LPP is a Src substrate required for invadopodia formation and efficient breast cancer lung metastasis - PubMed (original) (raw)

LPP is a Src substrate required for invadopodia formation and efficient breast cancer lung metastasis

Elaine Ngan et al. Nat Commun. 2017.

Abstract

We have previously shown that lipoma preferred partner (LPP) mediates TGFβ-induced breast cancer cell migration and invasion. Herein, we demonstrate that diminished LPP expression reduces circulating tumour cell numbers, impairs cancer cell extravasation and diminishes lung metastasis. LPP localizes to invadopodia, along with Tks5/actin, at sites of matrix degradation and at the tips of extravasating breast cancer cells as revealed by intravital imaging of the chick chorioallantoic membrane (CAM). Invadopodia formation, breast cancer cell extravasation and metastasis require an intact LPP LIM domain and the ability of LPP to interact with α-actinin. Finally, we show that Src-mediated LPP phosphorylation at specific tyrosine residues (Y245/301/302) is critical for invadopodia formation, breast cancer cell invasion and metastasis. Together, these data define a previously unknown function for LPP in the formation of invadopodia and reveal a requirement for LPP in mediating the metastatic ability of breast cancer cells.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing financial interests.

Figures

Figure 1

Figure 1. LPP is dispensable for mammary tumour growth but is required for the efficient formation of lung metastases.

(a,b) NMuMG-ErbB2 and NIC breast cancer cells harbouring shRNAs targeting LPP (LPP-shRNA) or LucA control (LucA-shRNA) were injected into the mammary fat pads of mice (_n_=10 per cohort). Primary tumour growth was monitored by weekly caliper measurement and tumours were resected after 28 days. Mice were euthanized 3 (NMuMG-ErbB2) or 4 weeks (NIC) post primary mammary tumour resection and lung tissue was collected at necropsy. (c,d) The number of macroscopic lesions on the lung surfaces were quantified at necropsy from cohorts of mice injected with NMuMG-ErbB2 LucA-shRNA (_n_=8) and NMuMG-ErbB2 LPP-shRNA, (_n_=10) (c) or NIC LucA-shRNA (_n_=9) and NIC LPP-shRNA (_n_=10) breast cancer cells (d). The area of metastatic burden was quantified from H&E stained lung sections from the same samples, and is expressed as a percentage of total lung surface area (c,d). *P<0.006. Representative images are shown (c,d). Scale bar, 3 mm and applies to all images in c and d. Error bars represent s.e.m. for all panels.

Figure 2

Figure 2. Loss of LPP diminishes breast cancer-derived circulating tumour cells.

Whole blood was collected by cardiac puncture from mice-bearing NMuMG-ErbB2 or NIC mammary tumours, harbouring LucA- and LPP- shRNAs, to isolate CTCs. The number of CTC-derived adherent colonies was determined 2 weeks post isolation and were quantified from (a) NMuMG-ErbB2 LucA-shRNA (_n_=9) and NMuMG-ErbB2 LPP-shRNA, (_n_=10) or (b) NIC LucA-shRNA (_n_=10) and NIC LPP-shRNA (_n_=10) cultures. Representative images of formalin-fixed and crystal violet stained samples are shown. Scale bar, 1 cm, and applies to all images. Error bars represent s.e.m.

Figure 3

Figure 3. LPP promotes breast cancer cell-mediated gelatin degradation.

(a) HCC1954 human breast cancer cells were transfected with either scrambled (control) or LPP targeting siRNAs. Cells were pre-treated with TGFβ for 24 h and then plated onto fluorescently labelled gelatin for an additional 24 h. The area of degradation was determined from images taken from three independent experiments (control _n_=34, LPP siRNA _n_=29), and error bars represent s.e.m. Immunoblot analysis was performed to assess the level of LPP knockdown. α-Tubulin was used as a loading control. Scale bar, 20 μm, and applies to both images. *_P_=1.11 × 10−6. (b) NMuMG-ErbB2 cells harbouring shRNA against LucA (LucA-shRNA) or targeting LPP (LPP-shRNA) were used to express an EV (VC) or LPP rescue constructs (eGFP-LPP-WT, eGFP-LPP-mLIM1 and eGFP-LPP-ΔABD). The indicated breast cancer cells were pre-treated with TGFβ for 24 h, and subsequently plated onto fluorescently labelled gelatin for an additional 24 h. The area of degradation was quantified from images of LucA control (_n_=26), VC (_n_=25), WT (_n_=27), mLIM1 (_n_=26) and ΔABD (_n_=24) cells from four independent experiments and error bars represent s.e.m. *P<0.05. Scale bar, 20 μm, and applies to all images.

Figure 4

Figure 4. LPP co-localizes with Tks5 and actin at sites of ECM degradation.

