Anti-LRP-1 α-chain (8G1), anti-LRP-1 β-chain (5A6), anti-human IgGs used as negative control for immunoprecipitation experiments (HP6030), specific inhibitors for calpain (Calpeptin) and PKA (KT5720), and fluorigenic Calpain Substrate II were obtained from Calbiochem (distributed by VWR International, Strasbourg, France). Anti-calpain-2 (H240) used for immunoprecipitation experiments, anti-β1 (M-106), anti-calpastatin (C-20) and anti-β-actin (sc-1616) antibodies were purchased from Santa-Cruz Biotechnology (distributed by Tebu-Bio, Le Perray en Yvelines, France). Anti-phosphothreonine (9381) and HRP-conjugated anti-rabbit antibodies were from Cell Signaling (distributed by Ozyme, Saint Quentin Yvelines, France). Anti-calpain-2 (ab39165) and anti-talin (8D4) were from Abcam (Cambridge, UK). Anti-talin antibody used for immunocytochemistry (MAB1676) and anti-calpain-1 (MAB3104) were from Chemicon (distributed by Millipore, Molsheim, France). HRP-conjugated anti-mouse antibody (NA931V) was from Amersham Biosciences (Buckinghamshire, UK). Anti-pan β1 integrin (clone Mab13) and anti-active β1 integrin (9EG7) antibodies were from BD Transduction Laboratories (Mississauga, Canada). 7-amino-4-chloromethylcoumarin, t-BOC-L-leucyl-L-methionine amide (CMAC, t-BOC-Leu-Met), anti-mouse AlexaFluor 488 (A11001), AlexaFluor 568-phalloidin (A12380) and Prolong Gold antifade reagent with DAPI (P36935) were from Molecular Probes (Cergy Pontoise, France). Anti-phosphoserine (PSR-45), HRP-conjugated anti-goat secondary antibodies, leupeptin and other chemicals were purchased from Sigma-Aldrich (Steinheim, Germany).

The FTC133 human follicular thyroid carcinoma cell line was grown in Dulbecco’s Modified Eagle Medium-Ham’s F12 (Invitrogen, Cergy Pontoise, France) with 10% FBS, as previously described (44). SiRNA mediated silencing of LRP-1 was described elsewhere (9). Specific LRP-1 targeting sequences were designed by Dharmacon (distributed by Perbio Science, Brebiere, France) as follows: GACUUGCAGCCCCAAGCAGUU (sense), CUGCUUGGGGUGCAAGUCUU (antisense). Control transfection experiments were achieved by using SiGENOME Non-targeting siRNA #1 (D00121001-20) from Dharmacon. For transient transfection assays, siRNA were transfected for four hours by using Lipofectamine 2000 (Invitrogen) according to manufacturer instructions.

Total RNAs isolated using the RNeasy Mini kit (Qiagen, Courtaboeuf, France) were reverse transcribed to cDNA with the ABsolute Blue verso 2-step kit and subjected to quantitative real-time PCR by using the Absolute SYBR Green mix (Thermo Fisher Scientific, Epsom, Surrey, UK) and a Chromo4 Real-Time PCR Detector system from Bio-Rad. Primers for LRP-1 were previously described (9). Primers for calpain 1, calpain 2 and β-actin were designed by Eurogentec (Seraing, Belgium) as follows: calpain 1: forward, CCTTGAGGATGATCTGGTAGA; reverse, AGCTAGTGTTCGTGCACTCTG; calpain 2: forward: CTGGAGATCTGTAACCTGACC; reverse: GGTACTGAGGGTTCATCCAGA; β-actin: forward: GTGTGACGTGGACATCCGC; reverse: CTGCATCCTGTCGGCAATG. The relative levels of expression were quantified by using Opticon Monitor software (Bio-Rad). For each specific gene, the amount of target RNA (2^-ΔΔCt) was normalized to the internal actin reference (ΔCt) and related to the amount of target RNA in control sample, set to 1. All experiments were performed at least in triplicate with internal duplicate for each sample.

