Four years after making its first foray into proteomics, and a year and a half after shifting the group’s focus to cell surface proteins, Celera Genomics has something to show for its efforts.
Last week, the company said it will collaborate with Abbott Laboratories and General Electric to develop cell surface antigens resulting from its proteomics efforts. The collaboration with Abbott is to identify cancer therapeutics, while the GE alliance centers on diagnostics, Celera said.
“Companies with [the] experience and prestige of Abbott and GE wish to partner with us; that’s a good vote of confidence for us,” said Robert Booth, Celera’s CSO.
The collaboration with Abbott resulted in part from an ongoing project in molecular diagnostics between Abbott and Celera Diagnostics, a joint venture between Celera Genomics and Applied Biosystems, Booth said.
Terms of the new alliance call for Abbott to receive an undisclosed number of pre-validated cell surface targets from Celera. Abbott will further validate them using RNA interference and other pre-clinical technologies.
Depending on the results, Abbott might decide to develop therapeutic monoclonal antibodies or small molecules against these targets in collaboration with Celera.
“At that point, we would come back into the picture as a co-funder for the clinical work,” said Steven Ruben, Celera’s vice president for protein therapeutics. Abbott would commercialize products from the collaboration.
What made Abbott attractive as a research partner was its antibody capabilities, Booth said. The com-pany developed and currently markets Humira, a monoclonal antibody for the treatment of rheumatoid arthritis.
GE’s Global Research division plans to develop radio-labeled antibodies or antibody fragments directed against Celera’s cell surface proteins to enable physicians to image tumors in patients in vivo. The main advantage of these imaging agents is that they would have greater specificity than positron emission tomography and magnetic resonance imaging, said Ruben.
Eventually, though Abbott and GE will not work on the same set of cell surface antigens, the same antigen could be targeted by a drug and used to monitor disease progression or regression using this diagnostic procedure, Booth said.
The cell surface antigens are the first fruits Celera has reaped from its proteomics effort, started in 2000 under Scott Patterson (see PM 11-5-01), who joined Celera as vice president for proteomics from Amgen. At the time, Patterson’s group conducted a pilot study to find differentially expressed proteins in pancreatic cancer cell lines by using multidimensional liquid chromatography coupled to mass spectrometry, as well as ICAT (see PM 1-6-03). He left the company in early 2003 and is now back at Amgen as director of early development.
At the end of 2002, Celera announced that its proteomics group would exclusively focus on cell surface proteins (see PM 12-23-02), since those “would provide the most straightforward and significant opportunities,” Ruben said.
In addition, over the last year and a half the company has been validating targets, according to Ruben, who came to Celera in early 2003.
To discover differentially expressed cell surface proteins, Celera researchers first enrich fresh tumor tissue samples for tumor cells using flow cytometry. They then chemically label and extract the cell membrane proteins, leading to further enrichment. This is followed by an isotopic cysteine labeling step to reduce complexity.
After that, the researchers quantify the differential expression of peptides and sequence selected peptides by mass spectrometry, using mostly ABI QStar instruments rather than the company’s ABI MALDI TOF-TOFs, Ruben said. “The QStars provide the robustness and dependability in terms of a platform that one can put samples through continuously,” he said. “The TOF-TOF was still a research-stage instrument, at least from our standpoint.”
Algorithms built in-house by Celera’s informatics group under John Reynders have been essential for analyzing the data.
Celera’s validation, which is still ongoing for some targets, involves flow cytometry — both on cell lines and on tumor tissues — to see whether proteins identified by mass spectrometry are actually differentially expressed on the surface of tumor cells versus normal cells. However, Ruben said the gold standard for validation is immunohistochemistry, which ensures that expression differences hold up in patient samples. This validation step — performed by contractors — involves comparing 10 patients in each tumor type against a panel of 24 normal tissues using two different antibodies with independent epitopes.
Though each target originates from a specific cancer type, they’re being studied in several cancer types. “What we are finding is that many of these targets validate in additional cancers,” Ruben said, potentially broadening their utility.
In functional validation studies involving cell-based assays, researchers are using RNA interference and, where available, neutralizing antibodies to the targets.
“What we are providing [to Abbott and others] is a set of targets that wouldn’t come out of some of the other platforms for differential targets such as the gene expression arrays,” Ruben said.
Besides looking for additional cell surface antigens, Celera’s proteomics group — which employs 35 people, not including informaticians — is currently evaluating shed proteins for protein-based diagnostics in collaboration with Celera Diagnostics. “It’s basically a derivative of the same technologies that are used for the cell surface proteins,” Ruben said, but the project, which currently focuses on cancer, studies other proteins as well.
According to a Celera spokesman, the company is looking for other potential partners besides Abbott and GE. It is “reasonable to expect one or more additional alliances down the road,” he said.