Introduction

The metastasis of cancer cells is a very complicated process, involving multiple steps. During this process, cancer cells interact with basement membranes (BM), invade and pass through them at least three times1. Laminin (LN), a major noncollagenous glycoprotein in the BM, plays important roles in cancer metastasis1, 2. In vitro, LN enhances the attachment, spreading1, migration3 of many kinds of cancer cells. LN also induces cancer cells to secret type IV collagenase and disrupts BM structures4.

The biological function of LN is mediated by LN receptors on the cell surface. A number of these receptors have the characteristics of carbohydrate binding proteins5, indicating the possible role of carbohydrate moieties of LN. It has been reported that LN carbohydrates participated in the cell attachment6 and cell spreading7 of cancer cells. Furthermore, we have found that certain sugars, including N-acetyl-neuranminic acid, α-methyl-D-mannose, lactose and N-acetyl-lactosamine, but not glucose, galactose and mannose, inhibited the binding of a various cancer cells to LN8. Therefore, it seems that carbohydrate chains may be involved in the recognition between cancer cells and the MB. On the basis of these studies, the glycopeptides were isolated from LN and biological experiments demonstrated that LN-GPs were powerful inhibitors for blocking the interactions between cancer cells and LN, and also for blocking the interaction between the isolated 67kDa LN receptor and LN9. Later, we further documented the ability of LN-GPs to prevent the cancer metastasis in an in vivo system, the experimental lung and liver metastasis of mouse melanoma cells10, 11.

In this paper, the in vitro effects of LN-GPs on metastasis-related behaviors of cancer cells, including cell attachment, spreading, migration, and invasiveness will be reported.

Materials and Methods

Materials

Cycloheximide (CH), HEPES, MTT, PI, Agarose and BSA were purchased from Sigma; RPMI-1640 was purchased from JR Scientific (USA); LN was prepared in our laboratory from mouse Engelbreth-Holm-Swarm (EHS) tumor as described by Timple12; Matrigel was prepared in our laboratory from EHS tumor as described by Kleinman13; LN-GPS was prepared in our laboratory from LN6. NIH3T3 chemotatic factor was prepared as described by Albini14.

Mouse sarcoma cell line S180 was obtained from the Department of Cell Biology, Beijing Normal University; Human gastric carcinoma cell line BGC-823 was obtained from the Department of Pharmacology, Beijing Medical University (BMU); Human pulmonary giant carcinoma cell line PG was obtained from the Department of Pathology, BMU. All cell lines were cultured in RPMI-1640 supplemented with 10% fetal calf serum, penicillin (100 unit/ml), streptomycin (100 μg/ml) and 2% HEPES.

Scanning Electronic Microscopy S-450 is from Hitachi (Jap); Flow Cytometer FACS-440 is from B.D. Co (USA) and 8 mm pore size Nucleopore Polycarbonate Filters without PVP from Poretics Co. (USA) were used in this work; Boyden Chamber was made in China according to the model purchased from USA.

MTT assay

4μg LN was coated onto the bottom of 96-well plastic plates as previously reported by Zhou et al1. Control wells contained no LN. 1% BSA in PBS was then added (overnight at 4 °C) to block the unsaturated plastic surface. S180 cells were washed 3 times with PBS and incubated with serum free CH (25 μg/ml)-RPMI-1640 at 37 °C in 5% CO2 for 4 h to prevent the production of endogenous proteins. Then, S180 cells were harvested and resuspended in serum free CH-RPMI-1640 at a concentration of 4 × 105 cells/ml. Cell viability detected by Trypan Blue exclusion was found to be > 98%. 100 μl cell suspension in the presence or absence of LN-GPs (50 μg/ml) was added to wells and incubated at 37 °C in 5% CO2 for 18 h. MTT assay15 was performed as following: 20 μl MTT (5 mg/ml) was added to each well and incubated with cells at 37 °C for 2 h. 50 μl SDS (10%) was then added to each well and incubated further at 37 °C for 6 h. The value of OD 570 was measuned by ELISA reader, quadruplicates were measured in each sample.