Whole-cell extracts were prepared by scrapping cells in ice-cold lysis buffer (25 mM Tris-HCl, pH 7.7, 150 mM NaCl, 0.5% glycerol, 1% Triton X-114, 5 mM EDTA, 0.5 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM orthovanadate supplemented with proteinase inhibitor cocktail). After 30 min incubation on ice, extracts were centrifugated (10,000 X g for 10 min at 4°C) and pellets were discarded. The protein concentration in supernatants was quantified by the Bradford method (BioRad Laboratories). Proteins were separated by electrophoresis in a 10% sodium dodecyl sulfate-polyacrylamide gel, transferred onto a nitrocellulose membrane (Whatman GmbH, Dassel, Germany), and incubated overnight at 4°C with primary antibodies. Immunoreactive bands were revealed using an ECL plus chemiluminescence kit from Amersham Biosciences by using a ChemiDoc-XRS imaging station from Bio-Rad. Immunoblots presented are representative of at least three separate experiments. The specific signal of β-actin was used to ensure equal loading.

Separation of cell surface proteins was conducted by immunoprecipitation of biotinylated proteins. Briefly, after culture on gelatin-coated surfaces, cells were rinsed with ice cold PBS and incubated for 30 min at 4°C under slow agitation with 0,5 mg/mL EZ-Link Sulfo-NHS-LC-Biotin (Pierce). All further steps until cell lysis were made at 4°C to avoid cell surface protein endocytosis. After 3 PBS washes, reaction was quenched by 100 mM Glycine for 30 min under gentle agitation and cells were scrapped in lysis buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0,1% Triton X-100, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1mM orthovanadate supplemented with proteinase inhibitor cocktail). 500 µg of clarified cell extracts were then incubated for 1 hour with 40 µL of avidin beads (Pierce) under agitation. After three washes in lysis buffer, immunoprecipitated proteins were reduced and analyzed by immunoblotting. Supernatants were used to control selective fractionation of cell surface proteins. When indicated, cell surface specific integrin-containing complexes were subjected to a second immunoprecipitation step after competing biotin-avidin complexes overnight with 10 mM D-Biotin (Pierce). Procedure for immunoprecipitation is described in the following section.

The procedure was adapted from one previously described (9). Briefly, for the immunoprecipitation of cell surface integrin containing complexes, Biotinylated protein fraction was subjected to a second immunoprecipitation step in the same lysis buffer, as described above. Cell extracts for immunoprecipitation of calpain were prepared in a distinct lysis buffer (10 mM Tris-HCl, pH 7.4, 0.75% Brij, 75 mM NaCl, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM orthovanadate supplemented with proteinase inhibitor cocktail). After centrifugation (10,000 X g, 10 min at 4°C) and Bradford titration, 500 µg of total proteins were incubated for 4 hours at 4°C on an orbital agitator in presence of anti-calpain antibodies. Nonspecific IgGs were used for negative controls. Immunoprecipitation was performed with 40 µL protein G-Sepharose (Amersham Biosciences) for 2 hours at 4°C under agitation. The samples were washed three times in the corresponding lysis buffer and protein complexes bound to beads were solubilized under reducing conditions and analyzed by immunoblotting as described above.

PKA activity was assessed by using ProFluor PKA assay from Promega with minor modifications to quantify specific PKA activity in raw cell extracts. Briefly, cells were detached and cultivated on gelatin-coated for 3 hours. Cells were then scrapped 48 hours post transfection on ice in PKA buffer (25 mM Tris-HCl, pH 7.4, 0,5 mM EDTA, 0,5 mM EGTA), sonicated, clarified (10,000 X g for 10 min at 4°C) and protein concentration was determined by Bradford method. Equal amounts of proteins were then assessed for PKA activity following standard procedure, based on phosphorylation-dependent quenching of a fluorescent PKA substrate. Results were obtained from three separate experiments with internal quadruplicate, and related to control sample, set to 1. Specificity of reaction was ensured by subtracting residual PKA activity in the presence of 10 µM of specific inhibitors H-89 and KT5720.

