[2 Imaging Platforms for Measurement of Membrane Trafficking

By Rainer Pepperkok, Jeremy C. Simpson, Jens Rietdorf, Cihan Cetin, Urban Liebel, Stefan Terjung, and Timo Zimmermann


In this chapter we describe automated imaging methods used to measure the transport of an established membrane transport marker from the endoplasmic reticulum to the plasma membrane. The method is fast and significantly robust to be applied in systematic studies on a large scale such as genome-wide screening projects. We further describe the use of software macros and plugins in Image J that allow the quantification of the kinetics of membrane transport intermediates in fluorescence microscopy time-lapse sequences.


Membrane traffic enables cells to distribute proteins, lipids, and carbohydrates between membrane compartments and is thus vital for many cellular processes such as cell growth, homeostasis, and differentiation. Genetic and biochemical approaches have been very efficiently applied in the past to identify and characterize individual molecular components involved in the regulation of membrane traffic in the secretory and endo-cytic pathways. Central to many of these studies has been the reconstitution of the particular transport step of interest in vitro using purified components. Although this has led to an enormous body of information on how membrane traffic is organized at the molecular level, such simplified in vitro systems are lacking important regulatory elements relating to the spatial organization that occurs in living cells. More recently, systematic approaches, such as organelle proteomics or yeast two hybrid screening, have attempted to identify structural and regulatory components of membrane traffic with the goal of reaching a more complete description of its molecular regulation (see for example Bell et al., 2001; Calero et al., 2002; Monier et al., 2002). However, despite their great potential, these techniques have limitations, not least of which is their lack of demonstrating a functional involvement of the molecules identified in the particular trafficking step under investigation.

Functional microscope-based assays in intact living cells with the potential for large-scale analyses have been recently developed and applied to problems in membrane trafficking (Ghosh et al., 2000; Liebel et al., 2003; Pelkmans et al., 2005; Starkuviene et al., 2004). These techniques provide single cell or even subcellular resolution. They promise, in combination with genome-wide RNA interference (RNAi, Elbashir et al., 2001) or overexpression strategies (Starkuviene et al., 2004), to help to reveal comprehensively the regulatory networks underlying membrane traffic in intact cells. Here we describe such assays used to quantitatively monitor secretory transport to the plasma membrane and the subcellular kinetics of transport carriers in time-lapse series.


The problem of performing functional microscope-based assays to address membrane trafficking on a large set of proteins in cells is presently still a challenge. It requires automation and coordination of various steps such as sample preparation, image acquisition, the handling and analysis of large sets of image data, and the integration of the results with existing knowledge provided by bioinformatic databases. A number of automated microscope-based image acquisition systems are already available on the market. Limitations of such commercially available systems are often that they have been designed and optimized for specific applications, which restricts their adaptation to new assays. Systems with ultra-high throughput capacities are mostly lacking single cell or subcellular resolution and thus provide only specialized information. Therefore, it is important when setting up an imaging platform to choose equipment that is flexible enough to be compatible with several protocols for sample preparation, data handling, and analyses. The protocols we describe in this chapter are based on the use of cell arrays for high-content screening microscopy (see Chapter 1 of this issue). However, they will also work for other sample preparation methods with modifications.

Measurement of Transport from the Endoplasmic Reticulum to the Plasma Membrane

As a transport marker we use the temperature-sensitive membrane protein ts-O45-G from the vesicular stomatitis virus (Zilberstein et al. ,198(). This transmembrane protein has the feature of accumulating in the ER at 39.5°, but moves vectorially through the secretory pathway to the plasma membrane (PM) at the permissive temperature of 32°, where an antibody recognizing an external epitope can be used to detect it. This has the principal advantage that transport in individual cells is highly synchronized.

For the expression of this marker in cells we use a recombinant adenovirus encoding ts-O45-G tagged with either CFP or YFP (Keller et al., 2001). For automated image acquisition we use a ScanAR high content screening microscope (Olympus Biosystems, Munich, Germany, see also Liebel et al., 2003 for a detailed description of the features of this microscope). An outline of the rationale of the transport assay is shown in Fig. 1.

