Fuchsberger et al. (2023) quantified, using Shannon channel capacity, how C-type lectin receptors (CLRs) on monocytic reporter cell lines transmit information encoded in microbial glycans to NF-kappaB, finding that dectin-2 is a systematically inefficient (~0.7 bit) channel with a low signal-to-noise ratio whose capacity cannot be raised by enhancing sensitivity. The study's stated caveats and hypotheses — that its conclusions came from model cell lines used as surrogates for primary immune cells, that the molecular origin of dectin-2's signalling deficit remains unidentified, and that the inefficiency may reflect evolutionary avoidance of self-recognition — leave open a coherent set of questions that define the current problem: whether the same channel capacities and noise properties hold in primary immune cells; what molecular mechanism produces dectin-2's low signal-to-noise ratio; and whether dectin-2's inefficiency reflects evolutionary avoidance of self-recognition (i.e., whether a more sensitive dectin-2 would cause autoimmune pathology). Two further questions extend the study's results without having been posed by the study itself: whether channel capacities in primary cells couple to functional outputs such as cytokine profiles; and whether molecular interventions beyond MCL co-expression can increase the maximal information dectin-2 transmits. No work resolving any of these questions is known in the literature to date; the problem remains open.

Background

Glycans are branched sugar chains covalently attached to proteins and lipids at the cell surface; unlike nucleic acids and proteins they are not genomically templated, and subtle differences in linkage and branching encode information used in cell-cell communication, including self/non-self discrimination. Glycans are read by lectins, glycan-binding proteins that translate binding events into intracellular signal transduction. This problem concerns C-type lectin receptors (CLRs), a family of lectins on innate immune cells; the study at issue used DC-SIGN (dendritic-cell-specific ICAM-3-grabbing non-integrin, a mannose-binding CLR), MCL (macrophage C-type lectin, also called dectin-3), dectin-1, dectin-2, and MINCLE (macrophage-inducible C-type lectin), plus the TNF-alpha receptor (tumor necrosis factor alpha receptor, an inflammatory cytokine receptor) and TLR-1&2 (the toll-like receptor 1 and 2 heterodimer, a pattern-recognition receptor pair for microbial lipoproteins) as comparators. Dectin-2 and MINCLE both signal through the FcR-gamma chain adapter (the common Fc receptor gamma chain, a transmembrane adapter protein that relays activation signals from immune receptors) into the CARD9-BCL-10-MALT1 pathway (a downstream adaptor-protein cascade) that activates NF-kappaB (nuclear factor kappa-B), a transcription factor driving cytokine genes; dectin-1 uses a distinct intracellular pathway to NF-kappaB.

The quantitative framework comes from information theory: the mutual information I(X;Y)I(X;Y) between a stimulus (ligand concentration, XX) and a single-cell response (reporter output, YY), maximized over input distributions, gives the channel capacity in bits — the maximum information a signaling pathway can transmit. Prior measurements found roughly 1-bit capacities in mammalian signaling: the TNF-alpha pathway reached 1.64 bits in a comparable reporter system (Cheong et al., 2011), and receptor-level signal flow in other systems is similarly limited (Suderman et al., 2017).

Fuchsberger, Kim, Baranova, Vrban, Kagelmacher, Wawrzinek, and Rademacher (bioRxiv 2021; published in eLife 12:e69415, 2023) built monoclonal U937 human monocytic cell lines carrying an NF-kappaB-driven GFP reporter (GFP, green fluorescent protein, used here as the single-cell fluorescent readout of NF-kappaB activity) and overexpressing individual CLRs, stimulated them with defined glycosylated ligands (e.g., the model trisaccharide-bearing glycodendrimer FurFurMan, invertase, and mannan), and computed channel capacities from flow-cytometry dose-response data at single-cell resolution. Most receptors transmitted about 1 bit or more (TNF-alpha receptor ~1.35 bits, dectin-1 ~1.22 bits, MINCLE and TLR-1&2 ~1.00 bit), but dectin-2 transmitted only ~0.7 bits with FurFurMan. Strikingly, sensitivity and capacity dissociated: overexpressing the FcR-gamma co-receptor, co-expressing DC-SIGN, or stimulating with a high-affinity anti-dectin-2 antibody all improved sensitivity (lower EC50, the half-maximal effective concentration) without increasing channel capacity, whereas co-expression of MCL significantly increased dectin-2's channel capacity, but only at low maximum ligand concentrations. Co-expressing dectin-1 and dectin-2 yielded a channel whose capacity was approximately the average of the two — a compromise, not a sum. Decomposing signal and noise power showed that both are low for dectin-2 compared with the other receptors, with noise power exceeding signal power and hence a significantly lower signal-to-noise ratio; the authors concluded that the reduced channel capacity is dictated by the lower variation of the mean GFP expression (i.e., signal power), and explicitly noted that a similar conclusion cannot be drawn for the noise power, since it is inherently coupled to the signal power. The accompanying controls addressed stimulation time and steady-state reporter expression, clonal selection (re-stimulation of GFP-sorted clones gave the same broad response), and receptor-number variability, but did not identify a molecular origin of the deficit. The authors hypothesized that dectin-2's inefficiency is an evolutionary safeguard: because dectin-2 binds high-mannose structures that also exist on mammalian cells, a too-sensitive dectin-2 could drive permanent self-recognition and autoimmunity. All conclusions came from overexpression reporter cell lines, and the paper closes by stating that work is ongoing on the central questions of glycan-lectin-based cellular communication.

Problem Statement

The problem is to extend the quantitative, information-theoretic characterization of glycan-based communication from overexpression reporter cell lines to the real biological setting and to explain the anomalous dectin-2 channel. Its linked components consolidate questions raised or left open by Fuchsberger et al. (eLife 12:e69415, 2023); the coupling to cytokine profiles in component (1) and component (4) as a whole go beyond the questions the source itself posed and are retained as extensions motivated by its demonstrated sensitivity/capacity dissociation: (1) measure the channel capacities and noise properties of C-type lectin receptor signalling — dectin-2 in particular, alone and in combination with co-expressed lectins — in primary immune cells such as dendritic cells and macrophages, and determine how multi-receptor integration in these cells leads to distinct functional outcomes such as cytokine profiles; (2) identify the molecular mechanism underlying dectin-2's low signal-to-noise ratio and low signal power, i.e., discriminate among candidate origins such as the size and formation of receptor signalling clusters and background signalling from mannose structures on the cell surface (the source study's own candidates for the MINCLE/dectin-2 difference) or other mechanisms such as ligand rebinding and negative feedback, none of which the source study tested, its controls having addressed only stimulation time and steady-state reporter expression, clonal selection, and receptor-number variability; (3) test the hypothesis that dectin-2's inefficiency evolved to avoid self-recognition of mammalian high-mannose structures, specifically whether a more sensitive dectin-2 causes autoimmune defects; and (4) determine whether molecular interventions beyond MCL co-expression — altering receptor clustering, co-receptor stoichiometry, or downstream pathway components — can increase the maximal information dectin-2 transmits, given that sensitivity-enhancing manipulations (FcR-gamma overexpression, DC-SIGN co-expression, high-affinity antibody stimulation) demonstrably cannot. Any one of the four components is an independently sufficient resolution of the problem. Progress on any component should respect the overarching objective: a quantitative, mechanistic account of how immune cells decode glycan-encoded information, of which the reporter-cell-line findings (approximately 1-bit CLR channels, the dectin-2 capacity deficit, compromise-type integration of dectin-1/dectin-2, and MCL synergy at low dose) are the established baseline.

References

  1. Fuchsberger FF, Kim D, Baranova N, Vrban H, Kagelmacher M, Wawrzinek R, Rademacher C (2023). Information transfer in mammalian glycan-based communication. eLife 12:e69415. DOI: 10.7554/eLife.69415. https://doi.org/10.7554/eLife.69415
  2. Fuchsberger FF, Kim D, Kagelmacher M, Wawrzinek R, Rademacher C (2021). Information transfer in mammalian glycan-based communication. bioRxiv preprint. DOI: 10.1101/2021.05.10.443458. https://doi.org/10.1101/2021.05.10.443458
  3. Cheong R, Rhee A, Wang CJ, Nemenman I, Levchenko A (2011). Information Transduction Capacity of Noisy Biochemical Signaling Networks. Science 334(6054):354-358. DOI: 10.1126/science.1204553. https://doi.org/10.1126/science.1204553
  4. Suderman R, Bachman JA, Smith A, Sorger PK, Deeds EJ (2017). Fundamental trade-offs between information flow in single cells and cellular populations. Proceedings of the National Academy of Sciences 114(22):5755-5760. DOI: 10.1073/pnas.1615660114. https://doi.org/10.1073/pnas.1615660114
  5. Umeki N, Kabashima Y, Sako Y (2025). Evaluation of information flows in the RAS-MAPK system using transfer entropy measurements. eLife 14:e104432. DOI: 10.7554/eLife.104432. https://doi.org/10.7554/eLife.104432
  6. Tkačik G, ten Wolde PR (2025). Information Processing in Biochemical Networks. Annual Review of Biophysics 54:249-274. DOI: 10.1146/annurev-biophys-060524-102720. https://doi.org/10.1146/annurev-biophys-060524-102720
  7. Rabinovich GA, Rademacher C, Schattner M, Macauley MS (2026). Glycan-Binding Proteins in Immunity. Annual Review of Immunology 44:267-294. DOI: 10.1146/annurev-immunol-083024-030822. https://doi.org/10.1146/annurev-immunol-083024-030822
  8. Lefèbre J, Falk T, Ning Y, Rademacher C (2024). Secondary Sites of the C-type Lectin-Like Fold. Chemistry – A European Journal 30:e202400660. DOI: 10.1002/chem.202400660. https://doi.org/10.1002/chem.202400660
  9. Lefèbre J, Besch M, Csorba N, Garami K, Orgován Z, Schlosser G, Bermejo I, Ábrányi-Balogh P, Keserű GM, Rademacher C (2025). Covalent Activation of the C-type Lectin DC-SIGN. Angewandte Chemie International Edition 65:e20594. DOI: 10.1002/anie.202520594. https://doi.org/10.1002/anie.202520594
  10. Willinger O, Granik N, Salomon T, Goldberg S, Amit R (2026). High avidity phase-separated RNA-protein sialogranules sense lectins and inhibit influenza infection. bioRxiv preprint. DOI: 10.64898/2026.02.22.707264. https://doi.org/10.64898/2026.02.22.707264
  11. Joshi P, Satyajith A, Panda D, Rajeshwari B R, Thattai M, Balasubramanian N (2024). Cell-surface glycans are quantitative reporters of Golgi dysfunction in single cells. bioRxiv preprint. DOI: 10.1101/2024.06.17.599374. https://doi.org/10.1101/2024.06.17.599374