Origin and turnover of ECM proteins from the inner limiting membrane and vitreous body.

The inner limiting membrane (ILM) and the vitreous body (VB) are two major extracellular matrix (ECM) structures that are essential for early eye development. The ILM is considered to be the basement membrane of the retinal neuroepithelium, yet in situ hybridization and chick/quail transplant experiments in organ-cultured eyes showed that all components critical for ILM assembly, such as laminin or collagen IV, are not synthesized by the retina. Rather, ILM proteins, with the exception of agrin, originate from the lens or (and) ciliary body and are shed into the vitreous. The VB serves as a reservoir providing high concentrations of ILM proteins for the instant assembly of new ILM during rapid embryonic eye growth. The function of the retina in ILM assembly is to provide the cellular receptor proteins for the binding of the ILM proteins from the vitreous. The VB is a gelatinous ECM structure that fills the vitreous cavity of the eye. Its major structural proteins, collagen II and fibrillin, originate primarily from the ciliary body. Reverse transcription-PCR and western blotting show that the rate of synthesis of structural, monomeric ILM and VB proteins, such as laminin, collagen IV and II is very high during embryogenesis and very low in the adult. The downregulation of ILM and VB protein synthesis occurs during early postnatal life, and both ILM and VB are from then on maintained throughout life with minimum turnover. Our data explain why ILM and VB do not regenerate after vitrectomy and ILM peeling.

Detection and characterization of plasma cells in peripheral blood: correlation of IgE+ plasma cell frequency with IgE serum titre.

In atopic patients and patients with hyper-IgE syndrome (HIE) highly elevated IgE serum levels can be detected. Due to their very low frequency little is known about IgE-producing plasma cells (PC) in peripheral blood. We used CD138 MACS microbeads to enrich plasma cells from peripheral blood of normal donors, atopic patients and one HIE patient. CD138+ cells were mainly CD45+, CD44++, CD19dim, CD38++, CD27++, CD86+, HLA-DR+/++, CD71dim, VLA-4+, VLA-5-, CD28-, CD25-, CD69-, CLA-, CD20-, CD21- and CD22-. They show weak expression of surface Ig but high levels of intracellular Ig and they secrete Ig in culture. Thus CD138+ cells from peripheral blood show characteristics of early plasma cells. IgE+ CD138+ plasma cells could be detected in 19 of 24 normal donors with an average frequency of 0.06% IgE+ cells among CD138+ cells. Higher frequencies were detected in atopic patients, atopic patients with markedly elevated serum IgE levels and the hyper-IgE patient with an average of 0.32%, 7.21% and 6.54%, respectively. Additionally, using the recently developed cellular affinity matrix technology, we were able to detect IgE secreting plasma cells and thereby could demonstrate that most of the IgE secreting cells express CD138. The frequency of IgE+ CD138+ cells among PBMC correlated highly significantly with serum IgE titres (r = 0.8532***), indicating that IgE secreting CD138+ cells in peripheral blood are directly related to the plasma cell pool contributing to the IgE titre.

Increased deposition of fibulin-2 in solar elastosis and its colocalization with elastic fibres.

BACKGROUND: Fibulin-2 is a 195-kDa protein belonging to a novel family of extracellular matrix proteins that might be involved in microfibril and elastic fibre organization. OBJECTIVES: To determine the localization of fibulin-2 in relation to elastic fibres in normal skin and in solar elastosis characterized by increased elastotic material in the papillary dermis. METHODS: The expression and synthesis of fibulin-2 was investigated by means of in situ hybridization, immunohistochemistry and Western blot analysis in normal and photoaged skin. RESULTS: Immunohistochemistry and elastic tissue staining revealed that fibulin-2 deposition mainly colocalized with microfibrils and elastin fibres, with a marked staining of elastotic material in solar elastosis. Western blot analysis demonstrated that in photoaged skin fibulin-2 showed the same electrophoretic mobility as in sun-protected skin. However, in actinic elastosis the amount of fibulin-2 was significantly higher. In addition, smaller degradation products were detectable, presumably reflecting increased proteinase activity in photodamaged skin. CONCLUSIONS: This study shows that deposition of fibulin-2 and elastin is highly co-ordinated, indicating that this protein plays an important role in elastic fibre and microfibril formation in normal and actinically damaged skin.

BM-40 and MMP-2 expression are not coregulated in human melanoma cell lines.

