First published online October 22, 2003
doi: 10.1242/10.1242/jcs.00748
Migrating fibroblasts perform polarized, microtubule-dependent exocytosis towards the leading edge
Jan Schmoranzer1,2,
Geri Kreitzer3 and
Sanford M. Simon1,*
1 Laboratory of Cellular Biophysics, The Rockefeller University, 1230 York Avenue, New York, NY 10021, USA
2 Department of Biology, Chemistry, Pharmacology, Free University Berlin, 14195 Berlin, Germany
3 Department of Cell and Developmental Biology, Weill Medical College of Cornell University, New York, NY 10021, USA
View larger version (81K):
[in a new window]
|
Fig. 1. Fusion of a post-Golgi vesicle near the leading edge imaged by TIR-FM. NRK fibroblasts were microinjected with cDNA encoding LDLR-GFP and the newly synthesized proteins allowed to accumulate in the Golgi at 20°C. The cells were then imaged by epifluorescence (A) and TIR-FM (B) 10 minutes after the shift to 32°C. Scale bar, 10 µm. Using epifluorescence, the accumulated LDLR-GFP brightens the area of the Golgi complex. Using TIR-FM on the same cell, single vesicles can be seen that have arrived at the contact surface. (C) An enlargement of a vesicle moving toward the leading edge and (D) fusing with the plasma membrane. Scale bars, 2 µm (C), 1 µm (D). (E) Plots of the total intensity and the width of the fluorescence intensity of the fusion event in (D). Time is indicated relative to moment of fusion start at 0.00 seconds.
|
|
View larger version (105K):
[in a new window]
|
Fig. 2. Map of fusion sites of post-Golgi vesicles. NRK fibroblasts were microinjected with cDNA encoding LDLR-GFP and imaged using TIR-FM with a temporal resolution of  5 frames per second between 10 minutes and 60 minutes after the release of the Golgi block. Cells were outlined (yellow line) by thresholding the epifluorescence image against the background. Epifluorescence images indicate the LDLR-GFP load in the Golgi before the release of the Golgi block (grey level). The sites of fusion are indicated by either yellow crosses or cyan stars, depending on whether the fusions were complete or partial, respectively. For the migrating cells (C,D), the direction of the leading edge is indicated by white arrows. Stationary (A,B) and migrating (C,D) cells are either untreated (A,C) or treated with 10 µM nocodazole (B,D) during the last hour of the 3-hour Golgi block. Scale bars, 10 µm.
|
|
View larger version (151K):
[in a new window]
|
Fig. 3. Microinjected cells move during TIR-FM recording of exocytic events. (A,B) Monolayers of NRK fibroblasts were wounded. Cells at the edge of the wound were microinjected with cDNA encoding LDLR-GFP, incubated for 1 hour at 37°C and then shifted to 20°C for 2 hours prior to image acquisition. Wound-edge cells expressing LDLR-GFP were identified at low magnification by epifluorescence (black outlines). The cells were then imaged under transmitted light at the same magnification to localize the wound edge at the beginning of the experiment (green outline). At the end of the experiment, 90 minutes later, the same field was imaged again at low magnification (red outline). Cells at the edge of the wound advanced significantly (10-20 µm) during that time. All injected cells advanced with the wound edge. (C) High magnification contrast images were taken of the injected cell marked by the asterisk in (B). The outlines of the cell and the nuclei were traced just before the first TIR-FM recording (0 min, green) and after the last TIR-FM recording (30 min, red). (D) The cell shown in (C) was imaged by TIR-FM in eight intervals (each 500 frames, 5 frames per second). The complete (yellow crosses) and partial (cyan stars) fusion events were mapped. The fusion map was overlayed on top of an epifluorescence image taken at the beginning of the TIR-FM recording. Notice that the Golgi complex is oriented towards the direction of migration (white arrows). Scale bars, 50 µm (A,B), 10 µm (C,D).
|
|
View larger version (23K):
[in a new window]
|
Fig. 4. Distances of fusion sites of vesicles from the nearest Golgi element in untreated or nocodazole-treated stationary and migrating fibroblasts. The distances from each site of fusion (mapped in Fig. 2) to the closest Golgi element (before the release of the Golgi block) were measured for stationary (A) and migrating (B) cells. The data from untreated and nocodazole-treated cells are coded in black and grey, respectively. To normalize for cell size, the distance values are displayed as the percentage of the maximal distance between the centre of the Golgi and the furthest edge of the cell. The counts are normalized to the total counts (notice the axis break).
|
|
View larger version (19K):
[in a new window]
|
Fig. 5. Distribution of the fusion density and time in stationary and migrating NRK fibroblasts. For each cell, the long axis was defined as the furthest distance from edge to edge through the centre. The total region of the cell was then divided into three orthogonal parts with equal width along the long axis (A). To compare migrating and stationary cells, we chose region 1 to be at the end of each cell that was closest to the nucleus. The number of fusion sites were counted in each region and normalized by the total number of fusions per cell. The fusion density was determined by dividing the normalized number of fusions per region by the normalized area of each region. To account for the higher fusion rate in migrating cells (22 fusions per minute) compared with stationary cells (about eight fusions per minute), the values for the migrating cells were multiplied by 2.75. These fusion-density/time values of each cell [five stationary cells (176 fusions) and four migrating cells (375 fusions)] were averaged and plotted with the SEM as error bar (B).
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2003
