key: cord-0007898-wvuxhd0h authors: Komatsu, Toru; Inoue, Takanari title: A Method to Rapidly Induce Organelle-Specific Molecular Activities and Membrane Tethering date: 2014-04-15 journal: Exocytosis and Endocytosis DOI: 10.1007/978-1-4939-0944-5_16 sha: a08bf03bf38f2e0d9dc122c74330e4114733a3c2 doc_id: 7898 cord_uid: wvuxhd0h In this chapter we describe a technique for rapid protein targeting to individual intracellular organelles. This method enables a real-time imaging-based study of cellular behavior in response to controlled induction of signaling events in a specifically targeted cellular compartment. We provide rationales and a step-by-step protocol for probe design and imaging of protein targeting along with two different applications of this technique. One application involves organelle-specific activation of small GTPases, while the other application involves membrane tethering of two different organelles. In the former case, we activate Rac1 and Ras at distinct intracellular locations in order to study compartmentalization of their signaling pathways, and in the latter example, we induce membrane tethering of the endoplasmic reticulum and mitochondria in order to study organelle–organelle communication. The described technique allows to rapidly perturb molecular activities and organelle–organelle communications at precise locations with specified timing and represents a powerful strategy to dissect spatiotemporally complex biological processes. Many cellular signaling pathways are precisely localized and rapidly induced in response to environmental stimulation. Elucidation of the nature and functions of complex signaling networks requires development of cellular probes that operate in the same time scale as the signaling events themselves, act in well-defi ned cellular domains, and can be activated at defi ned time points [ 1 ] . A major strategy employed for this purpose involves a chemically induced heterodimerization to trigger rapid-onset and specifi c perturbations of various signaling molecules in living cells [ 2 -8 ] . Since most signaling events compose networks operating at multiple intracellular locations [ 9 ] , it is important to specifi cally induce signaling in distinct cellular organelles, such as plasma membrane, the Golgi complex, mitochondria, the endoplasmic reticulum (ER), or lysosomes. We achieved this goal in living cells via inducible dimerization of two engineered proteins triggered by addition of a chemical dimerizer and leading to the controlled organelle translocation of the proteins of interest. Specifi cally, we designed an anchoring unit that is attached to the cytoplasmic face of a specifi c organelle and an effector unit that is present in the cytosol and induces a signaling event upon translocation to the target membrane. Such translocation occurs via dimerization of the effector with the anchor, and it can induce site-specifi c signaling events on the organelle targeted by the anchoring unit ( see Fig. 1 , left). Furthermore, this technique can be used to dimerize two anchoring units (targeted to different organelles) in order to induce tethering of these organelles ( see Fig. 1 , right). One commonly used example of a chemically induced dimerization system involves FK506-binding protein (FKBP) and FKBP-rapamycin-binding (FRB) domain, which can be dimerized by addition of rapamycin or its analogs such as indole rapamycin ( see Note 1 ). Recent developments of this technique include photoactivatable dimerizers [ 10 , 11 ] , attempts to overcome a relative irreversibility of the standard FKBP-rapamycin-FRB system [ 12 ] , and approaches to eliminate possible off-target effects of the method [ 13 ] . In this chapter, we describe the basic organelletargeting technique and some of its cellular applications. Prepare all solutions using analytical grade reagents under sterile conditions. Ensure that all relevant waste disposal regulations are followed when disposing waste materials. 1. Vector DNAs: Plasmids for mammalian expression controlled by the cytomegalovirus (CMV) promoter. Single unit is encoded in a single plasmid ( see Note 2 ). 2. cDNAs of proteins for the anchor unit and the effector unit. 3. Reagents for standard subcloning. 