If you want to understand economics, study market crashes and depressions. If you want to understand immunology, study cancer. It is when systems are tottering on the brink of failure that you may observe stark differences between function and dysfunction. By seizing opportunities to see where systems go wrong, you may find ways to reverse systemic dysfunction—or even prevent it in the first place.

Whereas economic systems reflect the behaviors of individual actors, immune systems reflect the behaviors of individual immune cells. So, if immune systems are to be understood, immune cell types and subtypes must be identified, their numbers counted, their locations mapped, and their functions assessed.

Surveys help. Fortunately, the surveys that rely on immune profiling technologies are ranging more widely and delving more deeply. Conventional immune profiling technologies include immunohistochemistry (IHC), flow cytometry, and bulk transcriptomics. They have proven valuable in uncovering immune mechanisms and advancing biomarker discovery. However, these conventional technologies may amass information lacking in granularity. For example, they may fail to capture immune cell heterogeneity, identify rare immune cell subsets, or preserve spatial information.

Single-cell, spatial imaging enables researchers to build detailed network maps and analyze spatial interactions, leading to discovery of distinct cellular neighborhoods and niches within the microenvironment. This two-part image of a tissue sample from Akoya Biosciences demonstrates not just multiplexing and single-cell resolution, but the overlay of nearest-neighbor information.

Each of the conventional immune profiling technologies has an enhanced counterpart. This article will discuss three of them: multiplex IHC (mIHC), mass cytometry, and single-cell sequencing. In general, the enhanced counterparts profile immune cells more thoroughly. They may also help investigators uncover the hidden ways that individual immune cells contribute to the immune system’s collective response to cancer. They may even invest immunology with an econometric rigor, that is, the ability to combine the right data with the right models to generate accurate forecasts—in the case of cancer, forecasts about the outcomes of therapeutic interventions.

From IHC to mIHC

IHC is a form of immunostaining that uses enzyme- or fluorophore-linked antibodies to identify antigens in tissue sections. In conventional IHC, it is common to use just one or two markers per tissue section. Multiplex IHC, however, can be used to track multiple markers simultaneously without having to resort to laborious dye cycling procedures.

Some mIHC systems incorporate sophisticated optics and pattern-recognition software. Also, a form of mIHC is available that uses laser ablation to dislodge material from a tissue slide and direct it to a mass spectrometer, expanding the possibilities for spatially indexed protein analysis.

In general, mIHC systems can help users make finer distinctions among immune cells and obtain a fuller picture of tumor cell heterogeneity. Companies offering mIHC systems include Akoya Biosciences and NanoString Technologies.

Akoya has two platforms, the CODEX, which is for “hypothesis-free biomarker discovery,” and the Phenoptics, which is for “hypothesis-driven biomarker discovery and high-throughput translational research.” The company asserts that CODEX can image 40-plus protein markers across the whole tissue sample, and that Phenoptics can image up to 8 protein markers and 80 slides per run.

Akoya recently announced that CODEX was used by Stanford University scientists to uncover the organizing principles and spatial interactions between “neighborhoods” of cells in the colorectal cancer tumor microenvironment. The scientists, led by Garry Nolan, PhD, re-engineered CODEX for paraffin-embedded tissue microarrays, enabling simultaneous profiling of 140 tissue regions from 35 advanced-stage colorectal cancer patients with 56 protein markers. In Cell, the scientists reported that they identified nine “conserved, distinct cellular neighborhoods.”

“We were able to gain valuable insights about how tumors can disrupt immune functionality and how antitumoral immunity requires organized, spatially nuanced interactions between cellular neighborhoods in the tumor microenvironment,” Nolan indicated in a statement. “The results point to potential diagnostics and new targets for therapeutic intervention.”

NanoString also has two platforms: the nCounter, for “digitally detecting and counting large sets of molecules,” and the GeoMx Digital Spatial Profiler (DSP), for “high-plex and high-throughput spatial analysis of RNA and protein expression data.”

By hybridizing two probes—a biotin-containing capture probe and a fluorescent-barcode-containing reporter probe—directly to specific target molecules, the nCounter platform permits the nonamplified measurement of up to 800 targets within one sample. (Sample cartridges are scanned by an automated fluorescence microscope.) The nCounter may be used together with the GeoMx DSP, which runs assays that rely on antibodies coupled to photocleavable oligonucleotide tags.

After hybridization of tagged antibodies or ISH probes to slide-mounted tissue sections, oligonucleotide tags are released from discrete regions of the tissue via UV exposure,” the company notes. “Released tags are counted in a standard nCounter assay or sequenced using NGS, and counts are mapped back to tissue location yielding a spatially resolved digital profile of protein or RNA abundance.” Over 100 validated protein targets and up to the whole transcriptome may be spatial profiled with the GeoMx DSP.