Cells were scrapped on ice in lysis buffer (50 mM Hepes, pH 7.4, 150 mM NaCl, 1% Triton X-100, 0,5 mM DTT, 10 mM β-mercapto-ethanol, 20 µM leupeptin, 1 mM PMSF, 1 mM Na₃VO₄) and clarified (10,000 X g for 10 min at 4°C). In vitro calpain activity was then quantified through two different procedures. First, calpain activity was measured with Innozyme Calpain 1/2 Activity Assay Kit (Calbiochem) following manufacturer instructions. A second procedure was adapted from Goette et al. (45). Briefly, 100 µg of proteins were diluted in activation buffer (20 mM Tris-HCl pH 7.4, 5 mM CaCl₂) containing 2 µM calpain substrate II (Suc-Leu-Tyr-AMC). Background fluorescence was determined by negative control carried out for each sample in the presence of 10 mM EDTA, 10 mM EGTA and 50 µM of the specific calpain-1/2 inhibitor calpeptin. Reaction was developed in 96-wells plate at 37°C for 1 hour and fluorescence resulting from calpain-mediated release of AMC was measured at 380 nm (excitation)/430 nm (emission) wavelengths using a Perkin-Elmer LS50B spectrometer. Results were expressed as percentage of control after background subtraction. Each experiment was made at least four times with internal quadruplicate. Calpain activity was visualized in intact cells by using a previously described method (46). Cells were cultivated on gelatin-coated coverslips and incubated for 10 min in the presence of 50 µM BOC-LM-CMAC (t butoxycarbonyl-Leu-Met-chromomethylaminocoumarin), a cell permeant thiol-reactive substrate selective for calpains. Cells were then replaced in fresh medium for 1 hour, fixed on ice in 4% paraformaldehyde and mounted in aqueous antifading medium (Immunotech). Fluorescence resulting from Calpain-mediated CMAC release was observed under UV light with an optical microscope (BH2; Olympus). Specificity of the reaction was controlled in the presence of 50 µM calpeptin. Quantification of calpain activity was achieved by counting 300 cells from 3 distinct experiments. The number of positive cells exhibiting a clear fluorescence signal was normalized to a total cell number for each treatment and expressed as a percentage of control condition.

Matrigel invasion assay was performed using modified Boyden chambers in 24-well dishes with filter inserts provided with 8-µm pores (Transwell, Costar, Brumath, France), as described elsewhere (44). Matrigel-coated filters were used for invasion assays and uncoated filters were used for migration assays. Invasiveness was determined by counting cells in eight random microscopic fields per well, each seeded in triplicate.

Measurement of cancer cell adhesion to gelatin-coated surfaces and trypsinization assays were carried out according to a previously described method (9).

FTC133 cells were seeded onto gelatin-coated coverslips for 2 hours at 37°C, fixed with 4% paraformaldehyde for 5 min on ice and then permeabilized in ice-cold 0,1% Triton X-100 for 10 min. After three washes with PBS, cells were incubated for 30 min with PBS containing 2% bovine serum albumin to saturate nonspecific antigens. Coverslips were then incubated for 30 min with AlexaFluor 568-phalloidin or 60 min with anti-talin antibodies followed by three washes with PBS (Phosphate Buffered Saline). Talin staining was revealed after 30 min incubation with secondary antibodies conjugated to AlexaFluor 488. Coverslips were then mounted in Prolong Gold antifade medium with DAPI to obtain nuclei counterstaining. All acquisitions were made with an Olympus BH2-RFCA epifluorescence microscope (Olympus, Japan), equipped with a 100W mercury lamp (OSRAM, GmbH), using a SPlan achromat x 40 objective (Olympus) and fluorescein, rhodamin, DAPI filters for talin, actin and nuclei staining, respectively. Antibodies used for immunostainings of integrins were incubated overnight and revealed after 1 hour incubation with Alexafluor 568-coupled secondary antibodies. Images were acquired with Zen software on a Zeiss LSM 880 Airyscan confocal microscope (Oberkochen, Germany) using a x63 Plan Apochromat objective (oil immersion, 1.40 NA). Representative images from three separate sets of cultures were treated and merged with ImageJ software. The percentage of cells positive for focal adhesions was determined as previously reported (9), by examining 300 cells from three different experiments for each condition.

Bands from immunoblots or agarose gels were quantified by using Quantity One image-analyzer software (Bio-Rad Laboratories). All culture assays were normalized on the basis of cell viability by using the CellTiter-Glo assay from Promega. Each experiment was performed at least in triplicate and data were expressed as mean ± SEM. Comparisons between two or four samples were performed using Student’s t-test or one way analysis of variance (47) with post hoc Tukey’s test (Prism software, GraphPad Inc, San Diego, CA).

We previously reported that LRP-1 is involved in the regulation of cell-matrix interactions turnover by exerting an intimate control on the focal adhesion composition and dynamics (9, 28, 40). We demonstrated that LRP-1 functions as a molecular relay and provides an intracellular docking platform to regulate the incorporation and activation of MAPK into focal complexes. This work established a direct link between intracellular signaling pathways modulated by LRP-1 and dynamics of matrix attachment of migrating tumor cells. However, a complete overview of the intracellular connections of the receptor in that context is still lacking. We further investigated the still incomplete knowledge of intracellular functions of the receptor involved in the control of adhesion complexes stability. To achieve this goal, we used a validated method of invalidation of LRP-1 in FTC-133 thyroid carcinoma cells by siRNA (36) leading to a 70% decrease of its gene expression as evaluated by RT-qPCR ( Figure 1A ), and to a 66% decrease of its protein expression measured by western blot ( Figure 1B ). Calpains constitute a family of intracellular calcium-dependent cysteine-proteases which exert crucial actions on focal adhesion dynamics, and are able to cleave cytoskeleton components to mediate turnover and disassembly of focal complexes as well as the activation and clustering of extracellular matrix receptors (41, 48, 49). We tested the consequences of LRP-1 invalidation on calpain activity in FTC133 cells during the phase of attachment by performing a biochemical assay based on calpain substrate II cleavage ( Figure 1C ). We evidenced an increase of 34,5% of the capacity of LRP-1-silenced cells to degrade calpain substrate II in vitro, as compared to control tumor cells. A distinct fluorigenic assay was conducted to confirm this result using the quenching of DABCYL substrate. By using this complementary approach, we observed a 2.2 fold increase of calpain activity in LRP-1-depleted cells, as compared to control ( Figure 1D ). Of note, we were able to efficiently inhibit calpain activity by adding 50µM of their specific inhibitor calpeptin. A 26% inhibition was detected in control cells whereas the amplitude of this inhibition in LRP-1-silenced cells was 60%, reinforcing the idea that the baseline activity of this class of proteases is significantly higher in the absence of the receptor. Finally, we incubated thyroid carcinoma cells with the cell permeant substrate of calpains BOC-LM-CMAC in order to visualize their activities in living cells ( Figure 1E ) (46). We evaluated the proportion of cells exhibiting high calpain-dependent release of fluorescent CMAC probe ( Figure 1F ). In LRP-1 expressing control cells, 18,6% of the total cells were fluorescent when 43% of LRP-1 depleted cells had developed the calpain-dependent reaction. In conclusion, we showed by three independent biochemical assessments that LRP-1 silencing leads to a significant induction of intracellular calpain activity in thyroid carcinoma cells.

We reported that LRP-1 functionalities in invasive cancer cells are not restricted to the endocytic control of extracellular matrix degrading enzymes (9). We showed that this endocytic receptor can directly feed pro-motile signals by coordinating the stability and distribution of cell surface matrix receptors through its control of downstream intracellular signaling events (28, 36, 39). In order to measure the consequences of the LRP-1-mediated control of calpain activity, we performed Boyden chamber assays with FTC133 cells LRP-1-silenced or not, in the presence or absence of the specific calpains inhibitor calpeptin ( Figure 2A ). As expected, we observed that LRP-1 silencing reduced carcinoma cell migration by 29% as compared to migrated control cells. Interestingly, addition of 50µM calpeptin completely restored the migration of LRP-1-silenced cells to the level of control cells. The same treatment did not exert any significant effect on LRP-1 expressing cells, suggesting that calpain activity is low. We conducted complementary modified Boyden chamber assays to analyze the invasive behavior of thyroid carcinoma cells using Matrigel as matrix barrier ( Figure 2B ). In that context, the consequences of LRP-1 depletion were more drastic since the proportion of invading cells was decreased by 51% in LRP-1 deficient-cells. Inhibition of calpain activity in LRP-1-depleted cells restored invasive potential that recovered 71% of the level observed in control cells. In contrast, the calpain inhibitor failed to modulate the invasive ability of LRP-1 expressing cells. We therefore demonstrated that LRP-1 maintains a low level of calpain activity that contributes to the migratory and invasive capacities of carcinoma cells. We then conducted adhesion assays to test the likely involvement of LRP-1-dependent calpain regulation in the control of carcinoma cell attachment to matrix compounds ( Figure 3A ) (28). We evidenced an increase of the adhesion rate of cells lacking LRP-1, as 64% more cells spread compared to control cells. Inhibition of calpain activity reduced the adhesion rate of LRP-1-silenced cells to the level of control situation. By contrast, the adhesion of LRP-1 expressing cells adhesion was not affected by calpeptin treatment. In a complementary experimental set-up, we performed detachment assays of fully spread carcinoma cells in response to the enzymatic action of trypsin, using the proportion of cells remained adherent as a readout for cell-matrix interactions strength (36). The proportion of detached cells after LRP-1 knock-down was decreased by 51% as compared to control cells ( Figure 3B ). Adding calpeptin during the previous phase of attachment reduced the resistance of LRP-1-silenced cells to trypsin to recover the detachment rate detected in control cells. Although a similar trend could be observed in LRP-1 expressing control tumor cells, it was not significant. Together, those data illustrated that the pro-invasive action of LRP-1 in thyroid carcinoma cells relies in part on the maintenance of a low level of calpain activity leading to a moderate cell-matrix interactions strength.

Having identified calpains as new intracellular targets regulated by LRP-1 in carcinoma cells, we next set up an experimental design to decipher the mechanistic nature of the process controlling their activities. First we evaluated the differential expression of members of the ubiquitous calpain family by western blot ( Figures 4A, B ). Although we did not observe any difference of expression neither of calpain-1 (µ-calpain) nor of calpastatin, the endogenous inhibitor of activated calpains, we evidenced a significant 5-fold increase of calpain-2 (m-calpain) protein expression in LRP-1-silenced FTC133 cells. We then analyzed the level of the transcriptional expression of the two main proteases, namely calpain 1 and calpain-2, by RT-qPCR ( Figures 4C, D ). Consistently, calpain-1 transcriptional level was not affected by LRP-1 modulation, whereas calpain-2 gene expression was 4,4-fold increased under LRP-1 silencing. Since the expression level of calpastatin was not altered according to the expression of LRP-1, we then focused our attention on the regulation of the calpain activity by phosphorylation. In our screen for potential LRP-1 downstream signaling events, we measured PKA activity in cellular extracts from both LRP-1-silenced and control FTC133 cells. We evidenced that LRP-1 inhibition led to a 1.6-fold decrease of PKA activity ( Figure 5A ). Interestingly, PKA-dependent phosphorylation of calpain-2 had already been reported to inhibit the proteolytic activity of ubiquitous calpains (46). Accordingly, the treatment of FTC133 cells with H-89, a specific inhibitor of PKA, induced a 2-fold increase of calpain activity as measured by the cleavage of calpain substrate II ( Figure 5B ). H-89 treatment did not significantly affect this activity when LRP-1 was silenced. To support these data, we then used two distinct inhibitors of PKA activities, H89 and KT5720, and evaluated the impact on calpain activity using the fluorogenic calpain activity assay described in Figure 1D ( Figure 5C ). We evidenced that H-89 or KT5720 treatment respectively resulted in about a 2-fold and 1.7-fold increase of basal calpain activity in LRP-1 expressing carcinoma cells. Inhibition of PKA by H-89 was also efficient to increase the proportion of FTC133 cells positive for calpain activity ( Figure 5D ). Importantly, this effect was specific for controlling cells in which PKA activity was mainly detected. We then conducted pull-down experiments to study the phosphorylation status of calpain-2 ( Figure 5E ). We were able to immuno-precipitate the protein and we evidenced by western blot that its level of phosphorylation on serine residues was higher in control than in LRP-1 depleted cells. This signal was decreased in the presence of H-89, indicating that PKA-dependent mechanism was in part responsible for serine phosphorylation of calpain-2 in FTC133 cells. Adhesion assays were also conducted to demonstrate that PKA activity was functionally associated to the control of carcinoma cell attachment and spreading by LRP-1-mediated regulation of calpains ( Figure 5F ). Indeed, we specifically showed that inhibition of PKA by H-89 treatment in LRP-1 expressing cells enhanced their adhesion rate to gelatin to a similar level to that observed after LRP-1 inhibition. Our data indicated that PKA helped to limit calpain activity in a LRP-1-dependent manner in the frame of cell-matrix interactions established by thyroid carcinoma cells.

Considering the deep consequences of LRP-1-dependent regulation of calpain activity on the invasive potential and adhesion/deadhesion processes of tumor cells, we then analyzed the distribution and the morphology of cytoskeleton constituents. The fibrillar actin network of thyroid carcinoma cells after two hours of attachment was characteristic of spread and polarized cells such as mesenchymal cells ( Figure 6A ). Most of the fibrillar structures positively stained by phalloidin were concentrated at the cell periphery and protrusion sites. This specific pattern was not modified by calpains inhibition in control cells. In accordance with previous reports (9, 28), LRP-1 silencing induced drastic changes of carcinoma cells morphology and actin fibers distribution. Cells were overspread as compared to control cells and exhibited prominent radial and transverse fibers, widely distributed, both at tumor cell center and periphery. Under LRP-1 inhibition the calpain inhibitor induced a clear transition of cells toward the mesenchymal and elongated aspect observed in the control condition. We performed talin immuno-fluorescent staining to evaluate the consequences of calpains inhibition. In accordance with the morphological transitions described above, calpeptin did not affect the proportion of talin-positive structures in control cells, whereas LRP-1 inhibition reinforced the proportion of cells exhibiting obvious and stabilized talin-positive structures ( Figure 6B ). The occurrence of these complexes was specifically decreased in LRP-1-depleted cells under calpeptin treatment ( Figure 6C ). Our data illustrated that during the attachment of FTC133 cells to gelatin, calpains inhibition by LRP-1 sustained the polarization of thyroid carcinoma cells and constituted a limiting mechanism in focal adhesion maturation. We then measured the protein expression of talin by western blot using the 8D4 antibody, allowing the detection of the full length protein and its calpain-dependent cleavage fragment ( Figure 7A ). In accordance with the altered distribution of talin observed by immunocytochemistry ( Figure 6B ), we detected an accumulation of both native and cleaved forms of talin in LRP-1-silenced tumor cells. Interestingly, calpeptin treatment appeared to limit the accumulation of the cleaved form of talin. In order to assess the level of adhesion receptors expressed at cell surface, we performed biotin labelling of tumor cells spread on gelatin and treated with calpain inhibitor or vehicle. Precipitated protein membrane fractions were analyzed by immunoblotting ( Figure 7B ). Interestingly, we observed an increase of β1 integrin subunit at cell surface upon LRP-1 repression. This accumulation seemed partially decreased under calpeptin treatment. We then conducted immunostainings to qualify the distribution of β1 integrins in non-permeabilized carcinoma cells cultivated in the same conditions ( Figure 7C ). β1 integrins were detected in FTC133 cells and were mostly present at the periphery and at the front of spread and elongated cells. The presence of β1 integrins was accentuated in LRP-1 repressed cells and highly accumulated along the whole cell periphery. In the presence of calpeptin the polarized morphology of control cells did not show any clear critical transition, in opposition to LRP-1-silenced tumor cells which were more frequently polarized and spindle shaped and exhibited a discontinuous β1 integrin-positive margin that appeared more concentrated on distal sites of cell projections. We then conducted immunostainings by using an antibody specific for active forms of β1 integrins ( Figure 7D ). The localization of these forms of the receptors in non-permeabilized cells was more punctuated with some positive clusters at the leading edge of elongated control cells. These clusters were evenly observed across the ECM-attached area from overspread LRP-1-silenced cells. After inhibition of LRP-1 expression, both the number and size of theses clusters were diminished upon calpeptin treatment when LRP-1 was down-regulated. Altogether, these data showed that LRP-1-mediated control of calpain activity determined cell surface distribution of β1 integrins in thyroid carcinoma cells.