Reanalysis of a published cryo-EM dataset reveals domain-level variability in the Hrd1-Hrd3 complex.

The Hrd1-Hrd3 complex is a central component of the endoplasmic reticulum-associated degradation (ERAD) pathway; however, its conformational states during substrate recognition and retrotranslocation remain poorly characterized at the structural level. Conventional cryo-EM data processing pipelines typically aim to obtain a single high-resolution reconstruction, during which low-abundance or subtly distinct conformational states may be averaged out or lost. To explore potential structural heterogeneity in publicly available data, we reanalyzed the cryo-EM dataset of the Hrd1-Hrd3 complex (EMPIAR-10099). A side-view particle enrichment strategy combined with iterative 2D reclassification was applied to re-sort particles and reconstruct distinct subsets. Three density classes were obtained from the side-view-focused classification workflow, with overall resolutions of 7.24 Å, 6.80 Å, and 6.93 Å, respectively. Comparative analysis revealed consistent differences in the position of the Hrd3 luminal region relative to the Hrd1 transmembrane core, showing distinguishable Hrd3-associated density distributions among the classified particle subsets. These results indicate that, even at moderate resolution, a classification strategy focused on side-view particles and iterative 2D sorting can identify domain-level density variability among selected particle subsets from publicly available cryo-EM datasets. This study does not rely on new algorithm development and is fully implemented using existing cryo-EM analysis tools, providing a practical workflow for reanalysis of publicly available cryo-EM data.