UCSF

One third of the human proteome relies on the Sec61 translocon for proper folding and function. Each Sec61 client has a unique signal peptide or signal anchor, which plays an essential role in promoting its translocation into the endoplasmic reticulum. Cotransins are cyclic depsipeptides that inhibit the biogenesis of a subset of secreted and membrane proteins by preventing cotranslational translocation into the endoplasmic reticulum. Sensitivity to cotransins is determined by the protein’s N-terminal signal peptide or signal anchor, which intercalates into the lateral gate of Sec61 to initiate translocation. However, it has been difficult to predict Sec61 client sensitivity to cotransins due to the large diversity of signal peptides and signal anchors and the lack of high throughput assays to effectively survey all predicted human Sec61 clients. We developed a fluorescence-based reporter system with a library of 3880 human (3212) and mouse (668) signal peptides to interrogate which Sec61 clients are most affected by cotransin analogs. This pooled-cell screening platform allowed us to profile two cotransins with distinct effects on cancer cells, KZR-9873 and KZR-8445. We identified a total of 665 sensitive signal peptides, most of which were previously not known to be sensitive to cotransins. Among several validated targets, we discovered that the oncoprotein HER3, a coreceptor for HER2, is preferentially inhibited by the more selective cotransin, KZR-9873. Comparison of the human and mouse signal peptide paralogs within the library revealed a position-dependent role for Arg and Lys in conferring cotransin sensitivity. Our screening platform revealed distinct signal peptide sequence determinants that confer sensitivity to cotransins. We used mass spectrometry to profile the effects of KZR-9873 and KZR-8445 on endogenous proteins in three cancer cell lines, CAL27, BxPC3 and SW48 cells. This approach allowed us to interrogate Sec61 client sensitivity beyond SP-containing proteins, since type II and multi-spanning membrane proteins are understudied with respect to their potential cotransin sensitivity. Of the Sec61 clients identified across the three cell lines, there were 360 KZR-9873-sensitive and 659 KZR-8445-sensitive Sec61 clients. We identified a total of 51 and 81 type II membrane proteins that were sensitive to KZR-9873 and KZR-8445, respectively. In addition, we identified 34 and 45 multi-spanning N-cyt membrane proteins that were sensitive to KZR-9873 and KZR-8445, respectively. Our proteomic approach showed that cotransins can inhibit the biogenesis of proteins that utilize the lateral gate of Sec61 to initiate cotranslational membrane insertion.