Is Fimbrin Involved in Cancer

The potential of fimbrin to modulate cell behaviour is obviously quite considerable. There are references in the literature since the early 1980s that fimbrin expression increases with cell transformation, especially in fibroblast cells. However, little incisive investigation seems to have been carried out. More recent investigations have pursued the worthwhile aim of establishing a role for fimbrin in neoplastic transformation and tumour progression. The expression of L-fimbrin has been investigated in cell lines derived from carcinoma of the prostate. L-fimbrin has been found in many carcinoma cell lines, but not in normal epithelial cell lines derived from the prostate. L-fimbrin occurs at higher levels in prostate cancer tissue as compared with normal prostate tissue. Immunohistochemical staining has suggested that the increase of fimbrin occurs predominantly in the glandular epithelial cells of the carcinoma, whereas in normal prostate tissue, it is found mainly in the fibromuscular stroma. It would have been interesting to explore the fimbrin expression pattern of benign prostatic hyperplasia (BPH). Nonetheless, the differences shown to exist between normal prostatic epithelium and carcinomatous glandular epithelium are persuasive enough to accept that fimbrin expression may be related to the carcino-matous changes.

Interestingly, fimbrin expression was also reported to be rapidly up-regulated in LNCaP cells by dihydrotestosterone and oestradiol (J.P. Zheng et al. 1997). This observation does not per se contribute much to the argument that fimbrin is involved in the transformation of normal epithelium to carcinoma. Testicular hormones are the major factors that regulate the growth and development of the prostate. It is possible that changes in hormonal milieu are responsible for the development of BPH and onward to the development of prostatic carcinoma. Androgen receptors (ARs), oestrogen receptors (ERs), and progesterone receptors (PgR) occur in BPH and in prostate cancer (Srinivasan et al. 1995). Indeed, AR occurs in all histological types and all clinical stages of prostate gland cancer. Because androgens are involved in the development of the prostate itself, it would be reasonable to suppose that AR occurs in the developing prostate as well. Therefore, one would expect that the androgen-sensitive cell line LNCaP would respond by an altered level of fimbrin. What is crucially important is to find out whether normal epithelial cells respond differently from carcinomas. Testicular hormones bind to the appropriate receptors and stimulate the transcription of androgen-responsive genes, including those that regulate the growth of prostate cells. It should be borne in mind, further, that the progression of prostate cancer is associated with a change from androgen-dependent to androgen-independent state. The fimbrin gene may be a steroid-regulated gene, and it is important to determine whether this is the case, and whether androgen differentially stimulates fimbrin transcription in normal prostatic epithelium and prostatic carcinoma cells.

Another isolated observation, which may, nonetheless, be significant in the context of drug resistance of tumours, is the relationship discovered by Hisano et al. (1996) between cisplatin resistance and fimbrin expression. They observed that cisplatin-resistant cells possessed severalfold greater levels of T-fimbrin compared with sensitive cells. They transfected T-fimbrin antisense cDNA into cisplatin-resis-tant cells and showed that reduction of fimbrin expression resulted in increased sensitivity to the drug.

Modulation of Actin Dynamics and Cancer Cell Dissemination

Conceptually there can be no difficulties in accepting the proposition that the modulation of actin dynamics should be involved in some way with tumour dissemination. A reasonable body of experimental evidence, in the form of changes in the expression of gelsolin, thymosins, and fimbrin in cancer progression, has been adduced in support of this concept. In this context of actin-interacting proteins, the vasodilator-stimulated phosphoprotein (VASP) deserves a mention. VASP is approximately 45 to 50 kDa in size and occurs in domains of the plasma membrane that are involved in the formation of lamellipodia (Reinhard et al. 1992, 1995b). Together with the Drosophila- and murine-enabled (Ena) proteins, VASP belongs to a family of actin-binding proteins that are actively engaged in the regulation of the actin cytoskeleton. VASP has been shown to interact with actin stress fibres via its C-terminal region (Huttelmaier et al. 1999). VASP as well as Ena can interact with the actin CP profilin (see below) (Reinhard et al. 1995a) and with focal adhesion proteins (Reinhard et al. 1995b, Brindle et al. 1996; Lanier et al. 1999; Ahern-Djamali et al. 1998). This suggests that the VASP family members might be involved in signal transduction (Huttelmaier et al. 1998). Besides, during in vitro morphogenesis of human umbilical endothelial cells, VASP, profilin, and gelsolin are markedly up-regulated (Salazar et al. 1999). In view of the close association of all three proteins in actin dynamics, Salazar et al. (1999) have postulated that reorganisation of the actin cytoskeleton is required for the alignment of the endothelial cells during capillary morphogenesis. The expression of profilin or VASP in neoplastic transformation has not been studied in great detail. NIH3T3 fibroblasts that have been made VASP deficient by experimental manipulation have been reported to be able to form tumours in nude mice. In vitro, these cells show loss of contact inhibition and a deregulation of cell proliferation (Liu K.Y. et al. 1999). With the demonstration that these proteins are associated with capillary morphogenesis, it would be of much interest to see whether VASP and/or profilin are down-regulated in neoplastic transformation.

