Inter-alpha-inhibitor (IαI) is an abundant plasma protein whose physiological function is only now beginning to be revealed.1, 2 It consists of 3 polypeptides: 2 heavy chains and 1 light chain called bikunin.1 Bikunin, which has antiproteolytic activity, carries a chondroitin sulphate chain to which the heavy chains are covalently linked.1 Research from various directions has clarified many aspects of the biological function of proteins of the IαI family and provides a new vista on their interesting structure-function relationships, especially on their roles in inflammation2 and cancer.3 A growing amount of evidence demonstrates that IαI is an anti-inflammatory agent: in particular, circulating IαI levels were lower in patients with severe sepsis than in healthy volunteers.4 Furthermore, human IαI may be a useful predictive marker and potential therapeutic agent in sepsis.5, 6
Bikunin exhibits anti-metastatic functions in animals7, 8 and humans.3 The broad spectrum of the biological functions of bikunin suggests the existence of multiple molecular targets that mediate diverse responses to the compounds in cells.9, 10, 11 Identifying the target molecules can facilitate the design of better therapeutic agents for protection against certain conditions associated with cancer and inflammatory diseases. Current understanding of the mechanism of action by bikunin comes mostly from studies on the suppression of several signaling cascades.12 In cancers, administration of bikunin may block tumor cell invasion by a direct inhibition of tumor cell-associated plasmin activity as well as by inhibiting uPA expression at the gene and protein levels, possibly through suppression of the MAP kinase signaling cascade.3, 12 Upon exposure to bikunin, CD44 associates with an as-yet-unidentified receptor for bikunin and modulates the transcription of target genes, including uPA and uPAR, through MAP kinase- and PI3 kinase-dependent transcription.10 Thus, suppression of signaling activation can account for prevention by bikunin against cancer. Treatment of cancer patients with bikunin may be beneficial in the adjuvant setting to delay the onset of metastasis development and/or in combination with cytotoxic agents to improve treatment efficacy in patients with advanced ovarian cancer.3
To explore the role of endogenous bikunin, we used bikunin knockout (Bik−/−) mice.13, 14 In our study, we investigated whether the increased sensitivity of Bik−/− mice to lung metastasis in vivo is due to a lack of circulating proteins of the IαI family, especially bikunin, in plasma. Our findings provide new insights into a mechanism of protection against metastasis by endogenous bikunin.
Mice deficient in bikunin and their genotyping have been described elsewhere.13, 14 Mice were obtained from Mochida Pharmaceutical Co., Ltd. (Gotenba, Shizuoka, Japan). EC clones containing the correct targeting event were microinjected into the blastocoel cavity of embryos of C57BL/6 mice at the blastocyst stage. Injected embryos were then transferred into the uterus of pseudopregnant female ICR mice. Chimeric mice were mated with C57BL/6 female mice to produce heterozygous mice. Heterozygous mice were maintained by crossing them to C57BL/6 mice and intercrossed to obtain homozygous bikunin-deficient mice. To determine the genotype of the pups resulting from these crosses, genomic DNA was extracted from the tail tips of 3-week-old mice and analyzed by Southern blotting. The Bik−/− mice have a normal plasma level of alpha 1-microglobulin (α1M) protein. As predicted, IαI (230 kDa) and PαI (130 kDa) were absent in Bik−/− mice. Bikunin is essential for the biosynthesis of IαI and PαI. The Bik+/+ mice showed normal plasma levels of Bikunin, ITI and PαI. The Bik+/− mice, with less bikunin transcripts, showed normal plasma levels of ITI and PαI.13, 14). Bikunin deficiency is not neonatal lethal in mice. Bik−/− mice survived at the expected Mendelian ratio. Under specific pathogen-free conditions, Bik−/− mice remain healthy and survive to adulthood.13, 14 There was no difference in blood cell composition between Bik+/+ and Bik−/− mice. During ovulation, the cumulus oophorus, an investing structure unique to the oocyte of higher mammals, had a defect in forming the extracellular hyaluronan-rich matrix during expansion. The ovulated oocytes were completely devoid of matrix and were unfertilized, leading to severe female infertility.13, 14 Necropsy and microscopic examination of major tissues revealed no significant pathology in Bik−/− mice.
