Large-scale single-molecule analysis of tau proteoforms – Figure 4 with Vivek Budamagunta

In this video, Vivek Budamagunta presents Figure 4 from our recent Nature Methods paper titled, “Large-scale single-molecule analysis of tau proteoforms.” This figure covers the examination of the landscape of tau proteoforms in models of neurodegenerative disease.

Read the full paper.

Three takeaways from the video

“Phospho-tau is elevated” is not molecularly precise statement. When researchers report elevated phospho-tau in a disease, it’s unclear whether they mean more molecules carry random phosphorylation events, more molecules carry specific phosphorylations, or more molecules carry multiple phosphorylations at once. Nearly all available tools average signal across many molecules, quantifying the presence of an epitope in a population while discarding the single-molecule context. This makes them incapable of determining whether multiple phosphorylations occur together, incapable of proteoform analysis. As Vivek argues, the field has been building disease models on that blurry foundation for decades, and without knowing which proteoforms are actually present in a given model, researchers can’t tell which models best represent the disease they’re studying. Against a clinical trial success rate of roughly 10%, that’s a costly uncertainty.

Running one assay across many major research models reveals how different those models really are. The team pointed the 12-antibody Tau Proteoforms Assay at iPSC-derived neurons, cortical organoids carrying MAPT mutations, MiBrain models, mouse brain extracts, and human frontal cortex, analyzing all of them with a single assay so they could be compared at the molecular level. The heat map in Figure 4a captures 130 distinct proteoforms of the 768 the assay can measure, and only 11 of those 130 were shared across all models. Cellular models, organoids, MiBrains, and iNeurons, were dominated by 0N tau isoforms with heavy multi-site phosphorylation, resembling what’s seen in fetal tissue, while mouse and human brain looked distinct. Organoids carrying the IVS10+16 mutation associated with frontotemporal dementia carried the largest number of tau proteoforms at three months of culture. Reassuringly, the data also recapitulated known developmental biology; maturing organoids showed the expected 3R-to-4R isoform shift, with IVS10+16 organoids showing a 13-fold increase in 4R at three months that grew to 30-fold by six months, and miBrains shifted from the short 0N3R isoform toward longer 1N and 2N isoforms much as human brain development does.

At single-molecule resolution, tau phosphorylation is clearly not random. Collapsing the data to the isoform or single-PTM level — as other tools force you to — loses the molecular story: two models can show identical pT181 levels while looking entirely different at the proteoform level, because in one the pT181 sits alone on a molecule and in the other it co-occurs with pT217, pS396, or other modifications. Keeping the resolution reveals structure. The team observed up to six co-occurring modifications on a single tau molecule, and across all samples saw 18 of the 21 possible double-phosphorylation combinations but fewer than half of the possible triples, meaning not every combination that could be made is being made. Certain phosphorylations also appear to raise the likelihood of others on the same molecule: pT181 makes pT217 considerably more likely, while pT217 depends strongly on other sites being phosphorylated without those sites depending on it in return. That asymmetry suggests pT217 is either downstream in the cascade or that phosphatase activity is shaping the landscape. This asymmetry is uniquely clear with single-molecule resolution. For tau researchers, the question shifts from “is phospho-tau up?” to which proteoforms are up, which isoforms they derive from, and which PTMs co-occur.

Up next

In the next video in this series, James Joly shares Figure 5, walking through data showing how Iterative Mapping is already revealing new biology in samples from people with Alzheimer’s disease — biology that may one day be targetable with novel biomarkers and therapeutics.

If you want to access the full video series, please fill out the form here:

Access the full series

If you’d like to use the Nautilus Tau Proteoforms Assay or our new AKT1 and EGFR Proteoforms Assays in your research, please reach out through the form found on the Iterative Mapping Early Access Program webpage.