Identification of Titin Domains

Uniqueness of fit of titin domains into cryo-EM map

We used AlphaFold to predict the structures of the individual Ig and Fn domains in super-repeat 4 of the 11-domain super-repeat of the C-zone of the human cardiac thick filament. See text of bioRxiv preprint doi: https://doi.org/10.1101/2023.04.11.536274. We then fitted these domains to the 11 domains observed in the two titin strands of the cryo-EM map. The fit appears to be unique (ED Fig 5 of manuscript). To test this further, we compared the fit of the Ig domain 8 model into the map density we have interpreted as Ig domain 8, with the fit into an Fn domain (domain 9, i.e. a shift of one domain). This tests whether an Ig domain fits into an Fn map (Fig. 1). Clearly, fitting the Ig model into the Fn map does not work, as the new “best fit” is substantially worse: the model projects outside the map, and other parts of the map are not filled with the model. We also did the inverse, comparing the fit of an Fn domain into an Fn map (domain 7) with the fit into an Ig domain (domain 8) (Fig. 2); the fit is best when the Fn model is in the Fn map. This is shown for Titin A (Figs 1 and 2) and Titin B (Figs 3 and 4).

Please zoom in for details.

This is best appreciated in 3 dimensions, as demonstrated in the corresponding movies, numbered 1-8 (see red typeface in above figures).