Tumour-associated tenascin-C isoforms promote breast cancer cell invasion and growth by matrix metalloproteinase-dependent and independent mechanisms

Abstract

Introduction: The stromal microenvironment has a profound influence on tumour cell behaviour. In tumours, the extracellular matrix (ECM) composition differs from normal tissue and allows novel interactions to influence tumour cell function. The ECM protein tenascin-C (TNC) is frequently up-regulated in breast cancer and we have previously identified two novel isoforms - one containing exon 16 (TNC-16) and one containing exons 14 plus 16 (TNC-14/16).

Methods: The present study has analysed the functional significance of this altered TNC isoform profile in breast cancer. TNC-16 and TNC-14/16 splice variants were generated using PCR-ligation and over-expressed in breast cancer cells (MCF-7, T47D, MDA-MD-231, MDA-MB-468, GI101) and human fibroblasts. The effects of these variants on tumour cell invasion and proliferation were measured and compared with the effects of the large (TNC-L) and fully spliced small (TNC-S) isoforms.

Results: TNC-16 and TNC-14/16 significantly enhanced tumour cell proliferation (P < 0.05) and invasion, both directly (P < 0.01) and as a response to transfected fibroblast expression (P < 0.05) with this effect being dependent on tumour cell interaction with TNC, because TNC-blocking antibodies abrogated these responses. An analysis of 19 matrix metalloproteinases (MMPs) and tissue inhibitor of matrix metalloproteinases 1 to 4 (TIMP 1 to 4) revealed that TNC up-regulated expression of MMP-13 and TIMP-3 two to four fold relative to vector, and invasion was reduced in the presence of MMP inhibitor GM6001. However, this effect was not isoform-specific but was elicited equally by all TNC isoforms.

Conclusions: These results demonstrate a dual requirement for TNC and MMP in enhancing breast cancer cell invasion, and identify a significant role for the tumour-associated TNC-16 and TNC-14/16 in promoting tumour invasion, although these isoform-specific effects appear to be mediated through MMP-independent mechanisms.

Figure 1

Schematic diagram of tenascin-C. The domain structure of tenascin-C comprising N-terminal tenascin assembly (TA) domain followed by 14.5 epidermal growth factor (EGF)-like repeats, the fibronectin type III (FN III)-like repeats and the carboxy fibrinogen-like domain. The FN III region consists of eight conserved repeats, designated 1 to 8, and up to nine alternatively spliced FN III repeats designated as letters A to D, AD1 and AD2 (shown in shading). The exon organisation of the tenascin-C gene is shown below.

Figure 2

PCR-ligation strategy for generation of TNC-16 and TNC-14/16 clones. (a) Primers designed incorporating unique restriction sites were used to link exons 9 to 16 in a three-step polymerase chain reaction (PCR)-mediated ligation strategy. This allowed directional cloning from the Bcl1 and Sfi1 sites into the TNC-S sequence and this sequence was subsequently transferred into the mammalian expression vector pCMV script. The gel image shows exon 9 and exon 16 products and the combined 9–16 amplicon. (b) The same PCR-mediated ligation strategy was used to link exons 9, 14 and 16 prior to directional cloning into the TNC-S sequence for expression. The gel image shows the multi-step process used to link exons 9, 14 and 16.

Figure 3

Confirmation of expression of tenascin-C isoforms. (a) Endogenous expression of tenascin-C (TNC) isoforms in untransfected cell lines. Normalised relative expression of TNC isoforms were determined by reverse transcriptase polymerase chain reaction (RT-PCR) using primers and probes to invariant exon 17/18 boundary (Total TNC), the 9/16 (TNC-16) and 14/16 (TNC-14/16) exon boundary for breast cell lines ZR-75-1, MCF-7, T-47D, GI-101, Hs578T, MDA-MB-231, MDA-MB-436, MDA-MB-468 and melanoma cell line SKMel-28. Relative expression was calculated by comparison with standard curves derived from TNC-9-16 and TNC 9–14–16 recombinant clones to correct for differences in PCR efficiency for each TNC probe set normalised using level of housekeeping gene expression to correct for any differences in cellularity. (b) RT-PCR of primary fibroblasts and MCF-7 cells transiently transfected with TNC-S, TNC-L, TNC-16, TNC-14/16 and vector-only control (vector), using primers spanning the FN III alternatively spliced region (8F/18R). This shows appropriately sized bands in each of the cell populations, with no product in vector-only and non-transfected MCF-7 controls, although there was evidence of low level expression of TNC-S in vector only and non-transfected fibroblast controls. (c) Immunohistochemistry for anti-Flag M2 antibody in MCF-7 cells transfected with TNC-L (left image) and for TNC (Monoclonal, BC24) in MCF-7 transfected with vector control (right image). An average transfection efficiency of 35% was determined for each TNC isoform and staining confirmed that MCF-7 cells do not express TNC. (d) Western blot analysis for TNC in transiently transfected MCF-7 cells. This demonstrated a single species of TNC present in the whole cell lysate (WCL; i) and conditioned media (CM; ii) of transfected cells for each isoform. TNC-S is seen as a band at about 250 kDa, with slightly larger bands detected for TNC-16 and TNC 14/16, while TNC-L is detected at about 350 kDa.

Figure 4

Direct effects of tenascin-C on tumour cell invasion. (a) Invasion of MDA MB 231, MCF-7, T47D, MDA MB 468 and GI101 cell lines transfected with four different tenascin-C (TNC) isoform constructs and the vector alone. MDA-MB-231 and T47D cells transfected with TNC-L show a higher mean invasion index (MII) than controls (P < 0.05). All cell lines exhibited significantly higher MII with TNC-16 and TNC-14/16 compared with the vector alone (*P < 0.05, **P < 0.01, ***P < 0.001), other than GI101, which showed significantly higher invasion with TNC-14/16 only (P = 0.01). (b) Effect of co-transfection of TNC-S with either TNC-16 or TNC-14/16 isoforms on MCF-7 cell invasion. Both TNC-16 and TNC-14/16 led to an increased MCF-7 MII (*P < 0.01) compared with vector alone and other isoforms. However, there was no further enhancement in MII when the MCF-7s were co-transfected with TNC-S and either TNC-16 or TNC-14/16 (ns = not significant). (c) Proliferation of MDA MB 231, MCF-7 and T47D cell lines transfected with four different TNC isoforms constructs and the vector alone. MBA-MB-231 cells exhibit no significant changes in proliferation when transfected with any of the isoforms. MCF-7 and T47D both show increased proliferation when transfected with TNC-14/16 compared with vector alone (*P < 0.05). MCF-7 exhibited an increase in proliferation with TNC-16 compared with vector alone, whereas T47Ds did not show any changes with TNC-16 but did have a significant (P < 0.05) increase in proliferation with TNC-L. (d) Effect of co-transfection of TNC-S with either TNC-16 or TNC-14/16 isoforms on MCF-7 cell proliferation. Both TNC-16 and TNC-14/16 lead to an increased MCF-7 proliferation (*P < 0.01) compared with vector alone and other isoforms. However, there was no further enhancement in proliferation when the MCF-7s were co-transfected with TNC-S and either TNC-16 or TNC-14/16 (ns = not significant). In all cases, the bars indicate the mean of at least three experiments with standard errors shown.

Figure 5

Effect of fibroblast-associated tenascin-C expression on MCF-7 tumour cell invasion. (a) Mean invasion index (MII) for MCF-7 tumour cell co-cultured with primary breast fibroblasts (n = 5) transfected with tenascin-C (TNC) isoforms constructs or vector alone. Fibroblasts expressing TNC-16 or TNC-14/16 significantly increased MII compared with TNC-L (P = 0.05 and P = 0.001, respectively), TNC-S and vector alone (***P < 0.001). Higher MII was also seen with fibroblast expressing TNC-L (P < 0.001) compared with TNC-S and vector alone. Bars indicate the mean of five donors, each measured in triplicate. (b) MCF-7 cell MII using conditioned media from primary breast fibroblasts transfected with TNC isoforms constructs or the vector alone in the presence of blocking TNC mouse monoclonal antibody BC-24 or equivalent IgG control. The blocking antibody significantly reduced MII (*P < 0.05) independent of the TNC isoform. In all cases, the bars indicate the mean of three experiments with standard errors shown. (c) GI101 cell MII using conditioned media from fibroblasts transfected with TNC isoforms constructs or the vector alone in the presence of blocking TNC mouse monoclonal antibody BC-24 or equivalent IgG control. The blocking antibody significantly reduced MII (*P < 0.05, **P < 0.01) in all the cases except for TNC-16. In all cases, the bars indicate the mean of three experiments with standard errors shown. (d) MCF-7 cells transfected with different TNC isoforms or vector only all exhibited significant reduction in MII in the presence of the matrix metalloproteinase inhibitor GM6001 (***P < 0.001). Bars indicate mean of three experiments.