Expression and Transport of ts-O45-G to the Plasma Membrane

1. Plate 1.25 x 105 actively growing HeLa cells in 2.5 ml culture medium (DMEM containing 10% heat-inactivated fetal calf serum, 2 mM glutamine, 100 U/ml penicillin, and 100 yg/ml streptomycin) on one Lab-Tek culture dish containing spotted siRNAs or plasmid DNAs as described in Chapter 1 of this issue.

2. Incubate the cells for 24 h (plasmid DNAs) or 36 h (siRNAs) at 37°.

3. Overlay the cells with recombinant adenovirus encoding the secretory marker protein ts-O45-G tagged with CFP or YFP and incubate for further 1 h at 37°.

Transfect and express Infect and hold Release cDNA-CFP Ts-045-G-YFP/39.5° Ts-045-G-YFP/32°

Fig. 1. Rationale of the ts-O45-G transport assay. The assay consists of three steps. (I) cells are transfected with either siRNAs or, as shown in the figure, with plasmid DNAs tagged with a GFP version complementary to the tag of the ts-O45-G. (II) 24 h later cells are infected with adenovirus encoding GFP-tagged ts-O45-G and incubated at 39.5° to accumulate ts-O45-G in the ER. After a further 16 h of incubation at 39.5°, (III) the temperature is shifted to 32° to induce transport of ts-O45-G to the plasma membrane, where it is detected after fixation by immunostaining using a monoclonal antibody recognizing an external ts-O45-G epitope. In this way the relative amount of total ts-O45-G (e.g., YFP-signal) transported to the plasma membrane (a-VG staining using, e.g., a Cy3 conjugated secondary antibody) can be determined by the ratio of Cy3/YFP signals. Transfected cells can be distinguished from nontransfected cells by their CFP signal.

4. Wash cells three times for 2 min with culture medium to remove unbound virus.

5. Transfer cells to 39.5° and incubate at this temperature for 16 h in order to accumulate YFP- or CFP-tagged ts-O45-G in the endoplasmic reticulum.

6. Transfer cells to 32° in the presence of 100 yg/ml cycloheximide (Calbiochem, San Diego, USA) to release ts-O45-G from the endoplasmic reticulum.

7. Fix cells at various time-points after transfer to 32° with 3% paraformaldehyde. Saturation of the arrival of ts-O45-G at the plasma membrane typically occurs 1 h after the temperature shift to 32°.

8. Detect the arrival of ts-O45-G at the plasma membrane by immunofluorescence using a monoclonal antibody recognizing an extracellular epitope of ts-O45-G (obtained as a gift from Kai

Simons, Dresden, Germany; see Pepperkok et al., 2000 for detailed protocols for immunostaining).

9. Stain cell nuclei with Hoechst 33342 stain (final concentration 0.1 yg/ml) for 5 min at room temperature.


The stained samples are stored at 4° either embedded in Mowiol or in PBS solution containing 0.01% azide after a brief post-staining fixation of the samples with 3% paraformaldehyde for 2 min. The initial cell density after plating of the cells on LabTek dishes is critical for the success of the experiments and needs to be adjusted when different cell types are used. If cell densities are too high, the efficiency of adenovirus transfection decreases considerably and the number of cells expressing the transport marker is low. In cell cultures that are too sparse, the number of cells transfected per spot (see Chapter 1) with either siRNA or plasmid DNAs are too low for significant statistical analyses of the results.

Measurement of ts-O45-G Transport to the Plasma Membrane

For automated image acquisition we use a ScanAR system from Olympus Biosystems (Munich, Germany, see also Liebel et al., 2003). This microscope is equipped with standard filter sets for discriminating between DAPI, CFP, GFP, YFP, and Cy3 in sequential imaging mode. Typically we use a 10 x /0.4 air PlanApo objective (Olympus, Melville, NY) to image one entire spot on the LabTek array. For data analysis we use our own software packages (see Liebel et al., 2003) or the analysis software supplied with the ScanAR system by Olympus Biosystems. Both packages give identical results. The individual steps of data analysis described below can also be executed by using the freeware image processing software Imag e J (available at: http://rsb.info.nih.gov/ij/).

