Mara C. Duncan

Epiblast lumen formation is a critical step that occurs during human embryo implantation. It begins with the apicosome, an apical compartment partly derived from endocytosed material. However, how the endo-lysosomal system contributes to apicosome formation and later lumen development remains unclear. Using a human pluripotent stem cell-derived model of epiblast formation, we show that apicosome formation is accompanied by transient expansion of early endosomes, late endosomes, and lysosomes, as well as the formation of hybrid compartments with features of both early and late endo-lysosomal stages. These changes depend on the RAB GTPases RAB35 and RAB7, which are required for proper apicosome formation and lumen morphology of the human epiblast model. Our findings identify RAB35- and RAB7-dependent endo-lysosomal remodeling as a key mechanism driving apical membrane morphogenesis during human epiblast development.

Epitope tags are commonly used for various purposes in research labs. The DYKDDDDK-peptide epitope, trademarked as the FLAG epitope, is a commonly used epitope tag. It is often used for monitoring protein levels and for affinity chromatography. Multiple DYKDDDDK-binding antibodies are available; however, the mouse monoclonal anti-FLAG M2 is widely used due to its commercial availability in several formats. Many laboratory Saccharomyces cerevisiae strains, including the BY4741 strain that was used in multiple systematic deletion and tagging libraries, have a high-intensity band that cross-reacts with the FLAG-M2 antibody. The presence of this high-intensity cross-reactive band can be problematic in some applications. Here, we show that despite high-intensity in immunoblots, the cross-reacting band is not enriched by FLAG-M2 affinity beads under native conditions. We also report the fortuitous identification of a strain closely related to BY4741 that lacks the high-intensity cross-reactive band. Finally, contrary to anecdotal reports, we determined that the high-intensity cross-reacting band is not Rtf1. These findings and resources should assist other researchers using the DYKDDDDK-epitope for immunoblots and affinity chromatography.

Numerous biomedically important cargoes depend on adaptor protein complex-1 (AP-1) for their localization. However, controversy surrounds whether AP-1 mediates traffic from or to the Golgi. Robinson et al. (https://www.doi.org/10.1083/jcb.202310071) present compelling evidence that AP-1 mediates recycling to the Golgi.

The highly conserved HEATR5 proteins are best known for their roles in membrane traffic mediated by the adaptor protein complex-1 (AP1). HEATR5 proteins rely on fast-evolving cofactors to bind to AP1. However, how HEATR5 proteins interact with these cofactors is unknown. Here, we report that the budding yeast HEATR5 protein, Laa1, functions in two biochemically distinct complexes. These complexes are defined by a pair of mutually exclusive Laa1-binding proteins, Laa2 and the previously uncharacterized Lft1/Yml037c. Despite limited sequence similarity, biochemical analysis and structure predictions indicate that Lft1 and Laa2 bind Laa1 via structurally similar mechanisms. Both Laa1 complexes function in intra-Golgi recycling. However, only the Laa2–Laa1 complex binds to AP1 and contributes to its localization. Finally, structure predictions indicate that human HEATR5 proteins bind to a pair of fast-evolving interacting partners via a mechanism similar to that observed in yeast. These results reveal mechanistic insight into how HEATR5 proteins bind their cofactors and indicate that Laa1 performs functions besides recruiting AP1.

The highly conserved HEATR5 proteins are best known for their roles in membrane traffic mediated by the adaptor protein complex-1 (AP1). HEATR5 proteins rely on fast-evolving co-factors to bind to AP1. However, how HEATR5 proteins interact with these co-factors is unknown. Here, we report that the budding yeast HEATR5 protein, Laa1, functions in two biochemically distinct complexes. These complexes are defined by a pair of mutually exclusive Laa1-binding proteins, Laa2 and the previously uncharacterized Lft1/Yml037c. Despite limited sequence similarity, biochemical analysis and structure predictions indicate that Lft1 and Laa2 bind Laa1 via structurally similar mechanisms. Both Laa1 complexes function in intra-Golgi recycling. However, only the Laa2-Laa1 complex binds to AP1 and contributes to its localization. Finally, structure predictions indicate that human HEATR5 proteins bind to a pair of fast-evolving interacting partners via a mechanism similar to that observed in yeast. These results reveal mechanistic insight into how HEATR5 proteins bind their co-factors and indicate that Laa1 performs functions besides recruiting AP1.

The ability to complete an experiment successfully without mistakes is at the core of the scientific enterprise. Some scientists seem more able to avoid mistakes and missteps than others. In the field of cell biology, their success is often attributed to having naturally good hands. Here I propose that rather than possessing “good hands,” an innate and unlearnable aptitude for science, such scientists possess expert learning skills that can be effectively mastered. I share a straightforward approach to gaining and teaching expert learning skills in the research lab environment. I also discuss ongoing efforts by others to develop curricula that teach expert learning skills in laboratory courses.

Membrane traffic at the Golgi and endosomes plays many critical roles in the polarization and the morphogenesis of epithelial tissues. Studies into the roles of traffic in morphogenesis in mammals are often complicated by early embryonic lethality of mutations in membrane traffic as well as the inherent difficulty in imaging developing embryos posed by their size and location. Increasingly, human pluripotent stem cell (hPSC)-derived embryo- and organ-like systems (e.g., embryoids, organoids) provide a useful platform to illuminate the requirements of traffic in human embryonic tissue morphogenesis because these in vitro models are highly amenable to fluorescence microscopy and provide the ability to examine the role of essential genes not possible with animal studies. Here, we present a method to generate hPSC-cysts, a 3-D hPSC-based model of human epiblast lumen formation. This system provides unique opportunities to examine the role of membrane traffic during epithelial morphogenesis. We also present methods to process hPSC-cysts for immunofluorescence and staining with commonly used fluorescence labels useful for detecting defects in polarization and morphogenesis caused by defects in membrane traffic. Key words Membrane traffic Golgi Endosomes Stem cells hPSC hiPSC Immunofluorescence Method Lumen Polarity

APEX2-based spatial proteomics identifies cell polarity proteome of hPSC-cysts, a 3D model of human epiblast cavity formation.