Cleavage of extracellular matrix in periodontitis: gingipains differentially affect cell adhesion activities of fibronectin and tenascin-C

Gingipains are cysteine proteases that represent major virulence factors of the periodontopathogenic bacterium Porphyromonas gingivalis. Gingipains are reported to degrade extracellular matrix (ECM) of periodontal tissues, leading to tissue destruction and apoptosis. The exact mechanism is not known, however. Fibronectin and tenascin-C are pericellular ECM glycoproteins present in periodontal tissues. Whereas fibronectin mediates fibroblast adhesion, tenascin-C binds to fibronectin and inhibits its cell-spreading activity. Using purified proteins in vitro, we asked whether fibronectin and tenascin-C are cleaved by gingipains at clinically relevant concentrations, and how fragmentation by the bacterial proteases affects their biological activity in cell adhesion. Fibronectin was cleaved into distinct fragments by all three gingipains; however, only arginine-specific HRgpA and RgpB but not lysine-specific Kgp destroyed its cell-spreading activity. This result was confirmed with recombinant cell-binding domain of fibronectin. Of the two major tenascin-C splice variants, the large but not the small was a substrate for gingipains, indicating that cleavage occurred primarily in the alternatively spliced domain. Surprisingly, cleavage of large tenascin-C variant by all three gingipains generated fragments with increased anti-adhesive activity towards intact fibronectin. Fibronectin and tenascin-C fragments were detected in gingival crevicular fluid of a subset of periodontitis patients. We conclude that cleavage by gingipains directly affects the biological activity of both fibronectin and tenascin-C in a manner that might lead to increased cell detachment and loss during periodontal disease.

Fetal lungs of tenascin-C-deficient mice grow well, but branch poorly in organ culture

Tenascin-C (TNC) is a multidomain extracellular matrix protein that contributes to organogenesis and tumorgenesis. To elucidate its developmental function in the context of TNC deficiency, lung lobes of TNC null mice were obtained at Embryonic Days E11.5 and E12.5 and cultured for 3 d. In lung explants of homozygote TNC-deficient embryos (E12.5) the number of future airway branches was reduced by 36% as compared with wild-type. In heterozygote explants only half of the reduction (18%) was observed. No significant alteration, neither of the explant growth nor of the pattern of airway branching, was noticed in TNC-null explants. However, the terminal endbuds of the transgenic explants were enlarged. The results are supported by a morphologic investigation at Postnatal Day P2, where the airspaces of TNC-deficient lungs appeared larger than in wild-type lungs. Taken together, our results represent the first developmental phenotype of TNC-null mice. We conclude that TNC takes part in the control of fetal lung branching, and that not only the presence of TNC but also its amount is important. Because TNC is predominantly expressed at the growing tip of the future airways, we hypothesize that TNC promotes the penetration into the surrounding mesenchyme and the branching of the growing airways.

Tenascin-C is required for normal Wnt/β-catenin signaling in the whisker follicle stem cell niche.

Whisker follicles have multiple stem cell niches, including epidermal stem cells in the bulge as well as neural crest-derived stem cells and mast cell progenitors in the trabecular region. The neural crest-derived stem cells are a pool of melanocyte precursors. Previously, we found that the extracellular matrix glycoproteins tenascin-C and tenascin-W are expressed near CD34-positive cells in the trabecular stem cell niche of mouse whisker follicles. Here, we analyzed whiskers from tenascin-C knockout mice and found intrafollicular adipocytes and supernumerary mast cells. As Wnt/β-catenin signaling promotes melanogenesis and suppresses the differentiation of adipocytes and mast cells, we analyzed β-catenin subcellular localization in the trabecular niche. We found cytoplasmic and nuclear β-catenin in wild-type mice reflecting active Wnt/β-catenin signaling, whereas β-catenin in tenascin-C knockout mice was mostly cell membrane-associated and thus transcriptionally inactive. Furthermore, cells expressing the Wnt/β-catenin target gene cyclin D1 were enriched in the CD34-positive niches of wild-type compared to tenascin-C knockout mice. We then tested the effects of tenascins on this signaling pathway. We found that tenascin-C and tenascin-W can be co-precipitated with Wnt3a. In vitro, substrate bound tenascins promoted β-catenin-mediated transcription in the presence of Wnt3a, presumably due to the sequestration and concentration of Wnt3a near the cell surface. We conclude that the presence of tenascin-C in whiskers assures active Wnt/β-catenin signaling in the niche thereby maintaining the stem cell pool and suppressing aberrant differentiation, while in the knockout mice with reduced Wnt/β-catenin signaling, stem cells from the trabecular niche can differentiate into ectopic adipocytes and mast cells.

