How far are we from single cell immunopeptidomics?

Mapping the diversity of HLA-presented peptides is key to understanding immune recognition and designing more effective precision immunotherapies.

Over the last three decades, immunopeptidomics has progressed dramatically—from the first detection of just 9 HLA peptides from over a billion cells (Hunt et al., 1992) to the identification of 10,000–15,000 peptides from 10 million cells using modern workflows such as conventional methods or Chip-IP (Li et al., 2023). However, despite advancements in detecting an ever-higher number of peptides, these approaches do not guarantee the reliable detection of all mutation-derived neoepitopes, instead favoring the detection of the most abundant ones.

In single-cell analysis, these limitations intensify—where ultra-low peptide abundance, inefficient purification, and LC-MS sensitivity constraints make rare neoepitopes nearly undetectable.

Recent advances in miniaturized workflows in single cell proteomics and emerging single-molecule sequencing technologies, such as nanopore sequencing, offer potential solutions to these limitations. While these methods hold promise for both proteomics and immunopeptidomics, challenges in handling complex peptide mixtures, signal-to-noise ratios, and post-translational modifications still hinder their full deployment. For further insight into emerging single-molecule technologies in relation to MS-based proteomics, see Alfaro et al. (2021) and MacCoss et al. (2022).

Immunopeptidomics has grown exponentially, but does this progress translate to true single-cell resolution? Table 1 and Figure 1 highlight key trends and remaining bottlenecks.

According to Table 1, while HLA peptide detection has grown faster than Moore’s Law, increasing detection numbers alone does not guarantee functional relevance—especially at the single-cell level. Table 1 places this growth in context with theoretical models (Moore, 1965).

While the limit of detection (LOD) depends on a peptide’s physicochemical properties, the sample matrix (Salek et al., 2024), the LC-MS platform, and whether detection is targeted or untargeted, its overall trend has mirrored the exponential increase in the number of detected peptides over the years (Beattie & Jones, 2023).

In the following, for the sake of argumentation, we consider the current limit of detection in Figure 1, as 1 attomole. To estimate how this translates into HLA-peptide detection per cell and the required input cell number, and to assess how close we are to achieving single-cell immunopeptidomics by LC-MS, I considered a few key parameters:

• The efficiency or yield of immunoprecipitation (IP), currently the only available enrichment method.

• The number of input cells in the experiment.

• HLA-peptide expression per cell, which defines the available signal.

• The current LC-MS detection limit, estimated at 1 attomole.

  
  

Since 1 attomole corresponds to approximately 602,300 molecules, this sets a fundamental sensitivity threshold for peptide identification.

Figure 1. The heatmap shows the number of HLA peptides per cell (y-axis) as a function of total cell input (x-axis), with color representing the total number of detectable peptides at 50% immunoprecipitation (IP) yield (log scale). Dashed white and purple lines indicate sensitivity thresholds for 1%, 50%, and 100% IP yields, marking where peptide detection meets the minimum LC-MS limit. Labeled points correspond to key detection limits, including 6 × 10⁵, 1.2 × 10⁶, and 6 × 10⁷ cells, depending on peptide expression levels per cell and IP efficiency.This heatmap provides insight into the minimum number of cells required to detect peptides at different expression levels and purification efficiencies.

The Next 10 Years: What to Expect vs. What Is Needed

According to Table 1, LC-MS sensitivity is expected to improve by an order of magnitude every 5 years. While this would significantly lower detection limits, even a decade of progress will not be enough to achieve comprehensive single-cell immunopeptidomics.

These advancements would already improve detection in small biopsies, but achieving true single-cell resolution for low-abundance peptides requires more than incremental sensitivity gains—it demands breakthroughs in both peptide enrichment and detection efficiency. 

ImmunoEdge: Moving Beyond Incremental Gains

While significant advances have been made in HLA peptidomics, the ultimate goal is not just detecting more peptides from less cells—it is about identifying functionally relevant, immunocompetent neoepitopes that can drive effective immune responses.

The PRM-based method developed at the DKFZ (German Cancer Research Center) remains one of the most sensitive approaches for detecting mutation-derived neoepitopes, enabling the identification of a neoepitope from 2.5 million patient-derived xenograft (PDX) cells and another from a 17.5 mg tumor biopsy (Salek et al., 2024). However, despite these advancements, we remain several orders of magnitude away from achieving functional single-cell immunopeptidomics.

At ImmunoEdge, we are not just waiting for incremental LC-MS sensitivity improvements- we are redesigning peptide isolation and detection strategies to prioritize functional immunogenic targets, starting with small biopsies before expanding beyond. Our focus is on:

✔  Enhancing selective enrichment to isolate functionally relevant neoepitopes for better therapeutic targeting.

✔  Integrating next-generation LC-MS methodologies with optimized workflows to enhance both sensitivity and selectivity beyond expected technological progress.

✔  Exploring alternative and hybrid approaches to overcome HLA-peptide detection constraints.

By shifting the focus from sheer detection numbers to identifying functionally relevant neoepitopes, ImmunoEdge is accelerating the development of high-sensitivity, high-selectivity workflows- paving the way for precise, actionable insights into tumor immunogenicity and therapeutic interventions.