Preparation of bikunin-depleted fetal bovine serum
Fetal bovine serum (FBS) preincubated with nonimmune IgG or rabbit anti-bikunin antibody15 at 4°C for 2 hr was incubated with protein G-Sepharose with gentle mixing. Incubation was continued for another 2 hr at 4°C with gentle mixing. The samples were centrifuged at 4°C for 1 min at 1,200 rpm in a Microfuge, each supernatant was recovered and again incubated with proteinG-Sepharose with gentle mixing for additional 2 hr at 4°C. Then, each supernatant was used as bikunin-depleted FBS or control FBS.
Two clonal sublines of the murine Lewis lung carcinoma 3LL cells showed distinct patterns of organ-selective metastasis. Subline M-27 is highly metastatic to the lung and does not form liver metastases, while subline H-59 is highly metastatic to lymph nodes and liver, but not to lung. H-59 cells secrete high levels of a 72 kDa gelatinase, while M-27 cells produced low levels of this gelatinase and significantly higher levels of uPA.16 In our study, subline M-27 cells were used for further studies. It has been established that G-CSF can stimulate the invasive capacity of lung cancer cells17 and that G-CSF markedly enhanced upregulation of uPA in 3LL cells.15 The 3LL (subline M-27) cells were cultured and stimulated for different periods of time in RPMI 1640 medium containing 10% FBS or bikunin-depleted FBS as described.15 Cell activation was performed in complete medium containing serum or bikunin-depleted serum and G-CSF (10 ng/ml; CosmoBio, Tokyo, Japan) in the absence or presence of bikunin at the concentrations indicated in the text for a different period of time at 37°C in 5% CO2. Cells were counted using a hemocytometer, and viability was assessed by trypan blue staining.
Cells were trypsinized and counted, and 1 × 105 cells were placed in upper migration chambers,15 which were then rested in wells containing 10% FBS or bikunin-depleted FBS in the both chambers. Fibroblast-conditioned media were placed in the lower compartment as a source of chemoattractants. Incubation was carried out for 24 hr in a humidified CO2 chamber. Nonmigratory cells in the upper chamber were removed by scraping, and migrated cells on the lower surface were fixed in methanol and stained with hematoxylin. The number of migrated cells were then counted from a total of 9 regions of the filter and calculated as numbers/field. Cell attachment assay and the chemotactic assay were conducted as described previously.8
In vivo lung metastasis assay
Bik+/+ and Bik−/− mice matched for gender (female) and age (7–8 weeks) were used in the following spontaneous metastatic experiments. Bik−/− mice were fed ad libitum with regular chow. In an in vivo lung metastasis experiment, 3LL cells (106 cells/0.05 ml of PBS per mouse) were injected subcutaneously into the abdominal wall of Bik+/+ and Bik−/− mice as described.15 Two, 3 and 4 weeks later, the animals were sacrificed, and the lungs were fixed in formalin. Lung sections were processed, stained with hematoxylin and eosin, and observed for tumor cell colony formation. Animal experiments in the present study were performed in compliance with the guidelines of the Institute for Laboratory Animal Research, Hamamatsu University School of Medicine.
Western blot analysis
After histopathological procedures, the remaining tumor biopsy specimens and control normal tissues were immediately frozen in liquid nitrogen after surgical removal and stored at −80°C. The control tissues included normal abdominal wall and normal lung tissues. For the Western blot analysis of uPA, PAI-1 (American Diagnostica, Greenwich, CT), β-actin (Cosmobio), phosphorylated ERK1/2, or ERK1/2 (Calbiochem, San Diego, CA), whole cell lysates were prepared as described.15 Proteins (50 μg) were size-separated in gels of 10% SDS-polyacrylamide gel electrophoresis. Gels were blotted onto a polyvinylidene fluorescein membrane (Amersham Bioscience, Tokyo) and processed as recommended by the supplier of the antibodies: rabbit polyclonal antibody specific for phospho-ERK1/2 as well as anti-uPA and anti-PAI-1 antibodies. As a loading control, the filter was probed with a rabbit polyclonal antibody to detect total ERK1/2 and β-actin. Proteins recognized by the antibodies were revealed following the ECL technique (Amersham Bioscience, Tokyo). Autoradiographs were quantified by laser densitometry, and several time expositions were analyzed to ensure the linearity of the band intensities.
Results are expressed as the mean ± S.D. of the indicated number of experiments. The n value is indicated for each experiment. Statistical differences in uPA and PAI-1 production were determined using a 2-tailed Student's t-test, and statistical significance was accepted when p < 0.05. Statistical differences in Kaplan-Meier survival curves among the groups of mice were analyzed by log rank test. All statistical analysis was performed using StatView for Macintosh.
Bikunin deficiency enhances lung metastasis of 3LL cells
The 3LL cells, lacking bikunin expression, are highly metastatic.15 To investigate the role of circulating proteins of the IαI family in metastasis, 3LL cells were injected subcutaneously into the abdominal wall of Bik+/+ and Bik−/− mice and monitored for tumor growth in the primary site and metastasis to the lungs. Three and 4 weeks after injection, 3LL cells were found to be metastatic to the lung and formed large tumor clusters in Bik+/+ mice, whereas 3LL cells had markedly enhanced numbers and size of tumor colonies in Bik−/− mice, indicating that circulating proteins of the IαI family may inhibit lung metastasis (Fig. 1). In contrast, there were no significant differences in primary 3LL tumor growth between Bik+/+ and Bik−/− mice (data not shown).
Increased susceptibility of Bik−/− mice to lung metastasis-induced lethality
The relative contributions of endogenous proteins of the IαI family to the increased susceptibility of Bik−/− mice to lethality were assessed by survival. At 34 days, survival rate was 0% for Bik−/− mice compared to 100% for Bik+/+ mice (Fig. 2). Bik−/− and Bik+/+ mice induced 50% lethality by 32 days and 38 days after tumor inoculation, respectively. Bik−/− mice were found to exhibit significantly higher mortality than Bik+/+ controls after s.c. tumor inoculation. These data suggest that Bik−/− mice would be more susceptible to lung metastasis than Bik+/+ mice. Therefore, we speculate that deficiency of proteins of the IαI family increases the sensitivity of mice to death.
Bikunin protects against lung metastasis-induced lethality
It has previously been shown that bikunin inhibits 3LL lung metastasis in vivo.7). Therefore, we were interested in investigating whether bikunin, but not IαI, can protect against lung metastasis-induced lethality in vivo. No treatments with bikunin resulted in death of 100% of the Bik+/+ mice within 45 days after inoculation (Fig. 2). In contrast, 0/10 and 4/10 Bik+/+ mice were rescued by the i.p. administration of bikunin (5 [p=0.521] and 50 mg/kg [p=0.028]), respectively. A higher dose of bikunin reduced lung metastasis-induced death in Bik+/+ mice. This significant difference in mortality demonstrates that exogenous bikunin protects against lung metastasis-induced death.
We also investigated whether administration of bikunin would cause a significant reduction of the lung metastasis-induced lethality in Bik−/− mice (Fig. 2). The susceptibility of Bik−/− mice was almost comparable with that of Bik+/+, when Bik−/− mice were pretreated with the i.p. injection of bikunin (5 [p=0.791] or 50 mg/kg [p=0.041]). These results suggest a critical role of exogenous bikunin in protecting mice from lung metastasis-induced death.
Elevation of uPA and PAI-1 expression in primary and lung metastatic site of Bik+/+ and Bik−/− mice
Metastasis is regulated by pathways that involve uPA expression resulting in ECM degradation. To confirm whether this increased death is the result of increased uPA-dependent metastasis, we investigated possible differences in uPA levels by Western blot in primary and lung metastatic sites between Bik−/− and Bik+/+ mice. As shown in Figure 3, all of the primary and lung metastatic tumor samples contained detectable levels of uPA and PAI-1. Tissue uPA levels rose in the primary (lanes 1 and 2) and lung metastatic sites (lanes 5 and 6), but to a significantly higher level in Bik−/− mice vs. Bik+/+ mice (lane 1 vs. lane 2; and lane 5 vs. lane 6) (p<0.05). In a parallel experiment, tissue PAI-1 was also measured in the primary (lanes 1 and 2) and metastatic tumors (lanes 5 and 6) of Bik+/+ (lanes 2 and 6) and Bik−/− mice (lanes 1 and 5). The levels of PAI-1 was also significantly higher in Bik−/− mice than in Bik+/+ mice (lane 1 vs. lane 2; and lane 5 vs. lane 6) (p<0.05). These findings suggested that bikunin deficiency results in upregulation of the production of various invasion-related compounds, including uPA and PAI-1, which play crucial roles in the pathogenesis of metastasis.16 The control tissues exhibited markedly decreased levels of uPA and PAI-1 as compared to tumor lesions. Only a barely detectable uPA and PAI-1 signals resulted from 4 of the control specimens [abdominal wall (lanes 3 and 4) and normal lung tissues (lanes 7 and 8)]. The remaining 8 control specimens did not exhibit any uPA and PAI-1. Furthermore, there were no differences in the uPA and PAI-1 levels of normal tissues between Bik−/− mice and Bik+/+ mice (lanes 3 vs. lane 4; and lane 7 vs. lane 8).
uPA and PAI-1 levels in lung metastatic tumors in Bik+/+ and Bik−/− mice treated with G-CSF and/or bikunin
Immunoblot analysis revealed that bikunin12 and another Kunitz-type trypsin inhibitor (KTI)15 specifically reduced expression of uPA protein as well as phosphorylation of MAP kinase in some cells stimulated with agonists (G-CSF for 3LL cells or TGF-beta1 for ovarian cancer HRA cells). To determine the effect of exogenous bikunin on G-CSF-induced uPA and PAI-1 levels in lung metastatic tumors, Bik−/− mice were injected with G-CSF (10 μg/kg) in the absence or presence of bikunin (i.p.; 50 mg/kg), and the uPA and PAI-1 levels were determined 12 hr after G-CSF injection. The i.p. injection of bikunin significantly reduced G-CSF-induced upregulation of uPA and PAI-1 expression (Fig. 4, lanes 1–4). Administration of 50 mg/kg bikunin together with 10 μg/kg G-CSF largely inhibited the G-CSF-induced expression of uPA and PAI-1 for 32% and 40%, respectively. Furthermore, there were no significant differences in uPA and PAI-1 expression between the primary tumors and lung metastatic tumors (data not shown). We also showed that exogenous bikunin inhibits G-CSF-induced up-regulation of uPA and PAI-1 expression in Bik+/+ mice (Fig. 4, lanes 5–8). In contrast, only barely detectable uPA and PAI-1 signals resulted from 6 of the control specimens. The remaining 6 control specimens did not exhibit any uPA and PAI-1 expression (not shown). Inhibition of the G-CSF-induced tissue levels of these factors by concomitant administration of bikunin is therefore expected to render rodents less susceptible to lung metastasis. In a separate experiment, we confirmed that bikunin did not directly affect G-CSF stimulation, when cells were challenged with G-CSF preincubated with bikunin for 2 hr (unpublished data).
Effect of bikunin on invasion of 3LL Cells in response to G-CSF
3LL cells were also used in an in vitro assay for invasion using Matrigel-coated chambers. Nontreated FBS (Fig. 5, upper panel) and FBS pretreated with an irrelevant antibody (data not shown) were used as controls. The cells cultured in nontreated FBS or FBS pretreated with an irrelevant antibody, which contains endogenous proteins of the IαI family, demonstrated considerably reduced invasiveness across Matrigel compared to cells cultured in bikunin-depleted FBS (Fig. 5). We confirmed that, with experiments using bikunin-depleted FBS, additional factors were not removed in a nonspecific fashion during the bikunin depletion. In addition, we found that G-CSF (10 ng/ml) markedly enhanced 3LL cell invasion. The in vitro invasion assay was also performed in the presence of exogenous bikunin. Bikunin can reduce G-CSF-induced 3LL cell invasion, indicating inhibition of invasion by endogenous bikunin present in the FBS.
The cell chemotactic response was also tested to determine whether the inhibitory effect of bikunin on cell invasion of basement membranes was due to an inhibition of chemotaxis. The cells tested showed good chemotactic migration in the presence of bikunin (data not shown). The lack of negative effects on chemotaxis is consistent with an absence of toxicity of bikunin. In addition, we examined the effects of bikunin used in our study on cell attachment. No inhibition of attachment to Matrigel was seen with bikunin (data not shown).
G-CSF-induced ERK1/2 activation is enhanced in 3LL cells cultured in bikunin-depleted FBS
3LL cell invasion is regulated by pathways that involve uPA expression resulting in ECM degradation.15. These cellular activities are controlled predominantly by MAP kinase such as MEK and ERK.12 The cells were initially stimulated with G-CSF (10 ng/ml), and the lysates were immunoblotted with anti-phospho-ERK1/2 and anti-ERK1/2 antibodies. This resulted in the rapid (within 5 min) phosphorylation of ERK1/2, which peaked at 15 min, and was still apparent after 60 min (data not shown). To investigate whether functional changes in cell invasion by endogenous bikunin in lung cancer cells may be related to the activity of ERK1/2, 3LL cells cultured in bikunin-containing FBS or bikunin-depleted FBS were stimulated with G-CSF for 15 min and were subjected to an assay for ERK1/2 activation (Fig. 6). The results showed that even though equivalent amounts of total ERK1/2 protein were present, cells cultured in bikunin-containing FBS exhibited reduced expression of phosphorylated ERK1/2 compared to cells cultured in the bikunin-depleted FBS, when cells were stimulated with G-CSF (Fig. 6, lane 1 vs. lane 3). These results indicate that endogenous bikunin is able to prevent ERK1/2 activation.
To further investigate whether G-CSF-mediated signaling is regulated by exogenous bikunin, we examined the effect of bikunin on G-CSF-stimulated signaling activation. The cells preincubated with bikunin-containing FBS (left panel) or bikunin-depleted FBS (right panel) were treated with bikunin (0.4, 2 and 10 μM) for 1 hr, stimulated with G-CSF for 15 min and then the phosphorylated bands were analyzed by Western blot (Fig. 6). The ERK1/2 phosphorylation was down-regulated by bikunin in a dose-dependent manner; 10 μM bikunin was sufficient to suppress activation of ERK1/2. Thus, circulating bikunin may play an important role in suppression of G-CSF-induced activation of ERK1/2 signaling pathway. These data allow us to speculate that the increased sensitivity of Bik−/− mice to lung metastasis-induced death in vivo is due to a lack of circulating bikunin in plasma.
Bikunin is thought to play a vital role for the controlled extracellular proteolysis during tumor invasion and metastasis.3 In our study, we have studied the effect of bikunin deficiency in a transgenic mouse model of metastasizing lung cancer. Bikunin-deficient mice exhibited significantly enhanced spontaneous lung metastasis and shorter survival after 3LL tumor inoculation and exogenous bikunin overcomes this phenomenon, suggesting a potential involvement of the bikunin pathway in the suppression of tumor metastasis. We showed that the 3LL tumors inoculated in the bikunin-deficient mice produce significantly more uPA and PAI-1 synthesis than those inoculated in Bik+/+ mice. We then investigated the mechanism leading to the excessive elevation of uPA and PAI-1 in tumors of Bik−/− mice. As the first step, we examined the hypothesis that plasma bikunin can at least serve to downregulate the upregulation of G-CSF-induced uPA and PAI-1 expression. Indeed, G-CSF-induced production of uPA and PAI-1 was enhanced in tumor cells cultured in bikunin-depleted serum. Second, exogenously applied bikunin reduced G-CSF-stimulated uPA and PAI-1 synthesis in tumor cells in a dose-dependent manner. Therefore, plasma bikunin appears to play a role in regulation of uPA and PAI-1 synthesis to avoid invasion and metastasis. The mechanism by which uPA and PAI-1 are decreased may involve either bikunin-dependent suppression of de novo protein synthesis as induced by G-CSF or release from existing intra- and extra-cellular pools. Our in vitro and in vivo studies demonstrated that exogenous bikunin may decrease the G-CSF-induced production of uPA and PAI-1, probably at least by decreasing the synthesis of uPA and PAI-1 by tumor cells. These data are in full accordance with our findings that bikunin altered the in vitro kinetics of G-CSF-dependent activation of ERK1/2 signal cascade and largely decreased the phosphorylation of signaling molecules in tumor cells in response to G-CSF. The absence of bikunin from the plasma and extracellular matrix (bikunin-deficient mice) led to a higher sensitivity of the mice for treatment with G-CSF than Bik+/+ mice, because in the absence of bikunin, a 1.5–2.0-fold increase in uPA and PAI-1 levels was observed. These findings suggest that bikunin promotes survival of tumor inoculated mice possibly through suppression of uPA and PAI-1 expression, indicating that bikunin deficiency in plasma and extracellular matrix could increase susceptibility to metastasis.
Enzymes putatively mediating extracellular matrix dissolution (proteases instrumental to invasion and metastasis) may be derived from multiple sources. In most tumors uPA was localized in malignant cells as well as in surrounding stromal cells. Stromal cells also contribute to cancer growth and metastasis through the production of extracellular matrix modifiers such as uPA, its receptor uPAR, its inhibitors (PAI-1 and PAI-2), matrix metalloproteinases (MMPs) and growth factors, including the fibroblast and insulin-like growth factors.18 Thus in Figure 4, we may actually observe a stromal response to G-CSF. The metastatic capacity and/or clinical aggressiveness of carcinomas may reflect overall proteolytic enzyme expression, suggesting that cooperative enzyme interaction (e.g., uPA-MMP) may be required for invasive growth and metastasis. In our study, we selected M-27 cells to focus on uPA and simplify the results; the cells produced very low levels of MMPs and significantly higher levels of uPA. Therefore, no attention was given to MMPs nor to the inter-dependent pathway of activation between serine- and metallo-proteinases which could be hindered by bikunin.
uPA induces proliferation, invasion or metastasis, activates various kinases of the MAP kinase family and induces various transcription factors.19 uPA is overexpressed in various malignancies including breast, ovarian and prostate cancers and has been clearly demonstrated to be essential in the maintenance of invasive and metastatic phenotypes.20 Expression of uPA can be up-regulated by mitogen, growth factors, oncogenes and ligation of integrin with ECM protein.21 The PAI-1 blocks the activation of uPA, an extracellular protease vital to cancer invasion. Paradoxically, high levels of PAI-1 as well as uPA are equally associated with poor prognosis in cancer patients.22 The binding of uPA to its cell surface receptor (uPAR) promotes cell adhesion by increasing the affinity of the receptor for both vitronectin (VN) and integrins. PAI-1 can detach cells by disrupting uPAR-VN and integrin-VN interactions. Thus, PAI-1 induces cell detachment. The novel “deadhesive” activity of PAI-1 toward a variety of cells growing on different ECM may begin to explain why high PAI-1 levels often are associated with a poor prognosis in human metastatic disease.23 Taken together, Bik−/− exhibited high levels of uPA and PAI-1 expression in tumors, which resulted in poor prognosis in animals.
Bikunin is a member of the IαI family, which consists of at least 3 closely related molecules.1, 2 Until recently, bikunin has been assigned a classical protease inhibitory role. Recent data indicate that bikunin may have other functions that are unrelated to protease inhibition. Indeed, Balduyck et al.24 reported that urinary bikunin was significantly higher in severe inflammation or septic shock. In cancers, administration of bikunin may block tumor cell invasion and metastasis by a direct inhibition of tumor cell-associated plasmin activity as well as by inhibiting uPA and uPAR expression at the gene and protein levels, possibly through suppression of the MAP kinase signaling cascade.3 We have shown that bikunin binds to cancer cells and macrophages to suppress a variety of biological responses predominantly through MAP kinase signaling pathway.12 The present observations strongly suggest that we have now identified a protective role of bikunin in cancer metastasis. The pharmacokinetic observations from our study replicate those predicted from preclinical models. Oral administration of bikunin displays longer disposition profiles on the order of 12 hr in tumors. Most importantly, achievable bikunin tumor concentrations (∼0.4 μM), exceeded that required for in vitro effective concentration of bikunin (0.25 μM).25
It was remarkable that deficiency of proteins of the IαI family in Bik−/− mice significantly alters susceptibility to cancer metastasis. Since decreased systemic levels of both bikunin and IαI have been observed in Bik−/− mice, circulating IαI might also play a role in suppression of progression of tumor metastasis. For this, we must examine whether IαI administration reduces lung metastasis and prolongs the survival of Bik−/− mice or Bik+/+ mice with a model of spontaneous metastasis. By these experiments, we will be able to confirm whether, in addition to bikunin, IαI also plays a role in suppression of cancer metastasis. Unfortunately, we have no purified IαI concentrates and there is no information about IαI-dependent signaling pathway at present. This finding will be important in investigating genetic factors affecting susceptibility to metastasis in cancer patients.
In conclusion, our findings presented here suggest that endogenous proteins of the IαI family, especially bikunin, plays a role in suppressing spontaneous cancer metastasis. We postulate that at least plasma bikunin forms a defense mechanism against the development of cancer metastasis. Administration of bikunin concentrates may be of highly therapeutic significance to overcome failure of this endogenous defense mechanism. Study of the underlying mechanisms will aid our understanding of the function and signal transduction of bikunin in the uPA and PAI-1 production during cancer metastasis. The bikunin-deficient mice is a valuable tool for further elucidation of the in vivo role of bikunin in various disease conditions in which bikunin might be involved.13, 14 Finally, the present study should be of great value for research on the development of antimetastatic agents.
We are thankful to Dr. H. Morishita and Dr. H. Sato (BioResearch Institute, Mochida Pharmaceutical Co., Gotenba, Shizuoka) for their continuous and generous support of our work. This work was supported by a grant-in-aid for Scientific Research from the Ministry of Education, Science and Culture of Japan (to H.K.), by grants from the Fuji Foundation for Protein Research (H.K.), the Kanzawa Medical Foundation (H.K.), Sagawa Cancer Research Foundation (H.K.), and Aichi Cancer Research Foundation (H.K).