Impaired intercellular adhesion and immature adherens junctions in merlin-deficient human primary schwannoma cells.

Schwannomas that occur spontaneously or in patients with neurofibromatosis Type 2, lack both alleles for the tumor suppressor and plasma membrane-cytoskeleton linker merlin. We have shown that human primary schwannoma cells display activation of the RhoGTPases Rac1 and Cdc42 which results in highly dynamic and ongoing protrusive activity like ruffling. Ruffling is an initial and temporally limited step in the formation of intercellular contacts like adherens junctions that are based on the cadherin-catenin system. We tested if there is a connection between Rac1-induced ongoing ruffling and the maintenance, stabilization and functionality of adherens junctions and if this is of relevance in human, merlin-deficient schwannoma cells. We show intense ongoing ruffling is not limited to membranes of single human primary schwannoma cells, but occurs also in membranes of contacting cells, even when confluent. Live cell imaging shows that newly formed contacts are released after a short time, suggesting disturbed formation or stabilization of adherens junctions. Morphology, high phospho-tyrosine levels and cortactin staining indicate that adherens junctions are immature in human primary schwannoma cells, whereas they display characteristics of mature adherens junctions in human primary Schwann cells. When merlin is reintroduced, human primary schwannoma cells show only initial ruffling in contacting cells and adherens junctions appear more mature. We therefore propose that ongoing Rac-induced ruffling causes immature adherens junctions and leads to impaired, nonfunctional intercellular adhesion in aggregation assays in merlin-deficient schwannoma cells that could be an explanation for increased proliferation rates due to loss of contact inhibition or tumor development in general.

Differential gene expression between human schwannoma and control Schwann cells.

The NF2 gene encodes the tumour suppressor protein merlin. The mutation of a single allele of this gene causes the autosomal dominantly inherited disease neurofibromatosis type 2 (NF2), which is characterized mainly by vestibular schwannoma carrying a second hit mutation. Complete lack of merlin is also found in spontaneous schwannomas and meningiomas. As the events leading to schwannoma development are largely unknown we investigated the differences in gene expression between schwannoma cells from NF2 patients and normal human primary Schwann cells by cDNA array analysis. We identified 41 genes whose expression levels differed by more than factor 2. Most of these clones were corroborated by real-time reverse transcription polymerase chain reaction analysis. By this method a total of seven genes with increased and seven genes with decreased mRNA levels in schwannoma compared with normal Schwann cells could be identified. Regulated clones, some of which not been described in Schwann cells earlier, included matrix metalloproteinase's, growth factors, growth factor receptors and tyrosine kinases.

A functional association between merlin and HEI10, a cell cycle regulator.

Merlin and ezrin are homologous proteins with opposite effects on neoplastic growth. Merlin is a tumor suppressor inactivated in the neurofibromatosis 2 disease, whereas upregulated ezrin expression is associated with increased malignancy. Merlin's tumor suppressor mechanism is not known, although participation in cell cycle regulation has been suggested. To characterize merlin's biological activities, we screened for molecules that would interact with merlin but not ezrin. We identified the cyclin B-binding protein and cell cycle regulator HEI10 as a novel merlin-binding partner. The interaction is mediated by the alpha-helical domain in merlin and the coiled-coil domain in HEI10 and requires conformational opening of merlin. The two proteins show partial subcellular colocalization, which depends on cell cycle stage and cell adhesion. Comparison of Schwann cells and schwannoma cultures demonstrated that the distribution of HEI10 depends on merlin expression. In transfected cells, a constitutively open merlin construct affected HEI10 protein integrity. These results link merlin to the cell cycle control machinery and may help to understand its tumor suppressor function.