Transcriptional regulation of tenascin genes

Extracellular matrix proteins of the tenascin family resemble each other in their domain structure, and also share functions in modulating cell adhesion and cellular responses to growth factors. Despite these common features, the 4 vertebrate tenascins exhibit vastly different expression patterns. Tenascin-R is specific to the central nervous system. Tenascin-C is an "oncofetal" protein controlled by many stimuli (growth factors, cytokines, mechanical stress), but with restricted occurrence in space and time. In contrast, tenascin-X is a constituitive component of connective tissues, and its level is barely affected by external factors. Finally, the expression of tenascin-W is similar to that of tenascin-C but even more limited. In accordance with their highly regulated expression, the promoters of the tenascin-C and -W genes contain TATA boxes, whereas those of the other 2 tenascins do not. This article summarizes what is currently known about the complex transcriptional regulation of the 4 tenascin genes in development and disease.

The C-type lectin domains of lecticans, a family of aggregating chondroitin sulfate proteoglycans, bind tenascin-R by protein– protein interactions independent of carbohydrate moiety

The lecticans are a family of chondroitin sulfate proteoglycans including aggrecan, versican, neurocan, and brevican. The C-terminal globular domains of lecticans are structurally related to selectins, consisting of a C-type lectin domain flanked by epidermal growth factor and complement regulatory protein domains. The C-type lectin domain of versican has been shown to bind tenascin-R, an extracellular matrix protein specifically expressed in the nervous system, and the interaction was presumed to be mediated by a carbohydrate–protein interaction. In this paper, we show that the C-type lectin domain of brevican, another lectican that is specifically expressed in the nervous system, also binds tenascin-R. Surprisingly, this interaction is mediated by a protein–protein interaction through the fibronectin type III domains 3–5 of tenascin-R, independent of any carbohydrates or sulfated amino acids. The lectin domains of versican and other lecticans also bind the same domain of tenascin-R by protein–protein interactions. Surface plasmon resonance analysis revealed that brevican lectin has at least a 10-fold higher affinity than the other lectican lectins. Tenascin-R is coprecipitated with brevican from adult rat brain extracts, suggesting that tenascin-R and brevican form complexes in vivo. These results demonstrate that the C-type lectin domain can interact with fibronectin type III domains through protein–protein interactions, and suggest that brevican is a physiological tenascin-R ligand in the adult brain.

The Fibrinogen Globe of Tenascin-C Promotes Basic Fibroblast Growth Factor-induced Endothelial Cell Elongation

To investigate the potential role of tenascin-C (TN-C) on endothelial sprouting we used bovine aortic endothelial cells (BAECs) as an in vitro model of angiogenesis. We found that TN-C is specifically expressed by sprouting and cord-forming BAECs but not by nonsprouting BAECs. To test whether TN-C alone or in combination with basic fibroblast growth factor (bFGF) can enhance endothelial sprouting or cord formation, we used BAECs that normally do not sprout and, fittingly, do not express TN-C. In the presence of bFGF, exogenous TN-C but not fibronectin induced an elongated phenotype in nonsprouting BAECs. This phenotype was due to altered actin cytoskeleton organization. The fibrinogen globe of the TN-C molecule was the active domain promoting the elongated phenotype in response to bFGF. Furthermore, we found that the fibrinogen globe was responsible for reduced cell adhesion of BAECs on TN-C substrates. We conclude that bFGF-stimulated endothelial cells can be switched to a sprouting phenotype by the decreased adhesive strength of TN-C, mediated by the fibrinogen globe.

