The cell interaction sites of fibronectin in tumour metastasis

Regulation of tissue fibrosis by the biomechanical environment

The biomechanical environment plays a fundamental role in embryonic development, tissue maintenance, and pathogenesis. Mechanical forces play particularly important roles in the regulation of connective tissues including not only bone and cartilage but also the interstitial tissues of most organs. In vivo studies have correlated changes in mechanical load to modulation of the extracellular matrix and have indicated that increased mechanical force contributes to the enhanced expression and deposition of extracellular matrix components or fibrosis. Pathological fibrosis contributes to dysfunction of many organ systems. A variety of in vitro models have been utilized to evaluate the effects of mechanical force on extracellular matrix-producing cells. In general, application of mechanical stretch, fluid flow, and compression results in increased expression of extracellular matrix components. More recent studies have indicated that tissue rigidity also provides profibrotic signals to cells. The mechanisms whereby cells detect mechanical signals and transduce them into biochemical responses have received considerable attention. Cell surface receptors for extracellular matrix components and intracellular signaling pathways are instrumental in the mechanotransduction process. Understanding how mechanical signals are transmitted from the microenvironment will identify novel therapeutic targets for fibrosis and other pathological conditions.

Gene expression profiling identifies Fibronectin 1 and CXCL9 as candidate biomarkers for breast cancer screening

BACKGROUND

There is a need to develop blood-based bioassays for breast cancer (BC) screening. In this study, differential gene expression between BC samples and benign tumours was used to identify candidate biomarkers for blood-based screening.

METHODS

We identified two proteins (Fibronectin 1 and CXCL9) from a gene expression data set that included 120 BC samples and 45 benign lesions. These proteins fulfil the following criteria: differential gene expression between cancer and benign lesion, protein released in the extracellular medium and stable in the serum, commercially available ELISA kit, ELISA accuracy in a feasibility study. Protein concentrations were determined by ELISA. Blood samples were from normal volunteers (n=119) and early BC patients (n=133).

RESULTS

Seventy-three per cent of patients had cT1-T2 tumour. Patients had higher CXCL9 and Fibronectin 1 concentrations than volunteers. CXCL9 mean concentration was 851 and 635 pg ml(-1) for patients and volunteers respectively (P=0.013). CXCL9 concentration was significantly higher in patients with estrogen receptor (ER)-negative compared with volunteers (P=0.003), data consistent with gene expression profile. Fibronectin 1 mean concentration was 190 microg ml(-1) for patients and 125 microg ml(-1) for volunteers (P<0.001). Areas under the curve for BC diagnosis were 0.78 and 0.62 for Fibronectin 1 and CXCL9 respectively. A combined score including Fibronectin 1 and CXCL9 dosages presented 53% of sensitivity and 98% of specificity. Similar performances were observed for ER-negative tumours.

CONCLUSIONS

This study suggests that Fibronectin 1/CXCL9 dosage in serum could screen a significant rate of BC, including ER-negative, and that differential gene expression analysis is a good approach to select candidate biomarkers to set up blood assays cancer screening.

The monoclonal antibody SM5-1 recognizes a fibronectin variant which is widely expressed in melanoma

Background

Previously we have generated the monoclonal antibody SM5-1 by using a subtractive immunization protocol of human melanoma. This antibody exhibits a high sensitivity for primary melanomas of 99% (248/250 tested) and for metastatic melanoma of 96% (146/151 tested) in paraffin embedded sections. This reactivity is superior to the one obtained by HMB-45, anti-MelanA or anti-Tyrosinase and is comparable to anti-S100. However, as compared to anti-S100, the antibody SM5-1 is highly specific for melanocytic lesions since 40 different neoplasms were found to be negative for SM5-1 by immunohistochemistry. The antigen recognized by SM5-1 is unknown.

Methods

In order to characterize the antigen recognized by mAb SM5-1, a cDNA library was constructed from the metastatic human melanoma cell line SMMUpos in the Uni-ZAP lambda phage and screened by mAb SM5-1. The cDNA clones identified by this approach were then sequenced and subsequently analyzed.

Results

Sequence analysis of nine independent overlapping clones (length 3100–5600 bp) represent fibronectin cDNA including the ED-A, but not the ED-B region which are produced by alternative splicing. The 89aa splicing variant of the IIICS region was found in 8/9 clones and the 120aa splicing variant in 1/9 clones, both of which are included in the CS1 region of fibronectin being involved in melanoma cell adhesion and spreading.

Conclusion

The molecule recognized by SM5-1 is a melanoma associated FN variant expressed by virtually all primary and metastatic melanomas and may play an important role in melanoma formation and progression. This antibody is therefore not only of value in immunohistochemistry, but potentially also for diagnostic imaging and immunotherapy.

Cellular fibronectin concentration in the plasma of patients with malignant and benign diseases: a comparison with CA 19-9 and CEA

EDAcFN enzyme immunoassay (EIA) is a new tumour marker assay measuring the extra domain A-containing isoform of cellular fibronectin (cFN), a component mainly found in extracellular matrices. The concentration cFN was measured in plasma and serum from 468 patients with malignant and benign diseases. The concentrations of cFN were higher in plasma than in serum. Using receiver operating characteristic (ROC) curve analysis, determination from plasma was superior to serum at specificity levels higher than 78% and was chosen for further analysis. The highest frequencies of elevated cFN values were seen in patients with hepato-pancreato-biliary malignancies (50-67%). In pancreatic and bile duct cancers, cFN provided little further information to that obtained by CA 19-9. The greatest advantage over CA 19-9 and CEA was seen in patients with local colorectal cancer and in hepatocellular carcinomas. Four out of nine patients with Dukes' stage B colorectal cancer had an elevated cFn level, but only one had an abnormal CEA level. In hepatocellular carcinomas, cFN was also compared with alpha-fetoprotein. The sensitivity of cFN (72%) was superior to that of AFP (61%), and concomitant use of cFN and AFP raised the sensitivity to 83%. The highest frequencies of elevated values in patients with benign diseases were observed in those with severe liver disease (32%) and biliary (17%) and pancreatic (24%) diseases. A combination of cFN and CA 19-9 showed the highest overall sensitivity of 47%, compared with 31% for cFN and 33% for CA 19-9. The corresponding specificities were 76% for cFN +/- CA 19-9, 85% for cFN and 83% for CA 19-9. The accuracy of a combination of cFN and CA 19-9 or CEA (60% respectively) was higher than that of cFN (55%), CA 19-9 (55%) or CEA (45%) alone. In conclusion, the results of the new cFN test are encouraging and further studies on larger patient materials have been started.