NEW YORK (GenomeWeb) – With its recently released cellular identification and extraction platform, liquid biopsy startup RareCyte is now turning its eyes to the tissue biopsy space as well. RareCyte hopes to offer its new PickSeq workflow as a method for RNA-driven biomarker discovery in cancer tissues, as well as any situation that requires correlation between protein biomarkers with DNA- or RNA-level analysis.
Seattle-based RareCyte believes it can also help researchers identify rare cells in solid cancer tissues using the PickSeq platform, which combines the firm's CyteFinder —a set of automated channel fluorescence microscopes that quickly image slides and use machine learning to find rare cells — and CytePicker —a needle-based tool to isolate tissue microregions for molecular analysis.
In addition to providing products for imaging and cell retrieval, the 60-employee firm said it will aim to work with software, labeling, and sequencing partners to streamline the PickSeq workflow.
"Tissue is something relatively new for us, but PickSeq is a natural extension of our liquid biopsy platform to analyze tissue samples," RareCyte Chief Medical Officer Eric Kaldjian explained. However, he emphasized that the firm still plans to expand and develop technology for the liquid biopsy space as well.
RareCyte began as a spin off in 2011 from Applied Precision, which was purchased by GE Healthcare at the time. In 2015, the firm launched CyteFinder— to initial research and biopharma customers.
In 2017, the firm raised $30 million in a private financing round, with the goal of advancing sales efforts and inking new companion diagnostic partnerships. Earlier this year, RareCyte released its CyteFinder II tool for its liquid biopsy-based applications.
In a feasibility study presented in a poster at the American Association for Cancer Research annual conference in Atlanta last week, a group of researchers led by Harvard Medical School systems biology professor Peter Sorger and RareCyte measured the platform's ability to identify and extract cells in formalin-fixed, paraffin-embedded tonsil and frozen breast cancer tissues.
According to Kaldjian, the PickSeq workflow integrates the firm's CyteFinder and CytePicker before users perform downstream sequencing on the extracted material. Using CyteFinder II or CyteFinder II HT scanning microscopes, researchers can perform multi-parameter imaging with microregion cell retrieval for molecular analysis in up to seven fluorescent channels. Kaldjian noted that users can autoload and scan up to 80 slides on CyteFinder II HT, which has a barcode-driven workflow for sample tracking.
Initially extracting the target sample from a patient, researchers can stain the tissue with one of RareCyte's kits, an independent lab-developed method, or third party-based kits.
Researchers then scan and analyze the stained tissue sections with the CyteFinder system. After developing an image file, Kaldjian explained that researchers can process the picture in two ways — they can export the file into other image analysis systems, performing the analysis in their own labs, or they can use one from a third-party vendor. Researchers can also use RareCyte's independent image analysis tools built within the CyteFinder instrument, allowing a certain amount of independent analysis directly performed at the workstation.
While Kaldjian couldn't disclose the exact amount of analysis, he said that the platform could provide enough to "get figures into a paper, but probably not enough if you want to do really in-depth multiple parameter types of definition."
Kaldjian noted that the overall process — including both imaging the tissue sample and targeted retrieval — requires roughly thirty minutes before standard RNA isolation, amplification, and sequencing.
After finding microregions of interest, researchers remove them using the CytePicker, which Kaldjian noted is also built into the CyteFinder II instrument. Users then place the section in a tube and lyse it as the first step in an RNA sequencing process, which includes whole-transcriptome amplification and library preparation.
Despite having not yet applied downstream methods used for isolating circulating tumor cells in its liquid biopsy platform — such as PCR, Sanger sequencing, and mass spectrometry — for microregions from tissues, RareCyte anticipates that the applications will be straightforward to develop because the microregions contain a group of cells rather than a single cell.
In the study, the researchers at RareCyte manually stained the FFPE sections using antigen retrieval with a panel of antibodies to CD3, CD4, CD8, CD20, and Sytox Orange (nuclear dye). The group then performed whole-slide six-color scanning and region of interest identification with the CyteFinder system.
For frozen sections, the researchers directly retrieved microregions from the antibody-stained section using CytePicker. For FFPE tonsil sections, they identified regions of interest on the antibody-stained sample, retrieving microregions from the same location on an adjacent section stained with DRAQ5 nuclear dye.
The researchers then amplified RNA retrieved from microregions and performed differential gene expression analysis on the samples. From the RNA-Seq data, the team then used structurally different but morphologically similar genes in tonsil cells to develop PickSeq informed immunofluorescence panels.
The team saw that tonsil staining identified crypt lining, T-cells, and B- cell compartments. In frozen breast carcinoma tissue, PickSeq differentiated tumor and T-cell microregions. In FFPE tonsil tissue, the group noticed that PickSeq highlighted differential gene expression between adjacent follicles.
"Out of 20 selected genes preferentially expressed in the germinal center, 16 were confirmed by interrogation of the Human Protein Atlas to be specific or enriched in germinal centers by immunohistochemical staining," the study authors noted.
In addition, the team saw that later immunofluorescence staining for biomarkers selected using the RNA-seq results verified differential expression of follicle center proteins.
Although CytePicker retrieves microregions that contain a small number of cells (about five to 10 depending on size), Kaldjian argued that the system can indirectly detect evidence of the cell within a group of non-related cells.
"For example, if we pick a region in the tumor with a single CD8-positive T-cell, and the rest around them are cytokeratin-positive breast cancer cells, we can interrogate RNA that is T-cell specific," Kaldjian explained. "In effect, what we're getting is a single T-cell's worth of RNA and therefore perform a single T-cell analysis without selecting a single cell."
Kaldjian noted that RareCyte recently received a $223,000 STTR Phase I grant from the National Cancer Institute, which is part of a collaboration with Sorger's lab to develop PickSeq applications with multiparameter imaging. The firm also plans to eventually apply for Phase II funding to extend the work further.
According to Kaldjian, RareCyte has a portfolio of patents related to the methods used in the PickSeq platform.
"RareCyte has invested heavily in [its] patent portfolio, resulting in 34 issued and 23 pending applications as of early April," Kaldjian said. "Our PickSeq application utilizes many elements of our platform technology and consequently a significant portion of our patent portfolio supports the PickSeq application."
While RareCyte will perform some of its own proof-of-concept studies in its own labs, the firm aims to pursue a distributor model by offering the CytePicker, CyteFinder, and associated kits to laboratories so that "researchers can run the PickSeq platform and have complete control of their tissue samples and downstream workflow," Kaldjian said.
While Kaldjian declined to reveal the price of the integrated platform, he argued that "RareCyte is very competitive in the marketplace."
Kaldjian said that RareCyte will continue to develop use cases for the technology with outside collaborators by further highlighting the scientific results that the firm believes can stem from using PickSeq. He argued that researchers can apply the platform to cancers as well as "immune infiltrates" associated with cancers.
"We can now pick morphologically different regions in cancers like ovarian and prostate cancer, perform RNA sequencing, and ask questions about the gene expression differences associated with their morphologies," Kaldjian explained. "Since we now have a panel of up to seven markers, we can pick cells from regions defined by biomarker phenotype from the panel and see what else is different in there, either at an RNA level DNA clonal level."
Over the next few years, Kaldjian believes that RareCyte will develop specific use cases, where researchers can potentially apply a multi-parameter approach for diagnostic samples previously limited by rare cell issues.
"We're very enthusiastic about this as a technology for being able to get a somewhat high-throughput method to identify cell populations in immuno-oncology, develop spatial information, and then go back to that tissue and select microregions of interest based off the spatial information you found with the imaging," Kaldjian said.