NEW YORK (GenomeWeb) – As the technologies for single-cell analysis mature, more institutions are centralizing their capabilities.
Harvard and the University of California, San Francisco have in the last year established separate core facilities for single-cell research. At the Mayo Clinic, single-cell has become a prominent component of what the Medical Genome Facility offers researchers there.
The technology and applications offered at each differ greatly – Harvard only offers inDrop transcriptome sequencing, while the Mayo Clinic offers single-cell capture from both Fluidigm and 10x Genomics as well as downstream analysis like qPCR or sequencing on several Illumina platforms.
But a unifying theme is that consultation is one of the most important functions they offer.
"One of the first things we do before anything starts is to have an initial consultation," said Verna Simon, lead technician for single-cell research at Mayo's Medical Genome Facility. "We go over all the options for bulk cells and single cells."
In addition to explaining the limits of single-cell technology, the consultation also helps guide researchers to get the best results that they can. "Once the cells have been captured and lysed, then it's pretty straight forward," she said. "That's not the difficult part, it's the front-end stuff. Their shape, their viability, and whether you can capture enough of them."
At Harvard Medical School's Single Cell Core, staff scientist Sarah Boswell won't even meet with a prospective researcher until they've filled out an online application. "If the application is too much work for you, then you don't get how serious these experiments are," she said. Launched last year, her core has more users than it can handle and has to book projects a month in advance.
At UCSF, the consultation is more of a matchmaking session, Flow Cytometry Core Director Mike Lee said. The Single Cell Analysis Center (SCAC), which opened in February 2016, is an offshoot of his project to unite all the flow cytometry cores at the Parnassus campus, of which there were originally six. "A lot of times, researchers would ask me, 'Do you know of a genomics core than can do RNA-seq?'" he said. "Rather than me becoming a genomics expert, they're all right down the hall, so why not utilize what we have," he said.
While single-cell cores are popping up around the country to meet demand from researchers, it hasn't been a cakewalk to put it all together.
"Now I realize why nobody did this in the first place," Lee said. "It's a full-time job getting the appropriate info, meeting all the cores and users."
The diversity of origin stories for these three single-cell cores reflects the embryonic state of the field.
For Mayo, single-cell was taken up into their existing genomics core, with the addition of single-cell capture equipment starting around 2013 or 2014, Simon said. Now, they offer two Fluidigm C1 instruments and, as of last year, a 10x Genomics Chromium instrument. "We were early, but maybe not the first," she said.
She's already seen a shift in the types of cells researchers are bringing into the consultation. "In the beginning, it was mainly cells that were already in suspension, blood cells or cell culture cells that had been well prepared. Now we're seeing more and more people asking about flow-sorted and bead-sorted cells, things that could narrow down and capture a small fragment of a specific population they're interested in."
Harvard's Single Cell core came directly as a result of the inDrop technology developed by Allon Klein's lab. "Too many people were coming to the lab trying to form collaborations," Boswell said. The Systems Biology department paid for a technician to learn the platform and brought Boswell, who has experience with a bevy of RNA sequencing methods, on part time to run the core.
Starting in January 2016, she led a six-month pilot phase working with some former Klein lab collaborators before becoming an official Harvard core in July. Now, she's added a postdoc to help push the method even further.
Because her core only does inDrop, sometimes the result of the consultation is that the project is better handled by someone else. Because inDrop can only handle one sample at a time, she's told researchers that they're better off looking for a collaborator with a 10x Chromium platform.
For first-time users, there's often a lot of optimization that happens prior to library creation. "A lot of times we'll work with three to four samples in the beginning and have biological duplicates as well as some kind of perturbation. If you have a mutant that disrupts one population, that's a great preliminary experiment," Boswell said. "You can get a biological duplicate and look for the loss of a population. That's a great way to know if your sample prep worked and if you got enough reads per cell to find what you're looking for."
Just because they only work with one type of technology, doesn't mean they're not in contact with other core facilities, in fact, quite the opposite. With plenty of sequencing cores in the Boston area, Boswell doesn't offer to do the sequencing. By now, she's worked with most of the sequencing cores, so they know how to handle inDrop libraries, she said. She's also worked very closely with the Chan bioinformatics core at Harvard.
"They have a dedicated person for single-cell analysis, and they've added another because of us."
At UCSF, the need for collaboration was baked into their operating model. Cassandra Adams manages the SCAC and has open office hours for researchers looking into single-cell. After consulting with researchers, she can point them to resources in the flow cytometry, microscopy, genomics, library prep, and bioinformatics cores.
At the moment, the SCAC doesn't have a revenue model. "We're doing it for the greater good," Lee said. "I'm hoping the university sees the value because we're providing more money for the core facilities. A lot of users don't even know the cores exist."
Simon said that the next step for her core is to partner with her institution's flow cytometry cores, to help narrow down cell populations for researchers. Her core only charges by the number of successful captures, so increasing the yield, even after cells get roughed up by flow cytometry, is an area for improvement. "We're working with other groups on campus that have the equipment we could use now, to see how much we could get out of that," she said.
New equipment is always a consideration, too. Though it currently has no shortage of users, Boswell has been thinking about ways to expand. "I'd love to have a 10x [Chromium], if I were confident it was going to be capturing 50 percent of cells on every sample without any issues," Boswell said. For now, she's working on building a second inDrop instrument. She's also considered adding sample prep for single-cell mass spectrometry-based proteomics, notably SCOPE-MS, a method described in February on the bioRxiv preprint server.
While these cores are all there to serve the researchers at their respective institutions, Boswell has plans to expand the geographical footprint of single-cell analysis. Her postdoc is working on methods to be able to freeze cells prior to capture. "That's one of the biggest improvements that we could do," she said. Not only would it allow samples to come in from all over the country, it would help reduce the time constraints she faces on a daily basis.
They've already started testing freezing on the cell types that they see coming in every day. "For some cells it will never work," she said, "but maybe we can get to the point where some cells can be brought in in that way."