Seven months after emerging from stealth mode by announcing plans to map the entire immune system, Immunai has launched a partnership with 10x Genomics that will give it many of the tools it will need to fulfill its ambitious goal.
Immunai said yesterday it will apply its proprietary artificial intelligence and machine learning algorithms along with 10x Genomics’ single-cell technologies in order to map the hundreds of cell types and states within the immune system.
The collaboration—whose value was not disclosed—aims to help biopharmas accelerate their drug discovery and development efforts by uncovering biomarkers and generating insights to distill the data into meaningful and clinically relevant information, using data analysis and computational tools many of them now lack.
“In our database, we have patients that have different cancer indications, autoimmune indications, Alzheimer’s, COVID-19, and healthy people. And we are studying them together, including their immune response to different immunomodulators,” Noam Solomon, PhD, CEO and Co-founder of Immunai, told GEN. “We are applying machine learning algorithms to help us understand or predict how patients will respond to different therapies, and what is the difference in slight perturbations of the immune system; how they interact chemically.”
For example, Solomon said, Immunai is applying transfer and multi-task learning approaches to essentially transfer insights gathered for an immune checkpoint inhibitor in melanoma to better understand the immune activity for other immune indications.
“Essentially, every disease has an immune component, and wherever inflammation is involved, we are there,” Solomon added.
Since its inception, Immunai has established collaborations with undisclosed Fortune 100 pharmaceutical companies, as well as with several top-tier academic institutions, which include Memorial Sloan-Kettering Cancer Center, University of Pennsylvania, Massachusetts General Hospital, and Baylor College of Medicine.
Natural Killer vs. Neuroblastoma
Baylor and Immunai are partnering to study of natural killer T (NKT) cells that can be genetically engineered into immunotherapies targeting solid tumors and other cancers. In October, researchers from Immunai, Baylor, and the University of North Carolina at Chapel Hill published a study in Nature Medicine presenting interim results from an ongoing, Baylor-sponsored Phase I clinical trial (NCT03294954) evaluating genetically modified human NKT cells with a chimeric antigen receptor (CAR) that enabled them to recognize and attack neuroblastoma. Expressed with the CAR was interleukin-15 (IL-15), which supports NKT cell survival.
The study showed the modified cells were safe, localized to tumors, and induced an objective response in one of three patients with regression of bone metastatic lesions. The NKT platform has been licensed to Kuur Therapeutics, which is using it to develop KUR-501, a GD2-CAR NKT autologous product for the treatment of neuroblastoma.
“In terms of what we’re being asked to unlock, it can range from, can you find us a biomarker that helps us to segment patient populations: Responders, non-responders?” said Mark Jacobstein, Immunai’s Chief Business Officer. “Can you help us identify mechanisms of action and/or mechanisms of resistance in the cell therapy space, as in the Baylor paper? Can you characterize the infusion product, which in in their case they thought was homogenous until we did our work, and we were able to show that it actually wasn’t. Can you characterize what’s happening in the infusion product? And then help us think about how that correlates with response?”
“There’s a variety of different questions,” Jacobstein added. “At the end of the day, they’re all asking one fundamental question, which is: Can you help us make a more effective therapy?”
Immunai says its end-to-end computational AI pipeline customized for single-cell methods allows researchers at cell therapy developers and other biopharmas to better understand how immune cells operate with both granularity and scale. In turn, Immunai will help 10x’s customers answer clinical and translational questions related to the immune response to therapies.
The combined technologies of the companies, Solomon and Jacobstein said, enables a multi-omic approach that will look at not only gene expression and surface proteins but also T cell receptors (TCRs) or BCR (B cell receptor) sequences.
“For every cell that we measure, we are talking about over 40,000 different genes, hundreds of surface proteins, and the TCR-and-BCR sequencing data,” Solomon said. “We essentially map these 40,000 different genes and surface proteins into the identity of the cell type. We can say, this is a T cell of type CD4 and the CD4 cell is in sector Th17, an effective memory cell.”
Said Jacobstein: “This multi-omic approach doesn’t just sort of get you a little bit more data about itself. It actually allows us to answer questions that no one else can answer.”
For example, he said, carrying out the sequencing of BCRs allows researchers to do clone tracing, to see over time which clones were expanding and which ones were contracting, which ones are still present, and which ones are not.
“If you couple that with the gene expression and cell surface proteins, you’re able to actually characterize the phenotypes of these cells. So, that sort of combination that we have because of the multi-omic approach is allowing us to answer, we hope, very important questions, fundamental biological questions about why therapies work or don’t work for our customers,” Jacobstein said. “We are very, very grateful to 10x for the equipment and assays that we rely on to help make that possible.”
Single Cell Certified Provider
As part of the collaboration, Immunai has joined 10x Genomics’ Certified Service Providers program, becoming one of the first companies in North America to do so. Immunai has been granted the Single Cell Immune Profiling certificate through the program, which connects customers to service providers that according to 10x have met a high standard of technical and service requirements, giving its customers access to advanced sequencing solutions from a provider validated by 10x.
“Our customers can leverage insights from Immunai’s AI-powered technologies while using the 10x Genomics products that they’re engaged with today,” Brad Crutchfield, Chief Commercial Officer, 10x Genomics, said in a statement.
The arrangement is expected to add to Immunai’s customer base: “We’re a new company. Not everyone in the field has heard of us yet. We’re hopeful that this collaboration and 10x’s reputation will help us to reach more customers,” Jacobstein said.
Immunai has agreed to pair its immune cell atlas with the phenotypic clinical data that hospitals, biopharma, and biotech companies will derive from 10x Genomics’ technology. 10x offers a variety of technology solutions designed to assist in single cell research, all marketed under its Chromium brand:
- Chromium S/C ATAC = (Assay for Transposase Accessible Chromatin) – Analysis of chromatin accessibility at the single cell level, providing insights into cell types and states, and deeper understanding of gene regulatory mechanisms.
- Chromium Single cell (S/C) gene expression = Single cell transcriptome 3′ gene expression and multiomic capabilities designed to profile tens of thousands of cells. Designed to enable study of cellular heterogeneity, novel targets, and biomarkers with combined gene expression, surface protein expression, or CRISPR edits in each cell.
- Chromium S/C multiome ATAC+Gene expression = Simultaneous profiling of gene expression and open chromatin from the same cell; characterization of cell types and states; discovery of gene regulatory programs.
- Chromium S/C immune profiling = Multi-omic solution to immunology questions. Analysis of full length paired B-cell or T-cell receptors, surface protein expression, antigen specificity, and gene expression, all from a single cell.
- Chromium S/C CMV = Designed to provide a comprehensive, scalable solution for revealing genome heterogeneity and understanding clonal evolution. Single-cell study of disease pathogenesis or characterization of neuronal mosaicism.
“A New Era”
“The world of oncology, and essentially of personalized medicine, has been transformed by the introduction of singe-cell technologies. And in the past five years, you see that things are changing so rapidly that there is probably no area in medicine that will not be affected and not be using singe-cell technologies,” Solomon said. “In order to get the most interesting discoveries and understand how our drugs are affecting us on the single cell level, you have to use the technologies.”
“Until we had single-cell technologies, we only did what we could,” he added. “But now, I think it’s going to be a new era in personalized medicine.”
Solomon co-founded Immunai in December 2018 with the company’s Chief Technology officer Luis Voloch, a former Palantir machine learning engineer who did his undergraduate and graduate studies at MIT. Solomon is a double PhD who carried out postdoctoral research in the Mathematics department at MIT, and in mathematical sciences and applications at Harvard University.
Solomon and Voloch were joined soon after by founding scientists Dan Littman, MD, PhD, a professor of molecular immunology at New York University; Ansuman Satpathy, MD, PhD, a professor of cancer immunology at Stanford University; and Danny Wells, PhD, founding data scientist and current member of the Parker Institute for Cancer Immunotherapy. Satpathy and Wells were co-authors of a July 2019 study published in Nature detailing the origin of tumor-fighting T cells following PD-1 blockade.
Headquartered in New York City, with offices in San Francisco and Tel Aviv, Immunai has grown to 60 people between all three locations, and is “growing fast,” Solomon added.
Immunai emerged from stealth mode in May, when it disclosed it received $20 million in seed funding led by Viola Ventures and TLV Partners. If the company is pursuing Series A financing, it isn’t saying—for now.
“Call me in a month, and I may able to share that with you,” Solomon quipped.