[CANCER RESEARCH 52. 3267-3272, June 15. 1992] Effect of Thrombin Treatment of Tumor Cells on Adhesion of Tumor Cells to Platelets in Vitro and Tumor Metastasis in Vivo1
Mary Lynn R. Nierodzik, Francis Kajumo, and Simon Karpatkin2
New York University Medical School, New York, New York 10016 [M. L. R. N., F. K., S. K.J, and Department of Veterans Affairs Medical Center, New York, New York 10010 ¡M.L. R. N.¡
ABSTRACT liunit concentrations of thrombin results in a 2- to 5-fold enhancement of adhesion to 6 different tumor cell lines from 3 Seven different tumor cell lines (human melanoma SK MEL 28; different species in vitro; infusion of thrombin in vivo results in hamster melanoma HM29; murine melanomas B16F10 and amelanotic melanoma B16a; human colon carcinoma 11("IX:murine colon carcinoma a 4- to 413-fold enhancement in pulmonary metastasis with CT26; and murine Lewis lung carcinoma) were treated with thrombin at two different tumor cell lines (5). 0.5-1 unit/ml and examined for their ability to bind to adherent platelets; Because of the marked effect of thrombin on platelet adhe HM29 was studied for its ability to bind to fibronectin and von Willebrand siveness to tumor cells in vitro and tumor metastasis in vivo, as factor;
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research. THROMBIN-TREATED TUMOR CELLS ENHANCE METASTASIS
Stefan Niewiarowski, Temple University Medical School, Philadelphia, assuming a sphere. The mean nodule volume for each mouse was PA. calculated by dividing the total nodule volume for each group of mice Thrombin and Anti-Thrombin Reagents. Human thrombin (3000 NIH units/mg) was purchased from Sigma, DAPA,1 a potent, highly specific by the number of mice in each group. Tumor mass was calculated by multiplying the mean nodule number by the mean nodule volume per competitive inhibitor of thrombin (10) was a gift of Dr. Michael lung. Nesheim, Queens University, Ontario, Canada. Preparation of Platelets. Human platelets were separated from cit- Microscopic examination was performed in some experiments. The rated platelet-rich plasma (0.38% final concentration) on a 10 and 20% fixed lungs were embedded with paraffin in a unicassette (Miles Labo discontinuous Stractan gradient (arabino-galactan polymer) dissolved ratories, Inc., Elkhart, IN), and three coronal slices, 3 ^m thick, were in an isotonic balance salt solution, as described in detail, previously made through the lung. The pathological specimen was stained with (3, 4). hematoxylin-eosin and the largest slice was enumerated for number In Vitro Platelet-Tumor Cell Adhesion Assay. Platelets (3 x IO7) and surface area of tumor foci. Surface area of each metastatic focus were applied to 96-well flat bottomed microtitre plates (Falcon Labware was calculated with a micrometer by multiplying the largest two diam 3072, Oxnard. CA) for I h at room temperature to develop a platelet eters at right angles to each other. lawn as described previously (3, 4). Platelets were then treated with Statistics. Paired two-tailed t tests (or one-tailed, when indicated) buffer, thrombin, DAPA, or their combination and incubated overnight were used for statistical evaluations. at 4°C.Plates were then incubated with 0.01 M PBS plus 1% BSA for l h at 37'C, washed once with the above, and three times with PBS- BSA plus 0.9 mM CaCl;, 0.9 rriMMgCl2 before the addition of 1 x 10s RESULTS naive or thrombin-treated tumor cells in PBS-BSA-Ca-Mg for l h at 37°C.In some experiments, blocked and washed platelets were treated Comparison of Effect of Thrombin Treatment of HM29 Tumor with various monoclonal antibodies and inhibitors prior to the addition Cells versus Thrombin Treatment of Platelets on Adhesion of of tumor cells. Adherent tumor cells were removed with trypsin-EDTA HM29 Tumor Cells to Platelets (GIBCO Laboratories) and enumerated in a hemocytometer by phase microscopy. Tumor cells were incubated with or without thrombin for l h at Fig. 1 compares the effect of thrombin treatment of HM29 37°Cwith gentle rocking on a shaker prior to their addition to the melanoma tumor cells on their adhesion to platelets with platelet lawn. In some experiments, thrombin or sham-treated tumor thrombin treatment of platelets on their adhesion to tumor cells were sedimented and resuspended with RODS or monoclonal antibody 10E5 for l h at 37°Cand then washed in PBS-BSA-Ca-Mg cells. Note the 2.7-fold enhancement of adhesion of thrombin- treated tumor cells to naive platelets (Fig. 1, center) versus the prior to their addition to the platelet lawn. 3.1-fold enhancement of adhesion of untreated tumor cells to In Vitro Tumor Cell Adhesion to Fibronectin and von Willebrand Factor. Microtiter plates were coated with optimum concentrations of thrombin-treated platelets (Fig. 1, left). Note the lack of an fibronectin (0.6 /^g/well) or von Willebrand factor (1.2 ¿ig/well)over additive effect of the combination of thrombin-treated tumor night at 4°Cfor maximum adhesion. They were then washed and cells plus thrombin-treated platelets on tumor platelet adhesion blocked with PBS-BSA-Ca-Mg prior to addition of tumor cells for l h at 37°C.Non-adherent tumor cells were removed by washing with PBS- (Fig. 1, right, 2.8-fold enhancement). BSA-Ca-Mg. Adherent cells were removed with trypsin-EDTA and Effect of Thrombin Treatment of Other Tumor Cells on Their enumerated by phase microscopy. Adhesion to Platelets Protein Synthesis Inhibition with Cycloheximide. Tumor cells were incubated as described above in the presence and absence of 5 MCi/ml ['H]leucine (New England Nuclear NET-460, 37 MBq/ml) in leucine- Fig. 2 demonstrates the enhanced adhesion effect of throm free RPMI (GIBCO). Following incubation, cells were washed 3 times bin-treated tumor cells on naive platelets with 5 of 7 different with PBS-Ca-Mg. Cells were then centrifuged. the pellet was lysed with tumor cell lines from 3 different species. 1 ml of N NaOH for 15 min at 37°C,and protein was precipitated with 5% trichloroacetic acid at -20°C for 10 min. The supernatant was removed, the pellet was dissolved in 0.5 ml of N NaOH, 4 ml of scintillation fluid were added, and radioactivity was monitored. In Vivo Metastatic Studies. Experiments were performed as described previously (2-4). Tumor cells were removed from subconfluent tissue culture dishes with trypsin-EDTA, washed in PBS-Ca-Mg, and exam ined for viability with trypan blue. Tumor cells (1 x 106)were suspended CONTROL in 500-1,000 milliunits of thrombin in PBS-BSA-Ca2*-Mg2*, or PBS- BSA-Ca-Mg vehicle, incubated for l h at 37°C,sedimented, washed T 50mL7ml twice in PBS-BSA-Ca-Mg, and 25,000-50,000 viable tumor cells were injected i.v. into the tail vein of syngeneic BALB/c or C57B1/6J mice T SOOmU/ml in a volume of 200 /ul.In one experiment, thrombin-treated tumor cells were incubated with hirudin, 1 anti-thrombin unit/ml for 15 min, washed twice, and injected into the tail vein. Control and experimental 0 mice were injected alternately to avoid the bias of possible changes in in vivo metastatic potential after in vitro storage at 4°C(11). Animals PLATELETS PLATELETS + were killed on day 21 and the lungs were removed and fixed in Perfix TUMOR CELLS (Fisher Scientific Co., Springfield, NJ) for 24 h before enumeration of Fig. 1. Effect of thrombin-treated platelets versus thrombin-treated cells on métastasesmacroscopically or microscopically. For macroscopic ex tumor cell-platelet adhesion. Platelets (3 x 10') were applied to microtiter plates. Left, platelets were treated with thrombin or buffer overnight and then incubated amination, surface number and diameter of nodules were classified into with 1 x 10' naive HM29 cells in PBS-BSA-Mg-Ca. Middle, naive platelets were five categories, <0.1 mm, >1<2 mm, >2<3 mm, >3<4 mm, and >4 incubated with 1 x 10s HM29 cells previously treated with buffer or thrombin mm. The volume of each nodule was calculated from its diameter. for I h at 37°C,followed by washing. Right, thrombin-treated platelets were incubated with thrombin-treated HM29 cells as in Left and Middle. In all 'The abbreviations used are: DAPA, A'-(3-ethy!-1.5-pentanediyl)amide; PBS- experiments cells were washed with buffer, eluted with trypsin-EDTA. and BSA-Ca-Mg, phosphate-buffered saline plus 1% bovine serum albumin, plus 0.9 enumerated under phase microscopy. Columns, mean of 4 experiments performed mM Cad: and 0.9 mM MgClj. in triplicate; bars, SEM. 3268
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research. THROMBIN-TREATED TUMOR CELLS ENHANCE METASTASIS
bin is preincubated with DAPA (specific competitive inhibitor KOI•—¿ 1COM of thrombin) and lack of effect of DAPA if added l h after -,. Blta,„H,E52 preincubation of tumor cells with thrombin (probability of DAPA inhibiting after l h (P> 0.1)). ,,.400-Vf tlCIKin:.tiD OrCONTROLADIIKSION Role of Platelet GPIIb-GPIIIa Integrin in Adhesion of l.l.2ln=h1=1 MQ•00 Thrombin-treated Tumor Cells to Naive Platelets IIMMin-«iCD 111.in1 Fig. 6 demonstrates inhibition of thrombin-activated tumor SKMM. 2Xin =!i cell binding to platelets with previously determined, optimum concentrations of inhibitors of the platelet GPIIb-GPIIIa inte- 1 1 llll,H(Hn 1,1 grin receptor: RGDS (P = 0.03), monoclonal antibody 10E5 0^IW.1 (P = 0.02), and the disintegrin, albolabrin (P = 0.04). Isotype- Fig. 2. Effect of seven different thrombin-treated tumor cells on adhesion to naive platelets. Tumor cells (1 x 10s) were incubated with buffer or 1 unit/ml specific control murine IgG2a and irrelevant peptide GRGES thrombin for l h at 37'C. washed, and incubated with naive platelets as in Fig. had no effect. I. Number of experiments is given in parentheses. Effect of Monoclonal Antibody 10E5 and Tetrapeptide RGDS on Thrombin-treated Tumor Cells Effect of Thrombin-treated Tumor Cells on Their Adhesion to Fibronectin and von Willebrand Factor Fig. 7 demonstrates the inhibitory effect of 10E5 (P = 0.02) and RGDS (P = 0.002) on thrombin-treated tumor cells prior Fig. 3 demonstrates the enhanced adhesion effect of throm to their incubation with platelets, suggesting the presence of a bin-treated tumor cells on fibronectin (P < 0.001) and von "GPIIb-GPIIIa-like" integrin on thrombin-activated tumor Willebrand factor-coated plates (P = 0.002), adhesive ligands cells. which bind to platelet GPIIb-GPIIIa. Effect of Thrombin-treated Tumor Cells on in Vivo Metastasis Kinetics of Thrombin Preincubation of Tumor Cells on Their Adhesion to Platelets Table 2 demonstrates the effect of thrombin-treated tumor cells on experimental pulmonary metastasis. Four different Fig. 4 demonstrates a time course for the enhanced adhesion tumor cell lines increased metastasis (pulmonary tumor mass) of thrombin-activated tumor cells to platelets with the HM29 10- to 156-fold by increasing both their mean nodule number tumor cell line. Note the optimum effect at l h with HM29 and volume. Thus CT26 (2 experiments), B16F1, B16F10, and (>control at 60 min, P < 0.05; 60 min >15 min, P < 0.04; 60 B16a increased their mean nodule number 1.5-, 1.5-, 7.6-, 7.4-, min >30 min, P = 0.05; 60 min > 120 min, P < 0.03, one-tailed and 7.0-fold, respectively, as well as their mean nodule volume t test). 6.5-, 30-, 22-, 9.2-, and 5.4-fold, respectively. Hirudin did not Effect of Cycloheximide. Table 1 demonstrates that de novo inhibit the effect of thrombin-treated tumor cells on enhanced protein synthesis is not responsible for the thrombin effect. metastasis, indicating that active thrombin was not required on With these experimental conditions, cycloheximide at 5 and 10 the tumor cell surface (P > 0.1). /¿g/mlinhibited protein synthesis 88 and 92%, respectively. Effect of Thrombin-treated Tumor Cells on in Vivo Growth Specificity of Thrombin Reaction with Tumor Cells and Lack of Requirement of Active Thrombin on Tumor Cell Surface Once Because of the markedly 10- to 156-fold enhancement effect Tumor Cells Have Been Exposed to Thrombin of thrombin-treated tumor cells on in vivo metastasis compared to its 2- to 3-fold effect in vitro on tumor-platelet adhesion, we Fig. 5 demonstrates the specificity of the thrombin effect on postulated that thrombin may be affecting tumor growth as tumor cells as well as the lack of requirement of active thrombin well. Accordingly, 1 x IO5 CT26 cells, treated and untreated on the tumor surface, once the tumor cells have been treated with 1 unit/ml thrombin for l h at 37°C,followed by washing, with thrombin. Note inhibition of the thrombin effect if throm were injected s.c. into the flank of BALB/c mice, and their
30 CO CO
20 Fig. 3. Effect of thrombin-treated tumor cells on adhesion to fibronectin and von Vvïl- lebrand factor-coated plates. Optimum con LU centrations of fibronectin (0.6 ¿ig/wcll)andvon Willebrand factor (1.2 fig/well) were applied en 10 to microtiter plates overnight at 4"C. Plates _ were then blocked with PBS-BSA-Ca-Mg prior uu LU to the addition of I x IO5 HM29 cells previ U O ously treated with buffer or 0.5 unit/ml throm bin for l h at 37'C. Tumor cells were incubated with ligand-coated plates for I h at 37'C prior von WILLEBRAND FACTOR 1.2ng/WELL FIBRONECTIN 0.6pig/WELL to washing and enumeration of adherent tumor (n=4) (n=6) cells as in Fig. I.
CU CONTROL TUMOR CELLS TUMOR CELLS + THROMBIN 500mU/ml 60'37 °C
3269
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research. THROMBIN-TREATED TUMOR CELLS ENHANCE METASTASIS
ment of pulmonary metastasis in vivo. In a previous study (5) we noted the ability of thrombin to activate platelets, rendering | [CONTROL their surface 2- to 5-fold more adherent to 6 different tumor IÃœTHR15MIN cell lines. Thrombin injected in vivo enhanced pulmonary me tastasis 4- to 413-fold with 2 different tumor cell lines. The •¿ ITHR 30 MIN mechanism for the enhanced in vitro adhesiveness required active thrombin on the platelet surface, and was shown to [¿¿]THR120MIN require the interaction of tumor cells to the platelet integrin GPIIb-GPIIIa, presumably via fibronectin and von Willebrand [THR SOOmUml) factor bridging. The direct adhesive affect of thrombin on tumor cells appears to be via both a similar as well as dissimilar mechanism as its Fig. 4. Effect of time on thrombin pretreatment of tumor cells prior to their action on platelets. It is similar with respect to the requirement incubation with adherent platelets. SK-Mel HM29 tumor cells (1 x 10s) were of the platelet integrin GPIIb-GPIIIa for binding to thrombin- preincubated with buffer or 0.5 unit/ml thrombin for various time intervals prior to their incubation with adherent platelets on microtiter plates for l h at 37'C, activated tumor cells (i.e., inhibition of enhanced adhesion by as in Fig. 1. Except for the 120-min thrombin preincubation, thrombin-treated preincubation of platelets with monoclonal antibody 10E5 but tumor cells were first incubated with buffer for various intervals of time, prior to the addition of thrombin at various time intervals, so that all tumor cells were not 6D1; inhibition by the tetrapeptide RGDS but not GRGES; incubated with or without thrombin for the same period of time (120 min). inhibition by the disintegrin, albolabrin), as well as the enhanced Adherent tumor cells were enumerated as in Fig. 1. Data are from three experi adhesion of thrombin-activated tumor cells to the adhesive ments performed in triplicate. ligands fibronectin and von Willebrand factor. Indeed, throm bin-activated tumor cells appear to have a GPIIb-GPIIIa-like Table 1 Effect of cyclohexamide on binding of thrombin-treated HM29 tumor cells to naive platelets receptor on their surface, since their binding to platelets is HM29 cells (1 x 10') were incubated with varying concentrations of cyclo- blocked by preincubating tumor cells with monoclonal antibody hexamide for 10 min prior to the addition of buffer or 1 unit/ml thrombin for 1 h at 37'C. Cells were washed and then applied to a lawn of adherent platelets for 10E5 or the tetrapeptide RGDS. This observation is compatible l h at 37'C. Nonadherent cells were washed away and adherent cells were eluted with reports of other investigators who have noted GPIIb- with trypsin-EDTA and enumerated via phase microscopy. GPIIIa-like integrins on tumor cells [i.e., reactivity with mono Adherent cells x 10~3 clonal antibody l OE5 (12,13)], as well as other integrins capable of binding fibronectin, von Willebrand factor, vitronectin, lam- Cyclohexamide (ÃÃg/ml)0 buffer1.6 (1 unit/ml)5.0 increase3.1
2.5 1.43.9 4.3 3.12.7 PLATELET PRF.TRF.ATMF.NT 5.0 10.6 10.0Control 2.1Thrombin 6.2Fold 3.1 I I BUFFER ALONF. •¿ im 111 K ALONE X 1 J20-J VTA RGDS 50|i(/ml
^ GRGES SOng/ml ESSa M.. M, IOF.5 -Cl..,, ml JL I ICONTROL u SS MoAb 6DI «III.u ml u Fxl AI.ROI.ABRIS StiR/ml Hi THROMBIN Ill/ml CONTROL THROMBIN TREATED TIMOR CK1.I.S THROMBIN SOOmlVml 60' J7° DAPA luM Fig. 6. Effect of thrombin-treated tumor cells on adhesion to adherent platelets (TOT); preincubated with inhibitors of the platelet GPIlb-GPIIla integrin. Adherent THEN DAPA platelets were preincubated with 50 jig/ml of the RGDS tetrapeptide or the control peptide GRGES; 50 /jg/ml of monoclonal antibody 10E5 or the control TOÕT+DAPA) monoclonal antibody 6DI; or 5 »jg/mlofthe disintegrin albolabrin for I h prior PRE-MIXTURE to washing. HM29 tumor cells (1 x 10s), pretreated with 0.5 unit/ml thrombin for l h at 37°C,were then incubated with adherent platelets for l h at 37'C. Adherent tumor cells were enumerated as in Fig. 1. Data are from nine experi Fig. 5. Effect of the DAPA on thrombin-stimulated tumor cells (7"C) prior to ments (except for albolabrin, 4 experiments) performed in triplicate. their incubation with adherent platelets. HM29 cells (1 x IO5)were preincubated with either buffer, 1 unit/ml thrombin, or 1 JIM DAPA (specific thrombin inhibitor) for 1 h; or 1 unit/ml thrombin for l h (7T + 7"),followed by washing plus addition of 1 nM DAPA; or a mixture of DAPA + thrombin (7" + DAPA) 300 (preincubated for 5 min) prior to incubation with tumor cells for 1 h at 37"C. Preincubated tumor cells were then reacted with adherent platelets for 1 h at CD CONTROL 37"C prior to washing and enumeration of adherent tumor cells as in Fig. 1. Data M THROMBIN 500mU/ml are from 14 experiments performed in triplicate. ^ T+10E550ug/ml ESZ3 T+IgG 2a SOng/ml volume was measured at 21 days. Thrombin-treated tumor cells had a 2.1-fold increased in mean volume compared to sham- R^l T+RGDS 50ng/ml treated tumor cells. (6.4 ±1.3 versus 3.1 ±0.75, cm3, n = 9, P IÕÕÕÕÕÕ1T+GRGES50ug/ml = 0.02). Fig. 7. Effect of preincubation of thrombin-treated tumor cells with RGDS DISCUSSION and monoclonal antibody 10E5 on their adhesion to naive platelets. Tumor cells (1 x 10s) were incubated with buffer or 0.5 unit/ml thrombin for 1 h, followed These data indicate that thrombin can directly activate tumor by washing and addition of either buffer, 50 Mg/ml RGDS. GRGES (control peptide), monoclonal antibody 10E5 or IgG2a (isospecific control antibody) for cells to enhance their binding to adherent platelets, fibronectin, 1 h prior to their addition to adherent naive platelets for 1 h at 37°C.Data are and von Willebrand factor in vitro and enhance their develop from eight experiments performed in triplicate. 3270
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research. THROMBIN-TREATED TUMOR CELLS ENHANCE METASTASIS
Table 2 Effect of washed thrombin-treated tumor cells on pulmonary metastasis Fifty thousand viable CT26, B16F1, or B16F10 cells and 25,000 B16a cells were treated with or without 0.5-1 unit/ml for 60 min at 37'C. Control and thrombin- treated cells were then washed three times with PBS-Ca2*-Mg!*-0.1% BSA before i.v. injection of tumor cells. Pulmonary métastaseswereexamined macroscopically on day 21. The Tirsi /' value refers to mean nodule difference between control and thrombin-treated tumor cells; the second value refers to mean nodule volume difference. no. of nodule TumorCT26CT26BI6FIHIM nodules8.0 volume6.9 mass55.2 increase9.7M411567838P0.0080.045° ±0.8 ±1.6 95 ThrombinControl 11.9±2.12.2 44.9 ±14.91.1 5342.162.7
±0.5 ±0.5 5 Thrombin 3.2 ±1.0 19.6 ±7.9 0.040"0.003 5111212 ThrombinhirudinControl + 4.2 ±0.60.7 20.4 ±8.60.8 85.70.693.80.5
±0.3 ±0.4 ThrombinControl 5.3 ±1.30.5 17.7±6.30.1 0.0110.009
MlBiteN12 ±0.33.7 ±0.1 1220 ThrombinControl ±1.00.1 9.2 ±4.9Mean 34.0Total 0.0460.03* surface tumor area(jiiti2)165 area/section16.5
20GroupControlThrombinMean 0.7Mean 895Tumor 627Fold ' Paired one-tailed t test. ' Difference in total tumor area of control and thrombin-treated tumor cells on a slide section. inin, and collagen type IV (14-19). Using specific monoclonal induces a metastatic phenotype by enhancing tumor cell adhe antibodies, we have identified integrin receptors for fibronectin sion to platelets, subendothelial matrix (fibronectin, von Wil- (a3/31) and vitronectin (av/83) (5). lebrand factor, other RGD ligands, collagen), and as we have The mechanism is dissimilar with respect to the lack of recently noted, endothelial cells4; or that thrombin renders the requirement of active thrombin on the tumor cell surface as is tumor cell more oncogenic. the case with active thrombin on the platelet surface (5). We The in vivo enhancement of pulmonary métastasesinduced have demonstrated that thrombin binds to the tumor cell surface by i.v. thrombin (4- to 413-fold) (5) or thrombin-activated in a saturation-dependent manner." The absence of active tumor cells (10- to 156-fold) is considerably greater than the in thrombin on tumor cells rendered more adhesive implies a vitro 1- to 5-fold enhanced adhesion (5) of platelet-tumor cell secondary event on the tumor cell surface. This is supported by or adhesive ligand interaction. Indeed two tumor cell lines the presence of a lag period of approximately l h before responsive to thrombin in vivo (CT26 and B16F10) did not achieving the optimum thrombin effect. The secondary event enhance their in vitro adhesion to platelets or fibronectin fol does not require protein synthesis as it is not inhibited by lowing thrombin activation. These observations would suggest cycloheximide. We postulate the formation of a more adhesive that thrombin may be having an additional effect on tumor integrin either by an increase in density or change in confor cells, possibly unrelated to platelet or adhesive ligand interac mation of a GPIIb-GPIIIa-like integrin already present on the tion, at least with some tumor cell lines. This is supported by tumor cell surface or the development of a new adhesive recep the enhanced s.c. growth of thrombin-treated tumor cells versus tor, perhaps secondary to thrombin proteolysis. naive cells. We propose that thrombin may enhance the met The requirement of the platelet integrin GPIIb-GPIIIa as astatic phenotype by activating the oncogenic potential of tumor well as RGD ligands (fibronectin and von Willebrand factor) cells. for both the thrombin-activated platelet (5) as well as the A precedent for thrombin-induced enhanced oncogenesis is thrombin-activated tumor cell arm of the enhanced platelet- supported by the following observations. Thrombin can act as tumor cell interaction suggests that thrombin may be having a mitogenic agent (21-27). Thrombin stimulates cell division the same effect on both platelets and tumor cells (i.e., mobili in nonproliferating fibroblast cultures of chick and mouse em zation and/or conformational activation of GPIIb-GPIIIa and bryo, human diploid foreskin, and Chinese hamster lung in the GPIIb-GPIIIa-like integrins on platelets and tumor cells with absence of added serum or additional growth factors (21, 22, RGD ligand bridging). This is supported by our observation 24). Thrombin potentiates the response to fibroblast growth that enhanced adhesion of thrombin-activated platelets and factor of human vascular endothelium (23). Thrombin pro thrombin-activated tumor cells is not additive. Thus, we pro motes cell transformation in Balb 3T3/a31-l-13 cells (27). The pose that active thrombin on the platelet surface interacts with proliferative events initiated by thrombin require receptor bind a thrombin receptor on tumor cells to enhance tumor cell ing as well as proteolysis (25, 26, 28). integrin reactivity. This is supported by the well-described Thrombin generates some of the same signals induced by greater efficiency and reactivity of procoagulants on the platelet oncogenes. Thrombin activates phosphoinositol production via membrane surface (20), as well as the greater thrombin sensi a specific phospholipase C through a pertussis toxin-sensitive tivity of thrombin-activated platelet adhesion to naive tumor GTP-binding regulatory G-protein. Several laboratories have cells [10-100 milliunit reactivity (5)] than thrombin-activated recently established that activated ras oncogenes, transfected tumor cell adhesion to naive platelets (500-1000 milliunit into NIH-3T3 cells, can induce pulmonary metastasis of the reactivity). nonmetastatic parental line (29-32). The ras gene encodes a M, The 10- to 156-fold increase in pulmonary metastasis of 4 21,000 protein on the inner surface of the plasma membrane different thrombin-activated tumor cells is particularly impres with structural similarity to the G-proteins. Thus, it is likely sive. This indicates that thrombin activation of tumor cells that the ras protein initiates aberrant signals and/or interferes with normal signal transduction across the cell membrane, ' A. Klepfish and S. Karpatkin. unpublished data. presumably by binding to GTP. A recent report has demon- 3271
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research. THROMBIN-TREATED TUMOR CELLS ENHANCE METASTASIS
that thrombin induced c-fos mRNA expression in cul glycoprotein Ilb-IIIa complex. Blood. 74: 658-663, 1989. 14. Plantefaber. L. C.. and Hynes, R. E. Changes in integrin receptors on tured human umbilical vein endothelial cells (33). c-fos and c- oncogenically transformed cells. Cell. 49: 393-402. 1989. 15. Cheresh. D. A., and Spiro, R. C. Biosynthetic and functional properties of jun can form a heterodimer complex corresponding to the an Arg-Gly-Asp-directed receptor involved in human melanoma cell attach transcription factor API which regulates transcription of the ment to vitronectin. fibrinogcn and von Willebrand factor. J. Biol. Chern.. ras gene through an API site (34). Thus thrombin could be 262: 17703-17711, 1987. 16. Rieber. M..Gross, A..and Rieber. M. S. Relationship of a A/, 14D fibronectin activating raÃthrough the induction of c-fos and other possible receptor and other adhesion-related glycoproteins to tumor cell-cell interac oncogenes. tion. Cancer Res.. 47: 5127-5131, 1987. In summary, thrombin directly activates four different tumor 17. Cheresh, D. A., Smith. J. W., Cooper, H. M., and Quaranta, V. A novel vitronectin receptor integrin (a.fix) is responsible for distinct adhesive prop cell lines to induce a markedly enhanced metastatic phenotype erties of carcinoma cells. Cell, 57: 59-69, 1989. 18. Kramer. R. H.. McDonald, K. A., Crowley, E., Ramos. D. M.. and Damsky, in vivo. The precise mechanism(s) for this induction remains to C. H. Melanoma cell adhesion to basement membrane mediated by integrin- be established. In at least 5 of 7 tumor cell lines it may be related complexes. Cancer Res.. 49: 393-402, 1989. related to the in vitro enhancement of adhesion of tumor cells 19. Akiyama, S. K., Larjava, H., and Yamada. K. M. Differences in the biosyn thesis and localization of the fibronectin receptor in normal and transformed to platelets, subendothelial matrix, or endothelial cells; in other culture human cells. Cancer Res., 50: 1601-1607. 1990. cell lines it may be acting as a growth factor. 20. Mann, K. G., Nesheim, M. E., Church. W. R.. Haley P., and Krishnaswamy. S. Review: surface-dependent reactions of the vitamin K-dependent enzyme Extrapolating from these data, we propose that low grade complexes. Blood, 76: 1-16, 1990. thrombin generation may foster metastatic progression in some 21. Chen. L. B., and Buchanan, J. M. Mitogenic activity of blood components. patients with some tumors. Indeed, many tumor cells activate I. Thrombin and prothrombin. Proc. Nati. Acad. Sci. USA, 72: 131-135. 1975. the coagulation system with generation of thrombin (35). Low- 22. Carney, D. H., Glenn, K. C.. and Cunningham, D. D. Conditions which grade intravascular coagulation has been observed in most affect initiation of animal cell division by trypsin and thrombin. J. Cell patients with solid tumors (36-39). One study reported elevated Physiol.. 95: 13-22, 1978. 23. Gospodarowicz, D., Brown, K. D.. Birdwell. C. R.. and Zetter, B. R. Control fibrinopeptide A levels in 60% of cancer patients at time of of proliferation of human vascular endothelial cells. Characterization of the response of human umbilical vein endothelial cells to fibroblast growth factor, presentation with increasing levels noted with progression of epidermal growth factor, and thrombin. J. Cell. Biol.. 77: 774-788. 1973. disease. Persistent elevation was associated with a poor prog 24. Simmer, R. L., Baker, J. B.. and Cunningham. D. D. J. Direct linkage of nosis (39). Thus, it is proposed that anti-thrombin and anti- thrombin to its cell surface receptors in different cell types. Supramol. Struct.. 12: 245-257. 1979. platelet adhesive agents may be helpful in the prevention of 25. Glenn. K. C.. Carney. D. H.. Fenton, J. W.. and Cunningham. D. D. Thrombin active site regions required for fibroblast receptor binding and tumor metastasis. initiation of cell division. J. Biol. Chem., 255: 6609-6616, 1980. 26. Carney, D. H., Stiernberg, J.. and Fenton, J. W. Initiation of proliferative events by human «-thrombin requires both receptor binding and enzymic REFERENCES activity. J. Cellular Biochem.. 26: 181-195. 1984. 27. Morris, D. L., Ward, J. B.. Jr., and Carney. D. H. Thrombin promotes cell 1. Gasie, G. J.. Gasic. T. B.. and Stewart, C. C. Antimetastatic effects associated transformation in Balb 3T3/A 31-1-13 cells. Carcinogenesis(Lond.), 13:61- with platelet reduction. Proc. Nail. Acad. Sci. USA. 61: 46-52. 1968. 73, 1992. 2. Pcarlstein. E., Ambrogio, C, and Karpatkin. S. Effect of anti-platelet anti 28. Vu, T-K. H., Hung, D. T.. Wheaton, V. I., and Coughlin. S. R. Molecular body on the development of pulmonary métastasesfollowing injection of cloning of a functional thrombin receptor reveals a novel proteolytic mech CT26 colon adenocarcinoma. Lewis lung carcinoma, and B16 amelanotic anism of receptor activation. Cell, 64: 1057-1068. 1991. melanoma tumor cells into mice. Cancer Res., 44: 3884-3887, 1984. 29. Greenberg, A. H.. Egan. S. E., and Wright. J. A. Oncogenes and metastatic 3. Karpatkin. S., Pearlstein. E., Ambrogio, C., and Coller, B. S. Role of adhesive progression. Invasion Metastasis, 9: 360-378. 1989. proteins in platelet tumor interactions in vitro and metastasis formation in 30. Bernstein. S. C.. and Weinberg, R. A. Expression of the metastatic phenotype vivo. J. Clin. Invest..«/: 1012-1019. 1988. in cells transfected with human metastatic tumor DNA. Proc. Nati. Acad. 4. Humphries. M. J., Olden, K., and Yamada, K. M. A synthetic peptide from Sci. USA, 82: 1726-1730, 1985. fibronectin inhibits experimental metastasis of murine melanoma cells. Sci 31. Greig, R. G.. Koestler. T. P.. Trainer, D. C., Corwin, S. P.. Miles. L.. Kine. ence (Washington DC). ¿W:467-470, 1986. T., Sweet. R.. Yokoyama. S.. and Poste, G. Tumorigenic and metastatic 5. Nierodzik, M. L., Plotkin, A., Kajumo, F., and Karpatkin, S. Thrombin properties of normal and ius-transfected NIH/3T3 cells. Proc. Nati. Acad. stimulates tumor-platelet adhesion in vitro and metastasis in vivo. J. Clin. Sei. USA. Ä2:3698-3701, 1985. Invest., 87: 229-236, 1991. 32. Thorgeirsson. U. P., Turpeenniemi-Hujanen, T.. Williams. J. E., Westin, E. 6. Coller, B. S., Peerschke, E. I., Scuddler, L. E., and Sullivan. C. A. A murine H., Heilman, C. A., Talmadge, J. E.. and Liotta. L. A. NIH/3T3 cells transfected with human tumor DNA containing activated ras oncogenes monoclonal antibody that completely blocks the binding of fibrinogen to express the metastatic phenotype in nude mice. Mol. Cell. Biol., 5: 259-262, platelets, produces a thrombasthenic-like state in normal platelets and binds to glycoproteins lib and/or Illa. J. Clin. Invest.. 72: 325-328. 1983. 1985. 33. Lampugnani, M. G., Lolotta, F., Polentarutti. N.. Pedenove. M.. Mantovani. 7. Coller, B. S., Peerschke, E. I., Scudder. L. E., and Sullivan, C. A. Studies A., and Dejana. E. Thrombin induces c-fos expression in cultured human with a murine monoclonal antibody that abolishes ristocetin-induced binding endothelial cells by a Ca** dependent mechanism. Blood, 76: 1173-1180. of von Willebrand factor to platelets: additional evidence in support of GPlb 1990. as a platelet receptor for von Willebrand factor. Blood. 61: 99-110, 1983. 34. Rauscher. F. J., Cohen. D. R.. Curran. T.. et al. yôs-associatedprotein p39 is 8. Thorell, L., and Blomback, B. Purification of the factor Vili complex. the product of they'un-oncogene. Science (Washington DC). 240:1010-1016. Thromb. Res., 35: 431-450. 1984. 9. Soszka, T., Knudsen. K. A., Beviglia, L.. Rossi, C., Poggi. A., and Niewig- 1988. 35. Honn. K. V.. Sloane. B. F.. and Cavanaugh, P. G. Role of the coagulation rowski. S. Inhibition of murine melanoma cell-matrix adhesion and experi system in tumor-cell-induced platelet aggregation and metastasis. Haemosta- mental metastasis by Albolabrin. an RGD-containing peptide isolated from sis, 8: 37-46. 1988. the venom of Trimeresurus albolabris. Exp. Cell. Res.. 196: 6-12. 1991. 36. Yoda. Y., and Abe, T. Fibrinopeptide A (FPA) level and fibrinogen kinetics 10. Nesheim, M. E., Prendergast. F. G., and Mann. K. G. Interactions of a in patients with malignant disease. Thromb. Haemostasis. 46: 706-709, fluorescent active-site-directed inhibitor of thrombin: dansylarginine-AI-(3- 1981. ethyl-l,5-pentanediyl)amide. Biochemistry, 18: 996-1003, 1979. 37. Merskey. C., Johnson, A. J., Harris, J. U., Wang. M. T., and Swain, S. 11. Karpatkin, S., Ambrogio, C., and Pearlstein, E. Lack of effect of in vivo Isolation of fibrinogen-fibrin related antigen from human plasma by immune- prostacyclin on the development of pulmonary métastasesinmice following affinity chromatography: its characterization in normal subjects and in defi- intravenous injection of CT26 colon carcinoma. Lewis lung carcinoma, or brinating patients with abruptio placentae and disseminated cancer. Br. J. B16 amelanotic melanoma cells. Cancer Res., 44: 3880-3883, 1984. Haematol.. 44: 655-670, 1980. 12. Grossi, 1. M.. Hatfield. J. S.. Fitzgerald, L. A., Newcombe, M., Taylor. J. 38. Peuscher, F. W., Cleton, F. J.. Armstrong. L.. Stoepman-van Dalen, E. A., D.. and Honn, K. V. Role of tumor cell glycoproteins immunologically van Mourik, J. A., and van Aken, W. G. Significance of plasma fibrinopeptide related to glycoproteins Ib and Ilb/IIIa in tumor cell-platelet and tumor cell- A (FPA) in patients with malignancy. J. Lab. Clin. Med.. 96: 5-14, 1980. matrix interactions. FASEB J.. 2: 2385-2395, 1988. 39. Rickles, F. R.. Edwards, R. C., Barb, C.. and Cronlund. M. Abnormalities of 13. Boukerche, H., Berthier-Vergnes, O., Tabone. E., Doré,J.F., Leung, L. K., blood coagulation in patients with cancer: fibrinopeptide A generation and and McGregor, J. L. Platelet-melanoma cell interaction is mediated by the tumor growth. Cancer (Phila.). 5/: 301-307, 1983.
3272
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research. Effect of Thrombin Treatment of Tumor Cells on Adhesion of Tumor Cells to Platelets in Vitro and Tumor Metastasis in Vivo
Mary Lynn R. Nierodzik, Francis Kajumo and Simon Karpatkin
Cancer Res 1992;52:3267-3272.
Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/52/12/3267
E-mail alerts Sign up to receive free email-alerts related to this article or journal.
Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].
Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/52/12/3267. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.
Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1992 American Association for Cancer Research.