5. Fluorescence microscope ( see Notes 12 and 13 ): In our laboratory, live cell measurements are performed using a spinningdisc confocal microscope. Cyan fl uorescent protein (CFP) and yellow fl uorescent protein (YFP) are excited with a heliumcadmium laser and an argon laser (CVI-Melles Griot, NM, USA), respectively. These two lasers are fi ber-coupled to the spinning disc confocal unit (CSU10; Yokogawa, Japan) equipped with dual-CFP/YFP dichroic mirrors. Appropriate fi lter sets for CFP and YFP are used to enable capture of fl uorescence images with a CCD camera (Orca ER, Hamamatsu Photonics, Japan). Images are taken using a 40× objective lens mounted on an inverted microscope (Axiovert 200, Carl Zeiss, Germany). The microscope is operated by the MetaMorph software package (Molecular Devices, CA, USA). 6. Data analysis software. The dimerization system is composed of an anchoring unit and an effector unit. The anchoring unit is designed to localize on the cytosolic face of a specifi c organelle, and the effector unit is designed to induce specifi c signaling events on the surface of the organelle targeted by the anchor. Plasmids for the anchoring and the effector units should be prepared by using standard subcloning protocols. 1. Construct design for the anchoring unit: The design of the anchoring unit is based on proteins that are known to specifi cally localize in the target organelle. This unit also includes a fl uorescent protein and one of the two dimerizing ; however, it may be important to design novel organelle anchors. Here, we describe how to develop such a novel unit using the Golgi anchoring motif as an example. Since it is impossible to predict the cellular localization of a particular protein based on its structural features, new anchors should be developed using a trial-and-error approach involving structural analysis of known organelle-resident proteins. To obtain motifs that specifi cally target the Golgi complex, we tried four different Golgi proteins and found that constructs derived from giantin work the best (Fig. 4 ). Giantin is a Golgi-specifi c structural protein [ 14 ] that is known to interact with multiple proteins to maintain the Golgi structure [ 15 ] . Since we intended to eliminate unwanted interactions between the anchoring unit and other proteins, we focused on the giantin sequence around its transmembrane domain (TMD). We found that a minimum sequence around the TMD is successfully targeted to the Golgi without disrupting morphology of this organelle. Therefore, this sequence was selected as the Golgi anchoring unit. Next, we selected the dimerization domain and optimized the order of the dimerization domain and the fl uorescent protein in the designed construct. These two parameters are known to affect the expression level of the anchor, the heterodimerization rate, and the structural features of the targeted organelle ( see Note 14 ) . Specifi cally, we swapped FKBP with FRB and placed FKBP or FRB at the N-terminal or the C-terminal ends of either the fl uorescent protein or the anchoring motif. A construct with the following order, FKBP-( fl uorescent protein)-(giantin TMD), appears to be the most suitable for our purposes. The effector unit can be designed to perturb cellular signaling mediated by small GTPases [ 1 , 2 , 16 , 17 ] or phosphoinositides [ 3 , 16 , 18 ] . Here we describe the design of the effector construct aimed to perturb GTPase signaling. GTPases are known to cycle between an active, GTP-bound and inactive, GDP-bound forms. The activation step is controlled by guanine nucleotideexchanging factors (GEF), whereas the inactivation step is accelerated by GTPase-activating proteins (GAP) (Fig. 5 ). Based on this activation-inactivation cycle two different strategies for modulating GTPase signaling have been designed. One strategy is to create a constitutively active GTPase that is not regulated by GEF and GAP, and the other strategy is to use a specifi c GEF for the targeted GTPase. The fi rst approach was used to recruit constitutively active forms of Rac1, Cdc42, and RhoA to the cellular membranes [ 2 ] . The second, more natural approach to activate endogenous GTPases was used to target Rac1 [ 2 , 17 ] , Ras [ 1 ] , and Arf6 [ 16 ] . The design of the effector unit involves removing domains that determine effector compartmentalization to ensure its localization in the cytosol prior to dimerization. Afterwards, dimerization domains and fl uorescent proteins are attached in various orders and pairings, and the activity of the constructs is tested. Prepare the plasmids according to standard subcloning protocols using the design strategy described above. Validate all plasmids by sequencing. 1. To study organelle-specifi c cellular events and membrane tethering, select plasmid pairs containing appropriate anchoring and effector units. To study the morphology and functions of the targeted organelle, prepare required protein biosensors ( see Note 15 ), antibodies for immunolabeling ( see Note 16 ), and small molecular fl uorescent probes, such as calcium sensors and fl uorescent lipids ( see Note 17 ). 5. In the meantime, prepare glass cover slips coated with poly-D -lysine. 6. Trypsinize cells, transfer them into 15 mL tube using 10 mL of the culture medium, and spin them down (1,000 × g for 3 min). 7. Resuspend the cells into 1 mL culture media for transfection (per condition). 8. Add 1 mL of resuspended cells to 100 μL of the transfection mixture, and mix them well by gentle pipetting. 9. Add 50 μL of the cell suspension onto each cover slip. 10 . Leave the cover slips in the incubator for 1 h ( see Note 18 ). 11. Aspirate the media, and add 1 mL of fresh culture media for transfection. 12. Acquire images 24 or 48 h post-transfection. 1. Wash cover slips twice with PBS, and place them into metal frames fi lled with the imaging medium. The medium should not contain chemical dimerizer that should be added only during the fl uorescence imaging. 2. Use confocal microscopy for image acquisition. Usually a single Z-point (preferably around the center of the cells), where the translocation and following signaling events can be clearly visualized, is selected for the imaging. For short imaging times (<30 min), neither incubation at 37 °C nor CO 2 are necessary, but these conditions are desirable for longer imaging in order to preserve cell viability. 3. Under the microscope, select the cell(s) for monitoring in the live imaging mode. These cells should contain all the constructs required for the perturbation and monitoring, which should be confi rmed by checking appropriate fl uorescence signals. Cells that are unusually dim or bright, or that show aberrant localization of the construct, should not be selected ( see Note 19 ). 4. Start the image acquisition in a time-lapse mode. The time frame should be determined depending on the event to be monitored. Usually, translocation is completed within 10 s after addition of chemical dimerizer, so images are generally acquired every 10 or 15 s. After time-lapse imaging has been started, add the chemical dimerizer. It is important to minimize the focus drift during addition of chemical dimerizer. We achieve this by removing half of the medium (500 μL in the case of imaging in 1 mL volume) from the dish by mechanical pipette, mixing it with the chemical dimerizer, and then gently returning it back into the dish ( see Note 20 ). Then, the dimerizer rapidly reaches cells by diffusion, so the translocation should begin immediately. The microscope operating software should be used to monitor the timing of stimuli for subsequent data analysis. 6. Acquire images until the event of interest is completed. 1. Use the image acquisition software or other image analysis software for the data analysis. Evaluate translocation of constructs and biosensors by taking regions of interest (ROIs) and calculating the fl uorescence signal intensity in these regions (Fig. 6 , top and middle). 2. Alternatively, quantify the probe translocation by measuring the decrease of its fl uorescence intensity in the cytosol ( see Fig. 6 , bottom). The latter approach is useful to examine protein translocation to irregular compartments or small or thin organelles, where quantifi cation may be diffi cult due to cellular movement. For example, translocation of Ras-binding protein to the Golgi complex can be examined by monitoring the increase in fl uorescence intensity of the Golgi. On the other hand, translocation to the plasma membrane is better analyzed by monitoring the decrease in the cytosolic fl uorescence, because the plasma membrane is thin, and its fl uorescence measurements suffer from high noise level due to ROI drift. Ras localizes on various membranes of living cells, and its activation has been observed at the plasma membrane and/or the Golgi complex in T cells, depending on the type of stimulus [ 9 ] . To distinguish between different membrane pools of Ras, we designed the organelle-specifi c anchoring units together with the effector unit consisting of a guanine nucleotide exchange factor for Ras (RasGEF). This allowed us to investigate the output of compartmentalized Ras signaling. We detected formation of membrane ruffl es a few minutes after targeting of active Ras to the plasma membrane (Fig. 7 ) . Oppositely, targeting of active Ras to the Golgi complex did not cause changes in plasma membrane dynamics. This data suggests that local activation of Ras at the plasma membrane, but not at the cell interior (Golgi), regulates remodeling of the cortical actin fi laments. The same system has been used to study downstream signaling, such as ERK phosphorylation. To rapidly link mitochondria and ER (thus mimicking mitochondria-associated membranes, MAMs [ 19 ] ) in living cells, we expressed the anchoring units for both ER and mitochondria and then induced their heterodimerization to connect the two membranes. As shown in Fig. 8 , addition of the dimerizer to cells co-expressing two anchoring units induced a rapid transformation of the typical meshwork-like structure of ER to a tubular shape that is more typical of mitochondria. The resulting synthetic MAMs showed an accumulation of fl uorescent dye-labeled phosphatidylserine [ 20 ] at the interface between the two organelles, indicating the formation of the unique membrane structures. Fig. 6 Quantifi cation of the probe translocation to the cellular membranes. Due to limited expression of the constructs containing fl uorescent protein, their translocation from the cytosol to the target compartment results in a decrease of fl uorescence in cytosol. In case of the thin plasma membrane, such translocation can be quantitated by using the average of multiple ROIs to reduce the error or by monitoring the decrease in fl uorescence intensity in the cytosol 1. Dimerization systems that are orthogonal to the rapamycinbased system have been reported, such as gibberellin-based heterodimerization. When using the rapamycin-induced system, one should be aware that rapamycin can inhibit mTOR activity, thereby perturbing downstream signaling events [ 21 ] . However, these downstream effects of mTOR activation require hours to develop; therefore, they will unlikely interfere with cellular events that occur on a time scale of seconds to minutes. There are alternative chemical dimerizers, such as iRap, which does not affect the mTOR signaling [ 4 ] . 2. In order to visually confi rm the transfection and quantify the translocation, each construct should contain a fl uorescent protein, such as CFP, YFP, or mCherry, in addition to the dimerization domain. Since both the anchoring and the effector units are required for perturbation, cells are transfected with a mixture of two constructs that contain the respective dimerization domains and differently colored fl uorescent proteins. If the signaling events need to be monitored with fl uorescent protein-based biosensors, these sensors are also expressed in the cells. In this case, the anchoring unit is designed without a fl uorescent protein to allow tagging this fl uorescence protein with the biosensor. 3. The system can be applied to any cultured cells as long as the constructs can be expressed. For example, the tethering systems appeared to be useful in model neutrophils to study the role of phosphoinositide signaling in chemotaxis [ 15 ] . In this chapter, we focus on commonly used mammalian cell lines, such as HeLa, NIH3T3, HEK 293, and COS7 cells, for all of which the system has been successfully used. 4. The use of the lipofection reagents with antibiotics can cause cytotoxicity due to the internalization of antibiotics. Therefore, transfection should be performed in the antibiotic-free medium. 5. The optimal concentration of trypsin is determined by the tightness of cellular adhesion. Most cell types can be detached with 0.05 % trypsin, but highly adherent cells such as MDCK require 0.25 % trypsin. 6. Alternatively, commercially available glass-bottomed dishes or chambered cover slips pre-coated with poly-D -lysine can be used. A confocal microscope is more suitable than an epifl uorescence microscope to monitor and quantify protein translocation. Any confocal microscope can be used as long as it can detect differently colored fl uorescent proteins independently, and the image acquisition speed is fast enough relatively to the targeted cellular response. 13 . A multi-positioning scanner on the microscope makes it possible to monitor signaling events in multiple cells on the dish simultaneously. Monitoring a larger number of cells is desirable in order to increase the reliability of results and identify both false-positive and false-negative events. 14. These effects can be evaluated by measuring fl uorescence intensities of the constructs, the responses to chemical dimerizers, the fl uorescence images of co-expressed organelle markers, and viability of the transfected cells. Different responses may be due to differences in protein stability and/or the steric environment of the FKBP-FRB interaction. 15. Fluorescent biosensors can be used to visualize signaling events induced by the organelle-specifi c manipulations described above. Several biosensors are available to monitor live cell activity of small GTPases. For example, Ras activation can be examined by a fl uorescent protein attached to the Ras-binding domain (RBD) that specifi cally interacts with active Ras [ 22 ] . Additionally, several FRET-based biosensors for small GTPases have been developed [ 23 ] . 16 . If the appropriate biosensor is not available, the downstream signaling events can be studied by immunofl uorescence labeling with phospho-specifi c antibodies. Since transfection effi ciency can vary from cell to cell, and some cells might show either high or low level of construct expression, the signaling events should be analyzed in the cells with moderate level of expressed reporters. A key question in studying organelle-to-organelle communications is how the interaction causes redistribution of the molecules of interest. Phospholipids are important messengers, and their translocation during organelle-organelle interactions can be studied by using the membrane-tethering system and commercially available fl uorescently conjugated lipids [ 20 ] . For example, we used fatty acid-labeled phosphatidylserine (Avanti Polar Lipids, Inc., AL, USA) to monitor phospholipid translocation at junctional sites of ER and mitochondria upon organelle tethering. 18. The incubation time should be exactly 1 h. 19 . During the search for cells to image, care should be taken to limit light exposure, and avoid photobleaching of the constructs or photodamages of the cells. 20. This also helps to preserve the temperature before and after addition of the chemical dimerizers to avoid the focus drift arising from temperature changes. Organelle-specifi c, rapid induction of molecular activities and membrane tethering An inducible translocation strategy to rapidly activate and inhibit small GTPase signaling pathways Rapid chemically induced changes of PtdIns(4,5) P2 gate KCNQ ion channels Rapid and orthogonal logic gating with a gibberellin-induced dimerization system Membrane recruitment of Rac1 triggers phagocytosis Inducible membrane recruitment of small GTP-binding proteins by rapamycin-based system in living cells Three-part inventions: intracellular signaling and induced proximity Manipulating signaling at will: chemicallyinducible dimerization (CID) techniques resolve problems in cell biology Compartmentalized Ras/MAPK signaling A photocleavable rapamycin conjugate for spatiotemporal control of small GTPase activity Spatio-temporal manipulation of small GTPase activity at subcellular level and timescale of seconds in living cells Rapidly reversible manipulation of molecular activity with dual chemical dimerizers Rapidly relocating molecules between organelles to manipulate small GTPase activity Giantin, a novel conserved Golgi membrane protein containing a cytoplasmic domain of at least 350 kDa The amino-terminal domain of the Golgi protein giantin interacts directly with the vesicletethering protein p115 Triggering actin comets versus membrane ruffl es: distinctive effects of phosphoinositides on actin reorganization Synthetic spatially graded Rac activation drives directed cell polarization and locomotion Synthetic activation of endogenous PI3K and Rac identifi es an AND-gate switch for cell polarization and migration MAM: more than just a housekeeper Transport of exogenous fl uorescent phosphatidylserine analogue to the Golgi apparatus in cultured fi broblasts Isolation of a protein target of the FKBP12-rapamycin complex in mammalian cells Analysis of Ras activation in living cells with GFP-RBD Design and optimization of genetically encoded fl uorescent biosensors: GTPase biosensors We are grateful to Robert DeRose for constructive comments. This work was supported in part by the National Institute of Health (NIH) (GM092930 to TI) and by JST (10216 to TI and 10602 to TK) and JSPS (24655147 to TK). TK is a recipient of a research grant from the Mochida Memorial Foundation for Medical and Pharmaceutical Research.