- Top of page
- Materials and Methods
The pH-sensitive green fluorescent protein (GFP) variant pHluorin is typically fused to the extracellular domain of transmembrane proteins to monitor endocytosis. Here, we have turned pHluorin inside-out, and show that cytoplasmic fusions of pHluorin are effective quantitative reporters for endocytosis and multivesicular body (MVB) sorting. In yeast in particular, fusion of GFP and its variants on the extracellular side of transmembrane proteins can result in perturbed trafficking. In contrast, cytoplasmic fusions are well tolerated, allowing for the quantitative assessment of trafficking of virtually any transmembrane protein. Quenching of degradation-resistant pHluorin in the acidic vacuole permits quantification of extravacuolar cargo proteins at steady-state levels and is compatible with kinetic analysis of endocytosis in live cells.
Although considerable advances have been made in understanding the mechanisms and regulation of vesicular transport, the development of novel techniques to visualize and quantify trafficking events will undoubtedly enhance our understanding of these processes. In particular, the Aequorea victoria green fluorescent protein (GFP) and its subsequent variants have been used extensively for studying protein, membrane and cytoskeletal dynamics in living cells. However, two important factors limit the use of GFP derivatives for quantitative analysis of cargo movement via vesicular trafficking: resistance to proteolysis in the lysosome (or vacuole in organisms such as Saccharomyces cerevisiae, which will be used in the present study) and insensitivity to changes in pH (1).
Persistent vacuolar fluorescence of GFP-cargo chimeras is readily apparent when monitoring trafficking through the endocytic pathway. Internalized cargo proteins are frequently targeted to the lysosome or vacuole for degradation, and the stable GFP tag often remains detectable for extended periods following degradation of the cargo protein. Thus, persistence of vacuolar GFP can confound whole-cell quantitative analysis of internalized cargo chimeras. Other fluorescent markers used in studies of endocytic trafficking, such as the lipophilic dye FM4-64 and the fluid-phase marker Lucifer yellow (2), have been used to follow endocytosis quantitatively. However, both dyes are used as general markers of endocytosis, and cannot be used to monitor trafficking of specific cargos. It is also worth noting that transmembrane cargo proteins that are commonly used for studying anterograde transport from the endoplasmic reticulum (ER) to the plasma membrane are frequently targeted for vacuolar proteolysis, thus GFP-tagged markers of numerous trafficking pathways besides endocytosis can render quantitative analyses difficult.
Several variants of GFP have been generated that, in theory, could reduce the accumulation of detectable fluorescence in the vacuole. For example, protease-sensitive variants, in which cleavage sites have been introduced into the sequence of GFP, are degraded upon delivery to sites of protease activity (3–5); however, their usefulness might be limited by the efficiency of cleavage. Instead, pH-sensitive GFP variants, known as pHluorins, display changes in excitation and/or emission properties depending on the pH of their local environment (6). Upon lowering of pH, ratiometric pHluorin shows a decrease in excitation at 395 nm with a concomitant increase in excitation at 475 nm, while ecliptic and superecliptic pHluorin show a pronounced dampening of excitation at both wavelengths (6,7). In wild-type (WT) yeast, cytoplasmic pH has been determined to lie between 6.5 and 7 when cells are grown in media with a pH of 5.5, whereas vacuolar pH has been reported at 5.45 under similar growth conditions (8–10). Although ratiometric pHluorin fluorescence is readily detectable within the above pH range, ecliptic pHlourin as well as the superecliptic variant, which has enhanced fluorescence (7), are detectable at cytosolic pH but not at the lower pH within the vacuole.
In the present study, we sought to use superecliptic pHluorin as a tool for quantitative analysis of endocytosis in yeast. Past studies using pHluorin in mammalian cells have placed the tag on the extracellular face of transmembrane proteins such as synaptobrevin or the transferrin receptor (6,11,12). This approach is not amenable for use in yeast because GFP does not fold correctly in the ER lumen, which can lead to degradation, loss of function or mistargeting (13,14). Instead, we reasoned that placing a cytoplasmic pHluorin tag on a cargo protein that reaches the vacuole via incorporation into multivesicular bodies (MVBs) would result in quenching of the fluorescent tag within the vacuole lumen. As shown in Figure 1, a cytoplasmic fluorescent tag remains exposed to the cytosol while at the plasma membrane, but is packaged into the lumen of internal vesicles within MVBs by the endosomal sorting complex required for transport (ESCRT) machinery (15). Upon fusion with the acidic vacuole, the contents of the MVB are degraded, and although GFP fluorescence remains detectable, pHluorin is expected to lose fluorescence. In this study, we show that chimeric cargo proteins with a cytoplasmic pHluorin tag can be used for quantitative analysis of endocytosis in yeast cells. We also show that the pHluorin tag can be used to study the properties and dynamics of endocytic pathway intermediates by showing that the cytosolic contents of luminal MVB vesicles are acidified prior to MVB fusion with the vacuole.
Figure 1. Transit of cargo bearing a cytoplasmic tag through the endocytic pathway. During endocytosis, transmembrane proteins retain their topology with respect to the cytoplasm and the extracellular space or vesicle lumen. However, membrane-associated proteins destined for degradation in the vacuole are often packaged into internal vesicles within the MVB via the ESCRT machinery. Upon fusion of MVBs with the vacuole, the internal vesicles are exposed to vacuolar hydrolases, and their contents are degraded. Although a cytoplasmic GFP tag (left) can be delivered to the vacuole in such a manner, it retains detectable fluorescence within the vacuole (denoted as a green tag) because of resistance to degradation by vacuolar proteases and changes in pH. In contrast, a cytoplasmic pHluorin tag (right), which is similarly resistant to degradation but is sensitive to changes in pH, loses its fluorescence (denoted as a black tag) upon incorporation into internal MVB vesicles and delivery to the acidic vacuole.
Download figure to PowerPoint
- Top of page
- Materials and Methods
Previous studies have made use of an extracellular pHluorin tag for analyzing endocytic events such as transferrin receptor internalization in mammalian cells (6,11,12). Such approaches require modulation of extracellular pH in order to quench cell surface-localized receptors, leaving only newly internalized receptors detectable. pHluorins have also been used to study synaptic vesicle fusion and recycling, as well as for measurement of pH in intracellular compartments (6). In yeast and in some other organisms, extracellular or luminal GFP tagging can be problematic because of inefficient folding of GFP within the ER lumen or subsequent mistargeting (13,14). In our system, we have instead engineered chimeric Ste3p and Mup1p receptors with superecliptic pHluorin on the cytoplasmic face. As these chimeras become incorporated into MVBs, the pHluorin signal is dampened without a need for rapid changes in pH of the medium, which could cause unanticipated cellular responses.
To date, quantitative analysis of endocytosis in yeast has relied primarily upon biochemical assessment of internalization using radiolabeled markers such as uracil or the α-mating pheromone, or by assessing the stability of endocytic cargo proteins such as Ste3p in cycloheximide-treated cells (23). While these approaches can be used to accurately detect changes in the rate of ligand uptake for mutants with reduced endocytosis, they are not compatible with real-time visualization and analysis of live cells. Our pHluorin-based approach allows us to quantify changes in the steady-state level of the constitutively internalized endocytic cargo protein Ste3p and to monitor rates of internalization during receptor-mediated endocytosis of the methionine permease Mup1 in living cells; we envision its use as an additional tool for higher throughput analyses such as flow cytometry and fluorescence-activated cell sorting (FACS). One distinct advantage of using a receptor, such as Mup1p, instead of the more widely used biochemical α-factor uptake assay is that the α-factor receptor Ste2p is expressed only in haploid MATa cells, while Mup1p is expressed in both haploid mating types as well as in diploid cells.
Our ability to observe prevacuolar Ste3-pHluorin in Vps4E233Q-expressing cells that are defective in MVB biogenesis shows that the cytoplasmic pHluorin-cargo chimeras can be used to study defects at specific trafficking stages prior to vacuole fusion, and our finding that Ste3-pHluorin is quenched in Ist1-labeled late endosome/MVB structures highlights the possible applications of a cytoplasmic pHluorin tag in studying MVB sorting. Notably, an in vitro reconstitution of MVB biogenesis has recently been described (24), thus the cytoplasmic pHluorin tag can provide an in vivo approach that is complementary to the reported biochemical technique.
Overall, we report a novel and versatile approach to quantitative analysis of endocytosis using cargo proteins bearing a cytoplasmic pHluorin tag to overcome the known proteolytic resistance and pH insensitivity of GFP in the vacuole. The ability to use PCR-based approaches for tagging of genes by homologous recombination in yeast will facilitate quantitative analysis of a wide variety of cargo proteins. Although our studies were performed in yeast, the cytoplasmic pHluorin tag can also be employed in other systems. We envision the use of cytoplasmic pHluorin chimeras in diverse applications, including studies of specific cargo trafficking requirements, MVB biogenesis, vacuole/lysosome acidification, mutagenic screens in yeast and RNAi-based screens in mammalian cells.