NEW YORK – Genemo, a California startup, has launched an extracellular RNA sequencing service that it hopes to develop into companion diagnostics.
The firm, a University of California, San Diego spinout, began offering its small-input liquid volume extracellular RNA sequencing (SILVER-seq) service last week.
"It measures extracellular RNA from a droplet of blood" or saliva, according to Sheng Zhong, a professor of bioengineering at UCSD and Genemo's founder.
The launch coincided with a proof-of-concept study published in the Proceedings of the National Academy of Sciences this week, demonstrating that extracellular RNAs — both microRNAs and fragments of longer ones — detected by the method were associated with human biology and disease states.
SILVER-seq adapts single-cell RNA sequencing steps to deal with small amounts of input material. Unlike other cell-free RNA-seq methods, it does not start with RNA purification and instead adds library prep reagents directly into the original sample. This prevents the loss of RNA; ribosomal RNA is depleted during DNA library construction.
The initial results suggest SILVER-seq could be useful in both early cancer screening and monitoring patients for cancer recurrence, Zhong said. His team, which also included Irene Su and Shu Chien of UCSD, analyzed blood samples from patients with breast cancer and healthy donors.
"The difference in terms of extracellular RNA in peripheral blood is so pronounced that in fact, you don't require a precise gene panel, nor would you require sophisticated machine learning algorithms," Zhong said.
The firm is also pursuing development of companion diagnostics using SILVER-seq and another molecular technology, based on fluorescence in situ hybridization, that it licensed from UCSD.
"Their data looks strong," said Raghu Kalluri, a researcher at the University of Texas MD Anderson Cancer Center who has studied extracellular RNAs and exosomes in cancer. "There's no reason to believe they're not able to detect [extracellular RNAs.]" Though the study looked at cancer samples, it probably had more value as a proof-of-concept study than as a translational one, he noted.
Compared with previously described methods for detecting extracellular RNAs, "it seems like it's marginally better, but not standout superior to everything else," Kalluri said. SILVER-seq joins other next-generation sequencing-based methods to characterize extracellular RNAs in blood, such as Phospho-RNA-seq, a method published in May by researchers from the University of Michigan.
Zhong founded Genemo, which is based in San Diego, in 2017 and the firm initially licensed technology he developed at UCSD, including Lucid RNA, a fluorescence in situ hybridization-based RNA detection method. Zhong is now the firm's scientific advisor.
He declined to disclose any funding the firm has received and did not specify how many employees it has. SILVER-seq is something Genemo developed on its own, he said. "The company wanted to do personalized diagnostics based on RNA," and SILVER-seq is the result, he added.
The RNAs the method detects range from 20 nucleotides to 200, although most are small, Zhong said. "You have not only microRNAs but also fragments of long RNAs," he said, including messenger RNAs and long non-coding RNAS (lncRNAs).
Kalluri noted that the literature shows many extracellular RNAs are associated with extracellular vesicles such as exosomes, small packages of biological material. Where the RNAs are situated, where they come from, and their relation to biology are still outstanding questions in the field.
Once the researchers characterized the RNAs that SILVER-seq picked up, they calibrated the method to account for both technical and biological variation. Zhong said they looked at differences between drops, between people, and between SILVER-seq and existing RNA-seq methods.
After establishing the assay's performance, the researchers tested whether it could find RNAs correlated with sex and chronological age. "Sex was an easy tell," Zhong said. Age was more complicated, but 1,449 extracellular RNAs exhibited age-associated expression changes. The researchers built a regression model and came up with a classifier that could predict the age of the donor within two years more than 90 percent of the time. Moreover, "genes we observed in extracellular RNA had great overlap with tissue-specific age-related genes," Zhong said.
The researchers then moved on to looking at 96 samples from patients with breast cancer and 32 samples from healthy donors. "There was a clear difference between the cancer and non-cancer donors," Zhong said, "So clearly, you don't have to use a particular gene panel." Simply looking at all the microRNAs led to classifying cancer patients "almost perfectly," he said.
Zhong said the method may also be able to detect extracellular RNAs that predict cancer recurrence. "The difference between extracellular RNA expression between recurring and nonrecurring cancer patients were not as pronounced [as those] between cancer and healthy donors; however, still there is a relatively clear signal that would allow differentiating them," he said.
To say so more definitively would require more studies. The authors noted that analyzing only 96 breast cancer samples m,eant they could not assess the significance of a number of confounding factors, including the fact that there was differential expression between breast cancer subtypes and that there were chemotherapy-induced gene expression changes. They added that they would like to do double-blind prospective trials to build upon their retrospective study.
Kalluri suggested the study could have provided more clarity on how reliably the method can detect extracellular RNA. "If 150 patients came in, and they took a drop of blood, would they be able to detect RNAs in all of them? That's not clear to me," he said. He also suggested adding another control for the study, by digesting the samples with RNase to see if the method still gave a signal. "That's potentially one way to address whether they're actually detecting something specific."
SILVER-seq has already garnered interest from pharmaceutical companies as a basis for companion diagnostics, Zhong said, although he declined to name any companies. So far, Genemo only provides SILVER-seq as a service. One sample costs $399, while a batch run of 20 samples costs $6,384.
Genemo is also looking to build companion diagnostics targeting fusion genes using Lucid RNA. "Fusion proteins can be created by fusion RNAs" as well as fusion genes, Zhong said. "There are some fusions, including ALK fusions, in actual tumors that are only at the RNA level. If you're only detecting fusion genes, there could be a false negative."