1. Images of the samples prepared as described in 1.1 are acquired automatically using the ScanAR system. An example of one multicolor data set acquired is shown in Fig. 2(A-C).

2. All images are background corrected by using a user defined fixed gray value that is subtracted from the images.

3. Subsequently images are low-pass filtered by a 3 x 3 median filter to reduce image noise.

4. Cell nuclei are then identified in each image data set by simple thresholding of the image acquired with the DAPI filter set (D in

Fig. 2). The respective thresholding value is set by the operator at the beginning of an analysis session.

5. Cell nuclei touching each other or the image borders are disregarded from further analysis.

6. The image containing the thresholded nuclei is then used to generate a binary image mask with the nuclei having associated gray value ''1'' and background is set to "0'' (Fig. 2D).

7. The areas of the cell nuclei are then dilated to also extend into the cytoplasm of each cell (Fig. 2E). This image is then used to generate a second mask.

8. This mask is then multiplied with the image containing information of the cell surface staining (Fig. 2B) resulting in image Fig. 2H.

9. A cytoplasmic mask is generated by subtracting image Fig. 2D from image Fig. 2E, resulting in a mask shown in Fig. 2F.

10. This cytoplasmic mask is then multiplied with the image containing information on the total ts-O45-G expressed (Fig. 2C) resulting in the image shown in Fig. 2I.

11. The average intensities of each object in images Fig. 2H and Fig. 2I are then determined and the values corresponding to the same cell are used to determine the ratio R defined as the intensity at the plasma membrane (Fig. 2H) divided by the intensity representing the amount of YFP- or CFP-ts-O45-G expressed in the cell (Fig. 2I). This ratio is proportional to the relative amount of ts-O45-G transported to the plasma membrane.


It is important to test at the beginning of the image acquisition that images are not saturated as this distorts the quantification of fluorescence. Therefore, to adjust the systems parameters appropriately, the operator should take a few test images from different areas on the sample and check the images for saturation. Parameters should then be adjusted such that even the brightest signal can still be acquired at nonsaturating conditions.

Figure 2G shows the results of an example analysis of siRNA trans-fected cells. Should however GFP-tagged plasmids be expressed instead of siRNAs, then tranfected cells can be distinguished from nontransfected cells by multiplying the mask shown in Fig. 2E with the image containing the signal for the transfected cells. The average fluorescence intensity of the objects of the resulting image is then determined and transfected cells are discriminated from nontransfected ones by a threshold defined by the operator.

CO PI sec31p p26 contr.

Analyzing the Kinetics of Transport Carriers Using Kymographs

Kymograph ''wave drawer,'' an instrument to monitor signal changes over time, was first described in 1845. Since then a number of devices and programs to plot events over time have inherited this name. Here we describe a variant we developed to plot dynamic events along an arbitrary line region selection in time-lapse sequences to analyze the intracellular movements of fluorescently labeled structures.

In a time-lapse recording, fluorescently labeled transport intermediates often appear as small dots or tubules of variable size and intensity (see Fig. 3A). Tracks of moving transport carriers can be obtained from these time series in a two-step protocol. First subsequent images are subtracted from each other, to eliminate in the resulting images those structures, which are not moving. Second, images in the resulting sequence are projected onto each other highlighting the trajectories the moving structures in the time-lapse series have taken (e.g., Fig 3B). Monitoring the intensities along each of these trajectories for every time-point in the image sequence results in time-space plots (Fig. 3I). In these, the x-axis represents the intensity profile of the selected track and the y-axis relates to time points of the analyzed series. In these time-space plots a moving fluorescent structure is represented by a slope (see arrows in Fig. 3I) from which the velocity of the moving structure can be determined.

