Vinculin, and A-Actinin K.-L

Home , Gelsolin (cellular), Stress fiber, Talin (protein), Vinculin

Proc. Natl. Acad. Sci. USA Vol. 93, pp. 9182-9187, August 1996 Medical Sciences

The differential adhesion forces of anterior cruciate and medial collateral ligament fibroblasts: Effects of tropomodulin, talin, vinculin, and a-actinin K.-L. PAUL SUNG*tt§, LI YANG*t, DARREN E. WHITTEMOREt, YAN SHI*, GANG JIN t, ADAM H. HSIEH*t, WAYNE H. AKESON*, AND L. AMY SUNGt¶ *Departments of Orthopaedics and tBioengineering, tCancer Center, and TCenter for Molecular Genetics, Institute for Biomedical Engineering, University of California at San Diego, La Jolla, CA 92093-0412 Communicated by Y C. Fung, University of California at San Diego, La Jolla, CA, May 28, 1996 (received for review April 8, 1996)

ABSTRACT We have determined the effects of tropo- fashion along the grooves of the actin double helix, stiffening modulin (Tmod), talin, vinculin, and a-actinin on ligament the filament and regulating the interaction between actin and fibroblast adhesion. The anterior cruciate ligament (ACL), other actin-binding proteins (12). Tmod may, therefore, reg- which lacks a functional healing response, and the medial ulate the length and/or organization of actin filaments by collateral ligament (MCL), a functionally healing ligament, differential binding to TM. In fibroblasts, hTM5 (one of the were selected for this study. The micropipette aspiration TM isoforms) is present not only in the more stable structures technique was used to determine the forces needed to separate of stress fibers, but also in the ruffled regions where the F-actin ACL and MCL cells from a fibronectin-coated surface. De- structures are rapidly changing (14). We found previously in a livery of exogenous tropomodulin, an actin-filament capping monocytic cell line that Tmod increases cell adhesion strength protein, into MCL fibroblasts significantly increased adhe- to fibronectin (FN), whereas the antibody against Tmod sion, whereas its monoclonal antibody (mAb) significantly decreases adhesion strength (15). decreased cell adhesiveness. However, for ACL fibroblasts, In this study we investigate the differences between a Tmod significantly reduced adhesion, whereas its mAb had no nonhealing ligament, the adult anterior cruciate ligament effect. mAbs to talin, vinculin, and a-actinin significantly (ACL), and a functionally healing ligament, the medial col- decreased the adhesion of both ACL and MCL cells with lateral ligament (MCL) (16, 17) by characterizing the roles of increasing concentrations ofantibody, and also reduced stress several cytoskeletal proteins in ligament fibroblast adhesion. fiber formation and cell spreading rate as revealed by immu- The purpose of the study is to determine the effects of a new nofluorescence microscopy. Disruption of actin filament and actin-filament capping protein, Tmod, and three integrin/ microtubule assembly with cytochalasin D and colchicine, actin-filament bridging proteins (talin, vinculin, and at-actinin) respectively, also significantly reduced adhesion in ACL and on ACL and MCL fibroblast adhesion behavior (Fig. 1). MCL cells. In conclusion, both ACL and MCL fibroblast Experiments were performed using the micropipette single- adhesion depends on cytoskeletal assembly; however, this cell aspiration technique to compare the adhesion forces of dependence differs between ACL and MCL fibroblasts in normal ACL and MCL cells with those of ACL and MCL cells many ways, especially in the role of Tmod. These results add electroporated with monoclonal antibodies (mAbs) against yet another possible factor in explaining the clinical differ- tropomodulin, talin, vinculin, and a-actinin. In addition, we ences in healing between the ACL and the MCL. examined the effects of agents that disrupt actin filament and microtubule assemblage [cytochalasin D (C.D.) and colchicine Cell-cell and cell-extracellular matrix adhesion play fundamen- (C.C.), respectively] on adhesion behavior as well as the tal roles in physiological processes such as wound healing, differences in the organization of stress fibers between ACL immune surveillance, thrombosis, and hemostasis. In the and MCL fibroblasts by immunofluorescence microscopy. To wounded ligament, fibroblasts embedded in the amorphous determine the effects of cytoskeletal assembly on cell adhe- healing tissue matrix of ligaments have been found to migrate sion, experiments were performed using a FN base. FN was into damaged sites. This migration, which regulates wound chosen because of its significance in ligament healing, not only closure during the healing process, involves adhesion events by providing a provisional submatrix for fibroblast migration and the remodeling of the internal cytoskeleton through the and ingrowth but also by acting as a linkage for wound reorganization of actin filaments and other cytoskeletal pro- contraction in the healing ligament (6). teins (1, 2). Studies have shown the importance of cytoskeletal proteins in the adhesion and motility of different cell types MATERIALS AND METHODS (1-7). Recent evidence indicates that the capping protein gelsolin, at the barbed ends of actin filaments, has striking Cell Cultures. Human ligament fibroblasts were obtained effects on cell motility by influencing actin-filament assembly from ACL and MCL explants from five subjects (one female (8, 9). and four male adults, 30-52 years old). Fibroblasts were Tropomodulin (Tmod) is a capping protein found at the free harvested at autopsy within 6-24 hr after death and isolated (pointed) ends of actin filaments, and its cDNA sequence has according to a prescribed protocol (18) and showed no dif- been determined (10, 11). It is known to be involved in ferences in viability and adhesion behavior among subjects (19, actin-filament assemblage, and it has been shown that Tmod 20). The cells were grown in Dulbecco's modified Eagle's interacts with the N terminus of tropomyosin (TM) and medium (DMEM)/10% fetal calf serum, supplemented with inhibits TM binding to actin by blocking the association of TM along actin filaments (12, 13). TM polymerizes in a head-to-tail Abbreviations: Tmod, tropomodulin; FN, fibronectin; ACL, anterior cruciate ligament; MCL, medial collateral ligament; TM, tropomyosin; C.D., cytochalasin D; C.C., colchicine. The publication costs of this article were defrayed in part by page charge §To whom reprint requests should be addressed at: Departments of payment. This article must therefore be hereby marked "advertisement" inI Orthopaedics and Bioengineering, University of California at San accordance with 18 U.S.C. §1734 solely to indicate this fact. Diego, La Jolla, CA 92093-0412. 9182 Downloaded by guest on September 29, 2021 Medical Sciences: Sung et al. Proc. Natl. Acad. Sci. USA 93 (1996) 9183 Microfilament and Microtubule-Disrupting Agents and Treatment. The actin-filament disrupting agent, C.D. (29) and the microtubule-disrupting agent, C.C. (30), both obtained from Sigma, were used at final concentrations of 0.5 and 0.1 ,uM, respectively, sufficient to inhibit cytoskeletal assembly within the time course of our experiments. Fibroblasts were treated with C.D. and C.C. by either (i) pre-incubation-30- min incubation at 37°C prior to cell seeding or (ii) post- incubation-cell seeding in the micropipette chamber at room temperature for 15 min followed by addition of C.D. or C.C. to the medium and the measurement of adhesion strengths for 45 min thereafter. These two treatments are intended to elucidate the effects of actin-filament- and microtubule- disrupting agents before and after the formation of stress fibers and adhesion plaques during cell attachment and spreading. matrix Control experiments were conducted without any agent. FIG. 1. Schematic drawing of cytoskeleton-integrin-extracellular Micropipette Chamber/Adhesion Force Measurements. matrix complex. Tmod, located at the pointed end of actin filaments, The detailed coating procedure of 5 ,ug/ml FN on cover glasses and the bridging proteins (talin, vinculin, and a-actinin), located within micropipette chambers and the micropipette- between actin filaments and integrins, are shown. micromanipulation system used for the measurement of adhesion forces has been described (19, 31, 32). Glass micropi- nonessential amino acids (0.10 mM), L-glutamine (4 mM), pettes (prepared using a Flaming Brown Model P-87 puller, penicillin (100 units/ml), streptomycin (100 ,ug/ml), and fun- Sutter Instruments, Novato, CA), with an internal tip radius of gizone (0.25 ,ug/ml) (BioWhittaker). Cultures were main- 1.5-3.0 ,um, were used and the adhesion characteristics were tained at 37°C and 5% C02/95% air. At confluency, fibro- measured under direct microscopic observation in conjunction blasts were trypsinized, washed, and resuspended for experi- with a video recording system. The force (product of aspiration ments. Passages 2-6 of fibroblast cultures were used. pressure and the cross-sectional area of the micropipette tip) Electroporation of Fibroblasts. Electroporation (21) was required to separate the cell from the FN coat was measured used to deliver Tmod (25 and 50 ,tg/ml), mAbs against Tmod by stepwise increases in aspiration pressure followed by re- (mAb Tmod-204, 25 and 50 ,tg/ml), talin (200 and 385 ,ug/ml), traction of the pipette. Cells were chosen at random and vinculin (30 and 100 ,g/ml), and a-actinin (50 and 100 ,tg/ml) adhesion data were collected for 45 min following an initial into fibroblasts. mAbs against Tmod, talin (Sigma), vinculin 15-min seeding time at room temperature. (Sigma), and a-actinin (Sigma) are anti-human antibodies Stress Fiber Visualization. Immunofluorescence staining from mice and albumin antibody (control) is an anti-human and epifluorescence (Nikon Diaphot-TMD-EF) microscopy antibody developed in rabbits. The selection of antibody were used to characterize stress fiber formation during adhe- concentrations was based on cell concentrations, stoichiome- sion and spreading of control cells and cells introduced with try, and spatial distributions of the involved molecules avail- mAbs against talin (385 ,ug/ml), vinculin (100 gg/ml), and able in literature: (i) talin was found to aggregate with a-actinin (100 ,ug/ml) on FN-coated glass. Cells were fixed f31-integrin in a 1:1 ratio at focal contact sites during the very with 3.7% (vol/vol) formalin for 20 min, permeabilized with early stages of fibroblast adhesion, whereas vinculin (0.5:1 0.5% Triton X-100 in PBS for 30 min, and incubated for 20 min at room temperature with 0.17 rhodamine phalloidin (Mo- ratio with f31-integrin) and a-actinin (0.2:1 ratio with 13i- ,AM lecular Probes), integrin) were less abundant at focal contacts (2, 22, 23); (ii) which binds to F-actin in the microfilaments. Statistics. vinculin, which has a total endogenous concentration of 100 Comparisons between controls and treated groups were performed using the unpaired Student's t test to ,ug/ml (24), binds to talin in a 1:3 molar ratio (25); and (iii) (mean) determine if any significant differences (P < 0.05) exist. a-actinin, which has a total endogenous concentration of 1 mg/ml (24), is found to be more abundant in stress fibers as an actin-filament crosslinking protein than at focal contacts (26). RESULTS Electroporation was a performed using T820 electroporation Adhesion Experiments. The effects of mAbs against four system (BTX, San Diego) at a setting (1 pulse, 99 ,usec pulse important actin-associated proteins, Tmod (25 and 50 Ag/ml), length, 300 volts) that yields 90% cell viability (trypan blue talin (200 and 385 ,ug/ml), vinculin (30 and 100 ,ug/ml), and exclusion). A negative control in the absence of antibody or a-actinin (50 and 100 jig/ml), on the adhesion of ACL and protein was used to monitor the nonspecific effects of the MCL fibroblasts to FN (5 ,ug/ml)-coated surfaces were inves- treatment. Cells electroporated with anti-albumin antibody tigated. The normalized forces required to detach the cells (150 ,tg/ml, control for mAbs) or albumin (50 ,tg/ml, control from the chemically well-defined surfaces are shown in Fig. 2. for Tmod) were used to normalize values in subsequent Anti-Tmod (50 ,ug/ml) significantly reduced MCL cell adhe- adhesion studies. Cells were used 2-3 hr after electroporation, sion by 18% (P < 0.002), but had no effect on ACL cell allowing time for the cells to recover and for the exogenous adhesion. A lower concentration of anti-Tmod (25 p,g/ml) was proteins or mAbs to bind. unable to affect the adhesion of either ACL or MCL cells. Western Blot Analysis. Western blot analyses were per- Adhesion of MCL cells was significantly reduced by anti-talin formed on fibroblasts before and after electroporation with 25 at both concentrations of 200 Ag/ml (31% reduction, P < 0.01) and 50 ,ug/ml of Tmod into the cells to verify the successful and 385 jig/ml (20% reduction, P < 0.05), whereas adhesion delivery of protein. After electroporation, 106 cells were of ACL cells was significantly reduced only by 385 ,ug/ml of dissolved in an SDS/solubilization solution (1.4 M 2-mercap- anti-talin (29% reduction, P < 0.015) and not 200 ,ug/ml (10% toethanol/4% SDS/50 mM Tris, pH 6.8), and the total pro- reduction, P > 0.2), suggesting that a talin threshold level teins were separated by SDS/10% PAGE (27). Protein profiles exists. There was no significant difference in adhesion force were transblotted onto nitrocellulose paper (28) and probed between MCL cells introduced with 200 and 385 ,ug/ml of with an mAb against Tmod (mAb Tmod-204, unpublished anti-talin. ACL cell adhesion was significantly reduced by work), followed by a goat anti-mouse IgG secondary antibody anti-vinculin and anti-a-actinin at both concentrations (43% conjugated with horseradish peroxidase (1:3000) (Bio-Rad). reduction by 30 and 100 ,ug/ml anti-vinculin, P < 0.0001; 32% Downloaded by guest on September 29, 2021 9184 Medical Sciences: Sung et al. Proc. Natl. Acad. Sci. USA 93 (1996) Table 1. Effect of Tmod on ACL and MCL fibroblast adhesion to an FN-coated (5 j,g/ml) surface ACL normalized MCL normalized adhesion, force adhesion, force Treatment (± SEM) (± SEM) 0.6 Tmod (25 t,g/ml) 0.69 + 0.05* (n = 57) 1.14 + 0.09 (n = 87) Tmod (50 ,ug/ml) 0.70 ± 0.06* (n = 38) 1.79 + 0.12* (n = 71) n, Number of cells. *P < 0.05. Tmod into cells (Table 1). ACL cells electroporated with 25 or p0.48 50 ,ug/ml of Tmod showed a decrease in adhesion by '30% with respect to control. In contrast, MCL cells showed a 79% 0.6- increase in adhesion relative to control at 50 ,tg/ml of Tmod, but no significant change with respect to the control at 25 0.4 ,ug/ml. The differential effects of electroporation of Tmod on the adhesion between these two types of cells strongly suggest distinct molecular mechanisms acting within the ACL and 0.2- MCL cells. The effects of disrupting actin-filament assembly with 0.5 ,uM C.D. and microtubule assembly with 0.1 ,uM C.C. on the adhesion of ACL and MCL fibroblasts to FN is shown in Fig. 0- 3. Pre-incubation treatment with C.D. significantly reduced Ant-ibody Tmod talin vinculin oz-actinin ACL and MCL cell adhesion force by 91% and 84% (P < (tgIml)025 50 200 385 30 100 50 100 0.0001), respectively. Post-incubation treatment with C.D. was less effective, but still significantly reduced the adhesion force FIG. 2. Effect of mAbs against Tmod, talin, vinculi'n, and a-actinin for ACL and MCL cells by 27% and 34% (P < 0.01), on the adhesion of ACL (Upper) and MCL (Lower) fibroblasts to an respectively. Pre-incubation treatment with C.C. significantly FN-coated (5 tkg/ml) surface. Two concentrations of mAbs used were reduced the adhesion force of ACL and MCL cells by 20% indicated beneath each mAb. Adhesion forces were normalized to control at 150 Error bars are SEM. (anti-albumin antibody 4g/ml). 1.5 - Number of cells measured: control ACL (n = 48) and MCL (n =87), a. Cytochalasin D anti-Tmod ACL (n = 79) and MCL (n = 97), anti-talin ACL (n =49) and MCL (n = 93), anti-vinculin ACL (n = 69) and MCL (n =96), and anti-a-actinin ACL (n = 82) and MCL (n = 85). *, denotes ACL significant difference with respect to the control. 1.0 - * MCL * * reduction by 50 ~tg/ml anti-a-actinin, P < 0.005; 36% reduction by 100 jig/ml anti-a-actinin, P < 0.002), whereas MCL cell adhesion demonstrated a concentration dependence for anti- 0 0.5- vinculin and anti-a-actinin, showing reduced adhesion only at u: the higher concentration of 100 gg/ml (11% reduction by 30 0 *W ~tg/ml anti-vinculin, P > 0.4; 20% reduction by 100 txg/ml anti-vinculin, P < 0.05; 4% reduction by 50 anti-a- 0 jitg/ml "00 actinin, P > 0.5; 36% reduction by 100 ttg/ml anti-a-actinin, N- P < 0.005). There was no significant difference in adhesion c/i force between ACL cells electroporated with 30 and 100 jtg/ml zA of anti-vinculin, and also between ACL cells introduced with 50 and 100 iLg/ml of anti-a-actinin. Adhesion forces between the negative control cells that received no protein or antibody during electroporation and the control cells that received a control protein, albumin or its antibody, were not significantly different, indicating that neither albumin nor its antibody have any effect on ACL and MCL fibroblast adhesion to FN and that they may serve as a negative control for the adhesion studies. In addition to the mAb to Tmod, a recombinant human Tmod was also used in evaluating the effects of Tmod on 'the adhesion of the ligament fibroblasts. Western blot analysis of ACL and MCL cells before and after electroporation with pre-incubation post-incubation Tmod (25 pLg/ml and 50 .tkg/ml) demonstrated that electroporation was successful in delivering the exogenous Tmod into FIG. 3. Effect of C.D. (0.5 ,uM, a) and C.C. (0.1 ,tM, b) on the both types of cells. Human Tmod is a 40.6-kDa protein (11) adhesion of the two types of ligament fibroblasts to an FN-coated (5 with a Mr of 43,000 (33). Compared with control, densitometer ,ug/ml) surface. Adhesion forces were normalized to a control without measurements of the treated groups showed that Tmod con- reagent treatment. Error bars are SEM. Number of cells measured: tent was 21% and 64% in cells C.D.-control ACL (n = 67) and MCL (n = 57), pre-incubation; ACL (ACL cells) (MCL cells) higher (n = 37) and MCL (n = 40), post-incubation; ACL (n = 78) and MCL electroporated with 25 .tLg/ml of Tmod and 55% (ACL cells) (n = 25). C.C.-control ACL (n = 103) and MCL (n = 71), and 101% (MCL cells) higher in cells electroporated with 50 pre-incubation; ACL (n = 99) and MCL (n = 111), post-incubation; jitg/ml of Tmod. ACL and MCL cell adhesion was measured ACL (n = 57) and MCL (n = 31). *, Significant difference with respect and found to be affected differently after delivering exogenous to the control. Downloaded by guest on September 29, 2021 Medical Sciences: Sung et al. Proc. Natl. Acad. Sci. USA 93 (1996) 9185

Ii111 III lil iIIIIII Ib FIG. 4. Epifluorescence micrographs of ACL (I) and MCL (II) fibroblasts seeded to an FN-coated (5 j,g/ml) coverglass. Actin filaments were stained with rhodamine phalloidin to visualize the spatial organization and formation of actin filaments into bundles (stress fibers). Seeding times: (a) 1 hr, (b) 4 hr, (c) 24 hr, and (d) 1 hr with mAb against a-actinin (I-d) and talin (II-d). (P < 0.01) and 29% (P < 0.0005), respectively. Post-incubation remaining parallel after 4 hr (Fig. 4 I-b and II-b). Stress fibers treatment with C.C. significantly reduced MCL cell adhesion organized into nearly parallel lines after 24 hr (Fig. 4 I-c and force by 23% (P < 0.05), but had no significant effect on ACL II-c). The changes in stress fiber orientation in ACL cells cell adhesion. occurred more quickly than in MCL cells, but stress fibers in Visualization of Actin Filaments. Fig. 4 shows epifluores- MCL cells appeared thicker than in ACL cells. The spaces between stress fibers in MCL cells were also wider than in ACL cence micrographs of actin-filament bundles (stress fibers) in cells. The effect of mAbs against a-actinin in ACL cells (Fig. ACL (Fig. 4I) and MCL fibroblasts (Fig. 411) seeded to FN (5 4I-d) and against talin in MCL cells (Fig. 4II-d) are also shown. ,g/ml) for different time periods (1, 4, and 24 hr, labeled as Stress fiber formation and cell attachment area were reduced a, b, and c, respectively). At the beginning ofcell seeding to FN, in ACL and MCL cells introduced with mAbs against talin, the stress fibers at the edge of both ACL and MCL cells vinculin, and a-actinin compared with control cells. emanated radially, whereas the inner stress fibers were ran- domly organized. After 1 hr, the inner stress fibers appeared DISCUSSION circumferential and in parallel rings (Fig. 4 I-a and II-a). Due This study investigated the roles of various cytoskeletal pro- to cell elongation, the fibers then became elliptical, while teins-Tmod, talin, vinculin, a-actinin, actin, and tubulin-in Downloaded by guest on September 29, 2021 9186 Medical Sciences: Sung et al. Proc. Natl. Acad. Sci. USA 93 (1996) ACL and MCL fibroblast adhesion to an FN-coated surface. The role of Tmod in ACL and MCL fibroblast adhesion is The major findings of this study are as follows: (i) Adhesion of less clear. For the two concentrations of Tmod used, MCL cell ACL and MCL fibroblasts is influenced by the actin-filament adhesion depended on concentration (being affected only at assembly and cytoskeletal integrity, and requires that cells the higher concentration of 50 ,tg/ml Tmod), whereas ACL possess intact stress fibers for their normal attachment prop- cell adhesion was affected similarly at both concentrations. erties; (ii) the actin-filament pointed-end capping protein, This seems to indicate a threshold difference in Tmod con- Tmod, influences cell adhesion behavior differently between centration between ACL and MCL cells. Fibroblasts from the ACL and MCL fibroblasts; (iii) adhesion force and stress fiber MCL had increased adhesion with Tmod and decreased ad- formation are functions of the intact linkage between actin hesion with anti-Tmod, similar to the findings from other filaments and integrin receptors by the bridging proteins talin, adhesion studies performed on white blood cells in our labo- vinculin and a-actinin; and (iv) ACL and MCL fibroblasts ratory (15). It is assumed that the capping of actin filaments at differ in their sensitivity to mAbs against the same cytoskeletal the pointed end (nonpreferred end of polymerization) with proteins. Tmod serves in accelerating the polymerization of actin fila- We believe that this study is the first to characterize, ments at the barbed end near the focal contacts. This allows for quantitatively, adhesion forces of human ligament fibroblasts greater bundling of actin filaments into stress fibers which, in influenced by the assembly of cytoskeletal proteins. By exam- turn, results in greater adhesion. Fibroblasts from the ACL, on ining intrinsic differences in adhesion between ACL and MCL the other hand, showed a much different effect of Tmod on fibroblasts in response to the disruption of cytoskeletal pro- adhesion behavior compared with MCL fibroblasts. This find- teins, we seek to provide important insights to the ligament ing is consistent with past findings on the contrasting intrinsic healing process. Previous studies in several laboratories have properties between ACL and MCL fibroblasts regarding their shown differences in cell morphology (34), procollagen pro- adhesion and migration behavior (18-20) and also consistent duction rate (35), and adhesion characteristics between ACL with new findings on intracellular differences. During the early and MCL cells in culture (18-20, 36-39). Our study shows that period of adhesion, ACL fibroblasts lack the elevation in the cytoskeletal proteins have striking effects on ACL and intracellular calcium observed in MCL fibroblasts (19), and MCL cell adhesion and on the formation of stress fibers during this may account for the differences seen in stress fiber cell spreading. A disruption in the cytoskeleton-integrin bridg- formation rate and size between ACL and MCL fibroblasts ing complex results in significantly reduced adhesion. This (Fig. 4). Overall, we found ACL fibroblast adhesion to be less reduction in adhesion is most likely a result of decreased sensitive than MCL fibroblast adhesion to calcium fluctuations formation of stress fibers and cell attachment area seen in (unpublished data). Signal pathway experiments we conducted fibroblasts introduced with mAbs against talin, vinculin, and also reveal differences in the roles of A-kinase, G-kinase, and a-actinin. Our results indicate that the cytoskeletal proteins the Ca2+/phospholipid pathway in cell adhesion between ACL talin, vinculin, and ac-actinin all influence the adhesion of ACL and MCL fibroblasts (unpublished data). This may also have and MCL fibroblasts to FN; however, their roles appear to be an impact on actin polymerization, cytoskeletal protein phos- cell-type specific. By performing experiments using different phorylation, calcium- regulated stress fiber assembly, and concentrations of mAbs against these three proteins, we possibly the actin-filament capping function of Tmod during determined that ACL fibroblasts are more dependent on adhesion. vinculin and a-actinin for adhesion to FN than MCL fibro- This investigation concentrates only on cellular adhesion blasts. ACL fibroblasts require lower concentrations of anti- forces and the topographic distribution of actin filaments vinculin and anti-a-actinin than MCL fibroblasts to reduce (stress fibers), and, therefore, primarily reflects cell adhesion adhesion force significantly. On the other hand, talin plays a behavior. It would be of interest to examine the cell signal more prominent role in MCL fibroblast adhesion to FN than transmission through the integrin receptor during cell adhe- in ACL fibroblasts. sion to the extracellular matrix, the process of phosphorylation By disrupting the assembly of actin filaments with C.D., of cytoskeletal proteins, and the assembly of stress fibers. adhesion force can be drastically decreased by 80-90% when Further study on the signal pathways of ACL and MCL ACL and MCL cells are pre-incubated with the agent before fibroblasts would help establish the molecular basis of the cell attachment. A previous study strongly suggested the biophysical behavior of cell adhesion, migration, and prolif- involvement of the cytoskeleton in the regulation of integrin- eration. mediated cell adhesion after it was found that C.D. abolished chimeric receptor adhesion to immobilized fibrinogen (40). We thank Dr. Shu Chien for his constant support in our scientific Therefore, the reduction in adhesion force observed in our research. This research was supported by National Institutes of Health experiments may be caused by a loss of adhesion function of Grants AR34264 and HL43026 and American Cancer Society Grant the 131-integrin receptor for FN. It is possible that this loss in IM-648. function could stem from a conformational change in the integrin receptor brought on by the disruption in the associ- 1. Stossel, T. P. (1993) Science 260, 1086-1094. 2. Nuckolls, G. H., Romer, L. H. & Burridge, K. (1992) J. Cell Sci. ation between the actin filaments of the cytoskeleton and the 102, 753-762. integrin receptor. Another possibility is that actin filaments are 3. Gluck, U. & Ben-Ze'ev, A. (1994) J. Cell Sci. 107, 1773-1782. anchored to the integrin receptor by bridging proteins. The 4. Lokeshwar, V. B., Fregien, N. & Bourguignon, L. Y. W. (1994) assembly of many actin filaments into stress fibers requires that J. Cell BioL 126, 1099-1109. the barbed ends of the actin filaments be anchored to the 5. Cunningham, C. C., Gorlin, J. B., Kwiatkowski, D. J., Hartwig, integrin receptors. This condensed area of integrin receptors J. H., Janmey, P. A., Byers, H. R. & Stossel, T. P. (1992) Science and stress fibers is built up to form a functional adhesion 255, 325-327. plaque. Any disruption in stress fiber assembly will disrupt 6. Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K. & Watson, adhesion plaque formation. Microtubule assembly also ap- J. D. (1994) Molecular Biology of the Cell (Garland, New York), 3rd Ed., pp. 821-847, 997. pears to be an important factor in fibroblast adhesion to the FN 7. Wang, N., Butter, J. P. & Ingber, D. E. (1993) Science 260, molecule, but not as essential as actin-filament assembly. This 1124-1127. mechanism, however, is not yet completely understood. One 8. Witke, W., Sharpe, A. H., Hartwig, J. H., Azuma, T., Stossel, report showed a close functional correlation between micro- T. P. & Kwiatkowski, D. J. (1995) Cell 81, 41-51. tubules and microfilaments with respect to cytoskeletal activity 9. Hug, C., Jay, P. Y., Reddy, I., McNally, J. G., Bridgman, P. C., in leukocytes (41). Elson, E. L. & Cooper, J. A. (1995) Cell 81, 591-600. Downloaded by guest on September 29, 2021 Medical Sciences: Sung et al. Proc. Natl. Acad. Sci. USA 93 (1996) 9187 10. Weber, A., Pennise, C. R., Babcock, G. G. & Fowler, V. M. 26. Pavalko, F. M., Schneider, G., Burridge, K. & Lim, S.-S. (1995) (1994) J. Cell Biol. 127, 1627-1635. Exp. Cell Res. 217, 534-540. 11. Sung, L. A., Fowler, V. M., Lambert, K., Sussman, M., Karr, D. 27. Laemmli, U. K. (1970) Nature (London) 227, 680-685. & Chien, S. (1992) J. Biol. Chem. 267, 2616-2621. 28. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl. Acad. 12. Sung, L. A. & Lin, J. J.-C. (1994) Biochem. Biophys. Res. Com- Sci. USA 76, 4350-4354. mun. 201, 627-634. 29. Wodnicka, M., Pierzchalska, M., Bereiter-Hahn, J. & Kajstura, J. 13. Fowler, V. M. (1990) J. Cell Biol. 111, 471-482. (1992) Folia Histochem. Cytobiol. 30, 107-111. 14. Lin, J. J.-C., Hegmann, T. E. & Lin, J. L.-C. (1988) J. Cell Biol. E. D., McKeel, M. & Hays, T. (1984) J. Cell Biol. 99, 107, 563-572. 30. Salmon, 15. Sung, K.-L. P., Yang, L., Whittemore, D. & Sung, L. A. (1995) 1066-1075. Proceedings of the Fourth China-Japan-USA-Singapore Confer- 31. Kato, G. & Wakabayashi, K. (1988) In Vitro CellularDev. Biol. 24, ence on Biomechanics, pp. 373-376. 274-280. 16. Mow, V. C. & Hayes, W. C. (1991) Basic Orthopaedic Biome- 32. Sung, K.-L. P., Sung, L. A., Crimmins, M., Burakoff, S. J. & chanics (Raven, New York), 1st Ed., pp. 51-92, 199-244. Chien, S. (1986) Science 234, 1405-1408. 17. Schreck, P. J., Kitabayashi, L. R., Amiel, D., Akeson, W. H. & 33. Fowler, V. M. (1987) J. Biol. Chem. 262, 12792-12800. Woods, V. L., Jr. (1995) J. Orthop. Res. 13, 174-183. 34. Amiel, D., Frank, C. B., Harwood, F. L., Fronek, J. & Akeson, 18. Nagineni, C. N., Amiel, D., Green, M. H., Berchuck, M. & W. H. (1984) J. Orthop. Res. 1, 257-265. Akeson, W. H. (1992) J. Orthop. Res. 10, 465-475. 35. Wiig, M. E., Amiel, D., Ivarsson, M., Nagineni, C. N., Wallace, 19. Sung, K.-L. P., Kwan, M. K. & Akeson, W. H. (1994) J. Biomech. C. D. & Arfors, K. E. (1991) J. Orthop. Res. 9, 374-382. Eng. 116, 237-242. 36. Lyon, R. M., Akeson, W. H., Amiel, D., Kitabayashi, L. R. & 20. Sung, K.-L. P., Steele, L. L., Whittemore, D., Hagan, J. & Ake- Woo, S. Y. (1991) Clin. Orthop. 272, 279-286. son, W. H. (1995) J. Orthop. Res. 13, 266-273. 37. Newton, P. O., Woo, S. L., MacKenna, D. A. & Akeson, W. H. 21. Chakrabarti, R., Wylie, D. E. & Schuster, S. M. (1989) J. Biol. (1995) J. Orthop. Res. 13, 191-200. Chem. 264, 15494-15500. 22. Samuelsson, S. J., Luther, P. W., Pumplin, D. W. & Bloch, R. J. 38. Schmidt, C. C., Georgescu, H. I., Kwoh, C. K., Blomstrom, G. L., (1993) J. Cell Biol. 122, 485-496. Engle, C. P., Larkin, L. A., Evans, C. H. & Woo, S. L. (1995) 23. Mueller, S. C., Kelly, T., Dai, M. Z., Dai, H. N. & Chen, W. T. J. Orthop. Res. 13, 184-190. (1989) J. Cell Biol. 109, 3455-3464. 39. Geiger, M. H., Green, M. H., Monosov, A., Akeson, W. H. & 24. Kreis, T. E. & Birchmeier, W. (1982) Int. Rev. Cytol. 75, 209-227. Amiel, D. (1994) Connective Tissue Res. 30, 215-224. 25. Lee, S.-W., Wulfkuhle, J. D. & Otto, J. J. (1992) J. Biol. Chem. 40. Peter, K. & O'Toole, T. E. (1995) J. Exp. Med. 181, 315-326. 267, 16355-16358. 41. Keller, H. U. & Niggli, V. (1993) Cell Motil. Cytoskel. 25, 10-18. Downloaded by guest on September 29, 2021