NEW YORK — Recent publications indicate that new mass spectrometry platforms are significantly improving the depth of coverage achievable in single-cell proteomics experiments.
In work using Bruker's timsTOF Ultra instrument and Thermo Fisher Scientific's Orbitrap Astral, researchers have more than doubled the number of proteins commonly measured in single-cell experiments.
Both instruments debuted at the American Society for Mass Spectrometry annual meeting in June. The timsTOF Ultra is the latest in Bruker's timsTOF line. It improves upon the small sample analysis capabilities of the existing timsTOF SCP while also allowing for high-performance analysis of larger samples.
The Orbitrap Astral, meanwhile, marked Thermo Fisher's foray into a new technology, its Astral (for asymmetric track lossless) analyzer, which, like a time-of-flight (TOF) analyzer, measures the travel of ions along a track within the instrument and their arrival at the surface of a detector.
A BioRxiv preprint posted in November by researchers at the University of Copenhagen provides one of the first looks at the Astral's single-cell proteomics capabilities. In the study, the scientists were able to identify more than 5,000 proteins in individual HeLa cells, which they noted is more than twice the roughly 2,000 proteins per cell previous single-cell experiments have typically topped out at.
The Astral has been "a real game-changer," said Jesper Olsen, professor and vice director of the Novo Nordisk Foundation Center for Protein Research at the University of Copenhagen and senior author on the preprint, noting that the Astral analyzer "provides us with extremely high sensitivity."
"We are seeing very substantial increases in the performance of single-cell protein analysis, and the new instruments are clearly an important part of those increases," said Nikolai Slavov, associate professor of bioengineering at Northeastern University and director of the Parallel Squared Technology Institute (PTI), which focuses on single-cell proteomics research.
The new mass spec instrumentation has enabled both larger numbers of peptide and protein identifications and more accurate quantification of those peptides and proteins, Slavov said.
Karl Mechtler, head of the proteomics tech hub at Vienna's Research Institute of Molecular Pathology, said that he believes it will be difficult to keep up with advances in single-cell work without access to the newer instruments.
"I'm in discussion with other people in the single-cell field, and we all together agree that the new generation of instruments is a big step forward," he said. "If you don't have the new instruments, you cannot really do single-cell proteomics anymore. They are so much better; it is really amazing."
Mechtler's lab has the Ultra and plans to purchase an Orbitrap Astral, both of which it will use in single-cell experiments.
At ASMS he presented data generated using the Ultra at a Bruker facility in which he and his colleagues were able to measure roughly 6,000 protein groups with a median coefficient of variation of 10 percent in a 250-picogram standard (approximately equivalent to the amount of protein in a single cell) compared to around 5,000 proteins with a median CV of 12 percent using the older timsTOF SCP. Looking at actual single HeLa and K562 cells (as opposed to standards), they identified, respectively, 3,803 and 3,221 proteins using the Ultra.
Mechtler said that in his lab's work with the Ultra since installing it at the Research Institute of Molecular Pathology, they have managed a slightly lower depth of coverage, down 5 percent to 10 percent from the figures he reported at ASMS. He said that he and his colleagues are still working to optimize the performance of the system, having only had it in-house for around six months.
In a BioRxiv preprint published in November by Slavov and colleagues at Northeastern, the researchers used the timsTOF Ultra for single-cell experiments in which they quantified between 3,000 and 3,700 proteins per cell.
While both instruments offer substantial boosts in performance, Slavov suggested that the timsTOF Ultra has certain advantages for single-cell work.
He said that the Ultra's trapped ion mobility (TIMS) capability allows the instrument to isolate and fragment a larger proportion of the ions delivered to the instrument than can the Orbitrap Astral, which he said offers advantages in terms of sensitivity, particularly in the case of single-cell experiments in which relatively small numbers of ions are generated.
In the data-independent acquisition (DIA) experiments commonly used for both single-cell and bulk proteomics experiments, the mass spec cycles through a series of m/z isolation windows, fragmenting all the precursor ions present in the m/z window being analyzed at a given point in time. The Ultra is able to use its TIMS device to time the release of ions from the device to match the m/z window being fragmented at that point in time. In the Astral, on the other hand, release of ions into the instrument can't be timed in this way, and so those outside the m/z window being fragmented at a specific point in time go unanalyzed.
Slavov added that the ion mobility data collected by the Ultra provides another level of information that improves the specificity of detection.
Olsen likewise noted that the Astral analyzes a small proportion — around 1 percent — of the ion beam generated from the sample, though he said that despite this, the system "really has exquisite sensitivity compared to any instrumentation we have worked with before."
He said that he and his colleagues have adjusted both the isolation windows and ejection times in their single-cell experiments on the Astral in order to take in more ions for analysis. In bulk experiments, Olsen and his team typically use 2 Thompson isolation windows. In their single-cell experiments they double the size of that window while also doubling the maximum allowed injection times.
"That seems to be the sweet spot for single-cell proteomics, at least in our hands," he said.
In addition to a substantial expansion in depth of coverage, the Astral's sensitivity also allowed Olsen and his colleagues to analyze post-translational modifications at the single-cell level. This, Olsen noted, has traditionally been difficult as the small size of single-cell samples has made it difficult to identify PTMs without enrichment, but enrichment protocols often result in sample loss that makes single-cell PTM analysis challenging.
The field's gains have also been driven by improvements in sample preparation workflows.
In its preprint, Olsen's group used Evosep's recently released ProteoCHIP EVO 96 platform, which the company designed in collaboration with Cellenion for single-cell research. The platform allows researchers to use Cellenion's CellenOne X1 platform to isolate and dispense single cells into the EVO 96 platform where up to 96 cells can be processed in parallel and then transferred into Evosep's Evotip separation devices and run on the company's Evosep One LC system.
The integration of the system makes for a nearly lossless sample prep process, Olsen said, noting that this "is key to [achieving] maximum sensitivity in the mass spectrometry analysis."
Slavov's team at PTI also uses the CellenOne system for its sample prep process. Called nPOP, the approach performs sample prep of individual cells in droplets on glass slides allowing for the simultaneous preparation of thousands of cells. In their recent preprint, the researchers said that the approach could prepare more than 3,000 cells for analysis in one to two days.
Mechtler has also collaborated with Cellenion on a single-cell sample prep workflow enabling parallel processing of large sets of single cells with minimal sample loss.
Throughput remains perhaps the most significant challenge facing single-cell proteomics. The recent instrument releases have helped address this to an extent as their high performance allows researchers to detect large numbers of proteins even with short liquid chromatography gradients, but many workflows are still limited to analysis of around several dozen cells per day.
"We want to be able to analyze thousands of cells per day at a low cost per single cell," Slavov said. He said that the PTI continues to develop its plexDIA method, which combines sample multiplexing with data-independent acquisition mass spec to enable higher-throughput experiments.
"This is [the workflow] we are using on the timsTOF Ultra at the moment," he said.
Olsen similarly noted that "throughput is our main issue right now."
He said that using the approach detailed in the preprint, his lab can run roughly 40 single cells per day, but he believes hitting 100 cells per day is realistic with "a little bit of tweaking."
He added that he and his colleagues are exploring multiplexing approaches that could further boost their throughput but that it is "still early days to say how well it will work or not work."