One can envisage that perturbation of the actin cytoskeleton could make the cell membrane malleable and the cell more deformable and in this way aid invasion. Mechanistic perturbations in endothelial cell alignment could generate abnormal fenestration in tumour-associated microvasculature that might serve as ports of entry into the vascular compartment for tumour dissemination. These observations highlight the potential importance of CBPs in cancer invasion, and also provide a fertile ground for extracting information that might be highly relevant in assessing the degree of malignancy of cancers and in cancer management.


The actin cytoskeleton is regulated by several actin-binding and cross-linking proteins that are themselves regulated by free calcium. a-actinin is an EF-hand protein approximately 110 kDa in size and it forms a major component of the cytoskeleton in many cell types. It forms antiparallel, highly stable homodimers. Actinin isoforms can also form heterodimers, both in vivo and in vitro (Y.M. Chan et al. 1998). Y.M. Chan et al. (1998) suggest the possibility that heterodimers formed by different isoforms could potentially have new functional characteristics. The dimerisation of smooth muscle actinin is believed to be mediated by a segment of the C-terminal region of the molecule (Baron et al. 1987; Imamura et al. 1988; Kahana and Gratzer, 1991; A.P. Gilmore et al. 1994). This segment has two cross-linking sites (A and B), and the actinin molecules form cross-links in an antiparallel fashion by the binding of an A site of one molecule with the B site of the second molecule (Imamura et al. 1988). Several isoforms of a-actinin have been identified from vertebrate cytoskeletal, skeletal, and smooth muscle sources (Duhaiman and Bamburg, 1984; Imamura and Masaki, 1992; Landon et al. 1985; J.P. Bennett et al. 1984). Isoforms also have been isolated from invertebrate sources such as Dictyostelium discoideum (Noegel et al. 1987) and Drosophila melanogaster (Fyrberg et al. 1990). A cDNA has been cloned from the nematode Caenorhabditis elegans. The sequence of this clone predicts that it codes for an a-actinin, which possesses a high degree of sequence homology to actinin from other sources (Barstead et al. 1991).

Molecular Structure of o-Actinin a-actinin is an EF-hand calcium-binding protein. Two EF-hands are located at the C-terminal region of the molecule. An actin-binding domain occurs at the N-terminal end, followed by the rod domain, which consists of four spectrin-like repeat elements, followed by the EF-hand to the C-terminus (A. Blanchard et al. 1989; Flood et al. 1995). The rod domain is required for stable dimerisation of the molecule, and dimer-isation is reduced markedly if either of terminal repeats 1 and 4 is deleted (Flood et al. 1995). Although all the isoforms possess EF-hands, these have been found to be in a functional state only in the nonmuscle or cytoskeletal isoforms of a-actinin (Burridge and Feramisco, 1981; Duhaiman and Bamburg, 1984). Their significance has been elucidated by introducing point mutations to make them nonfunctional. The first EF-hand seems to regulate the Ca2+-dependent cross-linking activity of a-actinin (Janssen et al. 1996). Nonmuscle isoforms might themselves differ in calcium-binding ability (Imamura et al. 1994). The EF-hands of a-actinin isoforms 2 and 3 of human skeletal muscle are not capable of binding calcium, and therefore the binding of these isoforms to actin might not be calcium sensitive (Beggs et al. 1992). Calcium sensitivity or the lack of it can be viewed from a functional viewpoint. The EF-hands appear to undergo conformational changes from a closed position in the absence of calcium, to an open position when calcium is present (Trave et al. 1995). The open conformation may be conducive to protein-protein interactions, which are required for the transduction of signals via the actin cytoskeleton.

a-AcTiNiN Isoforms

Two skeletal isoforms of a-actinin have been designated as a-actinin-2 and -3 and two nonmuscle isoforms as a-actinin-1 and -4 (Millake et al. 1989; Beggs et al. 1992, 1994; Honda et al. 1998). The skeletal muscle isoform a-actinin-2 is found in both human skeletal and cardiac muscle, but a-actinin-3 occurs only in limb skeletal muscle (Beggs et al. 1992). The isoforms a-actinin-1 and -4 occur in the actin microfilament bundles and at the adherens junctions. In skeletal, cardiac, and smooth muscle, a-actinin is localised to the Z-discs and dense bodies, and it participates in the anchoring of the actin thin filaments and giant titin molecules to the Z-disc (Endo and Masaki 1984; Geiger et al. 1990). a-actinin has been shown to interact with an N-terminal domain of titin in vitro (Ohtsuka et al. 1997). In the Z-disc, the assembly of the actinin-titin complex seems to involve two types of interaction between titin and a-actinin. In the other regions of the Z-disc, titin binds by means of a single binding site with the outermost pair of a-actinin molecules, but in the middle of the Z-disc the titin binds several a-actinin molecules through binding sites at their C-terminal region (Young et al. 1998). Thus a-actinin seems to play a major part in anchoring actin thin filaments of the two halves of the sarcomere at the Z-disc (see Figure 15). Abnormal expression of a-actinin and other thin filament-associated proteins are known to interfere with the assembly of Z-discs, which could lead to the formation of so-called nemaline bodies. Two loci are known to be involved in nemaline myopathy: the tropomyosin-3 locus and the nebulin locus on 2q21.2-q22. Mutations of both genes have been implicated in this form of myopathy (Laing et al. 1995; Pelin et al. 1999). The formation of the complex of tropomyosin, titin, nebulin, a-actinin, and actin can conceivably be affected by these mutations, leading to abnormalities in Z-disc assembly and the formation of nemaline bodies.

Function of a-actinin a-actinin functions predominantly in F-actin bundling and anchoring of the filaments to specific sites within the cell as well as to the cell membrane and in the linking up of the cytoskeletal machinery to the ECM. a-actinin might form a bridge between the actin cytoskeleton and integrin receptors then function as a receptor for components of the ECM (Burridge et al. 1990; Geiger et al. 1990). The cytoplasmic tail of 0-integrins and intercellular adhesion molecule-I (ICAM-I) interact with a-actinin via binding sites occurring in the rod domain. Furthermore, a-actinin appears to link the cytoskeleton to the cell membrane (Baron et al. 1987). Kahana and Gratzer (1991) have identified binding sites for long-chain fatty acids in the spectrin-like repeats of the a-actinin domain rod. Han et al. (1997) recently have shown that bundling of actin filaments occurs if diacylglycerol is present in the membrane. It follows, therefore, that a-actinin would be an important component of the signal transduction pathway as well as an important link in the cytoskeleton-mediated changes in cell shape and motility. Miyamoto et al. (1995) have delineated the pathway of integrin-mediated signal transduction. They reported that integrin aggregation induced the accumulation of several signal transduction molecules, such as Rho A, Rac1, ras, raf, and others, of cytoskeletal components, such as vinculin, talin, and a-actinin. The tyrosine kinase inhibitor genistein inhibits accumulation of both the signal transduction molecules and the cytoskeletal signal transduction components.

Actinins in Cell Adhesion, Motility, and Signal Transduction

The cross-linking of actin filaments increases filament elasticity and viscosity and may be expected to affect the structural properties of actin filaments. It follows from this that actinin might, in this fashion, change the properties of the cell membrane, such as intercellular and cell-substratum adhesion, which in turn would be reflected in alterations in cell shape and motility. These changes will have serious implications for cancer invasion and secondary spread. Furthermore, there are strong indications that actin cross-linking in endothelial cells might result in metabolite transport and permeability. Although yet to be demonstrated, one would expect that alterations in the organisation of adherens junctions might effectively alter endothelial integrity and lead to enhanced tumour cell diapedesis across the endothelium. Changes in the structural properties of the actin cytoskeleton would impugn the link-up between it and the ECM. Such changes would affect seriously the physical pathway of signal transduction, and effectively lead to its deregulation, and eventually to the deregulation of intercellular adhesion and communication and of cell proliferation and growth.

The Cadherin-Catenin Complex in Signal Transduction and Cell Adhesion

A signal transduction complex that occurs in the cell membrane, consisting of the transmembrane glycoprotein cadherin and other components such as a-actinin, 0-catenin, and the adenomatous polyposis coli (APC) protein, has assumed considerable significance by virtue of its apparent dual function, namely in signal transduction and cellular adhesion. Intercellular adhesion mediated by cadherin is dependent on the integrity of the link-up with the actin cytoskeleton (Sherbet and Lakshmi, 1997b) (Figure 13). P-catenin subserves two functions, viz. as a signal transduction molecule and in the formation of adherens-type junctions. Its participation in signal transduction has become apparent with the demonstration that plakoglobin and P-catenin are homologues of the armadillo protein of Drosophila that is involved in segment polarity.

FIGURE 13 Schematic representation of the interaction of the transmembrane adhesion protein cadherin with the actin cytoskeleton, via P-catenin, a-catenin, and a-actinin. (Based on van Roy [1992] and Sherbet and Lakshmi [1997b]). Reprinted by permission of the publisher Academic Press, from The Genetics of Cancer, (Sherbet and Lakshmi, 1997b).

FIGURE 13 Schematic representation of the interaction of the transmembrane adhesion protein cadherin with the actin cytoskeleton, via P-catenin, a-catenin, and a-actinin. (Based on van Roy [1992] and Sherbet and Lakshmi [1997b]). Reprinted by permission of the publisher Academic Press, from The Genetics of Cancer, (Sherbet and Lakshmi, 1997b).

The armadillo (arm) protein is a component of the signal transduction pathway of the wg (wingless) molecule (Riggleman et al. 1990). The arm family of proteins, e.g., P-catenin, plakoglobin, and the p120ctn protein, are characterised by a central arm repeat domain (Riggleman et al. 1989). The p120ctn gene potentially can code for a large number of isoforms that are generated by alternative splicing, and these have been regarded as constituting a subfamily of signalling proteins (Keirsebilck et al. 1998b). Jou et al. (1995) have shown that P-catenin binds to a C-terminal 25 amino acid region in the cytoplasmic domain of E-cadherin and to the N-terminal domain of a-catenin. The p120ctn protein, on the other hand, seems to bind to a juxtamembrane domain of the cadherin cytoplasmic tail (Yap et al. 1998).

The armadillo and wg proteins exert similar effects on embryonic development (Peifer et al. 1991). A family of wnt proteins has been identified. The wnt protein is a secreted glycoprotein signalling factor. It is a vertebrate homologue of wg and has been shown to be a regulator of morphogenesis of Xenopus (Gumbiner, 1996; Miller and Moon, 1996). In a similar vein, P-catenin and plakoglobin are components of the signal transduction pathway of the wnt gene as well as being involved in the formation of adherens junctions and Ca2+-mediated cell-cell adhesion (Hinck et al. 1994; Peifer, 1995). Cadherin, as a part of this signalling complex, not only regulates intercellular adhesion, but also seems to negatively regulate the signalling function of P-catenin. These two functions can be dissociated. Fagotto et al. (1996) noticed that P-catenin deletion mutants affect cadherin-mediated adhesion but not its signalling function. Sanson et al. (1996) showed that a full length E-cadherin and a truncated form of cadherin that have opposite effects on cadherin-dependent adhesion, nonetheless function effectively in wg signalling.

Two further elements are involved in the signal transduction complex involving E-cadherin and catenins. One is the APC protein and the other component is axin or conductin. P-Catenin forms complexes with both axin and APC. The APC protein is known to compete with cadherin in its binding to the internal repeats of P-catenin. The P-catenin/APC complex is phosphorylated by the glycogen synthase kinase GSK3P. A consequence of this phosphorylation is that the degradation of P-catenin is greatly enhanced (Munemitsu et al. 1995; Hayashi et al. 1997; Hart et al. 1998). Axin also induces P-catenin degradation and is suggested to function downstream of APC (Behrens et al. 1998). The GSK3P-mediated phosphorylation is inhibited when the wnt signal transduction pathway is active. This results in the accumulation of P-catenin in the cytoplasm (Hinck et al. 1994; Giarre et al. 1998; Papkoff and Aikawa, 1998) and leads to the formation of a complex with the T-cell factor/lym-phoid enhancer factor (Tcf/Lef). The binding between them occurs via the arm repeats. This complex functions as a transcription factor in the wnt/wg signalling pathway (Behrens et al. 1996; Van de Wetering et al. 1997). The pathway might be deregulated by inhibition of P-catenin degradation, leading to a constitutive activation of the transcription complex (Figure 14).



P-catenin degradation



ß-catenin p-catenin/Tcf-Lef Transcription complex

■ Cell adhesion

FIGURE 14 The wnt signal transduction pathway and cell adhesion signal involving P-catenin, APC, and cadherin. (Based on references cited in the text.)

The deregulation that accompanies neoplastic changes seems to be more frequently due to mutation of P-catenin or APC protein than to inactivating mutations of E-cadherin (Morin et al. 1997; Efstathiou et al. 1999). These mutations seem to stabilise P-catenin. They occur at the phosphorylation sites that are essential in the ubiquitination and degradation of the protein. P-catenin mutations occur in approximately 50% of colorectal tumours. Again, these tend to occur in the serine/threonine phosphorylation sites. They seem to occur more frequently in small adenomas than in larger adenomatous lesions. Furthermore, a majority of the mutations have been found in adenomas rather than in carcinomas (Samowitz et al. 1999). A high rate (61% of 31 patients) of somatic mutation of P-catenin has been encountered in human anaplastic thyroid carcinoma (Garcia-Rostan et al. 1999). In a murine hepa-tocarcinoma model, mutations have been reported to occur in the carcinomas but not in adenomas. These findings suggest that deregulation of signal transduction occurs as an early event in the pathogenesis and progression of tumours. The P-catenin mutations in colorectal tumours reported by Samowitz et al. (1999) were not accompanied by APC mutations. The suggestion has been made, therefore, that mutations of APC and P-catenin might be functionally different. Mutations of APC leading to the loss of its suppressor function have been identified with the familial adenomatous polyposis, and are also closely related to the progression of the disease. Furthermore, mutated APC seems to lack the ability to regulate P-catenin levels. Therefore, APC may be regarded ipso facto as functioning upstream of P-catenin.

The adhesion function of P-catenin flows from its linkage to cadherin, which spans across the cell membrane, on the one hand, and to the cytoskeleton, on the other, by means of two other elements, namely a-catenin and a-actinin. In fibroblasts, this linkage has been demonstrated by Knudsen et al. (1995), who also noted that a-actinin specifically immunoprecipitates with P-catenin and cadherin. Niesset et al. (1997) have identified the binding sites involved in the interaction of a-catenin with P-catenin on the one hand and with a-actinin on the other. In normal thyroid epithelial cells, the intercellular adhesion junctions show the presence of cadherin together with the catenins. In contrast, CGTHW-2 cells derived from thyroid carcinoma show a marked alteration in the pattern of distribution of these linking proteins in the foci of intercellular adhesion.

Further evidence in support of the importance of the integrity of the cad-herin-catenin complex in intercellular adhesion comes from the apparent involvement of the Rho GTPase family in the regulation of cadherin-mediated cell adhesion. IQGAP1, which is an effector of two members of the family, namely Cdc42 and Rad, has been shown to be able to dissociate a-catenin from the cadherin/P-cate-nin/a-catenin complex. This results in the disruption of intercellular adhesion. Both Cdc42 and Rad counteract this process (Kaibuchi et al. 1999). The inhibition of Rho and Rac (another member of the Rho GTPase family) has been reported to lead to a disruption of E-cadherin localisation at keratinocyte-keratinocyte cell junctions. However, when the GTPases themselves are inhibited, the cadherin-mediated adhesive contacts are reestablished (Braga et al. 1999). Rho protein, when microinjected into four-cell blastomere-stage embryos, disrupts cortical microfilaments and reduces interblastomere adhesion. Cdc42 in the same way also disrupts the cortical cytoskeleton and interblastomere contacts (Clayton et al. 1999). But the modes of involvement of Rho and Cdc42 are clearly distinguishable in NIH3T3 cells transformed by the dbl oncogene, which codes for a supposed exchange factor for RhoA and Cdc42. The transformed cells respond to adhesion to fibronectin substratum by changing cell shape. This involves the activation of RhoA and the associated ROCK and CRIK, but not Cdc42. In nontransformed cells, however, change of cell shape seems to require Cdc42 activation (Olivo et al. 2000). Rho GTPases may also be involved in the regulation of endothelial cell adhesion involving cadherins. Rho

GTPases are also involved in the regulation of the function of other adhesion-mediating glycoproteins such as CD44. Therefore, prima facie there is a reasonable basis for investigating their role in maintaining the integrity of the vascular endothelia and possible implications for metastatic spread of cancer.

Other proteins that link the plasma membrane with the actin cytoskeleton may disrupt cadherin-mediated adhesion. The proteins ezrin, radixin, moesin, and merlin subserve such a linking function. Hiscox and Jiang (1999) used antisense nucleotides to inhibit ezrin expression in colon carcinoma cells and noticed that this resulted in a loss of intercellular adhesion and acquisition of motility. They further noticed that ezrin co-precipitated with E-cadherin and P-catenin. In cells treated with antisense ezrin nucleotides, this could result in a reduced interaction between ezrin and the cadherin complex. Such a reduction indeed occurs when ezrin is phosphorylated or when the cells are treated with hepatocyte growth factor.

Huang et al. (1998) found no cadherin or y-catenin at the adhesion junctions, and P- and a-catenins were distributed diffusely in the cytoplasm of most cells. However, P-catenin, when detected at intercellular junctions, was found to co-localise with a-actinin. Huang et al. (1998) have therefore suggested that the loss of intercellular adhesion could be due to an incorrect assembly of the linking components. Implicit also in this is the suggestion that such an incorrect assembly might lead to the acquisition of invasive ability. Although this is an attractive hypothesis, it should be mentioned here that Honda et al. (1998) have reported the occurrence of a novel actinin, namely actinin-4, whose expression is said to be up-regulated with enhanced cellular migration. This new isoform is said to occur in the cytoplasm. It is reported to be associated with cytoplasmic extensions and is found in peripheral migrating cells of cell clusters. Actinin-4 appears to be translocated from the cytoplasm to the nucleus, when actin is depolymerised. Honda et al. (1998) have also examined the expression of this novel actinin in breast cancers. It is expressed in infiltrating breast carcinomas, and the expression was found to correlate with poor prognosis.

Mamura and Masaki (1996) identified a 115-kDa a-actinin in vascular endothelial cells. This actinin differed from other known actinins of muscle or nonmuscle origin in its sensitivity to calcium. However, this isoform of actinin did occur in heart tissue, the pectoralis muscle, and the gizzard. What could be significant about this actinin in relation to cancer progression is its apparent location in the vascular endothelium. It is well known that several inflammatory agents induce a reversible change in endothelial cell shape that results in the formation of intercellular gaps. If cancer cells can reduce the endothelial integrity in this way, that could form ports of entry into the vascular system for cancer cells. There is a view that invading cancer cells take advantage of naturally occurring fenestration of the endothelium. There is no demonstration to date that they might secrete substances that could induce the formation of ports of entry. From this point of view it would be interesting to see if cancer cells induce any changes in the expression of actinins of the endothelial barrier.

A new facet that can be added to the story of actinins in invasion and metastasis is their apparent relationship to the expression of metalloproteinases. It was shown some years ago that the production of these enzymes often paralleled the invasive and metastatic abilities of cancers, which has suggested the possibility that metal-loproteinases might reconstruct the ECM and provide cancer cells with ECM properties conducive to invasion (Sherbet and Lakshmi, 1997b). Crawford et al. (1999) have made some interesting observations on the expression of matrilysin and 0-catenin. They found that 0-catenin and matrilysin mRNAs were expressed in parallel in murine intestinal adenomas. A further observation of considerable importance is that 0-catenin significantly up-regulated the matrilysin promoter. Because 0-catenin is a structural component of signalling pathway and intercellular adhesion machinery, these experiments bring together, in a unique fashion, transduction of an extracellular signal that can putatively remodel the ECM and change the invasive behaviour of cells.

Other cell adhesion molecules such as the ICAM similarly require interaction with the actin cytoskeleton. A peptide containing a cytoplasmic sequence of ICAM-2 has been shown to interact with a-actinin in vitro. Indeed, interaction occurs between several sites of a-actinin and ICAM-2. Besides, ICAM-2 co-localises with a-actinin (Heiska et al. 1996). The epithelial ICAM (Ep-ICAM), an adhesion molecule regarded as specific for epithelial cell adhesion, has a cytoplasmic domain that regulates the cellular adhesion function. It appears that Ep-ICAM achieves this via a-actinin, with which Ep-ICAM appears to interact directly by means of specific binding sites (Balzar et al. 1998).

Shigella flexneri, which causes bacillary dysentery, is known to induce changes in the cytoskeleton of epithelial cells. These changes result in the production of membrane protrusions that engulf the bacterium. The bacterium secretes a protein called IipaA. The cytoskeletal reorganisation, which is essential for this process of bacterial invasion, involves the recruitment by IipaA of the linking protein vinculin as well as a-actinin (Van Nhieu et al. 1997).

The effects of the loss of a-actinin on cell shape, aggregation, and motility have been studied in Dictyostelium. Rivero et al. (1996) generated mutants lacking a-actinin as well as the "gelation factor" and these mutants presented a rounded shape rather than the typical polarised morphology of aggregating cells. The mutants also showed considerable loss of motility. In rat bladder carcinoma cells, the experimentally induced loss of motility was associated with the reorganisation of F-actin and a-actinin (Morton and Tchao, 1994). The transmembrane integrin a20j functions as the receptor for the ECM components collagen and laminin. This integrin as well as a-actinin have been implicated in the invasive ability of melanoma cells in vitro. L.M. Duncan et al. (1996) reported that a-actinin was not detectable in benign melanocytic naevi and in laterally spreading lesions, but it occurred in all nodular melanomas and in metastatic melanomas. It would have been of much interest whether there was an association between a-actinin expression and the vertical growth of melanoma.

EGF is known to alter cell morphology as well as promote cell invasion in vitro (Shiozaki et al. 1995). These cellular responses seem to originate from the linkage of the activated EGFr to the cytoskeletal system (Roy et al. 1989, 1991; Van Bergen en Henegouwen et al. 1989). There is some recent evidence that EGFr regulates intercellular adhesion by interfering with the formation of a complex between the invasive suppressor transmembrane protein E-cadherin and the actin cytoskeletal elements, including a-actinin and P-catenin (Hazan and Norton, 1998; see also Figure 13). As alluded to earlier, the proliferative signal imparted to the cell by EGF is mediated by PKC. N.R. Murray et al. (1999) generated transgenic mice that express PKC-PII at high levels. The colonic epithelium showed high cell proliferation. Besides, the epithelium was more prone to preneoplastic changes. They found that these effects were also accompanied by increased expression of P-catenin and a reduction in glycogen synthase kinase 3P activity. These studies suggest, therefore, that EGFr-mediated signal follows the wnt/APC pathway of transduction leading to cell proliferation. This pathway also may be utilised in neoplastic changes of epi-thelia. Overall, these data now provide a physical basis for previous observations that high EGFr expression correlated with poor prognosis, and that EGFr status is a powerful marker of clinical aggressiveness and metastatic potential of tumours (Sherbet and Lakshmi, 1997b).

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment