NEW YORK (GenomeWeb) – Researchers this week described a novel method for single-cell sequencing that preserves an awareness of the spatial orientation or arrangement of cells, applying it to a study of early breast cancer.
Although single-cell sequencing has emerged as a powerful tool for untangling the heterogeneity that can be present within a single tumor, methods have largely not been able to link this information with analysis of where different cells are located in relation to one another.
In their report published in Cell today, MD Anderson investigators described the strategy they developed to address this challenge — sequencing individual cells while maintaining a record of their spatial orientation in the body. They also offered preliminary evidence for how such methods could be useful in analyzing the biology of early cancer genesis, growth, and spread.
The team's method, which they call topographic single-cell sequencing, involves laser dissection and something called laser catapulting to extract individual cancer cells from different areas of a tissue specimen so they can be sequenced individually.
Studying samples from women with an early form of breast cancer called ductal carcinoma in situ (DCIS), the researchers set out to test a hypothesis that the cells that begin to grow out into the breast as invasive ductal carcinomas (IDC) represent some of the same populations that arise initially in the ducts as DCIS.
Some prevailing theories have leaned more toward a concept of DCIS and IDC representing distinct cell lineages. But using their novel topographic sequencing approach, the MD Anderson researchers were able to show that invasive cells do appear to share a direct genomic lineage with those in situ. Moreover, at least in this small study, most mutations and copy number aberrations looked to have evolved within the ducts in DCIS, before cells then pushed out to become IDC.
Similar investigations in other cancers are likely to become attractive as incipient cancers, or pre-cancers, are detected more frequently and earlier using new technologies.
Nicholas Navin, the study's senior author and co-director of the sequencing and microarray core facility at MD Anderson, said this week that the methods he and his colleagues used to dissect and catapult individual cells have been used before to study cell populations and spatial positioning, but mainly in other areas like neurology.
However, he noted that technologies like this have huge potential in oncology in part because screening technologies like mammogram have resulted in overdiagnosis and overtreatment in certain cancers. In addition, newer screening tools may increase the number of early cancers or pre-cancers detected in the population.
A number of companies have now said, for example, that they intend to develop and commercialize tests that use circulating DNA and other biomarkers to identify incipient cancers of all types in healthy individuals. The looming ascendance of early detection technologies has even spurred calls for a pre-cancer genome atlas to follow in the footsteps of the successful TCGA effort.
"There is a lot of funding now [in this area,] and the next stage of TCGA is going to have this pre-cancer focus," Navin said. "So lots of us are applying for grants right now [and I think] methods like this will be really really useful because researchers will be dealing with small lesions masked by stroma ... where spatial resolution is very important."
In their report in Cell, Navin and colleagues used their topographic sequencing strategy to isolate and sequence almost 2,000 single cells from 10 patients who had synchronous DCIS and IDC based on histopathology.
Briefly, the topographic sequencing method involves staining tissue sections to identify different spatial categories, or areas of interest — in this case in situ versus invasive regions. Using laser-capture microdissection one can then cut around each cell to be sequenced, and then pop that particular cell out of the section using a pulse of UV energy (laser catapulting).
Navin explained that most laser-capture dissection systems use a polymer to pull back tissue or cells. This hinders single-cell sequencing, because it contaminates samples with microbial species that distract the polymerases used in whole-genome amplification.
Using a laser to pulse the cells out of the tissue specimen results in much purer samples that can be more successfully amplified and sequenced.
After excising cells, the MD Anderson approach involves whole-genome amplification and barcoded multiplex sequencing. With the resulting copy number profiles, as well as X and Y coordinates denoting the spatial positioning of each cell, researchers can then map the arrangement of the resulting clonal genotypes back to the tissue sections.
According to Navin and his coauthors, their study results offer two main takeaways. First, they provide evidence that breast cancer genome evolution occurs within the ducts, before tumor cells escape what is known as the basement membrane. In other words, "one or more clones appear to have co-migrated through the basement membrane into the adjacent tissues to establish the invasive tumor mass," the group wrote.
This runs contrary to other proposals for how breast cancer develops, and the relationship between aggressive tumors and DCIS.
"We refer to this model as 'multiclonal invasion' to distinguish it from the evolutionary bottleneck or independent lineage models that have previously been proposed for invasion," Navin and his colleagues explained.
Second, the data suggest that all subclones in the ducts arise from a single initiating cell, as evidenced by shared truncal mutations and CNAs.
Navin stressed that the study did not have a clinical focus, but he said that there is more distant clinical relevance for such investigations, in DCIS and in other cancers.
"The results we saw here fit well with our previous studies of something called punctuated copy number evolution," Navin said, a burst of mutation that leads to transformation of precancerous lesions to invasive tumors.
"We weren’t sure if it happens in the ducts or after the cells become invasive," but based on the team's results it appears to be the former, Navin said.
If it's true that cancer genotypes are preprogrammed early on it would be good news from a diagnostics standpoint, he added. "It means that even when cells are still in situ, you can analyze them and get information about which ones will expand."
The MD Anderson study didn't look for biomarkers of progression, but it is encouraging for others that are ongoing or may be initiated in the future.
According to Navin and colleagues, extending their findings in DCIS to things like colorectal adenomas, lobular carcinoma in situ, precancerous prostate and pancreatic lesions, and others — will likely be of interest to many groups.
"In these cancers, spatial resolution can provide new insights into the context, organization, and migration of tumor clones as they … invade the surrounding tissues [and can] begin to shed light onto the enigmatic question of why some premalignant cancers remain indolent for the lifetime of the patient while others progress to invasive disease and ultimately cause morbidity in patients," the study authors wrote.