From flow cytometry to mass cytometry

“An increased need to discern cell phenotype and function simultaneously in a multiplexed fashion from precious samples necessitates improvements in speed and resolution of single-cell technologies,” says Alexander Cherkassky, senior director of product management at Fluidigm. “Suspension-based cytometry or imaging modalities, refined to meet the required levels of sensitivity and reproducibility for adequate measurement, can be used to monitor the immune system in assessing efficacy of different treatments and gauge success of novel immunotherapies.”

Flow cytometry streams cells, one cell at a time, past detectors sensitive to the presence of surface antigens that have been labeled with antibody-linked fluorophores. There are not very many fluorophores, however, and the excitation and emission windows of different fluorophores sometimes clash. Also, true fluorophore signals can be obscured by a sample’s autofluorescence.

Other kinds of markers, however, needn’t be distinguished on the basis of light spectra. Metal-linked antibodies may be used if flow cytometry is combined with mass spectrometry. This is the essence of mass cytometry. Another alternative to conventional flow cytometry involves the use of antibodies that have been conjugated with DNA barcodes.

Fluidigm’s Imaging
Fluidigm’s Imaging Mass Cytometry™ (IMC™) on the Hyperion™ Imaging System combines the high-multiplex capabilities of mass cytometry with imaging using metal-tagged antibodies, allowing the simultaneous interrogation of 4 to 37 protein markers in tissues and tumors at subcellular resolution while preserving the information in tissue architecture and cell morphology. IMC was used to produce this visualization of colon adenocarcinoma.

A mass cytometry technology called cytometry by time of flight (CyTOF®) is at the heart of the Helios™ platform, which is offered by Fluidigm. Besides serving as a standalone system, Helios may be an integral part of Fluidigm’s Imaging Mass Cytometry™ (IMC™) platform, which is called the Hyperion™ Imaging System. Fluidigm asserts that these systems can be used to monitor the immune system and thereby assess the efficacy of different treatments and gauge the success of novel immunotherapies.

“CyTOF technology can resolve highly multiplexed protein markers (both surface and intracellular) and reveal systems-level biology at single-cell resolution,” notes Cherkassky. “IMC enables highly multiplexed immunohistochemistry and simultaneous analysis of multiple protein markers from a single scan.

“Multi-omic analysis using complementary technologies more readily allows in-depth characterization and analysis of a tumor microenvironment, where information on single-cell behavior, cell-cell interactions, and spatial context can be discerned in one study. This is an exciting new area to be a part of—developing tools to enumerate molecules and look at proteins, RNA, specific signaling molecules, and drug distributions.”

“Data from these approaches can create a true multi-omic environment at subcellular resolution, [helping to generate] insights on the role of signaling pathways, epigenetic changes, or nuclear transcription in immune cell function,” Cherkassky adds. “The adoption and harmonization of CyTOF assays for deep immune cell profiling by the Cancer Immune Monitoring and Analysis Centers and Cancer Immunologic Data Commons (CIMAC-CIDC) for biomarker immunoprofiling in studies of cancer immunotherapy demonstrates the recognition mass cytometry has gained in this area.”

From bulk sequencing to single-cell sequencing

“Traditional bulk sequencing profiles lack the ability to catalogue individual cell states and their interactions, making it hard to identify therapies and limiting drug development,” says Michael Schnall-Levin, PhD, senior vice president of R&D and founding scientist at 10x Genomics. “Advancements in single-cell and spatial technologies bring the resolution that is needed to interrogate cancer pathogenesis and develop effective treatments.

“Single-cell immune profiling brings a multi-omic high-resolution approach to deciphering the complexity of cancer-host interactions,” he elaborates. “Understanding the spatial relationship between immune cells and the tumor microenvironment will be key to developing effective immunotherapies for cancer.”

Perhaps best known for its single-cell RNA-seq solutions, 10x Genomics has leveraged its core technologies to take on additional applications. For example, 10x Genomics’ Chromium system and GemCode technology figure in the company’s Single Cell Immune Profiling solution, encapsulating single cells in barcoded Gel Beads. Within each Gel Bead in emulsion (GEM), reverse transcription occurs. GEMs are broken, and cDNA amplification and library construction are performed. Assays are available that enable immune profiling by enriching barcoded cDNA for V(D)J sequences of T or B cells. The company also offers Feature Barcoding, a technology that uses oligo-labeled antibodies to determine the expression of cell surface proteins.

“From the same cell, researchers can profile the transcriptome, tens to hundreds of cell surface proteins, B/T-cell clonal diversity through immune receptor sequencing, and their antigen specificity,” Schnall-Levin asserts. “The solution also helps uncover the diversity of malignant cells and immune cells in the tumor milieu.”

The importance of moving beyond the averaging effects of bulk sequencing techniques is also emphasized by Anjali Pradhan, vice president of product management at Mission Bio. “Single-cell approaches reveal cellular heterogeneity,” she says. “This includes being able to detect rare cell populations that may represent therapy response, disease progression, and even therapy resistance.”

“Our Tapestri Platform can provide both genotype and immunophenotype of a cell simultaneously across thousands of individual cells,” she continues. “Being able to pair both the mutational profile along with the protein expression of a cell enables researchers and clinicians to both measure the disease (mutated cancer cells) along with the immunophenotyping. For the first time, we can understand what oncogenic mutations are present in each cell and identify the tumor milieu to better understand the interplay between immune cells and cancer cells.”

Coping with perpetual novelty

Terms such as “adaptive agents,” “multiagent systems,” and “emergent behaviors” are perhaps best recognized in the social sciences, particularly in economics. But the terms can be relevant anywhere powerful computational tools are used to analyze immense data sets, discern patterns, and model tangled webs of cause and effect. For example, the terms correspond to concepts that are being explored in immunology, where artificial intelligence (AI) is being used to identify and characterize the adaptive agents known as immune cells. Granted, immune cells don’t make choices the way people decide economic matters, but they can, for example, distinguish between self and non-self, triggering local actions and contributing to systems-level effects including cancer suppression.

Immune cells may also be said to process past experiences and form adaptive expectations of a sort, namely, T- and B-cell receptors. To help researchers treat immune cells like adaptive agents, several companies are developing technologies to amass and sift through sequencing data about immune receptor genes. For example, Adaptive Biotechnologies has developed an immune medicine platform that looks at the data stored in individual receptors and gathers insights both at the individual and population levels.

“At one end of the spectrum are technologies that can be used to study single cells, while at the other end are bulk sequencing technologies that can sequence all cells in a given sample,” says Thomas Manley, MD, vice president of clinical development and medical affairs at Adaptive Biotechnologies. “Large-scale immunosequencing such as Adaptive’s immunoSEQ Technology lands in the middle, with benefits of both technologies allowing researchers the ability to study, quantify, scale, track, and monitor the T and B cells that are key players of the adaptive immune response.

“At the epicenter of our immune medicine platform is a massive and growing amount of immune receptor sequencing data. We currently have 58 billion immune receptors in our clinical immunomics database.

“On the T-cell side of the Adaptive platform, we partnered with Microsoft to use AI, machine learning, and cloud computing to map the T-cell immune responses to clinically relevant antigens at a population level. On the B-cell side of our platform, clonoSEQ is used to monitor minimal residual disease in clinical trials to determine treatment efficacy or to help a clinician monitor and manage patient care.

“Some of the new drug discovery offerings we are excited about in oncology include using our sequencing and TCR-antigen mapping technology, immunoSEQ T-MAP, in collaboration with AstraZeneca, to inform signatures of immune response (or resistance) to cancer therapies which may provide information to guide treatment decisions. And we have our ongoing collaboration with Genentech to identify optimal sequences to develop TCR-mediated cellular therapies. Adaptive also recently launched an antibody drug discovery capability where we can rapidly identify fully human antibody sequences and screen them for the best antibody candidates for therapeutic development.”

Another company that leverages single-cell technologies and computational resources to generate highly granular and far-reaching immune surveys is the recently launched Immunai. The company asserts that it has developed a “vertically integrated platform for multi-omic single-cell profiling that offers a broader view of the immune system in states of health, disease, and treatment to examine the body’s response to stimulus.” This platform, the company adds, can help pharmaceutical companies accelerate immunotherapy discovery and development.

“From a single blood sample, we derive over a terabyte of immune cell data,” notes Noam Solomon, PhD, co-founder and CEO of Immunai. “We then apply our proprietary neural network models and transfer learning techniques to create immune profiles based on differentiated elements, such as state-specific expression versus normal expression. These immune profiles allow us to identify subtle changes in cell types and understand how they interplay with other cells and proteins. This level of insight generation supports biomarker discovery and target validation, allowing our partners to accurately measure the efficacy of therapies in clinical trials.”

The company has formed partnerships with 10x Genomics and leading academic institutions such as Baylor College of Medicine, Stanford, and Harvard. Through these partnerships, Solomon states, Immunai has “mapped out millions of immune cells and their functions and built the largest proprietary data set for clinical immunological data in the world.”

“The combination of a multi-omic approach with our machine learning algorithms produces much more granular cell annotation and characterization than standard methods, which allows for accurate analysis of underlying biology,” Solomon continues. “All of these modalities are on the table, but the most critical new aspect of our platform that we’ll roll out this year is an extension into functional genomics.

“We will use CRISPR screens to help validate targets we’ve found using our reverse translational approach on our existing platform. We remain interested in not only the peripheral immune system but, in an oncology context, its nexus with the tumor and the tumor microenvironment, so we may also explore spatial.”