The time-space plotting is a very sensitive, accurate, and fast interactive method to determine velocities of moving carriers from time-lapse recordings. In the following we describe the use of a collection of macros and plugins (available at: www.embl.de/eamnet/downloads) we have developed for the freeware multiplatform image analysis softwar e ImageJ (http://rsb. info.nih.gov/ij/). These macros aim at the automation of the requir ed tasks to determine transport carrier kinetics by kymographs. In addition they offer a variety of functions to improve the image quality in time-lapse sequences.

Fig. 2. Image acquisition and processing steps to determine the transport of ts-O45-G to the plasma membrane. HeLa cells were transfected with siRNAs on LabTek arrays as they are described in Chapter 1 of this issue. The ts-O45-G transport assay was carried out as described in protocol 1.1 as described earlier in this chapter. Images were acquired sequentially using a 10X objective on a ScanAR system using filters to detect specifically DAPI stained nuclei (A), Cy3 stained ts-O45-G at the plasma membrane (B), and CFP-tagged ts-O45-G (C). Images D-I were generated as described in protocol 1.2. earlier in this chapter. ''R'' in (G) is the ratio of ts-O45-G at the plasma membrane (measured in H) to ts-O45-G expressed in cells (measured in I). Results for siRNAs targeting the COPI component ,3-COP, the COPII component Sec31p, and a p24 related membrane protein p26 are shown. The values are the average of two independent experiments (Bar = 50 ^m).

Fig. 3. Generation of a kymograph or time-space-plot to measure the dynamics of transport carrier molecules. (A) First frame of a time-lapse recording of a cell expressing YFP-tagged ts-O45-G at 20 min after temperature shift from 39.5° to 32° (the entire QuickTime movie of the experiment shown is available at: http://www.embl.de/eamnet/ downloads/vsv-tsO45G.mov). (B) Maximum intensity projection of 30 frames covering 6 sec of the time-lapse sequence. The arrow indicates the track of a carrier analyzed. (C-H) Zoom in of images in the region surrounded by the box in A and B. Images are 1 sec apart from each other. (I) Time space plot of the track indicated by arrows in B. The arrows in (I) point to the slope describing the kinetics of the moving particle. Scalebars in B and H:5 ^m, Scalebar in I: horizontal 1 ^m, vertical 1 sec.

Fig. 3. Generation of a kymograph or time-space-plot to measure the dynamics of transport carrier molecules. (A) First frame of a time-lapse recording of a cell expressing YFP-tagged ts-O45-G at 20 min after temperature shift from 39.5° to 32° (the entire QuickTime movie of the experiment shown is available at: http://www.embl.de/eamnet/ downloads/vsv-tsO45G.mov). (B) Maximum intensity projection of 30 frames covering 6 sec of the time-lapse sequence. The arrow indicates the track of a carrier analyzed. (C-H) Zoom in of images in the region surrounded by the box in A and B. Images are 1 sec apart from each other. (I) Time space plot of the track indicated by arrows in B. The arrows in (I) point to the slope describing the kinetics of the moving particle. Scalebars in B and H:5 ^m, Scalebar in I: horizontal 1 ^m, vertical 1 sec.

Installation of the Macros and Plugins in Image J

1. Install the Image J pr ogram (available at: http://rsb.info.nih.gov/ij/).

2. Copy the plugins ''WalkingAverage_.class, StackDifference_.class, MultipleOverlay_.class, MultipleKymograph_.class'' (all available at: www.embl.de/eamnet/downloads) to the Image J plugin folder. Alternatively the collection of macros tsp040421.txt may be loaded into Image J.

Analyzing the Kinetics of Transport Carriers

1. Load the time-lapse image sequence into Image J.

2. If necessary crop the image sequence using the "crop'' tool of Image J ( edit menu).

3. Reduce image noise in all images of the sequence by using the ''walking average'' plugin.

4. Generate tracks of moving objects in the image sequence by running the macro ''track'' or the plugin ''StackDifference.''