Trypan blue exclusion assay

BGC-823 cells were treated as S180 cells in MTT assay and resuspended in serum free CH- RPMI-1640 at 5 × 105 cells/ml. Cells in the presence or absence of LN (20 μg/ml) and LN (20 μg/ml) plus LN-GPs (100 μg/ml) were incubated at 37 °C in 5% CO2 for 24 h. Cell viability was detected either before or after incubation on the basis of trypan blue exclusion.

Cell attachment assay

20 μg LN in PBS or 50 μg Matrigel in PBS was added into glass vials with flat bottom (Φ2.2 cm) to prepare LN or Matrigel substrate coating1. Uncoated bottom of glass vials was used as control. 1% BSA in PBS was added to each glass vials (overnight at 4 °C). The cell suspension of S180 cells (2 × 105 cells/ml) was prepared as in MTT assay and the cell viability was > 98%. Cells (2 × 105) in the presence of various doses of LN-GPs (0, 12.5, 25, 37.5 μg) were added into glass vials in a final volume of 0.5 ml and incubated at 37 °C in 5% CO2 for 6 h. Unattached cells were discarded, the number of attached cells was determined by detecting the activity of LDH1. All experiments were performed in triplicates.

Cell spreading assay

This experiment was performed similarly to cell attachment assay. The difference was that S180 cells were incubated at 37 °C in 5% CO2 for 2, 4, or 6 h. The percentage of spreading cells was calculated under phase contrast inverted microscope.

Cell surface morphology

The surface of glass coverslip (11 × 11 mm) was coated with 10μg LN and put into glass vials with LN-coated surface upward. Uncoated coverslips were used as control. 1% BSA in PBS was added to each glass vials (overnight at 4 °C). Cell suspension of S180 cells (5 × 106 cells/ml) was prepared with viability > 98%. 5 × 105 cells in the presence or absence of LN-GPs (37.5μg) were added to each glass vials in a final volume of 0.5 ml and incubated at 37 °C in 5% CO2 for 6 h. Unattached cells were discarded, attached cells were treated as the method used to prepare SEM samples. Then, the cell surface was observed under SEM (S-450). Triplicates were prepared for each sample.

Cell migration assay

The method used here was referred to agarose drop method described by Varani et al16. S180 cells were washed 3 times with PBS and suspended in serum free RPMI-1640. The cell viability was > 96%. The cell pellet (108 cells/ml) was added to serum free RPMI-1640 containing 0.5% (W/V) agarose at the ratio of 1:3 (V/V). 1 ∼ 2 μl droplets of such cell suspension contained either LN (20 μg/ml), LN (20 μg/ml) plus LN-GPs (100 μg/ml), Matrigel (50μg/ml), or Matrigel (50 μg/ml) plus LN-GPs (100 μg/ml) were delivered into the wells of 24-well plates. In control group, the cell droplets contained no supplemented material(s) as just described. The 24-well plates were then placed at 4°C for 10 min. The overlay medium with the same serum free RPMI-1640 medium as the agarose droplets was added into each well. Each sample was performed in quadruplets. The 24-well plates were incubated at 37 °C in 5% CO2 for 18 h. The distance between the drop edge and the front of migrating cells was measured by phase contrast inverted microscopy via calibrated grid mounted with in the eyepiece.

Invasion assay

This experiment was performed as described by Albini et al14. The highly metastatic PG cells were washed 3 times with PBS and resuspended in serum free RPMI-1640 at 5 × 105 cells/ml. The cell viability was > 96%. Cell suspensions in the presence or absence of LN-GPs (100 μg/ml) were preincubated at 37 °C in 5% CO2 for 1 h before use. 200 μl NIH3T3 chemotatic factor was added into blind wells of Boyden chamber. Filters (Φ13mm) were overlayed onto the blind wells, the chambers were assembled. Matrigel (120 μg) was added onto filters at 0 °C and then placed at 37 °C for 30 min. Cells (2 × 105) were then put into top wells of Boyden chamber and incubated at 37 °C in 5% CO2 for 6 h. Then filters were removed from chambers and stained with hematoxylin. Each sample were in quadruplicates. The number of cells that invaded into the filter was counted under microscope for five high-power fields (hpf).

