[Cancer Research 61, 7627-7634, October 15, 2001]
© 2001 American Association for Cancer Research
Structural Fragility of Blood Vessels and Peritoneum in Calponin h1-deficient Mice, Resulting in an Increase in Hematogenous Metastasis and Peritoneal Dissemination of Malignant Tumor Cells1
Shun’ichiro Taniguchi2,
Michiko Takeoka,
Takashi Ehara,
Shigenari Hashimoto,
Hiroto Shibuki,
Nagahisa Yoshimura,
Hidekazu Shigematsu,
Katsuhito Takahashi and
Motoya Katsuki3
Department of Molecular Oncology and Angiology, Research Center on Aging and Adaptation [S. T., M. T., S. H.], Department of Pathology [T. E., H. S.], and Department of Ophthalmology [H. S., N. Y.], Shinshu University School of Medicine, Matsumoto 390-8621; Department of Medicine, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka 537 [K. T.]; The Graduate School of Pharmaceutical Science, Osaka University, SORST, Japan Science and Technology Corporation, Osaka 537 [K. T.] Japan; Institute of Medical Science, University of Tokyo, Tokyo 108-8639 [M. K.], Japan
 |
ABSTRACT
|
|---|
We have observed weak expression of calponin h1, which stabilizes the actin filament system, in blood vessels within human malignant tumors. This observation suggested that because of a deficiency in stabilization by calponin h1, the structure of blood vessels in malignant tumors is fragile compared with blood vessels in normal tissues. We therefore generated calponin h1-deficient (CN-/-) mice to examine the effect of calponin h1 on the integrity of the barrier system in blood vessels against cancer metastasis. The CN-/- mice exhibited morphological fragility of the tissues, including the uterus and blood vessels. In particular, we frequently observed bleeding into the surrounding tissue from blood vessels of the ocular fundus in CN-/- mice. In addition, mesothelial cells, which usually express calponin h1 in normal (CN+/+) mice, were retracted in the CN-/- mice. When fluorescein was injected i.v. into mice, the CN-/- mice exhibited a greater and more rapid leakage of fluorescein from the blood vessels of the ocular fundus compared with the CN+/+ mice. In the CN-/- mice receiving i.v. inoculations of B16 melanoma cells, significantly more metastatic nodules were formed in the lung than in the CN+/+ mice. When B16 melanoma cells were injected i.p., the severity of peritonitis carcinomatosa was greater in CN-/- than in CN+/+ mice. These results indicate that calponin h1 plays an important role in the regulation of the integrity of the blood vessels and peritoneum, which in turn is an important factor influencing the frequency of cancer metastasis. The CN-/- mice, which exhibit fragile blood vessels and peritoneum, could serve as sensitive and useful host models to investigate cancer metastasis.
|
INTRODUCTION
|
|---|
Cancer metastasis consists of multiple steps. Recent molecular biological approaches have greatly assisted in the identification and characterization of numerous molecules responsible for the metastatic phenotypes of cancer cells, such as proteases and cell adhesion molecules. In particular, we have concentrated on the cytoskeletal molecules, including actin (1, 2, 3)
, vinculin (4)
, tropomyosin (5
, 6)
, calponin h1 (7
, 8)
, and keratins (9)
, which play important roles in the regulation of cellular morphology, adhesiveness, motility, and/or growth. Because cancer metastasis occurs as a result of the interaction between cancer cells and the host, we have also directed our attention to the host factors relating to cancer metastasis in addition to the phenotypes of tumor cells.
Calponin h1 is an actin-binding protein, largely expressed in smooth muscle cells, that positively regulates and stabilizes actin polymerization and inhibits actomyosin ATPase (10)
. Calponin h1 has also been detected in the endothelial cells of blood vessels by RT-PCR,4
although the expressed amount is low compared with that in smooth muscle cells (11)
. Three calponin isoforms, calponin h1 and h2 and acidic calponin, have been identified and characterized (12
, 13)
. Calponin h1 is expressed mainly in the smooth muscle cells, whereas calponin h2 is a nonmuscle type and acidic calponin is present mainly in the brain. The differences in the biological functions among these isoforms remain to be determined. Calponin h1 is suppressed in transformed smooth muscle cells in leiomyosarcoma (7)
. Introduction of the calponin h1 gene into the transformed cells suppressed proliferation and induced the differentiated phenotype (8)
.
We previously reported the down-regulation of 
SMA in malignant human melanoma tissues compared with benign pigment tissues (14)
. When tumor tissues were immunologically stained with anti-SMA, the expression of SMA was decreased in the blood vessels of malignant tumors (15, 16, 17)
. We also observed that a melanoma cell line, M14, released PDGF-BB, a product of the sis oncogene, which suppressed SMA. These observations indicated to us that the low expression of SMA in the vessels of malignant tumors is induced by the paracrine effect of factors excreted from the tumor cells. Additionally, in the blood vessels of several malignant human tumors, we observed that the down-regulation of calponin h1, which is also induced by PDGF-BB (18)
, as seen with SMA, was greater than that of SMA: calponin h1 expression was not detected, even in blood vessels expressing SMA. We then assumed that a reduction in calponin h1, as well as in SMA, leads to fragile blood vessels and the subsequent enhancement of cancer metastasis because the suppression of calponin h1 results in destabilization of actin filaments, which probably weakens the cellular adhesion molecular mechanisms, such as the cadherin and integrin systems, that are regulated in connection with actin filaments.
