QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 30 January 2007
doi: 10.1242/jcs.000273

This Article Summary Full Text Full Text (PDF) Supplementary Material Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Cavanna, T. Articles by Zicha, D. Search for Related Content PubMed PubMed Citation Articles by Cavanna, T. Articles by Zicha, D.

Evidence for protein 4.1B acting as a metastasis suppressor

¹ Light Microscopy, Cancer Research UK London Research Institute, Lincoln's Inn Fields Laboratories, 44 Lincoln's Inn Fields, London, WC2A 3PX, UK
² Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, Flemingovo nám. 2, 166 37 Prague 6, Czech Republic

View larger version (8K): [in this window] [in a new window]   Fig. 1. Comparison of the metastatic potential among the sarcoma cell populations. Cell populations K2, T15, A297 and A311 were subcutaneously implanted (one-million cells per transplant), into inbred Lewis rats. Primary tumours developed in all cases. The metastatic potential – the percentage of subjects developing metastases – was calculated for each cell population. The K2 cells had no metastatic potential, whereas T15, A297 and A311 cells had metastatic potentials of 38%, 90% and 100%, respectively. The metastases were found in the lungs. The numbers of rats used to test the K2, T15, A297 and A311 cells were 20, 24, 10 and 10, respectively; 2 **P<0.01.

View larger version (33K): [in this window] [in a new window]   Fig. 2. Motile responses of the sarcoma cells to PDGF-IGF. (A) Distribution of directions of cell translocation in a gradient of PDGF-IGF in the Dunn chemotaxis chamber is represented by circular histograms. The circular histograms are aligned such that the concentration of PDGF-IGF in the gradient increases vertically towards the top. Translocation of a cell from its starting point was evaluated and its direction was taken on migrating 70 µm. Cells not migrating 70 µm were not included in this analysis. Significant unimodal clustering of directions according to the Rayleigh test is indicated by an arrow identifying the mean direction, and a shaded wedge representing its 95% confidence interval. K2 cells are weakly chemotactic to PDGF-IGF, whereas T15, A297 and A311 cells have strong chemotactic responses (P<0.005). Data from left to right are from K2, T15, A297 and A311 cells and represent 5, 4, 5 and 5 independent experiments, respectively, each recorded over a period of 16 hours. The total numbers (fractions) of K2, T15, A297 and A311 cells included in the analysis are 61 (36%), 81 (70%), 86 (49%) and 47 (42%), respectively. (B) Histograms to illustrate the distribution of speed of cell migration within the four cell populations. The non-metastatic K2 cells moved at 11.1±0.5 µm/hour, whereas the metastatic T15 and A297 cells moved significantly faster,at 17.1±0.7 µm/hour (ANOVA P<0.05) and 19.1±0.9 µm/hour (ANOVA P<0.01), respectively. There was no significant difference between the speed of cell migration in K2 and A311 cells, which migrated at 14.1±0.9 µm/hour. However, the proportion of cells moving faster than 20 µm/hour was significantly increased (2 P<0.01) in A311 cells (20%) compared with K2 cells (8%). The speed data were also derived from the experiments with the Dunn chemotaxis chamber without excluding any cells. The total numbers of K2, T15, A297 and A311 cells were 171, 116, 176 and 113, respectively, and numbers of 5-minute displacements were 21340, 13390, 20179 and 15436, respectively.

View larger version (32K): [in this window] [in a new window]   Fig. 3. Differences in F-actin organisation among the sarcoma cell populations. (A) These single confocal images of Rhodamine-phalloidin-stained cells reveal differences among the sarcoma cell populations. Lower magnification images (right panels) demonstrate that these phenotypes are representative across a larger number of cells. F-actin in K2 (non-metastatic cells) is arranged into stress fibres, which can be quite thick and can traverse the cell. T15, A297 and A311 cells typically have F-actin concentrated into ruffles at the leading edge of the cell; elsewhere it is arranged cortically and in a disordered manner. These cells are often polarised. Bars, 20 µm. (B) The percentage of actin stress fibre-containing cells in each population. Quantification of randomly acquired fields of cells revealed that 89±8% of K2 cells contain at least one stress fibre, compared with only 7±2%, 3±8% and 7±8% of T15, A297 and A311 cells, respectively. The numbers of cells scored for K2, T15, A297 and A311 were 34, 84, 58 and 69, respectively. The individual differences in stress fibre distribution between the T15, A297 and A311 compared with K2 cells are all significant (***P<0.001).

