4.10 Mass Cytometry (CyTOF) & Single-Cell Elemental Analysis

4.10 Mass Cytometry (CyTOF)

Mass cytometry (cytometry by time-of-flight — CyTOF, commercialized as Helios) replaces fluorescent dyes with metal isotope tags and replaces the photomultiplier/ADC signal chain with an inductively coupled plasma time-of-flight mass spectrometer.

Principle

Antibodies are conjugated to purified rare-earth metal isotopes (e.g., lanthanides, Maxpar chelation). The cell is stained, then introduced into the instrument; an argon plasma atomizes and ionizes it, and a time-of-flight mass analyzer separates ions by mass-to-charge ratio. Each isotope channel corresponds to one antibody specificity. A few dozen to 40+ markers can be measured per cell.

Advantages Over Fluorescence Flow

• No spectral overlap. Metal isotope channels are resolved by mass, so there is no compensation as fluorescence flow knows it. (A small isotope-purity spillover remains: a tagged isotope may contain trace adjacent isotopes that register in a neighboring mass channel.)
• No autofluorescence. Biological material does not fluoresce in the mass range, so background from dead/damaged cells is dramatically reduced.
• Many channels. 40+ channels are routine, allowing deep immunophenotyping in a single tube (e.g., a 40-marker immune profiling panel).

Trade-offs

Lower throughput and higher cell loss make mass cytometry less suitable for very rare events requiring millions of events, and sample introduction consumes more cells per acquisition. Cells are also destroyed in the plasma, so no sample can be sorted or re-analyzed after CyTOF.

Isotope Spillover

While there is no fluorescence compensation, isotope purity is the equivalent constraint. A metal tag is never 100% pure; a small fraction lands in adjacent mass channels and must be subtracted (spillover/compensation matrix analogous to fluorescence). Purity and spillover matrices are supplied by the reagent manufacturer and applied during preprocessing. Barcoding (by metal tags) pools many samples into one acquisition, reducing run-to-run variability.

High-Dimensional Analysis

Because panels routinely measure 30–40 markers per cell, mass cytometry is intrinsically high-dimensional. Conventional manual gating does not scale; analysis relies on:

• Dimensionality reduction: viSNE (t-SNE-based), UMAP, PCA — to visualize populations in 2D maps.
• Clustering: SPADE, FlowSOM, PhenoGraph, X-Cluster — to identify populations automatically.
• Differential analysis: comparing cluster abundance or marker expression across conditions.

These tools are shared with fluorescence flow but are the default workflow for mass cytometry because the marker count exceeds what manual gating can manage. A typical workflow: barcode samples, pool, stain with a 40-marker panel, acquire, de-barcode, run viSNE/FlowSOM, then compare cluster abundances across conditions.

Applications

Mass cytometry excels at deep immune profiling: comprehensive T-cell, B-cell, myeloid, and innate subsets in one tube; cytokine/stem/activation panels; and clinical research on cancer immunology and immunotherapy response. It is less commonly used for routine clinical enumeration (e.g., CD34) where fluorescence flow's throughput and quantitation dominate.

Worked Example: Isotope Spillover Correction

A panel uses a ¹⁵³Eu-tagged antibody in channel 153 and a ¹⁵⁴Eu-tagged antibody in channel 154. The ¹⁵³Eu reagent is 99% pure, so 1% of its signal spills into the adjacent 154 channel. If a cell has high ¹⁵³Eu (say 10,000 counts) and no true 154 signal, the spillover contributes ~100 counts in channel 154 — enough to look like low positivity. The spillover matrix (from single-stain metal-tag controls) subtracts this: 154_corrected = 154_observed − 0.01 × 153_observed. The principle is identical to fluorescence compensation; only the source (isotope impurity) and the channel (mass) differ.

Sample Introduction and Cell Loss

CyTOF sample introduction uses a nebulizer and spray chamber that lose a fraction of cells before they reach the plasma, so only ~30–50% of input cells are typically recorded. This cell loss, combined with lower throughput, raises the input requirement: a panel needing 100,000 acquired events may need 200,000–300,000 input cells. For precious or rare samples, this loss is a real constraint, and pre-enrichment may be required.

Exam Traps

• CyTOF uses metal isotope tags and ICP-MS, not fluorochromes and PMTs.
• No spectral overlap / no compensation (only isotope-purity spillover), and no autofluorescence.
• Throughput is lower and cell loss higher than fluorescence flow; cells are destroyed by the plasma.
• High-dimensional analysis (viSNE, FlowSOM, SPADE) is intrinsic because of the high marker count.