an obvious match for CE as it currently exists. He added that he expects continued technological improvement boosting the amount of sample CE can deal with "is going to broaden the application base very rapidly across the field of proteomics."
Thompson said he also sees potential for CE for analysis of protein post-translational modifications.
"Normally we're trying to dig these [modified] peptides out of the grass by doing sample pre-enrichment," he said. "What if we can do that a different way, by separating them not by hydrophobicity [as in LC], but by charge?"
Along those lines, Xia noted that CMP has internal research projects ongoing using its CE technology for glycoproteomic analyses looking for disease biomarkers.
In July, Marto and collaborators including researchers from Bruker and 908 Devices published a study in Analytical Chemistry using 908's on-chip sample preconcentration approach with the company's ZipChip CE system to profile deubiquitinases (DUB) on Bruker's timsTOF Pro mass spectrometer. The scientists developed a workflow on this system capable of quantifying the binding of DUB inhibitors across 49 different DUBs in under 15 minutes.
Based on his initial work with 908's on-chip preconcentration technique, Marto said the approach "looks very promising."
The coupling of the device to Bruker's timsTOF Pro, which is one of the fastest mass spec instruments currently in wide use for proteomics research, points to another trend that CE advocates believe favor the technology — ongoing improvements in mass spec speed and the shift toward shorter experiment run times.
"Electrophoretic-based separations are lightning fast. It's very high peak capacity per unit time," Marto said, noting that this fits into the larger move within proteomics toward increasing sample throughput through data-independent, label-free types of acquisition strategies.
For applications like clinical research and single-cell proteomics "people really see an advantage in speed," Thompson said. He noted that while CE has commonly used long capillaries that have not prioritized speed, "there's really no fundamental reason it can't be faster."
As with LC, longer CE separations provide more separation power than shorter ones, but Thompson said 908 sees a growing demand for speed that, along with increasing mass spec speeds, "plays into our strength."
CMP, on the other hand, is focused on longer CE separations, which Xia said he believes are well suited to proteomics experiments looking at thousands of proteins. The company's ECE-001 CE system uses capillaries with a minimum length of 50 cm. 908 Devices' ZipChip uses capillary lengths of either 10 cm or 22 cm.
Looking to the future, Marto, whose lab has long experience developing online multidimensional LC-based sample separation systems, said that CE's relative mechanical simplicity raises interesting possibilities for developing more complex, multiplexed separation workflows.
With more elaborate LC systems, researchers are "embracing [increased] mechanical complexity to drive higher peak capacity separations," he said. "But the cost of that is sometimes the brittleness of relying on ultra-high pressure LC pumps and fittings and all of that."
"The beauty of the electrophoretic approach is that you are driving things with high voltage rather than high pressure," he said. "And I think it is interesting to start thinking about building more complex separation schemes around that, as opposed to stringing together multiple high pressure LC pumps."