View larger version (78K): [in this window] [in a new window]   Fig. 4. RT-PCR was performed for ten of the candidate genes to validate the microarray data. PCR products were resolved on a 1% TBE agarose gel supplemented with ethidium bromide, and visualized by UV trans-illumination. We observed the expression patterns that had been indicated by the microarray. Gapdh was used as control and was expressed uniformly.

View larger version (25K): [in this window] [in a new window]   Fig. 5. Expression of Epb41l3, encoding 4.1B protein, in the sarcoma cells. (A) Normalised expression levels of the Epb41l3 gene are shown for the cultured cells and the primary tumours. The normalisation was based on median values from individual microarrays. Each dot and bar represents the mean value ± s.e.m. of the expression levels obtained from multiple microarrays. Epb41l3 is significantly downregulated in the metastatic cells (P<0.05), and this expression pattern is closely matched by that of the tumours. (B) The difference in expression also occurs at the protein level, as seen in this western blot, in which very little 4.1B is detectable in cultured T15, A297 and A311 cells. Actin was used as a loading control.

View larger version (90K): [in this window] [in a new window]   Fig. 6. Cellular localisation of GFP-4.1B fusion protein. To investigate the localisation of 4.1B, cells were transfected with GFP-4.1B fusion construct and, 24 hours later, fixed and stained with Rhodamine-phalloidin. (A-D) Sarcoma cells K2. (E-L) HeLa cells. (A) GFP-4.1B is expressed in the cytoplasm and is enriched at the plasma membrane. (B,C) There is also some colocalisation with F-actin especially at the plasma membrane compartment enriched with GFP-4.1B. The brighter area of GFP-4.1B between the two cells in A is probably a result of the two cells overlapping, suggested by the phase-contrast image (D). (E,F) Localisation of 4.1B in HeLa cells is similar, with cytoplasmic expression and some colocalisation with F-actin (G), especially in peripheral areas of ruffling and protrusive activity, as seen in the enlarged region (H; boxed region in G). There is also an enrichment of GFP-4.1B at the membrane (I-K). This is clearer still in the close-up (L) of the boxed region in (I). All images are single confocal images and representative of three independent experiments. Bar, 20 µm.

View larger version (76K): [in this window] [in a new window]   Fig. 7. Distribution of GFP-4.1B and focal adhesions in a T15 cell. The cell was microinjected with GFP-4.1B fusion construct, fixed after 5 hours and analysed under a laser scanning confocal microscope. (A-D) We simultaneously acquired an image of GFP-4.1B at the substrate level (A) and an interference reflection microscopy image (B) in which focal adhesions appear dark. C and D are enlarged images of boxed regions in A and B, respectively. They reveal enrichment of GFP-4.1B in many but not all focal adhesions. Arrows indicate examples of focal adhesions with an increased concentration of GFP-4.1B. Bar, 20 µm.

View larger version (73K): [in this window] [in a new window]   Fig. 8. Depletion of 4.1B by RNAi causes disruption of the F-actin cytoskeleton. (A) K2 cells injected with Cy3-labelled siRNA targeting rat Epb41l3 (R-epb41l3 siRNA) or control siRNA were fixed 48 hours later and labelled with Alexa Fluor-488–phalloidin. Cells with siRNA targeting rat Epb41l3 displayed a disrupted F-actin cytoskeleton, with fewer stress fibres and less total F-actin. Cells transfected with Cy3-labelled control siRNA were unchanged. (B) Similarly, HeLa cells transfected with three independent siRNAs targeting human EPB41L3 (H-epb41l3 siRNA1, H-epb41l3 siRNA2 and H-epb41l3 siRNA 3) exhibited fewer stress fibres after 48 hours of incubation. Some fine stress fibres remained, but the majority of cells had a disordered F-actin cytoskeleton and few F-actin; images are representative of five independent experiments. F-actin was labelled with Rhodamine-phalloidin. (C) Western blot confirming that HeLa cells transfected with three independent siRNAs targeting human EPB41L3 expressed reduced levels of 4.1B after 48 hours. Densitometric quantification showed that the actin-normalised presence of 4.1B was reduced to 44%, 18% and 28% when using H-epb41l3 siRNA1, H-epb41l3 siRNA2 and H-epb41l3 siRNA3, respectivley. (D) EPB41L3 siRNA-treated HeLa cells transfected with GFP cDNA as a control exhibited lack of stress fibres similarly to the experiments with siRNA on its own shown in B. EPB41L3 siRNA-treated HeLa cells subsequently transfected with an RNAi-resistant 4.1B cDNA recovered stress fibres and organised F-actin structures to similar levels observed in control siRNA treated cells shown in B. This rescue experiment confirms that the siRNAs were specific; images shown in D are representative of three independent experiments. F-actin was labelled with Rhodamine-phalloidin. (E) Quantitative analysis of HeLa cells with actin stress fibres showed that siRNA targeting human EPB41L3 resulted in a significant reduction, from half of the cells to less then a quarter on average (ANOVA ***P<0.001; the numbers of cells evaluated for control siRNA were 48 and 427 for siRNA targeting human EPB41L3). EPB41L3 siRNA-treated HeLa cells with and without GFP cDNA transfection had similar significantly reduced numbers of cells with actin stress fibres (ANOVA ***P<0.001; 216 siRNA-treated cells transfected with GFP cDNA were evaluated). EPB41L3 siRNA-treated HeLa cells transfected with an RNAi-resistant 4.1B cDNA did not have significantly changed number of cells with actin stress fibres compared with the control siRNA (n.s., no significance in ANOVA; 766 siRNA-treated cells transfected with RNAi-resistant 4.1B cDNA were evaluated). Cells that lost actin stress fibres exhibited disordered F-actin, with F-actin puncta and cortical F-actin organisation.

View larger version (10K): [in this window] [in a new window]   Fig. 9. Effect of 4.1B on the motility of the sarcoma cells. (A) Non-metastatic K2 cells were injected with Cy3-labelled R-epb41l3 siRNA1 and recorded 48 hours later by low light level digital microscopy. Cell translocations were interactively tracked, and speeds of cell migration evaluated in Mathematica. Speed of cell migration was increased twofold in cells treated with siRNA targeting epb41l3 compared with those treated with control siRNA (ANOVA P<0.01; mean ± s.e.m. of normalised speed of cell migration was 1.0±0.1 and 2.2±0.2 for control siRNA and siRNA targeting Epb41l3, respectively; a total of 49 cells from ten time-lapse recordings was analysed). (B) GFP cDNA, or a mixture of GFP and 4.1B cDNA, were microinjected into metastatic T15 cells and after 5 hours cell movements recorded by low light level digital microscopy. Cells expressing both 4.1B and GFP migrated at half the speed of the cells injected with GFP alone (ANOVA P<0.01; mean ± s.e.m. of normalised speed of cell migration was 1.3±0.1 and 0.7±0.1 for GFP cDNA alone and for mixture of GFP and 4.1B cDNA, respectively; a total of 49 cells from nine time-lapse recordings was analysed).

View larger version (16K): [in this window] [in a new window]   Fig. 10. Effect of exogenous 4.1B expression on the chemotaxis of T15 sarcoma cells. The motility of cells in the Dunn chemotaxis chamber was recorded 5 hours after microinjection and evaluated in a way similar to experiments presented in Fig. 2A. (A,B) Cell tracks with starting points shifted to a common origin are for uninjected cells (A) and in cells expressing a mixture of 4.1B and GFP (B). Bar, 100 µm. The speed of cell migration of the expressing cells was significantly reduced 2.3 times (ANOVA P<0.02; numbers of time-lapse recordings and cells were 14 and 41, respectively, for uninjected cells, and 9 and 29, respectively, for cells expressing 4.1B and GFP). (C,D) Circular histograms of cells that migrated 10 µm for uninjected cells (C) and for cells expressing 4.1B and GFP (D). Arrow (mean direction) and a grey wedge (its 95% confidence interval) indicate unimodal significant clustering of the directions (Rayleigh test P<0.001). The chemotactic response, measured by the displacement in the gradient direction, is significantly different (ANOVA P<0.05; numbers of cells that migrated 10 µm were 41 for uninjected cells and 27 for cells expressing 4.1B and GFP).