Results

Effects of LN-GPs on cancer cell survival

The results of MTT Assay (Tab 1) showed that there were no obvious difference between S180 cells incubated for 18 h in the presence or absence of LN-GPs (50 μg/ml), suggesting that LN-GPs had no toxic effect on cancer cell survival. The trypan blue exclusion assay also demostrated that there was no difference in cell viability either before or after incubation of BGC-823 cancer cells with LN-GPs (100 μg/ml) for 24 h (data not shown). Thus, it seems that LN-GPs used have no cytotoxic effect on either S180 sarcoma or BGC-823 carcinoma cells.

Tab. 1 The OD570 values of S180 cells incubated in the presence or absence of LN-GPs (50 μg/ml) at 37 °C in 5 % CO2
Full size table

Effects of LN-GPs on metastasis-related behaviors of cancer cells

The results of cell attachment assay as shown in Fig 1 indicated that in the presence of LN-GPs for 6 h, the attachment of S180 cells on LN or Matrigel substrate was inhibited in a dose-dependent manner, however cell attachment on Matrigel substrate was inhibited much more strongly.

Fig. 1
figure 1

Does-dependent effects of LN-GPs on cell attachment of S180 cells on LN and Matrigel substrate. The cells were incubated in the presence of various concentration of LN-GPs (0,25,50,75 μg/ml) at 37°C in 5% CO2 for 6h.

Full size image

From the cell spreading assay we found that after the treatment of S180 cells with various concentrations of LN-GPs for a definite period of time, a dose-dependent and a time-dependent inhibition of cell spreading on LN substrate (Figs 2, 3 A) or Matrigel substrate (Figs 2, 3 B) were observed. Similar to cell attachment, cell spreading on Matrigel substrate was also much more greatly inhibited.

Fig. 2
figure 2

Does-dependent effects of LN-GPs on cell spreading of S180 cells on LN (A) and Matrigel (B) substrates. The cells were incubated in the presence of various concentration of LN- GPs(0,25,50,75 μg/ml) at 37°C in 5% CO2 for 2,4 and 6 h.

Full size image
Fig. 3
figure 3

Time-dependent effects of LN-GPs on cell spreading of S180 cells on LN (A) and Matrigel (B) substrate. The cells were incubated in the presence of various concentration of LN-GPs (-•0μg, -□25μg, -△50μg, -○-75μg) at 37 °C in 5 % CO2 for 2, 4 and 6 h.

Full size image

From the observation of cell surface under SEM, it was noticeable that the S180 cells attached onto bare glass showed round shape with numerous ruffles (Fig 4 A); while the cells attached onto LN substrate displayed polygonal and flattened shape with long pseudopodia and fewer ruffles (Fig 4 B). After the treatment with LN- GPs (50 μg/ml) for 6 h, cells attached onto LN substrate showed polygonal or round shape with shorter pseudopodia and numerous ruffles, just similar to the cell surface morphology of some cells attached onto bare glass (Fig 4 C).

Fig. 4
figure 4

Cell surface morphology of S180 cells observed under SEM. the cells were incubated at 37 °C in 5 % CO2 for 6 h. A: cells attached onto bare glass showed round shape with numerous ruffles (X6000); B: Cells attached onto LN substrate showed polygonal and flattened shape with long pseudopodia and fewer ruffles (X3000); C: Cells attached onto LN substrate in the presence of LN-GPs (75 μg/ml) showed polygonal or round shape with shorter pseudopodia and numerous ruffles (X4000).

Full size image

The results of cell migration assay demonstrated that in the control group, almost all cells maintained in the original position and only a few cells traversed the drop edge. LN (20 μg/ml) and Matrigiel (50 μg/ml) promoted the migration of S180 cells to the drop edge and increased greatly the number of migrating cells. However, LN-GPs (100 μg/ml) inhibited both the promoting effects of LN or Matrigel on cell migration (Fig 5).

Fig. 5
figure 5

Effects of LN, Matrigel and LN- GPs on the migration of S180 cells. The cells were incubated in the presence of either LN (20 μg/ml), or LN + LN-GPs (100 μg/ml); Matrigel (50 μg/ml), or Matrigel + LN-GPs (100 μg/ml), as compared with that of the control without such materials at 37 °C in 5 % CO2 for 18 h.

