The late endosomal p14-MP1 (LAMTOR2/3) complex regulates focal adhesion dynamics during cell migration

Abstract

Cell migration is mediated by the dynamic remodeling of focal adhesions (FAs). Recently, an important role of endosomal signaling in regulation of cell migration was recognized. Here, we show an essential function for late endosomes carrying the p14-MP1 (LAMTOR2/3) complex in FA dynamics. p14-MP1-positive endosomes move to the cell periphery along microtubules (MTs) in a kinesin1- and Arl8b-dependent manner. There they specifically target FAs to regulate FA turnover, which is required for cell migration. Using genetically modified fibroblasts from p14-deficient mice and Arl8b-depleted cells, we demonstrate that MT plus end-directed traffic of p14-MP1-positive endosomes triggered IQGAP1 disassociation from FAs. The release of IQGAP was required for FA dynamics. Taken together, our results suggest that late endosomes contribute to the regulation of cell migration by transporting the p14-MP1 scaffold complex to the vicinity of FAs.

Figure 1.

Loss of p14–MP1 endosomal signaling impairs cell migration and FA. (A) Migration of MEFs in wound-healing assay. Red lines depict edges of the wound. See also Video 1. (B) Calculation of migration speed of MEFs in wound-healing assay (µm/h, mean of cell migration speeds ± SD; **, P < 0.001, Student’s t test). (C) IF: anti-Paxillin antibody. Note formation of elongated peripheral FAs in p14^−/− MEFs (red arrows) compared with FAs and FCs at the leading edge of the control p14^f/− MEFs (red arrowheads). Expression of myc6-MP1 or Xpress-p14-CAAX does not restore FA pattern in p14^−/− MEFs (presence of myc6-MP1 is confirmed by anti-myc antibody showing cytoplasmic distribution of MP1, presence of p14-CAAX is confirmed by anti-Xpress antibody). Reduction of FA size can be achieved by expression of p14-GFP (blue arrowheads, presence and localization of p14-GFP is confirmed by GFP image). Bar, 10 µm. (D) Quantification of average FA length and detailed analyzes of FA populations (“Length distribution”) in MEFs. Left graph: mean in percent ± SEM compared with control p14^f/− MEFs (mean of FA length in control p14^f/− MEFs was taken as 100%). Right graph: FAs are clustered in four specified length groups (x-axis) and are presented in percentage of all adhesions (y-axis) in each type of MEFs. See also Table S1. (E) FACS analysis of activated β1 integrins (identified with 9EG7 antibody) reveals no difference in expression in p14^f/− MEFs (black line) and p14^−/− MEFs (blue line; dotted lines indicate negative controls). The data shown are from a single representative experiment out of three independent experiments. For complete integrin profile, see Fig. S2.

Figure 2.

Analysis of FA turnover. (A) IF of p14^f/− and p14^−/− MEFs with anti-Paxillin and anti-Zyxin antibodies. White arrows indicate leading edge formation areas in the control cells where no Zyxin-positive FAs are detected (magnification). In the p14 knockout MEFs the overall elongated FAs are also positive for Zyxin (magnification). (B) FRAP analysis of FAs in indicated MEFs. Image shows representative example of FAs (white arrows) before bleaching and during recovery time in p14^f/− MEFs. Graph shows curves of fluorescence recovery (mean of 10 cells per group, number of measured FAs = 60) after bleaching of indicated MEFs. (C) TIRF microscopy of MEFs expressing mRuby-Paxillin. See also Videos 2 and 3. Shown are mature FAs of magnified areas of control and p14 knockout MEFs at three different timeframes (1 frame = 10 min). Merging of the three color-coded frames indicates the development of single FAs over time. Note: white FAs are stable over the shown time period (20 min). (D) Quantification of FA turnover. The left graph displays the existence of single FAs (n = 168) as dots over a time period of 60 min. All existing FAs were included. The right graph shows dynamic FA events. New growing FAs were analyzed during a time period of 80 min (P < 0.001).

Figure 3.

p14–MP1-carrying late endosomes target FAs. (A) IF of p14–MP1-carrying endosomes colocalizing with FAs in cells coexpressing GFP-MP1 (green) and mCherry-Paxillin (red; see white arrowheads). (B) NIH3T3 cell coexpressing GFP-MP1 (green) and mCherry-Paxillin (red). White arrows: FAs repetitively targeted by p14–MP1-carrying endosomes. See Video 4. (C) TIRF microscopy. Overlay of MP1-GFP Z-projection (green) and FAs (mCherry-Paxillin is shown in red). White arrowheads on selected time-lapse images point to individual MP1 endosomes (green) that target FAs (red). See Video 5. (D) Colocalization of MP1 with Rab7 and Transferrin. Top: images from time-lapse series of HeLa cell coexpressing GFP-MP1 (green) and mCherry-Rab7 (red). GFP-MP1 colocalizes with mCherry-Rab7 in cell center and cell periphery. Deconvolved image from the time-lapse shows colocalization of MP1 (green) and Rab7 (red) on ring-like structures of endosomes. Bottom: uptake of Transferrin (red) in HeLa cell expressing GFP-MP1 (green). (E) Quantification of the colocalization (mean ± SEM) between Rab7 and MP1 versus Transferrin and MP1. Insets in A and D are taken from white boxes in main panels.

Figure 4.

MP1 endosomes target mature FAs, but neither FCs nor newly formed FAs. (A) HeLa cells expressing GFP-MP1 (green) and mCherry-Paxillin (red) are used for detailed quantification of FA targeting. Top: images from the video sequence showing overlay of FAs (red) and endosomal (green) masks. Bottom: center of each FA was identified as a round sphere (purple) and used for further FA tracking. FA tracks (black arrows) during time lapse are illustrated in a line of different colors, indicating temporal changes of FA positions (as shown on the color bar below). (B) Quantification of the FA targeting by p14–MP1-carrying endosomes (mean in percentage ± SEM) and of FA targeting (per frame) during the time lapse. Green arrows indicate parameter combination of highly dynamic FAs, blue arrows of mature stable FAs, and red arrows of sliding FAs. Experiments were repeated three times, n = 10 (n = the number of cells used for quantification). (C) Schematic segmentation of FAs dissected with respect to the change of its fluorescence signal over time (left image). Right graph depictures mean calculation of fluorescence MP1–Paxillin colocalization in the FA segments. Experiments were repeated three times, n = 10 (n = the number of cells used for quantification).

Figure 5.

Arl8b-dependent MT plus end–directed transport of late endosomes regulates FAs. (A) MP1 endosomes are transported along MTs. Time-lapse images of HeLa cells expressing GFP-MP1 (green), mCherry-Paxillin (red), and mCherry-tubulin (red) show colocalization of MP1 and MTs (white arrowheads). Representative individual endosome moves along MTs toward two FAs (bottom panels, white arrowheads). See also Video 7 and 8. (B) Nocodazole treatment of a cell transfected as in A results in MT depolymerization and “trapping” of few GFP-MP1 endosomes in FAs (white arrows and arrowheads). Time-lapse images of the same cell show that positions of GFP-MP1 endosomes do not change in time due to abolished MP1 transport. (C) Arl8b knockdown in MEFs. IF: anti-LAMP1 (green), anti-tubulin (red) antibodies, and Hoechst. LAMP1-positive late endosomes collapse to the perinuclear region upon Arl8b knockdown (white arrow). WB: anti-Arl8b antibody, anti-tubulin used as loading control. (D) The p14^−/−;p14-GFP MEFs treated with control and Arl8b RNAi. The late p14-GFP endosomes cluster in the Arl8b RNAi-treated cells (red arrow). See also Video 9. (E) The graph on the left shows the quantification of average FA length in MEFs. Mean in percent ± SEM compared with control p14^f/− MEFs treated with control RNAi (mean of FA length in control p14^f/− MEFs treated with control RNAi was taken as 100%). See also Table S1. The graph on the right shows the migration speed of p14^−/−;p14-GFP MEFs transfected with control (n = 26) and Arl8b siRNA (n = 66) in wound-healing assay (µm/h, mean of cell migration speeds ± SD). (F) Colocalization of Paxillin and Rab7 in MEFs. Images from time-lapse series of MEF cells coexpressing GFP-Rab7 (green) and mCherry-Paxillin (red). White arrows indicate FAs targeted by GFP-Rab7. See also Video 10.

Figure 6.

IQGAP1 and MP1 interaction. (A) Reciprocal co-immunoprecipitation of myc6-MP1 and IQGAP1. Cell extracts were prepared from myc6-MP1–expressing p14^f/− MEFs. IP: with anti-myc (lane 2) and anti-IQGAP1 (lane 3) polyclonal antibodies; IgG control (lane 1). WB: anti-IQGAP1 (top) and anti-myc (bottom) antibodies, respectively. Black arrow indicates myc6-MP1 signal. (B) Immunoprecipitation of endogenous MP1 and p18. Cell extracts were prepared from HeLa cells. IP: with anti-IQGAP1 (lane 3) polyclonal antibody; cell lysates (lane 1), IgG control (lane 2). WB: anti-IQGAP1 (top), anti-p18 (middle), and anti-MP1 (bottom) antibodies, respectively. (C) Top: HeLa protein lysates were incubated with GST-MP1 or GST alone immobilized on glutathione-Sepharose beads. WB: anti-IQGAP1 antibody. Lysates not subjected to GST pull-down were processed in parallel (HeLa lysate). Bottom: HeLa protein lysates were incubated with GST-IQGAP1 or GST alone immobilized on glutathione-Sepharose beads. WB: anti-MP1 antibody. Lysates not subjected to GST pull-down were processed in parallel (HeLa lysate). Black arrow indicates MP1 signal.

Figure 7.

Absence of p14–MP1 or blockage of Arl8b-dependent late endosomal transport causes IQGAP1 accumulation in FAs. (A) IF: anti-Paxillin and anti-IQGAP1 antibodies. IQGAP1 localizes to the leading edge in p14^f/− (white arrows) and colocalizes with Paxillin in FAs in p14^−/− MEFs (red arrows). (B) IF: anti-Paxillin and anti-IQGAP1 antibodies. Shown are different time points during spreading of p14^f/− and p14^−/− MEFs. White arrows indicate accumulation of IQGAP1 at the leading edge in control MEFs. Red arrows point at IQGAP1 localization at Paxillin-positive FAs in control and p14^−/− MEFs. (C) IF: anti-LAMP1 (green), anti-IQGAP1 (red) antibodies, and Hoechst (blue). IQGAP1 localizes to the leading edge in p14^f/− MEFs treated with control RNAi (white arrows) and localizes to FAs upon Arl8b depletion (red arrows). (D) p14^f/− MEFs treated as in C. IF: anti-Paxillin and anti-IQGAP1 antibodies. Note IQGAP1 localization to the leading edge in p14^f/− MEFs treated with control RNAi (white arrows) versus colocalization of IQGAP1 and Paxillin in FAs upon ARl8b depletion (red arrows).

Figure 8.

In the absence of p14–MP1 endosomal signaling, down-regulation of IQGAP1 by RNAi rescues FAs and cell migration. (A) IF: anti-Paxillin and anti-IQGAP1 antibodies. (B) Left: quantification of average FA length in MEFs transfected with control or IQGAP1 RNAi. Mean in percent ± SEM is compared with control p14^f/− MEFs transfected with control RNAi (taken as 100%). Right: calculation of speed of migration in wound-healing assays. Mean in µm/h ± SEM. Importantly, only cells that had strong down-regulation of IQGAP1 (no signal on IF using anti-IQGAP1 antibodies as in A) were used for the calculation. See also Table S1.