5. Select individual tracks with the ''segmented line'' tool ofImage J or use the plugin ''MultipleOverlay'' to select and store more than one track.

6. Execute the macro ''kymograph'' or the plugin ''Multiple Kymograph'' to generate a time-space plot (e.g., Fig. 3I).

7. Mark in the time-space plot the trace of a moving particle by using the ''segmented line'' tool of Image J.

8. Determine the speed of the moving particle at each time-point by using the macro ''read velocities from tsp.''


The speed of image acquisition is a critical parameter for the success of the time-space plot approach and needs to occur at a frequency high enough to result in a continuous trajectory in the projection image (refer to Step 4 in the protocol described earlier). The macro ''read velocities from tsp'' calculates a number of parameters such as, the entire length of the trajectory taken by the structure under view, average, and instantaneous speed. The speed is returned in units of pixels/frame, which needs to be converted to, for instance, ^m/sec using image acquisition calibration parameters (e.g., physical distance represented by one pixel and the time two subsequent images in the sequence are apart from each other).

Summary and Perspectives

We described two methods to quantify membrane transport in single cells using automated fluorescence microscopy and image analysis. The assay to measure the plasma membrane transport of ts-O45-G is robust and has already been used in systematic analyses to identify new proteins involved in the regulation of the secretory pathway (e.g., Starkuviene et al., 2004). Applying this approach to genome-wide siRNA screens in intact cells, for example, may help to reveal the machinery and interaction networks underlying the regulation of membrane traffic in the secretory pathway.


Bell, A. W., Ward, M. A., Blackstock, W. P., Freeman, H. N., Choudhary, J. S., Lewis, A. P., Chotai, D., Fazel, A., Gushue, J. N., Paiement, J., Palcy, S., Chevet, E., Lafreniere-Roula, M., Solari, R., Thomas, D. Y., Rowley, A., and Bergeron, J. J. (2001). Proteomics characterization of abundant Golgi membrane proteins. J. Biol. Chem. 276, 5152-5165.

Calero, M., Winand, N. J., and Collins, R. N. (2002). Identification of the novel proteins Yip4p and Yip5p as Rab GTPase interacting factors. FEBS Lett. 515, 89-98. Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl, T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494-498.

Keller, P., Toomre, D., Diaz, E., White, J., and Simons, K. (2001). Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nat. Cell. Biol. 3, 140-149. Liebel, U., Starkuviene, V., Erfle, H., Simpson, J. C., Poustka, A., Wiemann, S., and Pepperkok, R. (2003). A microscope-based screening platform for large-scale functional protein analysis in intact cells. FEBS Lett. 554, 394-398. Monier, S., Jollivet, F., Janoueix-Lerosey, I., Johannes, L., and Goud, B. (2002). Characterization of novel Rab6-interacting proteins involved in endosome-to-TGN transport. Traffic 3, 289-297. Pelkmans, L., Fava, E., Grabner, H., Hannus, M., Habermann, B., Krausz, E., and Zerial, M. (2005). Genome-wide analysis of human kinases in clathrin- and caveolae/raft-mediated endocytosis. Nature 436, 78-86. Pepperkok, R., Girod, A., Simpson, J. C., and Rietdorf, J. (2000). Imunofluorescence microscopy. In "Monoclonal Antibodies: A Practical Approach,'' pp. 355-370. Oxford University Press, New York. Starkuviene, V., Liebel, U., Simpson, J. C., Erfle, H., Poustka, A., Wiemann, S., and Pepperkok, R. (2004). High-content screening microscopy identifies novel proteins with a putative role in secretory membrane traffic. Genome Res. 14, 1948-1956. Zilberstein, A., Snider, M. D., Porter, M., and Lodish, H. F. (1980). Mutants of vesicular stomatitis virus blocked at different stages in maturation of the viral glycoprotein. Cell 21, 417-427.

0 0

Post a comment