(A) The plot from Figure 3A, with genes in the TRC pathway (GET pathway in yeast) labeled. All five genes tested in the sublibrary screens (WRB and CAMLG were omitted) produced significant increases in HiLITR activity in the tail-anchored (TA) screen (p=1.7e-7, hypergeometric test). (B) Schematic of the TRC pathway. (C) Quantitation of individual fluorescence-activated cell sorting (FACS) analysis of gene knockdown in the TRC pathway. The K562 TA and signal-anchored (SA) (top) and ER (bottom) HiLiTR cell lines were transduced with individual sgRNAs against TRC pathway genes. Log2-transformed ratio of high mCherry to low mCherry cells was calculated for each plot and normalized to that of nontargeting (NT) control. (D) Schematic showing possible membrane insertion pathway of TA protease. Most protein traffics to the outer mitochondrial membrane (OMM), but a subpopulation may be nonproductively handled by TRC pathway chaperones, resulting in rejection from ER insertion by the receptors, adaptor-mediated recruitment of ubiquitination machinery, and subsequent degradation. (E) Schematic showing possible membrane insertion pathway of mTA* protease. (F) Immunofluorescence microscopy analysis of TRC pathway knockdown. In HeLa cells, the localization of mTA* protease was compared to Golgi (GRASP65) and mitochondrial (TOMM20) markers. Scale bar, 10 µm. Note: knockdown of CAMLG in HeLa cells impaired cell adherence, preventing immunofluorescence analysis. (G) Quantification of data in (F), along with ~20 additional fields of view per condition (total ~50 cells per sample). For each cell, the mean intensity of Golgi-colocalized GFP was divided by the mean intensity of mitochondria-colocalized GFP. ***p<0.001, Student’s t-test. Full data in Figure 3—figure supplement 1—source data 1.
Figure summary - Analysis of TRC pathway genes in the CRISPRi sublibrary screens. The TRC pathway is the first pathway discovered for the targeting of ER-destined TA proteins and is well-characterized (Borgese et al., 2019). (B) shows the key players in the TRC pathway. ER-targeted TA proteins that are TRC pathway clients are handled directly by two chaperones, SGTA and TRC40 (Chang et al., 2010; Schuldiner et al., 2008; Stefanovic and Hegde, 2007). Three adaptor proteins (UBL4A, TRC35, and BAG6) (Mariappan et al., 2010) coordinate handoff between SGTA and TRC40 (Shao et al., 2017; Wang et al., 2010), while BAG6 additionally recruits the E3 ubiquitin ligase RNF126 (not shown) to degrade nonproductively associated proteins (Rodrigo-Brenni et al., 2014). At the ER membrane, WRB and CAMLG act as receptors to assist with insertion of the client protein (Schuldiner et al., 2008; Vilardi et al., 2011; Yamamoto and Sakisaka, 2012), but proteins with significant positive charge flanking the transmembrane domain (such as mitochondrial TA proteins) are rejected, either by the receptors or due to the energetic barrier posed by the ER membrane (Rao et al., 2016). The chaperones (SGTA and TRC40) and adaptors (TRC35, BAG6, and UBL4A) were included in our CRISPRi sublibrary, while WRB and CAMLG were not. We further explored the TRC pathway with individual knockdown of the chaperones SGTA and TRC40, the adaptor TRC35, and the receptors WRB and CAMLG. Interestingly, the profiles of HiLITR performance across the three HiLITR configurations segregated based on the function of each protein in the TRC pathway (C). Notably, knockdown of none of the proteins affected the SA screen HiLITR configuration, consistent with the fact that the TRC pathway acts only on TA proteins. In the TA screen configuration, knockdown of the chaperones and adaptors both increased HiLITR activation (C). If the chaperones are knocked down, there will be decreased mishandling of mitochondrial TA protein, and therefore an increase in normal topogenesis, localization to the mitochondria, and release of mitochondrial TF (D). Similarly, upon loss of adaptors, handoff of TA protein between SGTA and TRC40 is less coordinated. Adaptor-mediated degradation of uninserted TA protein will also be reduced. Both effects promote increased targeting of the TA protein to the mitochondrial membrane, increasing HiLITR activation (D). Knockdown of receptors in the TA screen configuration did not impact HiLITR activation (C). Since mitochondrial TA proteins that reach the ER are rejected on the basis of charge, loss of the receptors will have no additional impact (D). In the ER screen configuration, we used the mutant TA protein (mTA*-protease) that partitions between the OMM and ER membrane. We observed that knockdown of the chaperones and receptors decreased HiLITR activation (A, C). Loss of chaperones will mean less mTA* protease is routed to the ER, meaning less ends up colocalized with the ER-targeted TF, leading to reduce HiLITR activation (E). Similarly, if the receptors are disrupted, the mTA* protease cannot be inserted into the ER, decreasing HiLITR activation (E). In contrast to the chaperones and receptors, the adaptors gave mixed results, with two of the three adaptors producing no significant result in the ER screen (A, C). Knockdown of the adaptors will decrease handoff of the mTA* protease between SGTA and TRC40, providing opportunity for escape to the mitochondrial membrane (and decreased HiLITR activation). However, this effect is opposed by the fact that loss of adaptors will reduce degradation of the mTA* protease, providing more time for its insertion into the ER membrane (increasing HiLITR activity). As such, the impact of knocking down the adaptors is harder to predict for the ER screen (E). Lastly, we tested the effect of TRC protein knockdown on the localization of the mutant TA protease. Knockdown of SGTA, TRC40, and WRB decreased activation of HiLITR in the ER screen configuration, indicating reduced ER targeting of the mTA* protease. As expected, knockdown of any of these components reduced the fraction of mTA* protease colocalizing with the Golgi (G). TRC35 also decreased the fraction of mTA* protease colocalized with the Golgi, despite a neutral effect in the ER screen HiLITR configuration. It is likely that loss of TRC35 increases the fraction of mTA* protease which is rescued from the TRC pathway, while also increasing the efficiency by which unrescued mTA* protease is inserted into the ER (E). This would result in an increase in mitochondrial mTA* protease and neutral effect on ER mTA* protease, which would decrease the ratio of Golgi-localized mTA* protease without affecting total Golgi-localized protease or subsequent HiLITR activation in the ER screen configuration.