HCC1954 breast cancer cells pre-treated with TGFβ (24 h) were plated onto Alexa 405-conjugated gelatin. Cells were fixed after 24 h and stained with antibodies against LPP, Tks5 and actin (phalloidin). (a) Confocal images were acquired to visualize gelatin degradation, and the localization of actin, Tks5 and LPP. Scale bar, 20 μm, and applies to all images. (b) Linescan analysis was performed over areas of degraded gelatin (indicated by black arrow) and the signal intensity of actin, Tks5, LPP and gelatin are shown for five regions of interest. Scale bar, 10 μm. (c) _Z_-stack acquisition was performed over 5.20 μm depth at 0.20 μm intervals. Orthogonal views (y_–_z plane: red box; x_–_z plane: black box) are presented. Black arrows indicate areas of gelatin degradation where LPP, Tks5 and actin are co-localized. Scale bar, 10 μm.

Figure 5

Figure 5. Wild-type LPP localizes to invadopodia and promotes ex ovo breast cancer cell extravasation.

NMuMG-ErbB2 LPP knockdown (LPP-shRNA) breast cancer cells harbouring a Tks5-mCherry fusion protein and expressing eGFP-LPP-WT (a), eGFP-LPP-mLIM1 (b) and eGFP-LPP-ΔABD (c) were intravenously injected into the CAM and monitored using high-resolution time-lapse intravital imaging in an ex ovo chick embryo model. White arrows point to regions where cancer cells contact the vascular wall. Chick endothelium is labelled with A647-lectin (grey signal). Scale bar, 20μm, and applies to all images. (d) The percentage of breast cancer cells that have extravasated was quantified. The data represents quantification from ≥500 cells per cell line in _n_≥3 animals and error bars represent s.e.m. ***P<0.0001.

Figure 6

Figure 6. NMuMG-ErbB2 breast cancer cells expressing LPP mutants fail to efficiently establish breast cancer lung metastases.

NMuMG-ErbB2 breast cancer cells harbouring either a dox-inducible control shRNA against LucA or LPP-shRNA in which LPP rescue constructs were expressed (VC, LPP-WT, LPP-mLIM1, LPP-ΔABD), were employed in an experimental metastasis assay. The indicated breast cancer cells were injected into the tail vein of athymic mice (_n_=10 per cohort). The breast cancer cells and the cohorts of mice were pre-treated with doxycycline starting 1 week before tumour cell injection and oral doxycycline was maintained throughout the experiment. (a) Immunoblot analysis was performed before breast cancer cell injection to assess ErbB2 and LPP expression levels. α-Tubulin was used as a loading control. (b) The number of macroscopic lesions on the lung surfaces was quantified at necropsy (*_P_=0.007, **_P_=0.002). (c) The number of lung metastases in each cohort were quantified from four H&E step sections (*_P_=0.038, **_P_=0.038, ***_P_=0.016). The metastatic burden was determined from the same set of H&E samples and expressed as a percentage of the total lung tissue area (*_P_=0.005, **_P_=7.7 × 10−4, ***_P_=8.2 × 10−4). Quantification was performed from 4-step sections, taken at 100 μm between each step from the following samples: LucA control: _n_=9; VC: _n_=10; LPP-WT: _n_=9; LPP-mLIM1: _n_=9; LPP-ΔABD: _n_=10. (d) Representative H&E images are shown. Scale bar, 3 mm, and applies to all images in d. Error bars represent s.e.m. for all panels.

Figure 7

Figure 7. TGFβ-induced LPP phosphorylation is Src dependent.

(a) NMuMG-ErbB2, NIC, HCC1954 and BT549 breast cancer cell lines were stimulated with or without TGFβ, and incubated with inhibitors against Src Family Kinases (Dasatinib and PP2) or vehicle control (dimethylsulphoxide). Total cell lysates were immunoprecipitated with antibodies against LPP and immunoblotted for phosphotyrosine (pY) or LPP. (b) NMuMG-ErbB2 cells were infected with two independent Src-shRNAs, along with EV controls. Following 24 h of stimulation with or without TGFβ, LPP immunoprecipitates were immunoblotted for phosphotyrosine (pY) and LPP. (c) NMuMG-ErbB2 cells were transfected with a construct encoding a constitutively activated Src kinase and treated with or without TGFβ (24 h). Tyrosine phosphorylation (pY) and LPP levels were assessed by immunoblot from LPP immunoprecipitates. (d) Total cell lysates from three independent NIC; Src+/+ and three independent NIC; Srcfl/fl breast cancer cell lines were immunoprecipitated for LPP following treatment with or without TGFβ (48 h). LPP tyrosine phosphorylation (pY) and total LPP levels were assessed by immunoblot.

Figure 8

Figure 8. Tyrosine phosphorylation of LPP is required for TGFβ-induced breast cancer cell invasion and invadopodia formation but is dispensable for TGFβ-induced cell migration.

