We investigate the global principles underlying secretory membrane trafficking, sorting, and compartmentalization within eukaryotic cells. We are using live cell imaging of green fluorescent protein (GFP) fusion proteins in combination with photobleaching and photoactivation techniques to investigate the subcellular localization, mobility, transport routes, and binding interactions of a variety of proteins with important roles in the organization and regulation of membrane traffic and compartmentalization. To test mechanistic hypotheses related to protein and organelle dynamics, we use quantitative measurements of protein characteristics in kinetic modeling and simulation experiments. Among the topics currently under study are growth and maintenance of endoplasmic reticulum (ER) and Golgi morphology in mammalian cells and in developing Drosophila embryos; the mechanism(s) of secretory protein transport into and out of the Golgi apparatus; membrane binding/dissociation kinetics of trafficking machinery and its regulation; the generation and maintenance of cell polarity; and organelle breakdown and reassembly during mitosis. We have also recently developed a photoactivatable GFP whose mechanism of photoactivation is currently under investigation.

Development of a Photoactivatable GFP
Patterson
Photoactivation is the rapid conversion of photoactivatable molecules to a fluorescent state by intense irradiation; it is a useful method for marking and monitoring selected molecules within cells. Previous efforts to develop a photoactivatable protein capable of high optical contrast when photoactivated under physio-logical conditions have had limited success. By mutagenizing wild-type GFP (wtGFP), we have developed a variant of GFP that allows selective marking of proteins through photoactivation. The wtGFP normally exists as a mixed population of neutral phenols and anionic phenolates producing major 397 nm and minor 475 nm absorbance peaks, respectively. Upon intense illumination of the protein with ultraviolet or ~400 nm light, the chromophore population undergoes photoconversion and shifts predominantly to the anionic form, giving rise to an increase in minor peak absorbance. The result is an approximately three-fold increase in fluorescence upon excitation at 488 nm. The photoactivatable variant of wt GFP that we developed (called PA-GFP for “photo” and “activatable”) contains a histidine substitution at the T203 position of GFP, which results in a negligible minor absorbance peak. Photocon-version with irradiation of about 400 nm produces a large increase in absorbance at the minor peak and thus a more noticeable optical contrast under 488 nm excitation.

Upon photoactivation of living cells, PA-GFP exhibited an optical enhancement of nearly two orders of magnitude, making it suitable for marking specific protein or cell populations. Furthermore, the speed with which an optical signal was obtained and the absence of signal from newly synthesized proteins indicated that PA-GFP photoactivation was a labeling method preferable to photobleaching for studying the temporal and spatial dynamics of proteins. We used the photoactivatable GFP as both a free protein to measure protein diffusion across the nuclear envelope and as a chimera with a lysosomal membrane protein to demonstrate rapid interlysosomal membrane exchange. Our results suggest that the photoactivatable variant of GFP, PA-GFP, has the potential for addressing many fundamental questions in cell and developmental biology.

Dissection of COPI Dynamics in Vivo
Presley, Liu
Cytosolic coat proteins that bind reversibly to membranes have a central role in membrane transport within the secretory pathway. One well-studied example is COPI or coatomer, a heptameric protein complex that is recruited to membranes by the GTP-binding protein Arf1. Assembly into an electron-dense coat then helps in budding off membrane to be transported between the ER and Golgi. To study COPI dynamics in vivo, we fused variants of GFP to the carboxy terminus of the eCOPI subunit of coatomer for expression in ldlF cells, which contain a mutated eCOP that is degraded at 40°C, causing coatomer inactivity and growth inhibition. The cells grew indefinitely at 40°C, indicating that eCOP-GFP could substitute functionally for endogenous eCOPI in coatomer complexes in these cells.

We studied the identity and behavior of coatomer-containing membranes, assessed by eCOP-YFP labeling, in dual-color time-lapse experiments in ldlF cells co-expressing cyan fluorescent protein (CFP)-tagged secretory cargo markers. eCOP-YFP was present on juxtanuclear Golgi membranes and on pre-Golgi transport intermediates containing secretory cargo, and it was depleted from retrograde (Golgi-to-ER) transport intermediates. The pre-Golgi intermediates remained brightly labeled with eCOP-YFP as they tracked into the Golgi.

To gain insight into the role of COPI on anterograde (ER-to-Golgi) transport inter-mediates, we used photobleaching techniques to investigate the characteristics of COPI binding and release from membranes. Upon photo-bleaching either Golgi or pre-Golgi structures expressing eCOP-GFP in ldlF cells, fluorescence recovered onto these structures exponentially with a half-time of 35 seconds, thus indicating that coatomer binding and dissociation from membranes occur continuously. To test whether the activity is coupled to vesicle budding, we measured the kinetics of coat exchange on and off membranes, at low temperatures, in which vesicle transport is nonexistent. We found that at all temperatures down to 4°C, eCOP-GFP underwent binding and release from Golgi membranes, with no abrupt change in the kinetics on reaching temperatures at which vesicle budding is slowed or inhibited. The data thus indicated that the cycling of coatomer on and off membranes can be uncoupled from vesicle formation and that feedback from productive vesicle budding is not necessary for COPI dissociation. Based on these findings, we proposed that membrane binding and release of COPI serves to initiate and stabilize lateral sorting and segregation of cargo into membrane domains that progressively differentiate into pleiomorphic membrane transport intermediates and/or vesicles for membrane transport in the ER/Golgi system.

Arf1 Regulation of COPI Dynamics
Liu, Presley, Pfeffer
COPI is known to be recruited to membranes by the GTP-binding protein Arf1 while GTP hydrolysis of GTP-bound Arf1 is believed to be necessary for COPI release from membranes. To gain further insight into Arf1 regulation of COPI dynamics, we asked if Arf1 GTP hydrolysis is sufficient for COPI to dissociate from membranes; that is, whether Arf1 and COPI dissociate together from membranes. To address this question, we used a functional Arf1-CFP chimera that allowed us to visualize the behavior of Arf1 and coatomer in the same cell. FRAP experiments revealed that the half-time for Golgi membrane dissociation of Arf1-CFP (13s) was significantly faster than for eCOP-YFP (30s), suggesting that Arf1 and coatomer dissociate independently from Golgi membranes. These results, in turn, were supported by results from experiments using BFA, which inhibits Arf1 and COPI recruitment to membranes, allowing the dissociation of these proteins to be observed in the absence of rebinding. On treatment with BFA, Arf1-CFP dissociated significantly faster (T1/2 13s) than eCOP-YFP (t1/2 30s). The findings indicated therefore that Arf1 and coatomer dissociation from Golgi membranes are regulated differently. A mathematical formulation of COPI and Arf1 membrane binding and dissociation kinetics was able to fit simultaneously both Arf1 and COPI photobleaching data and the Arf1 and COPI release kinetics following BFA treatment, providing new quantitative estimates of the lifetime of these molecules on membranes and their binding and dissociation rates.

Constitutive Cycling of Golgi Proteins between ER and Golgi
Ward
A major controversy in the field of organelle biology has centered on which of two mechanisms, template versus self-organization, best describes the growth and inheritance of the Golgi apparatus. The template model views the Golgi as an autonomous organelle with stable components that provide a template or scaffold for its growth and division. The de novo model, by contrast, envisions the Golgi as a dynamic, steady-state structure whose maintenance and biogenesis depend on dynamic membrane input and outflow pathways. A key prediction of the template model is that stably associated proteins on Golgi membranes serve as a scaffold or template for Golgi growth and remodeling. To test such a hypothesis, we used a photobleaching assay to determine whether any Golgi proteins were permanently associated with the Golgi. We tested GFP-tagged members of four different Golgi protein classes (i.e., enzyme, itinerant, matrix, and coat). Our results revealed that none was stably associated with Golgi membranes; after selective photobleaching, fluorescence recovered into the Golgi area in all cases. For GalT-GFP and p58-GFP, which represent the glycosylation enzyme and itinerant classes of Golgi proteins, respectively, recovery occurred on a time scale of minutes and depended on microtubule-driven membrane delivery from the ER. For GRASP65-GFP and eCOP-GFP, which represent the matrix and coat classes of Golgi proteins, repectively, recovery occurred within 20 seconds of Golgi photobleaching and did not depend on the presence of microtubules. Thus, members of all Golgi protein classes localize dynamically to Golgi membranes through either association/dissociation with the cytoplasm or membrane cycling pathways through the ER.

Building the Golgi de novo
Ward
To understand what underlies the ability of the Golgi to maintain and organize itself in the absence of long-lived components, we examined how Golgi structure becomes disorganized under different treatments. The phenotype of Sar1[H79G]-expressing cells (in which Sar1 is locked in its GTP-bound state) was found to resemble that of BFA-treated or Arf1[T31N]-expressing cells (in which Arf1 is locked in its GDP-bound state). Under all three conditions, COPI dissociated from membranes, COPII components continued to exchange on/off ER export sites, and Golgi structure was lost as Golgi enzymes redistributed to the ER and matrix/itinerant proteins relocated to ER export domains. No stable Golgi remnant was found to persist under any of these conditions. Instead, structures containing Golgi matrix and itinerant proteins present in these cells represented ER export domains active for the recruitment of some (including Golgi matrix and itinerant proteins) proteins that populate the secretory pathway.

To test if the complete disruption of ER exit machinery would result in the absorption of all Golgi components into the ER, we expressed the constitutively inactive Sar1[T39N] mutant that is blocked in the GDP-bound state. In these cells, no ER exit sites were visible by EM, and ER export machinery, including Sar1 and Sec13-YFP, were diffusely localized to the ER, which appeared swollen due to accumulated secretory cargo. Golgi matrix proteins, including GRASP65 and GM130, and the recycling protein p58-GFP then redistributed into the ER, where Golgi enzymes were localized. Given that there are no non-ER structures with which Golgi components associate under these conditions, the data demonstrated that there is no stable scaffold or higher order architecture of the Golgi apparatus.

Taken together, our findings suggest that the Golgi apparatus exists as a steady-state system in interphase cells, with rapid and substantial recycling of Golgi proteins through the ER. At all times, a significant fraction of Golgi-resident proteins resides in the ER, available for export into pre-Golgi elements that subsequently become part of the Golgi complex. When such activities are disrupted, the Golgi structure breaks down and disappears within cells. Maintenance of the Golgi complex thus depends on continual membrane export from the ER, export that is controlled by the sequential activities of the Sar1/COPII and Arf1/COPI coat systems.

Arf1 Regulation of Mitotic Golgi Disassembly
Altan
Previous work from our laboratory has shown that during mitosis the Golgi apparatus disperses and reforms through the intermediary of the ER, exploiting constitutive ER/Golgi recycling pathways. Over the past year, we have focused on understanding how and why this occurs. An important clue for addressing these questions came with our finding that treatment of cells with the drug H89, known to prevent Golgi disassembly caused by BFA-induced Arf1 inactivation, also blocks mitotic Golgi disassembly. This finding led us to focus on Arf1, the target of BFA, as a regulator of mitotic Golgi disassembly. Live cell imaging of a single cell expressing Arf1-CFP showed that Arf1 dissociated from Golgi membranes and became cytoplasmic in prophase before Golgi fragmentation. Addition of H89 completely blocked Arf1 release from Golgi membranes. We observed a similar phenotype in cells expressing Arf1[Q71L], which cannot undergo GTP hydrolysis and release from Golgi membranes. Our observation suggested that both mitotic and BFA-induced Golgi breakdown depend on Arf1 inactivation. We used FRAP both to analyze the mechanism for Arf1 inactivation in mitotic cells and to measure the binding/dissociation properties of Arf1 in interphase and prophase cells. Our results suggested that Arf1 inactivation and depletion from Golgi membranes in BFA-treated and mitotic cells occurs through a common mechanism: the inhibition of Arf1-GEF. The results imply that Arf1 regulates Golgi structure in both interphase and mitosis.