Full size image

In the presence of LN-GPs (100 μg/ml), the number of highly metastatic PG cells that invaded into filters with Matrigel on the surface was 18.8 ± 3.96 (X ± SD)/hpf (Fig 6 A), in comparison with a value of 43.8 ± 10.71 (X ± SD)/hpf, for the control without LN-GPs (Fig 6 B), indicating the strong inhibition of PG cells invasion by LN-PGs.

Fig. 6
figure 6

The detection of invasion of PG cells. The cells were incubated at 37 °C in 5 % CO2 for 6 h. A: in the presence of LN-GPs (100 μg/ml); B: in the absence of LN-GPs. The invaded cells were marked by arrows.

Full size image

Discussion

It has been reported that the carbohydrate moieties of LN are involved in the cell attachment6 and spreading7 of different cancer cells. Matrigel is an artificial BM preparation extracted from mouse EHS tumor. The major component of Matrigel is LN, the others are collagen IV, heparin sulfate proteoglycans, and entactin /nidogen13. Because Matrigel is more like BM than LN, it is also used as substrate for cell attachment and spreading assays. It is evident from the present work, both Matrigel and LN can promote cell attachment, spreading and migration; LN-GPs can inhibit the promoting effects of LN and Matrigel. The inhibitory effects on S180 cells plated onto Matrigel substrate were found to be much more strongly than those onto LN substrate. The reason is not well understood.

Both the promotion of cell spreading by LN and its inhibition by LN-GPs are time-dependent. The explanation for this phenomenon is very complicated. It may involve not only the assembly and disassembly of cell cytoskeleton, but also signal transduction and gene expression. Recently, LN has been considered as a signal molecule17. The binding of the LN receptors to LN may trigger signal transduction, and it is possible that LN-GPs may bind competitively to LN receptors. Integrin mediated FAK-MAPK pathway is very important in signal transduction triggered by binding of extracellular matrix components to integrins17. Recently, we have found that the focal adhesion formation and FAK expression in focal adhesion can be promoted by LN and inhibited by LN-GPs, and the expression of a number of genes can be specifically induced by putting cancer cells onto LN, and the LN-GPs can inhibit some of such induced gene expression (unpublished data). Therefore, cell spreading is a phenomenon coupled not only to cytoskeleton assembling, but also to many other intracellular biochemicel events, including gene expression.

Moreover, cancer cell spreading on LN or Matrigel substrate was inhibited much more easily than cancer cell attachment. Since the first identification of the 67kDa LN receptor, a large number of LN receptors have now been reported18. The interaction of cancer cells with LN would be mediated differentially by the bindings of different LN receptors to different parts of LN molecule and carbohydrate chains are involved in some of the binding. It is possible that different combination of receptors may be involved in cell attachment and cell spreading; also different components of LN-GPs may interrupt differentially to cell attachment and cell spreading. To test such posibility, the LN-GPs should be further fractionated, and the biological effects of various active components should be studied separately.

Likewise, the different cell surface appearances of cells on LN in the presence or absence of LN-GPs under SEM may indicate that LN-GPs can alter the assembly and disassembly of the cytoskeleton, thus affecting not only cell shape, but also cell spreading, migration and invasiveness.

The mechanism of inhibition of the cancer cell invasiveness induced by LN-GPs is also difficult to explain at the moment. It was reported that the attachment onto LN can induce malignant cancer cells to secrete type IV collagenase and disrupt the basement membrane5. Possibly, LN-GPs, via its interfering with the binding between LN and LN receptors, can decrease the ability of PG cells to secrete type IV collagenase, thereby decreasing their invasion ability.

Taken together, the results of in vitro studies on the effect of laminin glycopeptides on metastatic behaviors of different cancer cells suggest that LN-GPs can manifest multiple anti-metastatic roles. It is hoped that further sustained studies on the basic mechanism of the effect of LN-GPs, on cancer cells may contribute from a different angle to our understanding towards cancer cell invasion and metatasis in vivo.