In the present study, we generated calponin h1-deficient (CN-/-) mice and examined the degree of sensitivity of these mice to cancer metastasis. We also examined the morphology of several tissues usually rich in calponin h1-positive cells and assessed the permeability of blood vessels together with the degree of hematogenous metastasis and peritoneal dissemination of cancer cells in CN-/- mice.
|
MATERIALS AND METHODS
|
|---|
CN-/- Mice.
CN-/- mice were generated as described previously (19)
. Targeting of the calponin h1 gene was achieved by use of ES cells from 129SV/J mice, and chimeric mice were generated by injecting the ES cells into C57BL6/J blastocysts. Offspring exhibiting germinal transmission were interbred to generate homozygous mice. The mice thus generated were mated with C57BL6/J mice to increase the genetic similarity of the offspring to C57BL6/J mice, in which B16 melanoma cells are transplantable. The genotypes of the mice used were determined by Southern blot analysis and/or PCR.
For the PCR, exons 4 and 5 of the calponin h1 gene were amplified to examine deletion of the fragment from the EcoRV site in the intronic site of exon 4–5 to the ApaI site of exon 7, using the primers 5'-CATACACAAGTTCAGTCCAC-3' (forward) and 5'-TCCTGCCTTCTCTCAACTTC-3' (reverse), the combination of which generated 297-bp DNA fragments. To confirm the replacement of the EcoRV-ApaI fragment with the neo-containing cassette, the neo-resistant gene was amplified using primers 5'-GGCACAACAGACAATCGGCT-3' (forward) and 5'-ACTTCGCCCAATAGCAGCCA-3' (reverse), which generated a 218-bp DNA fragment. Expression of the calponin h1 gene was determined by RT-PCR using the primers 5'-GTCTGTGTCATCTGCACCTC-3' (forward) and 5'-TCCCGTCGCAGGAATGGGGC-3' (reverse), which produced 1295-bp cDNAs of calponin h1 transcripts.
We compared various phenotypes between the mice homozygous null for the calponin h1 gene (CN-/-) and the mice carrying only the normal calponin h1 gene (CN+/+); these mice were generated by mating mice that were heterozygous null for the calponin h1 gene. In the present work, mice with five to seven back-crosses to the C57 BL6/J background were used. B16 mouse melanoma cells were transplantable to all of the animals used in the present study. All animal-handling procedures were in accordance with the ethics protocol approved by Shinshu University School of Medicine.
Western Blot and Immunostaining.
The synthetic COOH-terminal peptide, Leu281-Ala297, of the mouse calponin h1 protein was coupled to keyhole limpet hemacyanin with N-maleimidobenzoyl-N-hydroxysuccinimide ester. An antiserum against the peptide was generated in rabbits as reported previously (19)
. The antibody was used for the immunoblot analysis as well as for the immunostaining of mouse tissues. Mesothelial cells from the mouse mesentery were cultured to confluency in a 60-mm dish and lysed with sample buffer for SDS-PAGE analysis. The lysed samples from the mice were electrophoresed and used for Western blot analysis with the above antibody.
Immunohistochemistry.
Mice were euthanized under anesthesia with sodium pentobarbital. Various mouse tissues were resected, immediately fixed in Carnoy solution, and embedded in paraffin. Human tissues were supplied with the consent of the patients and fixed in the same manner as the mouse tissues. The tissue sections were cut into 3-µm slices and deparaffinized. Staining with H&E, together with Toluidine Blue, was performed for routine histological examination. For immunohistochemistry, the samples were microwave-treated prior to incubation with the antibodies. The antibodies used were rabbit anticalponin h1 antibody against mouse calponin h1 (19)
, already described above; mouse antihuman calponin h1 monoclonal antibody (DAKO); and mouse antihuman SMA antibody (anti-SMA). After the sections were washed with PBS, endogenous peroxidase was blocked by immersing the sections in methanol and H202 for 30 min. Primary antibodies were diluted 1:100 with PBS and incubated with the preparations overnight. After the sections were washed with PBS, secondary antibodies were applied. Diaminobenzidine was used as the colorimetric substance. The control sections were processed with nonimmunized rabbit or mouse serum (DAKO) as primary antibodies or without the primary antibodies.
Electron Microscopy.
A part of the resected uterus and parietal peritoneum was fixed in 2% glutaraldehyde for 2 h. After being washed in phosphate buffer, sections were post-fixed in 1.5% osmium tetroxide for 1 h, and then dehydrated and embedded in Quetol-812. Ultrathin sections were cut with a diamond knife, using an LKB ultramicrotome, and stained with uranyl acetate and lead citrate. The prepared samples were examined with a JOEL 1200 electron microscope.
Lung tissue for scanning electron microscopic observations was fixed in 1% glutaraldehyde for 12 h and post-fixed in 1% osmium tetroxide for 1 h. After dehydration in ethanol, the specimens were rinsed in isoamyl acetate and dried by a critical-point drying method. The dried specimens were mounted on copper plates and coated with osmium in an osmium plasma coater (OPC 40; Nippon Laser and Electronics Lab). The specimens were examined with a JSM-6000F scanning electron microscope (JEOL) at 15 kV.
Angiography Using Fluorescein.
For fluorescein angiography, 10% sodium fluorescein (1 ml/kg; Fluorescite; Alcon, Fort Worth, TX) was injected through the tail veins of anesthetized mice, and angiograms were taken with a Kowa fundus camera (GENESIS; Kowa, Tokyo) at appropriate intervals.
Assay of Metastasis.
Gently trypsinized mouse B16-melanoma, B16-F10, or B16-F1 cells were suspended in HBSS. The experimental metastasis was evaluated by injecting 1 x 105 viable cells in 0.5 ml of HBSS into the tail or femoral veins of the mice. The mice were euthanized 17–21 days later, and each organ was examined for the formation of metastatic tumor nodules.
To examine peritoneal dissemination, gently trypsinized B16-F10 melanoma cells were suspended in HBSS, and 2.5 x 105 viable cells/0.5 ml were injected into the peritoneal cavity. After 15 days, the mice inoculated with the tumor cells were euthanized and examined for the existence of ascites and the number of tumor nodules formed in the peritoneum and on the liver.
Statistical Analysis.
Statistical analyses between the groups (CN+/+ and CN-/- mice) in each experiment were performed by the Student t test or the Kaplan-Meier method. In all analyses, the differences were considered to be significant at P < 0.05.
|
RESULTS
|
|---|
Immunohistochemistry of Human Malignant Tumor Tissues.
We stained various human malignant tumors with anti-SMA and monoclonal antihuman calponin h1 antibodies. Fig. 1
shows representative images of the human malignant tumor tissues, such as uterus carcinoma, angiosarcoma, and prostate cancer. The blood vessels were detectable by immunohistochemical staining for SMA, where calponin h1 expression was very weak (Fig. 1A
, thick arrows) in the blood vessels of the malignant tumors compared with the surrounding normal tissues (Fig. 1A
, top row, thin arrows). To allow easy evaluation and highlight the RBCs in the lumens, enlarged figures are shown in the second row of Fig. 1A
. In both angiosarcoma (Fig. 1B)
and prostate cancer (Fig. 1C)
, the blood vessels that were recognizable by staining with anti-SMA were faintly stained with the anticalponin h1 antibody. We also observed the same phenomena in various human skin pigmentous tumors.5
Malignant melanoma blood vessels showed weaker staining with the anticalponin h1 antibody than the normal tissues and benign tumors. According to our immunofluorescence observations, the staining intensity with anti-SMA was also weaker in the blood vessels of the malignant tumor tissues than in the surrounding normal tissues (data not shown), which is consistent with previous reports (15, 16, 17)
.
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 1. Immunostaining of human uterus carcinoma, angiosarcoma, and prostate carcinoma. Shown are serial sections of human uterus carcinoma (A), angiosarcoma (B), and prostate carcinoma (C). Left panels, H&E staining; middle panels, immunostaining with anti-SMA (magnification, x100); right panels, immunostaining with monoclonal antihuman calponin h1 antibody (magnification, x100). In all of the tumors, calponin h1 was weakly expressed in the blood vessels. In A, thin arrows in the top row show blood vessels in the normal areas outside the tumors. The bottom row of A shows enlarged images of blood vessels to highlight the red cells. Thick arrows in A–C show blood vessels inside the tumors. Bars: 250 µm in the top row of A; 100 µm in bottom row of A and in B and C.
|
|
CN-/- Mice.
The genotypes of mice were confirmed by amplification of the genomic DNA using PCR. Mice that exhibited the neo fragment (218 bp) but lacked the exon 4–7 (EcoRV-ApaI) fragment were identified as the CN-/- genotype. Conversely, mice that exhibited the expected exon 4–7 fragment (EcoRV-ApaI; 297 bp) but no neo fragment were identified as the CN+/+ genotype; mice exhibiting both the 218- and 297-bp fragments were identified as the CN+/- genotype. Using RT-PCR, we detected cDNA fragments (1295 bp) derived from calponin h1 transcripts in the CN+/+ mice, but we found no expression of calponin h1 transcripts in the CN-/- mice (data not shown). There was no enhanced or ectopic expression of neutral and acidic calponin mRNAs (data not shown), which is consistent with our previous report (19)
. Immunostaining with antimouse calponin h1 polyclonal antibody showed calponin h1 expression in the blood vessels of the thigh and ocular fundus in the CN+/+ mice, whereas calponin h1 expression was not detected in the CN-/- mice (Fig. 2)
. As for the CN+/- mice, the expression of calponin h1 protein was much the same as in the CN+/+ mice, so that the various phenotypes were compared between the CN+/+ and CN-/- mice.