(a) Schematic diagram of eGFP-tagged LPP constructs. LPP is composed of a PRR (amino acids 1–413) and three LIM domains (amino acids 414–613). An ABD within LPP is located between residues 41–57 (indicated by black box). The positions of five tyrosine (Y) residues that were mutated to phenylalanine (F) residues within eGFP-LPP-5F are indicated. (b) Immunoblot analyses of LPP levels in a panel of NMuMG-ErbB2 cells treated with or without TGFβ (LucA-shRNA: expressing non-targeting shRNA against LucA, LPP-shRNA: expressing shRNA against the 3′-UTR of LPP, in which eGFP-LPP-WT or eGFP-LPP-5F were expressed). α-Tubulin was used as a loading control. (c,d) NMuMG-ErbB2 cell populations, treated with or without TGFβ, were subjected to migration and invasion assays. The data is expressed as the average pixel count obtained from four independent migration and three independent invasion experiments performed in duplicate (*P<0.004) and error bars represent s.e.m. (e) NMuMG-ErbB2 cells harbouring either a shRNA against LucA (LucA-shRNA) or targeting LPP (LPP-shRNA) were used to express eGFP-LPP-WT and eGFP-LPP-5F rescue constructs. The indicated panel of NMuMG-ErbB2 cells were pre-treated with TGFβ (24 h), and subsequently plated onto fluorescently labelled gelatin for an additional 24 h. The area of degradation was quantified from images of LucA-shRNA control (_n_=26), LPP-shRNA (_n_=11), eGFP-LPP-WT (_n_=14) and eGFP-LPP-5F (_n_=15) cells from four independent experiments, and error bars represent s.e.m. Scale bar, 20 μm, and applies to all images in e. (*P<0.01).

Figure 9

Figure 9. LPP phosphotyrosine sites at residues 245/301/302 are required for TGFβ-induced cell invasion.

(a) Schematic diagram of eGFP-tagged LPP constructs indicating the location of the tyrosine (Y) residues that were converted to phenylalanine (F) residues to create a panel of LPP phospho-mutants. (b) Immunoblot analyses of LPP levels in a panel of NMuMG-ErbB2 cells (control: LucA-shRNA; LPP knockdown: LPP-shRNA; LPP rescues: cells expressing LPP-shRNA, in which eGFP-LPP-WT or individual eGFP-LPP phospho-mutants were expressed). α-Tubulin was used as a loading control. (c) NMuMG-ErbB2 cell populations, treated with or without TGFβ, were subjected to migration and invasion assays. The data are expressed as the average pixel count obtained from three independent experiments performed in duplicate, and error bars represent s.e.m. (*P<0.0001).

Figure 10

Figure 10. Phosphorylation on LPP tyrosine residues 245/301/302 is required for efficient breast cancer lung metastasis.

NMuMG-ErbB2 breast cancer cells expressing LucA-shRNA and LPP-shRNA with LPP rescue constructs (LPP-WT, LPP-245F, LPP-C and LPP-D) were injected into the mammary fat pads of athymic mice (_n_=10 per cohort). (a) Mammary tumour growth was determined by weekly caliper measurements. (b) At the time of euthanasia, tumours were collected and weighed to determine the final tumour burden. (c) Whole blood was collected by cardiac puncture and the number of CTC-derived adherent colonies was assessed 2 weeks post isolation (***_P_=0.003, **_P_=0.02, *_P_=0.05). (d) Lungs were collected at necropsy and the number of macroscopic surface lesions was quantified (***_P_=0.006, **_P_=0.007, *_P_≤0.04). The area of metastatic burden was quantified from four H&E stained lung sections per animal, and is expressed as a percentage of total lung surface area (*_P_=0.03, **_P_=0.04, ns; _P_=0.07) (e) Representative H&E lung sections are shown. Scale bar, 2 mm and applies to all images in e. Error bars represent s.e.m. for all panels.

References

    1. Buccione R., Caldieri G. & Ayala I. Invadopodia: specialized tumor cell structures for the focal degradation of the extracellular matrix. Cancer Metastasis Rev. 28, 137–149 (2009). - PubMed
    1. Mizutani K., Miki H., He H., Maruta H. & Takenawa T. Essential role of neural Wiskott-Aldrich syndrome protein in podosome formation and degradation of extracellular matrix in src-transformed fibroblasts. Cancer Res. 62, 669–674 (2002). - PubMed
    1. Yamaguchi H. et al.. Molecular mechanisms of invadopodium formation: the role of the N-WASP-Arp2/3 complex pathway and cofilin. J. Cell Biol. 168, 441–452 (2005). - PMC - PubMed
    1. Clark E. S., Whigham A. S., Yarbrough W. G. & Weaver A. M. Cortactin is an essential regulator of matrix metalloproteinase secretion and extracellular matrix degradation in invadopodia. Cancer Res. 67, 4227–4235 (2007). - PubMed
    1. Bowden E. T. et al.. Co-localization of cortactin and phosphotyrosine identifies active invadopodia in human breast cancer cells. Exp. Cell Res. 312, 1240–1253 (2006). - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources