Transmembrane/cytoplasmic domain-mediated membrane type 1-matrix metalloprotease docking to invadopodia is required for cell invasion

Abstract

The invasion of human malignant melanoma cells into the extracellular matrix (ECM) involves the accumulation of proteases at sites of ECM degradation where activation of matrix metalloproteases (MMP) occurs. Here, we show that when membrane type 1 MMP (MT-MMP) was overexpressed in RPMI7951 human melanoma cells, the cells made contact with the ECM, activated soluble and ECM-bound MMP-2, and degraded and invaded the ECM. Further experiments demonstrated the importance of localization of the MT-MMP to invadopodia. Overexpression of MT-MMP without invadopodial localization caused activation of soluble MMP-2, but did not facilitate ECM degradation or cell invasiveness. Up-regulation of endogenous MT-MMP with concanavalin A caused activation of MMP-2. However, concanavalin A treatment prevented invadopodial localization of MT-MMP and ECM degradation. Neither a truncated MT-MMP mutant lacking transmembrane (TM) and cytoplasmic domains (DeltaTMMT-MMP), nor a chimeric MT-MMP containing the interleukin 2 receptor alpha chain (IL-2R) TM and cytoplasmic domains (DeltaTMMT-MMP/TMIL-2R) were localized to invadopodia or exhibited ECM degradation. Furthermore, a chimera of the TM/cytoplasmic domain of MT-MMP (TMMT-MMP) with tissue inhibitor of MMP 1 (TIMP-1/TMMT-MMP) directed the TIMP-1 molecule to invadopodia. Thus, the MT-MMP TM/cytoplasmic domain mediates the spatial organization of MT-MMP into invadopodia and subsequent degradation of the ECM.

Figure 1

Degradation/invasion of fibronectin-coated crosslinked gelatin substrata by melanoma cells overexpressing MT-MMP. RPMI-7951 cells transfected with MT-MMP were plated onto fluorescein-fibronectin coated gelatin films. After 1 h cells were photographed under epifluorescence (A) or by phase contrast microscopy (B). Open arrows indicate degradation spots (A) that correspond to invadopodial complexes at the leading edge of the transfected cell (B). The solid arrows mark a nontransfected cell in the same field that has not produced any degradation spots. RPMI-7951 cells transfected with pSG5 containing MT-MMP (C and D) or with pSG5 alone (E and F) were plated onto fluorescein-fibronectin coated gelatin films. After 6 h cells were photographed under epifluoresence (C and E) to reveal fluorescein-fibronectin degradation or by phase contrast microscopy (D and F). Cells overexpressing MT-MMP extensively degraded the area of film in contact with the cells (C and D) whereas cells transfected with vector alone did not (E and F). Bars: A and B, 10 μm; C–F, 50 μm; G, RPMI-7951 cells transfected with MT-MMP and parental cells were surface-labeled with biotin. Proteins were immunoprecipitated from cell lysates using mAb 113–5B7 against MT-MMP then detected using anti-biotin horse radish peroxidase.

Figure 2

Activation of MMP-2 and degradation/invasion of fibronectin-coated crosslinked gelatin substrata by melanoma cells treated with Con A or overexpressing MT-MMP. (A) Activity of MMP-2 released into the medium (Medium) or associated with cells (Cell lysate) in the presence (Con A) or absence (Parent) of Con A or transfected with pSG5-MT-MMP (MT-MMP) were analyzed by gelatin zymography. The three bands shown represent 72-kDa pro-MMP-2, the 62-kDa intermediate form, and the 59-kDa active form of MMP-2. (B) Activity of MMP-2 released into the medium (Conditioned medium) by the cells or associated with cells transfected with pSG5-MT-MMP (Cell lysate) in the absence (MT-MMP) or presence of the following protease inhibitors: 100 nM CT1847 (InMMP), 200 nM CT1847 (InMMP′), 320 μg/ml ɛ-amino caproic acid (ACA), 5 μg/ml pepstatin A (PSA), and 10 μg/ml leupeptin (LP). Gelatinase activities were analyzed by gelatin zymography. (C) Fibronectin degradation by cells transfected with pSG5 plasmid (Mock), treated with Con A, or transfected with pSG5-MT-MMP (MT-MMP). Cells were allowed to spread on fluorescein-fibronectin-coated crosslinked gelatin films and were photographed with fluorescence microscopy after 1, 3, 6, and 12 h. Wide black areas in the lower panels indicate the degradation of fibronectin-gelatin films by cells. Mock transfectants and the cells treated with Con A showed very little black area on the film, indicating background levels of fibronectin degradation by the melanoma cells, whereas cells transfected with MT-MMP show much higher levels of degradation. Bar = 25 μm. (D) The areas of degraded fibronectin substratum under the cells transfected with pSG5 plasmid (Mock) or transfected with pSG5-MT-MMP (MT-MMP) in the absence or presence of Con A were measured with Optimas image analyzer after 6 h. Each value represents the mean of three separate determinations (among 100 cells) ± SD. Duplicate experiments gave similar results. (E) The invasiveness of the cells transfected with pSG5 plasmid (Mock) or transfected with pSG5-MT-MMP (MT-MMP) in the absence or presence of Con A after 12 h was analyzed in terms of percentage of cells forming invadopodial extensions in the crosslinked gelatin film. Each value represents the mean of three separate determinations (among 100 cells) ± SD. (F) The areas of degraded fibronectin substratum under the cells in the presence of protease inhibitors as indicated above were measured with Optimas image analyzer. Each value represents the mean of three separate determinations (among 100 cells) ± SD.

Figure 3

The TM/cytoplasmic domain is required to localize MT-MMP at invadopodial complexes. Arrows in individual panels indicate sites of invadopodial complexes. (A and B) RPMI-7951 cells transfected with MT-MMP were plated on a fluorescein-fibronectin crosslinked gelatin film. Cells were stained for MT-MMP using mAb 113–5B7 (A) and areas of fibronectin degradation were visualized by epifluorescence (B). (C and D) RPMI-7951 cells transfected with MT-MMP were cultured on fibronectin-gelatin films for 3 h, and double immunostained with mAb113–5B7 against MT-MMP (C) and antibody A27 against invadopodia (D). Double labeling of MT-MMP transfected cells shows that MT-MMP (C) colocalizes with invadopodia as labeled by mAb A27 (D). (E and F) RPMI-7951 cells transfected with MT-MMP were cultured on fibronectin-gelatin films in the presence of Con A (25 μg/ml) for 3 h, and double immunostained with mAb113–5B7 against MT-MMP (E) and antibody A27 against invadopodia (F). MT-MMP loses its invadopodial localization in the presence of Con A. (G) RPMI-7951 cells transfected with TIMP-1/TM_MT-MMP were cultured on fibronectin-gelatin films for 3 h, and immunostained with antibody 7–6C1 against TIMP-1. A cell expressing TIMP-1/TM_MT-MMP localized this chimera at invadopodial complexes as revealed by labeling with an anti-TIMP-1 antibody (G). (H) RPMI-7951 cells transfected with chimeric ΔTM_MT-MMP/TM_IL-2R were cultured on fibronectin-gelatin films for 3 h, and immunostained with mAb113–5B7 against MT-MMP. A cell expressing ΔTM_MT-MMP/TM_IL-2R failed to localize MT-MMP to invadopodia (H) as shown by staining with mAb 113–5B7.

Figure 4

The TM/cytoplasmic and catalytic domains of MT-MMP are required for activation of ECM-bound MMP-2, local ECM degradation and subsequent invasion by melanoma cells. Gelatinase activities of MMP-2 released into the medium by the cells transfected with pSG5 lacking an insert (Mock), pSG5 plasmids containing MT-MMP (MT-MMP), truncated mutant lacking the TM domain (ΔTM_MT-MMP), chimeric ΔTM_MT-MMP/TM_IL-2R, TIMP-1, chimeric TIMP-1/TM_IL-2R, or chimeric TIMP-1/TM_MT-MMP, were analyzed by gelatin zymography. (B) Gelatinase activities of MMP-2 associated with the ECM and invadopodial membranes of cells transfected with chimeric MT-MMP molecules. Cells cultured on crosslinked gelatin films were fractionated as described (24) and ECM/invadopodial fractions were examined by gelatin zymography. (C) The fibronectin-degrading activities of transfectants were measured with Optimas image analyzer after 6 h. Each value represents the mean of three separate determinations (among 100 cells) ± SD. (D) The invasiveness of transfectants after 12 h was analyzed in terms of the percentage of cells forming invadopodial extensions in the crosslinked gelatin film. Each value represents the mean of three separate determinations (among 100 cells) ± SD.