Cell-Adhesive Responses to Tenascin-C Splice Variants Involve Formation of Fascin Microspikes

Tenascin-C is an adhesion-modulating matrix glycoprotein that has multiple effects on cell behavior. Tenascin-C transcripts are expressed in motile cells and at sites of tissue modeling during development, and alternative splicing generates variants that encode different numbers of fibronectin type III repeats. We have examined the in vivo expression and cell adhesive properties of two full-length recombinant tenascin-C proteins: TN-190, which contains the eight constant fibronectin type III repeats, and TN-ADC, which contains the additional AD2, AD1, and C repeats. In situ hybridization with probes specific for the AD2, AD1, and C repeats shows that these splice variants are expressed at sites of active tissue modeling and fibronectin expression in the developing avian feather bud and sternum. Transcripts incorporating the AD2, AD1, and C repeats are present in embryonic day 10 wing bud but not in embryonic day 10 lung. By using a panel of nine cell lines in attachment assays, we have found that C2C12, G8, and S27 myoblastic cells undergo concentration-dependent adhesion to both variants, organize actin microspikes that contain the actin-bundling protein fascin, and do not assemble focal contacts. On a molar basis, TN-ADC is more active than TN-190 in promoting cell attachment and irregular cell spreading. The addition of either TN-190 or TN-ADC in solution to C2C12, COS-7, or MG-63 cells adherent on fibronectin decreases cell attachment and results in decreased organization of actin microfilament bundles, with formation of cortical membrane ruffles and retention of residual points of substratum contact that contain filamentous actin and fascin. These data establish a biochemical similarity in the processes of cell adhesion to tenascin-C and thrombospondin-1, also an “antiadhesive” matrix component, and also demonstrate that both the adhesive and adhesion-modulating properties of tenascin-C involve similar biochemical events in the cortical cytoskeleton. In addition to these generic properties, TN-ADC is less active in adhesion modulation than TN-190. The coordinated expression of different tenascin-C transcripts during development may, therefore, provide appropriate microenvironments for regulated changes in cell shape, adhesion, and movement.

Interaction of voltage-gated sodium channels with the extracellular matrix molecules tenascin-C and tenascin-R

The type IIA rat brain sodium channel is composed of three subunits: a large pore-forming α subunit and two smaller auxiliary subunits, β1 and β2. The β subunits are single membrane-spanning glycoproteins with one Ig-like motif in their extracellular domains. The Ig motif of the β2 subunit has close structural similarity to one of the six Ig motifs in the extracellular domain of the cell adhesion molecule contactin (also called F3 or F11), which binds to the extracellular matrix molecules tenascin-C and tenascin-R. We investigated the binding of the purified sodium channel and the extracellular domain of the β2 subunit to tenascin-C and tenascin-R in vitro. Incubation of purified sodium channels on microtiter plates coated with tenascin-C revealed saturable and specific binding with an apparent Kd of ≈15 nM. Glutathione S-transferase-tagged fusion proteins containing various segments of tenascin-C and tenascin-R were purified, digested with thrombin to remove the epitope tag, immobilized on microtiter dishes, and tested for their ability to bind purified sodium channel or the epitope-tagged extracellular domain of β2 subunits. Both purified sodium channels and the extracellular domain of the β2 subunit bound specifically to fibronectin type III repeats 1–2, A, B, and 6–8 of tenascin-C and fibronectin type III repeats 1–2 and 6–8 of tenascin-R but not to the epidermal growth factor-like domain or the fibrinogen-like domain of these molecules. The binding of neuronal sodium channels to extracellular matrix molecules such as tenascin-C and tenascin-R may play a crucial role in localizing sodium channels in high density at axon initial segments and nodes of Ranvier or in regulating the activity of immobilized sodium channels in these locations.

Skin wounds and severed nerves heal normally in mice lacking tenascin-C.

A large number of functions have been demonstrated for tenascin-C by antibody perturbation assays and in vitro cell culture experiments. However, these results contrast sharply with the lack of any apparent phenotype in mice with a genetic deletion of tenascin-C. A possible explanation for the lack of phenotype would be expression of some altered but functional tenascin-C in the mutant. We report the generation of an independent tenascin-C null mouse and conclude that the original tenascin-C knockout, which is genetically very similar to ours, is also a true null. As found previously, the absence of tenascin-C has no influence on development, adulthood, life span, and fecundity. We have studied in detail two models of wound healing. After axotomy, the regeneration of the sciatic nerve is not altered without tenascin-C. During healing of cutaneous wounds, deposition of collagen I, fibulin-2, and nidogen is identical in mutant and wild-type mice. In contrast. fibronectin appears diminished in wounds of tenascin-C-deficient mice. However, the lack of tenascin-C together with the reduced amount of fibronectin has no influence on the quality of the healing process.

The versican C-type lectin domain recognizes the adhesion protein tenascin-R.

The core proteins of large chondroitin sulfate proteoglycans contain a C-type lectin domain. The lectin domain of one of these proteoglycans, versican, was expressed as a recombinant 15-kDa protein and shown to bind to insolubilized fucose and GlcNAc. The lectin domain showed strong binding in a gel blotting assay to a glycoprotein doublet in rat brain extracts. The binding was calcium dependent and abolished by chemical deglycosylation treatment of the ligand glycoprotein. The versican-binding glycoprotein was identified as the cell adhesion protein tenascin-R, and versican and tenascin-R were both found to be localized in the granular layer of rat cerebellum. These results show that the versican lectin domain is a binding domain with a highly targeted specificity. It may allow versican to assemble complexes containing proteoglycan, an adhesion protein, and hyaluronan.

Neurite outgrowth by the alternatively spliced region of human tenascin-C is mediated by neuronal alpha7beta1 integrin

The region of tenascin-C containing only alternately spliced fibronectin type-III repeat D (fnD) increases neurite outgrowth by itself and also as part of tenascin-C. We previously localized the active site within fnD to an eight amino acid sequence unique to tenascin-C, VFDNFVLK, and showed that the amino acids FD and FV are required for activity. The purpose of this study was to identify the neuronal receptor that interacts with VFDNFVLK and to investigate the hypothesis that FD and FV are important for receptor binding. Function-blocking antibodies against both alpha7 and beta1 integrin subunits were found to abolish VFDNFVLK-mediated process extension from cerebellar granule neurons. VFDNFVLK but not its mutant, VSPNGSLK, induced clustering of neuronal beta1 integrin immunoreactivity. This strongly implicates FD and FV as important structural elements for receptor activation. Moreover, biochemical experiments revealed an association of the alpha7beta1 integrin with tenascin-C peptides containing the VFDNFVLK sequence but not with peptides with alterations in FD and/or FV. These findings are the first to provide evidence that the alpha7beta1 integrin mediates a response to tenascin-C and the first to demonstrate a functional role for the alpha7beta1 integrin receptor in CNS neurons.

A novel bioreactor for ligament tissue engineering

Bioreactors are defined as devices in which biological and/or biochemical processes develop under closely monitored and tightly controlled environmental and operating conditions (e.g. pH, temperature, mechanical conditions, nutrient supply and waste removal). In functional tissue engineering of musculoskeletal tissues, a bioreactor capable of controlling dynamic loading plays a determinant role. It has been shown that mechanical stretching promotes the expression of type I and III collagens, fibronectin, tenascin-C in cultured ligament fibroblasts (J.C.-H. Goh et al., Tissue Eng. 9 (2003), S31) and that human bone marrow mesenchymal stem cells (hBMMSC) – even in the absence of biochemical regulators – could be induced to differentiate into ligament-like fibroblast by the application of physiologically relevant cyclic strains (G. Vunjak-Novakovic et al., Ann. Rev. Biomed. Eng. 6 (2004), 131; H.A. Awad et al., Tissue Eng. 5 (1999), 267; R.G. Young et al., J. Orthop. Res. 16 (1998), 406). Different bioreactors are commercially available but they are too generic to be used for a given tissue, each tissue showing specific mechanical loading properties. In the case of ligament tissue engineering, the design of a bioreactor is still an open question. Our group proposes a bioreactor allowing cyclic traction–torsion on a scaffold seeded with stem cells.

Role of Basement Membrane and Extracellular Matrix Proteins in the Adhesion and Spreading of Immortalized Human Corneal Epithelial Cells

The repair of corneal wounds requires both epithelial cell adhesion and migration. Basement membrane (BM) and extracellular matrix (ECM) proteins function in these processes via integrin and non-integrin receptors. We have studied the adhesion, spreading and migration of immortalized human corneal epithelial (HCE) cells and their interactions with the laminins (Lms), fibronectins and tenascins produced. Human corneal BM expresses Lms-332 and -511, while Lm-111 was not found in these experiments. HCE cells produced both processed and unprocessed Lm-332, whereas neither Lm-111 nor Lm-511 was produced. Because HCE cells did not produce Lm-511, although it was present in corneal BM, we suggest that Lm-511 is produced by stromal keratocytes. The adhesion of HCE cells to Lms-111, -332 and -511 was studied first by determining the receptor composition of HCE cells and then by using quantitative cell adhesion assays. Immunofluorescence studies revealed the presence of integrin α2, α3, α6, β1 and β4 subunits. Among the non-integrin receptors, Lutheran (Lu) was found on adhering HCE cells. The cells adhered via integrin α3β1 to both purified human Lms-332 and -511 as well as to endogenous Lm-332. However, only integrin β1 subunit functioned in HCE cell adhesion to mouse Lm-111. The adhesion of HCE cells to Lm-511 was also mediated by Lu. Since Lm-511 did not induce Lu into focal adhesions in HCE cells, we suggest that Lm-511 serves as an ECM ligand enabling cell motility. HCE cells produced extradomain-A fibronectin, oncofetal fibronectin and tenascin-C (Tn-C), which are also found during corneal wound healing. Monoclonal antibodies (MAbs) against integrins α5β1 and αvβ6 as well as the arginine-glycine-aspartic acid (RGD) peptide inhibited the adhesion of HCE cells to fibronectin. Although the cells did not adhere to Tn-C, they adhered to the fibronectin/Tn-C coat and were then more efficiently inhibited by the function-blocking MAbs and RGD peptide. During the early adhesion, HCE cells codeposited Lm-332 and the large subunit of tenascin-C (Tn-CL) beneath the cells via the Golgi apparatus and microtubules. Integrin β4 subunit, which is a hemidesmosomal component, did not mediate the early adhesion of HCE cells to Lm-332 or Lm-332/Tn-C. Based on these results, we suggest that the adhesion of HCE cells is initiated by Lm-332 and modulated by Tn-CL, as it has been reported to prevent the assembly of hemidesmosomes. Thereby, Tn-CL functions in the motility of HCE cells during wound healing. The different distribution of processed and unprocessed Lm-332 in adhering, spreading and migrating HCE cells suggests a distinct role for these isoforms. We conclude that the processed Lm-332 functions in cell adhesion, whereas the unprocessed Lm-332 participates in cell spreading and migration.

Tumour markers as prognostic factors of survival of gastric cancer patients

The incidence of gastric cancer in the last decades has declined rapidly in the industrialised countries. Worldwide, however, gastric cancer is still the second most common cause of cancer death. Although surgery is currently the most effective treatment, the rapid progress in adjuvant chemotherapy and radiation therapy requires a re-evaluation of prognosis assessment. The TNM staging system of the UICC is ubiquitously used; it groups patients by decreasing survival times from stage I to stage IV based on the spread of disease, i.e. depth of tumour penetration (T), extent of spread to lymph nodes (N), and the presence or absence of distant (M) metastases. This is by far the most consistent prognostic classification system today. However, even within the stage groups there are patients that follow a varying course of disease. Our knowledge of the molecular differences between tumours of the same stage and morphology has been accumulating over the years and methods for a more accurate assessment of the phenotype of neoplasias are of value when evaluating the prognosis of individual patients with gastric cancer. In this study, the immunohistochemical expression of tumour markers involved in different phases in tumourigenesis was examined. The aim was to find new markers which could provide prognostic information in addition to what is provided by the TNM variables. A total of 337 specimens from the primary tumour of patients who underwent surgery for gastric cancer were collected and the immunohistochemical expression of seven different biomarkers was analysed. DNA ploidy and S-phase fraction (SPF) was assessed by flow cytometry. Finally, all biomarkers and clinicopathological prognostic factors were combined and evaluated by a multivariate Cox regression model to elucidate which specific factors provide independent prognostic information. By univariate survival analysis the following variables were significant prognostic factors: epithelial and stromal syndecan-1 expression, stromal tenascin-C expression, expression of tumour-associated trypsin inhibitor (TATI) in cancer cells, nuclear p53 expression, nuclear p21 expression, DNA ploidy, and SPF. By multivariate survival analysis adjusted for all available clinicopathological and biomolecular variables, p53 expression, p21 expression, and DNA ploidy emerged as independent prognostic biomarkers, together with penetration depth of the tumour, presence of nodal metastases, surgical cure of the cancer, and age of the patient at the time of diagnosis.

Evaluation of the Biocompatibility of Poly (ortho ester), Copolymer of ε-caprolactone/D,L-lactide and the Composite of Copolymer of ε-caprolactone/D,L-lactide and Tricalciumphosphate as Bone Filling Material

The purpose of this series of studies was to evaluate the biocompatibility of poly (ortho) ester (POE), copolymer of ε-caprolactone and D,L-lactide [P (ε-CL/DL-LA)] and the composite of P(ε-CL/DL-LA) and tricalciumphosphate (TCP) as bone filling material in bone defects. Tissue reactions and resorption times of two solid POE-implants (POE 140 and POE 46) with different methods of sterilization (gamma- and ethylene oxide sterilization), P(ε-CL/DL-LA)(40/60 w/w) in paste form and 50/50 w/w composite of 40/60 w/w P(ε-CL/DL-LA) and TCP and 27/73 w/w composite of 60/40 w/w P(ε-CL/DL-LA) and TCP were examined in experimental animals. The follow-up times were from one week to 52 weeks. The bone samples were evaluated histologically and the soft tissue samples histologically, immunohistochemically and electronmicroscopically. The results showed that the resorption time of gamma sterilized POE 140 was eight weeks and ethylene oxide sterilized POE 140 13 weeks in bone. The resorption time of POE 46 was more than 24 weeks. The gamma sterilized rods started to erode from the surface faster than ethylene oxide sterilized rods for both POEs. Inflammation in bone was from slight to moderate with POE 140 and moderate with POE 46. No highly fluorescent layer of tenascin or fibronectin was found in the soft tissue. Bone healing at the sites of implantation was slower than at control sites with the copolymer in small bone defects. The resorption time for the copolymer was over one year. Inflammation in bone was mostly moderate. Bone healing at the sites of implantation was also slower than at the control sites with the composite in small and large mandibular bone defects. Bone formation had ceased at both sites by the end of follow-up in large mandibular bone defects. The ultrastructure of the connective tissue was normal during the period of observation. It can be concluded that the method of sterilization influenced the resorption time of both POEs. Gamma sterilized POE 140 could have been suitable material for filling small bone defects, whereas the degradation times of solid EO-sterilized POE 140 and POE 46 were too slow to be considered as bone filling material. Solid material is difficult to contour, which can be considered as a disadvantage. The composites were excellent to handle, but the degradation time of the polymer and the composites were too slow. Therefore, the copolymer and the composite can not be recommended as bone filling material.