View larger version (76K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 2. Confirmation of no expression of calponin h1 in the CN-/- genotype mice. Immunostaining with rabbit polyclonal antimouse calponin h1 antibody. A, femoral artery. The second antibody was a goat antirabbit IgG antibody conjugated with peroxidase. Bar, 100 µm. B, blood vessels in the ocular fundus. The second antibody was antimouse IgG conjugated with FITC. Bar, 100 µm.
|
|
Morphological Fragility of Smooth Muscle Cells in the Uterus of CN-/- Mice.
Because we expected that the suppression of calponin h1 would affect cell-to-cell and cell-to-extracellular matrix adhesiveness, we initially carried out morphological examination on the smooth muscle cells of the uterus in the CN-/- mice, using electron microscopy. As shown in Fig. 3
, the smooth muscle cells of the uterus of the CN-/- mice were morphologically round in structure, and the cellular adhesion was clearly decreased.
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 3. Electron microscope images of murine uterus. Smooth muscle cells showed round morphology in the CN-/- female mice. The cell-cell adhesion of the CN-/- mice (right) was clearly weaker than that of the CN+/+ mice (left). Bar, 2 µm.
|
|
Alteration of Blood Vessel Structure in CN-/- Mice.
Electron microscopy of the femoral artery showed that the adhesion of the endothelial cells to the elastic fiber was weaker in CN-/- than in CN+/+ mice (Fig. 4A)
. In the coronary artery, fibrosis between the endothelial and smooth muscle cells was frequently observed (Fig. 4B)
. Bleeding from the blood vessels of the ocular fundus was often detected (Fig. 4C)
.
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 4. Morphological differences in the blood vessels between the CN+/+ and CN-/- mice. A, electron microscope image of the femoral artery. The endothelial cell is detached from the elastic fiber in the CN-/- mouse compared with the CN+/+ mouse. Bar, 1 µm. B, electron microscope image of the coronary artery. Fibrosis is observed between the endothelial and smooth muscle cells. Bar, 2 µm. C, Toluidine Blue staining of the blood vessels in the ocular fundus. Bleeding into the surrounding area was frequently observed in the CN-/- mice (arrow). Bar, 100 µm.
|
|
In the lung veins and venules, the diameters of the lumina appeared to be larger in the CN-/- than in the CN+/+ mice (Fig. 5, A and B)
; the same phenomenon was also observed in the blood vessels of the ocular fundus. These results led us speculate that the walls of such blood vessels could become thin in association with the enlargement of the diameter of the blood vessels; we then observed by electron microscopy that the walls of the pulmonary arterioles clearly became thinner in the CN-/- mice compared with the CN+/+ mice (Fig. 5C)
. When examined by scanning electron microscopy, the surfaces of the capillaries in the lungs of the CN-/- mice appeared rough and fragile compared with capillaries in the CN+/+ mice (Fig. 5D)
.
Fluorescein Leakage from Blood Vessels of the Retina in CN-/- Mice.
On the basis of the structural fragility of blood vessels in the CN-/- mice, we examined the permeability of retinal blood vessels because leakage from these blood vessels is easy to assay. When fluorescein was injected i.v. into the mice, fluorescein leakage was greater in the CN-/- than in the CN+/+ mice. As shown in Fig. 6
, the leakage of fluorescein, which is recognizable by the image at the rim of the vessels, was observed 30 s after the injection of fluorescein; the diffusion of fluorescence was greatly expanded in CN-/- mice 180 s after the injection period, whereas there was comparatively little leakage detected in the CN+/+ mice 300 s after the injection.
View larger version (73K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 6. Leakage of fluorescein from the blood vessels of the retina. Fluorescein was increasingly and rapidly leaked from the retinal blood vessels of the CN-/- mice compared with the CN+/+ mice.
|
|
Increase in Hematogenous Metastasis of B16 Melanoma in CN-/- Mice.
When B16-F10 cells were administered i.v. to the mice, the metastatic tumor nodules detected were present mainly in the lungs. The metastatic nodules were observed microscopically at x6.4 magnification. Colonies with diameters larger than 
100 µm, which would correspond to clinically relevant metastases, were selected for counting. The average number of metastatic nodules counted was 116 in the CN+/+ mice and 180 in the CN-/- mice, respectively. As shown in Fig. 7, A and B
, the metastasis of B16-F10 cells to the lung, as indicated by the relative number of metastatic nodules in composite from replicate experiments, was significantly greater in the CN-/- mice than in the CN+/+ mice (P < 0.05). When we inoculated B16-F1 cells i.v. into the mice, more metastatic nodules were formed in the lungs of the CN-/- than the CN+/+ mice (data not shown). In addition to the metastasis in the lungs, metastatic nodules were also observed in the liver, skin, and diaphragm as well as other organs of the female mice, although the number of nodules was low compared with the lung. The number and frequency of metastases in various tissues also tended to be increased in the CN-/- mice (Table 1)
; consequently, the survival period for the CN-/- mice was significantly shorter than that for the CN+/+ mice (Fig. 7C)
.
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 7. Lung metastasis of i.v. injected B16-F10 cells (5 x 105 cells/0.5 ml/mouse). The data in A reflect a composite from triplicate experiments. We used mice back-crossed 6–7 times for these experiments. The average number of lung metastases of the CN+/+ mice group in each experiment was arbitrarily adjusted to 100, and the relative number of lung metastases of each mouse used in the experiment was calculated. This relative number corresponding to the lung metastases was used for the statistical comparison between the CN+/+ and CN-/- groups. A, the relative average number of tumor nodules of B16-F10 cells in the lung was significantly larger (P < 0.05) in the CN-/- mice (152.0 ± 15.2; n = 21) than in the CN+/+ mice (100.0 ± 8.6; n = 24). Bars, SE. B, metastatic tumor nodules on the lungs of CN-/- (bottom) and CN+/+ (top) mice. C, survival rates of mice receiving i.v. injections of B16-F10 cells (5 x 105 cells/mouse). In the experiment of panel C and Table 1
, female mice were used. Dashed line, CN-/- mice (n = 9); solid line, CN+/+ mice (n = 9). The survival period of the CN-/- mice (22.9 ± 0.3 days) was significantly shorter than that of the CN+/+ mice (24.9 ± 0.6 days; P < 0.05).
|
|
Fragile Peritoneum in CN-/- Mice.
We cultured mesothelial cells from the CN+/+ and CN-/- mice and examined the expression of calponin h1 in the mesothelial cells by Western blot analysis with a specific polyclonal antibody against mouse calponin h1. Fig. 8
shows that calponin h1 was clearly expressed in the cultured mesothelial cells of the CN+/+ but not the CN-/- mice.
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 8. Western blot analysis of the cultured mesothelial cells with rabbit antimouse calponin h1 polyclonal antibody. Mesothelial cells were prepared and cultured according to the method of Akedo et al. (26)
. Calponin h1 (molecular mass, 34 kDa) was detected in the cells from the CN+/+ but not the CN-/- mice.
|
|
We next examined the ultrastructure of the peritoneum. Fig. 9
shows that the mesothelial cells in the CN-/- mice had shrunken or retracted cytoplasm and that the subserosal connective tissue was thinner than in the CN+/+ mice. Furthermore, the extracellular matrix of the peritoneum appeared to be disorganized in the CN-/- mice compared with the matrix in the CN+/+ mice; an organized parallel arrangement of collagen fibers was seen in the CN+/+ but not in the CN-/- mice.
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 9. Electron microscope image of the peritoneum. The mesothelial cells (indicated by straight lines) were retracted in the peritoneum of the CN-/- mice compared with those of the CN+/+ mice. Collagen fibers of the peritoneum were thinner in the CN-/- mice. The thickness of the peritoneum shown is 28 µm in the CN+/+ and 17 µm in the CN-/- mouse. Bar, 10 µm.
|
|
Enhancement in Peritoneal Carcinomatosa in CN-/- Mice.
When we inoculated B16-F10 cells into the peritoneal cavities of the mice, the severity of peritonitis carcinomatosa was greater in the CN-/- than in the CN+/+ mice; a representative result is shown in Fig. 10
. Similarly, the quantity of ascites and number of tumor nodules on the peritoneum and on the liver were much greater in the CN-/- than in the CN+/+ mice (Table 2)
.
View larger version (76K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 10. Peritoneal dissemination of B16-F10 cells inoculated at 2.5 x 105 cells/0.5 ml/mouse. More widespread dissemination of B16-F10 cells was observed in the CN-/- mice compared with the CN+/+ mice.
|
|
|
DISCUSSION
|
|---|
In this study, we used immunohistochemical techniques to detect low expression of calponin h1 in the blood vessels of malignant human tumor tissues and discovered a fragile structure of the blood vessels and peritoneum in CN-/- mice, which possibly leads to an increase in hematogenous metastasis and peritoneal dissemination of cancer cells. The results in the CN-/- mice together with the phenomena found in the human tumor tissues indicate that the suppression of calponin h1 in the blood vessels or peritoneum is an important factor for increasing cancer metastasis.
Although there are a number of factors that determine metastasis of cancer cells, the results in the present study suggest that the integrity of the blood vessels and peritoneum of the host is an equally important factor. The process by which cancer cells intra- and/or extravasate through vessels and the peritoneum must be one of the rate-limiting steps of metastasis that depends on whether cells constituting vessels and the peritoneum retract or not. Such retraction could occur as a result of the disruption of intercellular adhesion and/or cellular attachment to the extracellular matrix. The mechanisms of cellular adhesion, such as the adherence junction and the tight junction, are connected and regulated by actin filaments. Thus, the reduction in cellular adhesion is probably induced by the fragility of the actin filaments as a result of the suppression of stabilizers, such as calponin, as shown in the present study, as well as the down-regulation or mutation of adhesion molecules in the cellular plasma membrane.
In the blood vessels, calponin h1 is detected mainly in the smooth muscle cells and pericytes; our present study thus provides information on the vessels containing smooth muscle cells and pericytes, such as arteries/arterioles and veins/venules. If, as is the general belief, metastatic cells rarely intra- and extravasate through arteries/arterioles, the enhanced hematogenous metastasis, especially pulmonary metastasis, of B16-F10 cells in the CN-/- mice (Fig. 7
and Table 1
) seemingly can be attributed to the fragility of veins/venules induced by the deletion of calponin h1. Actually, the diameters of the pulmonary veins and venules in the CN-/- mice were enlarged, and the walls of the arterioles were thinner in the CN-/- than in the CN+/+ mice (Fig. 5)
, indicating the fragility of the blood vessels in the CN-/- mice. We believe that the walls of the venules in the CN-/- mice should be thin compared with the CN+/+ mice, although it was technically difficult to show as clear an image for the venules as for the arterioles. In contrast, it has been reported that calponin h1 can be detected in the endothelial cells by RT-PCR (11)
; we therefore thought that suppression of calponin h1 in the endothelial cells as well as in the vascular smooth muscle cells leads to structural immaturity and weakness in all types of blood vessels, including capillaries. Because both the pulmonary arterioles and venules in CN-/- mice were morphologically different from those in the CN+/+ mice, the capillaries existing between them should be somehow changed in the CN-/- mice. As expected, when observed with scanning electron microscopy, the surfaces of capillaries in the lungs in the CN-/- mice appeared rough and irregular compared with those in the CN+/+ mice (Fig. 5D)
. These morphological changes in the endothelial cells suggested to us the fragility of the capillaries, although direct verification remains to be done.
How calponin h1 is suppressed in human tissues is an important issue, although the mechanism of gene regulation has been only partially clarified. The 5' flanking region of the human calponin h1 gene has been shown to have a cis-acting domain for interaction with a methylated DNA-binding transcription repressor (20)
. Moreover, the expression of smooth muscle cell-specific molecular markers is reportedly affected by growth factors, such as PDGF (17
, 21
, 22)
and transforming growth factor ß (23)
, as well as by cytokines, such as IFN-
(24)
. The regulatory mechanism of gene transcription, however, remains to be elucidated, as well as the main factors involved in the reduction of -actin and/or calponin h1 expression in malignant tumor tissues. Recently, it has been noted that VEGF and the angiopoietins function together during vascular development, with VEGF acting early during vessel remodeling and angiopoietin-1 acting later during vessel remodeling, maturation, and stabilization. In vascular development in malignant tumors, the expression of VEGF and angiopoietin-1 is probably not well balanced, resulting in the immaturity of the blood vessels (25)
.
From a gene-therapy point of view, calponin h1 could serve as a useful molecule. Although it can suppress the proliferation of transformed cells (8)
, calponin h1 would also be able to stabilize intercellular adhesiveness and attachment of the host cells to the extracellular matrix, thus protecting the blood vessels and the peritoneum from penetration by the tumor cells. We have observed that introduction of the calponin h1 gene into HT1080 cells suppresses the proliferation of tumor cells and enhances attachment to fibronectin.6
In our recent study, we found an unexpected overgrowth of the bone in the CN-/- compared with the CN+/+ mice (19)
, and the ectopic bone formation induced by Bone Morphogenetic Protein-4 was greater in the CN-/- than in the CN+/+ mice. Thus, CN-/- mice might represent a useful host to examine bone metastasis, making use of rapid ectopic bone formation.
In conclusion, the enhancement of hematogenous and peritoneal dissemination in CN-/- mice is seemingly attributable to the structural fragility of the blood vessels and peritoneum, respectively, caused by a reduction in cellular adhesiveness. The reduction in cell adhesiveness could have be induced by the fragility of the actin filament system as a result of a lack of the stabilizer calponin h1. Although the precise mechanism remains to be elucidated, such animal models will be useful in the assessment of cancer metastasis. On the basis of our present results, we suggest that weak calponin h1 expression in the blood vessels of human malignant tumors is strongly indicative of the immaturity and fragile structure of the blood vessels and serves as a useful molecular marker for the differential diagnosis of malignant from benign tumors. Finally, we believe that there might be a good possibility for gene therapy with calponin h1, not only to suppress the growth of tumor cells but also to strengthen the physical barriers of the host system against tumor formation.
|
ACKNOWLEDGMENTS
|
|---|
We thank K. Kametani, T. Nishizawa, K. Suzuki, M. Watanabe, and E. Sato for technical assistance.
|
FOOTNOTES
|
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants for Scientific Research from the Ministry of Education, Science and Culture of Japan (109254104 and 13218055). 
2 To whom requests for reprints should be addressed, at Department of Molecular Oncology and Angiology, Research Center on Aging and Adaptation, Shinshu University School of Medicine, 3-1-1 Asahi, Matsumoto 390-8621, Japan. Fax: 81-263-37-2724; E-mail: stangch{at}sch.md.shinshu-u.ac.jp
3 Present address: National Institute for Basic Biology, Nishigonaka 38, Myodaiji, Okazaki 444-8585, Aichi, Japan.
4 The abbreviations used are: RT-PCR, reverse transcription-PCR; SMA, smooth muscle -actin; PDGF, platelet-derived growth factor; VEGF, vascular endothelial growth factor.
5 K. Koganehira, M. Takeoka, T. Ehara, H. Murata, T. Saida, and S. Taniguchi. Reduced expression of actin-binding protein, calponin-h1 and h-caldesmon, in the vascular smooth muscle inside human malignant melanoma is closely related to the poor prognosis, manuscript in preparation.
6 M. Takeoka, T. Ehara, J. Sagara, S. Hashimoto, and S. Taniguchi. Calponin h1 enhanced adhesion and suppressed growth of human fibrosarcoma HT1080, manuscript in preparation.
Received 9/29/00.
Accepted 8/15/01.
|
REFERENCES
|
|---|
-
Shimokawa-Kuroki R., Sadano H., Taniguchi S. A variant actin (ßm) reduces metastasis of mouse B16 melanoma. Int. J. Cancer, 56: 689-697, 1994.[Medline]
-
Taniguchi S. Cellular and molecular function of a variant actin (ßm), which inhibits metastasis of mouse B16 melanoma Hori Y. Hearing V. J. Nakayama J. eds. . Melanogenesis and Malignant Melanoma: Biochemistry, Cell Biology, Molecular Biology, Pathophysiology, Diagnosis and Treatment, : 169-183, Excerpta Medica New York 1996.
-
Okamoto-Inoue M., Kamada S., Kimura G., Taniguchi S. The induction of smooth muscle actin in a transformed rat cell line suppresses malignant properties in vitro and in vivo. Cancer Lett., 142: 173-178, 1999.[Medline]
-
Sadano H., Inoue M., Taniguchi S. Differential expression of vinculin between weakly and highly metastatic B16 melanoma cell line. Jpn. J. Cancer Res., 83: 625-630, 1992.[Medline]
-
Miyado K., Kimura M., Taniguchi S. Decreased expression of a single tropomyosin isoform, TM/TM30nm, results in reduction in motility of highly metastatic B16–F10 mouse melanoma cells. Biochem. Biophys. Res. Commun., 225: 427-435, 1996.[Medline]
-
Miyado K., Sato M., Taniguchi S. Transformation-related expression of a low-molecular weight tropomyosin isoform TM5/TM30nm in transformed rat fibroblastic cell lines. J. Cancer Res. Clin. Oncol., 123: 331-336, 1997.[Medline]
-
Horiuchi A., Nikaido T., Ito K., Zhai Y., Orii A., Taniguchi S., Toki T., Fujii S. Reduced expression of calponin h1 in leiomyosarcoma of the uterus. Lab. Investig., 78: 839-846, 1998.[Medline]
-
Horiuchi A., Nikaido T., Taniguchi S., Fujii S. Possible role of calponin h1 as a tumor suppresser in human uterine leiomyosarcoma. J. Natl. Cancer Inst. (Bethesda), 91: 790-796, 1999.[Abstract/Free Full Text]
-
Morifuji M., Taniguchi S., Sakai H., Ohishi M. Alteration of cytoskeleton in tongue cancer exhibiting different metastatic ability with an orthotopic implantation. Am. J. Pathol., 156: 1317-1326, 2000.[Abstract/Free Full Text]
-
Kake T., Kimura S., Takahashi K., Maruyama K. Calponin induces actin polymerization at low ionic strength and inhibits depolymerization of actin filaments. Biochem. J., 312: 587-592, 1995.
-
Sakihara C., Nishimura J., Kobayashi S., Takahashi S., Kanaide H. Expression of calponin mRNA in porcine aortic endothelial cells. Biochem. Biophys. Res. Commun., 222: 195-200, 1996.[Medline]
-
Masuda H., Tanaka K., Takagi M., Ohgami K., Sakamaki T., Shibata N., Takahashi K. Molecular cloning and characterization of human non-smooth muscle calponin. J. Biochem. (Tokyo), 120: 415-424, 1996.[Abstract/Free Full Text]
-
Applegate D., Feng W., Green R. S., Taubman M. B. Cloning and expression of a novel acidic calponin isoform from rat aortic vascular smooth muscle. J. Biol. Chem., 269: 10683-10690, 1994.[Abstract/Free Full Text]
-
Taniguchi S., Inoue M., Nakayama J., Sadano H., Hori Y., Baba T. Differential expression of a smooth muscle -like actin between benign and malignant human pigment tissues. Cancer Lett., 47: 29-36, 1989.[Medline]
-
Inoue M., Nakayama J., Taniguchi S., Urabe K., Kuroki R., Hori Y. Malignant schwannoma in a case of type 1 neurofibromatosis with decreased immunoreactivity of smooth muscle -actin in tumor vessels. J. Dermatol. Sci., 8: 157-161, 1994.[Medline]
-
Kobayashi H., Tsuruchi N., Sugihara K., Kaku T., Saito T., Kamura T., Tsukamoto N., Nakano H., Taniguchi S. Expression of -smooth muscle actin in benign or malignant ovarian tumors. Gynecol. Oncol., 48: 308-313, 1993.[Medline]
-
Okamoto-Inoue M., Nakayama J., Hori Y., Taniguchi S. Human malignant melanoma cells release a factor that inhibits the expression of smooth muscle -actin. J. Dermatol. Sci., 20: 170-177, 2000.
-
Hayashi K., Saga H., Chimori Y., Kimura K., Ymanaka Y., Sobue K. Differentiated phenotype of smooth muscle cells depends on signaling pathways through insulin-like growth factors and phosphatidyl inositol 3-kinase. J. Biol. Chem., 273: 28860-28867, 1998.[Abstract/Free Full Text]
-
Yoshikawa H., Taniguchi S., Yamamura H., Mori S., Sugimoto M., Miyado K., Nakamura K., Nakao K., Katsuki M., Shibata N., Takahashi K. Mice lacking smooth muscle calponin display increased bone formation that is associated with enhancement of bone morphogenetic protein responses. Genes Cells, 3: 685-695, 1998.[Abstract]
-
Takahashi K., Tazunoki T., Okada T., Ohgami K., Miwa T., Miki A., Shibata N. The 5'-flanking region of the human smooth muscle cell calponin gene contains a cis-acting domain for interaction with a methylated DNA-binding transcription repressor. J. Biochem. (Tokyo), 120: 18-21, 1996.[Abstract/Free Full Text]
-
Li X., Van Putten V., Zarinetchi F., Nicks M. E., Thaler S., Heasley L. E., Nemenoff R. A. Suppression of smooth-muscle -actin expression by platelet-derived growth factor in vascular smooth-muscle cells involves Ras and cytosolic phospholipase A2. Biochem. J., 327: 709-716, 1997.
-
Blank R. S., Owens G. K. Platelet-derived growth factor regulates actin isoform expression and growth state in cultured rat aortic smooth muscle cells. J. Cell Physiol., 142: 635-642, 1990.[Medline]
-
Ueki N., Ohkawa T., Yamamura H., Takahashi K., Tsutsui T., Kawai Y., Yokoyama Y., Amuro Y., Hada T., Higashino K. Induction of calponin-h1 by transforming growth factor-ß1 in cultured human Ito cells, LI90. Biochim. Biophys. Acta, 1403: 28-36, 1998.[Medline]
-
Yokozeki M., Baba Y., Shimokawa H., Moriyama K., Kuroda T. Interferon- inhibits the myofibroblastic phenotype of rat palatal fibroblasts induced by transforming growth factor-ß 1 in vitro. FEBS Lett., 442: 61-64, 1999.[Medline]
-
Thurston G., Rudge J. S., Joffe E., Zhou H., Ross L., Croll S. D., Glazer N., Holash J., McDonald D. M., Yncopoulos G. D. Angiopoietin-1 protects the adult vasculature against plasma leakage. Nat. Med., 6: 460-463, 2000.[Medline]
-
Akedo H., Shinkai K., Mukai M., Mori Y., Tateishi R., Tanaka K., Yamamoto R., Morishita T. Interaction of rat ascites hepatoma cells with cultured mesothelial cell layers: a model for tumor invasion. Cancer Res., 46: 2416-2422, 1986.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
T. Ogura, H. Kobayashi, Y. Ueoka, K. Okugawa, K. Kato, T. Hirakawa, S. Hashimoto, S. Taniguchi, N. Wake, and H. Nakano
Adenovirus-Mediated Calponin h1 Gene Therapy Directed against Peritoneal Dissemination of Ovarian Cancer: Bifunctional Therapeutic Effects on Peritoneal Cell Layer and Cancer Cells
Clin. Cancer Res.,
September 1, 2006;
12(17):
5216 - 5223.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Burgstaller and M. Gimona
Podosome-mediated matrix resorption and cell motility in vascular smooth muscle cells
Am J Physiol Heart Circ Physiol,
June 1, 2005;
288(6):
H3001 - H3005.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Burgstaller, W. J. Kranewitter, and M. Gimona
The molecular basis for the autoregulation of calponin by isoform-specific C-terminal tail sequences
J. Cell Sci.,
May 15, 2002;
115(10):
2021 - 2029.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
ABSTRACT
INTRODUCTION
