NEW YORK — With $17 million from a recently closed Series A funding round, proteomics firm Nomic Bio (formerly Nplex Bioscience) plans to launch services using its nELISA platform in the second quarter of 2022.
Upon launch, the company, which is headquartered in both Montreal and Boston, will be able to analyze up to 500 proteins per sample and handle up to 100,000 samples per quarter, said Milad Dagher, Nomic's cofounder and CEO.
To get ready, the firm has begun working with early-access users including GlaxoSmithKline and other pharma companies. Nomic is also collaborating with researchers at the Broad Institute to provide protein measurements as part of the Joint Undertaking in Morphological Profiling-Cell Painting, or JUMP-CP, consortium. That project is led by the imaging platform at the Broad, which is working with a group of pharma and biotech firms as well as nonprofit partners to create a large-scale public cell-imaging database to drive drug discovery work.
Based on technology developed in the lab of David Juncker, a professor of biomedical engineering at McGill University, nELISA uses a miniaturized, bead-based ELISA format to multiplex hundreds to thousands of immunoassays rapidly and without the problem of antibody cross-reactivity.
In addition to shrinking the format, the company has changed the way the antibodies for protein detection are applied, which is key to eliminating cross-reactivity. A traditional sandwich ELISA uses a capture antibody bound to some sort of surface and a detection antibody that produces a signal indicating the presence of the target protein. In this system, the capture antibody is incubated with the sample of interest, binding to any target protein present, and then the detection antibody is added. In some cases, the detection antibodies can cross-react with proteins other than their target, producing a nonspecific signal. This issue becomes more challenging as the number of multiplexed ELISAs increases, which has limited the multiplexing capabilities of the approach.
Nomic has tackled this problem by preassembling the capture and detection antibodies on beads, using a DNA linker that binds them together, so they are near each other before the sample is added. This way, the detection antibody never has the opportunity to bind nonspecifically. When a sample is added, the capture and detection antibodies each bind to their target protein if it is present, forming a complex of the three molecules. If the target protein is not present, the antibodies remain apart. A user can then remove the DNA linker connecting the two proteins. Detection antibodies not bound to a target protein can then be washed away while those bound to their target will remain.
Each bead is fluorescently barcoded according to the nELISA it hosts, and the barcodes are read using a flow cytometer. The company is using a barcoding approach based on multicolor Förster resonance energy transfer that Dagher, Juncker, and colleagues detailed in a 2018 paper in Nature Nanotechnology. The reaction can be done in a microtiter plate containing thousands of different beads each hosting a different nELISA reaction.
Currently, Nomic has a 200-protein inflammatory panel focused on analytes like cytokines and chemokines. According to Dagher, the company is able to measure up to 1,500 samples per day per instrument.
He noted that while 200 proteins is a relatively modest panel, it "maps the cytokine and chemokine space very comprehensively such that folks can start using it to discover new biology in the inflammation and immune response context."
Nomic's next target is a 500-protein panel, which it expects to reach next year. Dagher said the company believes it has "a clear path to at least a 2,000-plex [panel] by onboarding many of the validated antibodies on the market today."
Both the 200 proteins Nomic currently measures and the 2,000-plex panel Dagher envisions are considerably smaller than the panels offered by SomaLogic and Olink, the two most prominent affinity reagent-based proteomic platforms on the market. Boulder, Colorado-based SomaLogic currently measures 7,000 proteins per sample on its SomaScan platform, which uses proprietary aptamer reagents called Somamers. Uppsala, Sweden-based Olink measures up to 3,000 proteins using its proximity extension assay (PEA) technology, which uses pairs of antibodies linked to DNA strands that are brought into proximity when the antibodies bind. The DNA strands are then extended by a DNA polymerase, creating a new sequence that can be used as a surrogate marker for the target protein.
Dagher said Nomic aimed to distinguish itself in part through its capacity for high sample throughput. Olink says it is currently able to analyze 4,600 samples per week per platform, while SomaLogic says it is able to analyze 1,000 samples per day. As Dagher noted, Nomic is currently able to process 1,500 samples per day per system, and, he added, the company hopes to double that throughput in the next few years while also expanding its content to around 3,000 proteins.
"We believe that on top of content, you also have to solve for throughput, cost, and versatility," he said. "Scientists today, whether they are doing drug or biomarker experiments, need to be able to run large-scale experiments going through tens of thousands of samples to uncover variances."
Anne Carpenter, principal investigator and senior director of the imaging platform at the Broad, highlighted Nomic's high throughput as a key factor in her interest in using the system as part of the JUMP-CP consortium's work.
The consortium is using a set of six fluorescent dyes to stain and image more than 1 billion cells exposed to some 140,000 genetic and small molecule perturbations, creating a massive library of cell morphologies that can then be analyzed computationally.
"Even though it is only six dyes … the information that comes out of these images can be just as powerful as a full transcriptional profile," Carpenter said, and Nomic's technology offers the possibility of adding protein-level information to the consortium's dataset.
Prior to learning about Nomic, "I had not been aware that there was any proteomic profiling that could be done anywhere close to the scale that you could do with either transcriptional or image-based profiling," she said. "Of course, proteins, I would say, are toward the top of my list of things I would love to be able to measure in high throughput and high dimension."
Carpenter said she was not familiar with the Olink or SomaLogic platforms, so couldn't say whether they might also have proved good fits for the project.
The JUMP-CP researchers will use the Nomic platform to measure the 200 secreted proteins in the company's inflammatory panel in a 10,000-cell subset of the cell library they are generating. The effort will help them determine if "adding the protein information to the morphology information is synergistic," Carpenter said. "Are these two data sources really complementary to each other? That's the sort of question we'll be answering here."
Carpenter said the Broad researchers would be analyzing the 10,000 cells sometime in the spring of 2022 but that they had already completed a smaller pilot looking at a few thousand cells.
"Those results were sufficient to make us excited to do the larger-scale [experiment]," she said.
In addition to expanding its platform's protein content, Nomic is also working to develop the system for detecting things like post-translational modifications, Dagher said, noting that the company has seen interest in this from its early collaborators.
Dagher said that Nomic planned to initially offer the platform as a service out of its labs in Montreal and Boston but aimed by the end of 2022 to make it available for "select research groups and pharma" to use in their own facilities.