As published recently, suppression of BM-40 (osteonectin, SPARC) expression reduced the in vitro adhesive and invasive capacities of melanoma cells. In parallel, a decreased production of matrix metalloproteinase-2 (MMP-2) was observed in these cells indicating a close coregulation of these two proteins. However, analysis of 12 independently established melanoma cell lines, exhibiting different metastatic capacities, did not reveal a positive correlation between BM-40 or MMP-2 expression and the metastatic capacities of these cell lines. In addition, direct comparison of BM-40 and MMP-2 mRNA levels did not provide evidence for coregulation of these two proteins. This was further corroborated by the finding that exogenous BM-40 failed to alter synthesis and activation of proMMP-2.

Membrane-type 1 matrix metalloproteinase-mediated progelatinase A activation in non-tumorigenic and tumorigenic human keratinocytes.

Elevated expression of type IV collagenases (MMP-2 and MMP-9) has been strongly correlated with tumour progression and metastasis in various tumours. Here, we analysed expression and activation of these MMPs in non-tumourigenic HaCaT cells and the malignant HaCaT variant II-4(rt). In monolayer cultures, both cell types secreted latent MMP-2 (proMMP-2) in comparable amounts, while MMP-9 production was clearly higher in II-4(rt)cells. Upon contact with fibrillar collagen type I the malignant II-4(rt)cells, but not the HaCaT cells, gained the capability to activate proMMP-2. This process is shown to be membrane-associated and mediated by MT1-MMP. Surprisingly, all membrane preparations from either HaCaT cells or II-4(rt)cells grown as monolayers, as well as within collagen gels, contained considerable amounts of active MT1-MMP. However, within collagen gels HaCaT cells showed significantly higher TIMP-2 levels compared to II-4(rt)cells. This indicates that TIMP-2 might play a central role for MT1-MMP-mediated gelatinolytic activity. Indeed, collagen type I-induced MT1-MMP-mediated proMMP-2 activation by II-4(rt)membranes could be completely abolished by an excess of TIMP-2. In conclusion, our data suggest that MT1-MMP-mediated proMMP-2 activation might be associated with malignant progression of epidermal tumour cells.

The absence of nidogen 1 does not affect murine basement membrane formation.

Nidogen 1 is a highly conserved protein in mammals, Drosophila melanogaster, Caenorhabditis elegans, and ascidians and is found in all basement membranes. It has been proposed that nidogen 1 connects the laminin and collagen IV networks, so stabilizing the basement membrane, and integrates other proteins, including perlecan, into the basement membrane. To define the role of nidogen 1 in basement membranes in vivo, we produced a null mutation of the NID-1 gene in embryonic stem cells and used these to derive mouse lines. Homozygous animals produce neither nidogen 1 mRNA nor protein. Surprisingly, they show no overt abnormalities and are fertile, their basement membrane structures appearing normal. Nidogen 2 staining is increased in certain basement membranes, where it is normally only found in scant amounts. This occurs by either redistribution from other extracellular matrices or unmasking of nidogen 2 epitopes, as its production does not appear to be upregulated. The results show that nidogen 1 is not required for basement membrane formation or maintenance.

Tissue inhibitor of matrix metalloproteinase-2 regulates matrix metalloproteinase-2 activation by modulation of membrane-type 1 matrix metalloproteinase activity in high and low invasive melanoma cell lines.

Activation of pro-matrix metalloproteinase (MMP)-2 on the surface of malignant cells by membrane-bound MT1-MMP is believed to play a critical role during tumor progression and metastasis. In this study we present evidence that MT1-MMP plays a key role for the in vitro invasiveness of malignant melanoma. Melanoma cell lines secreted latent MMP-2 when cultured on plastic. However, when cells were grown in floating type I collagen lattices, only high invasive melanoma cells activated proMMP-2. Activation could be inhibited by antibodies against MT1-MMP, by addition of recombinant tissue inhibitor of metalloproteinases (TIMP)-2 and by inhibition of MT1-MMP cleavage. MT1-MMP protein was detected as an inactive protein in all cell lines cultured as monolayers, whereas in collagen gels, active MT1-MMP protein was detected in the membranes of both high and low invasive melanoma cells. Production of TIMP-2 was about 10-fold higher in low invasive cells as compared with high invasive melanoma cells and was further increased in the low invasive cells upon contact to collagen. Thus, in melanoma cells TIMP-2 expression levels might regulate MT1-MMP-mediated activation of proMMP-2. High invasive melanoma cells displayed increased in vitro invasiveness, which was inhibited by TIMP-2. These data indicate the importance of these enzymes for the invasion processes and support a role for MT1-MMP as an activator of proMMP-2 in malignant melanoma.

Selective loss of PMA-stimulated expression of matrix metalloproteinase 1 in HaCaT keratinocytes is correlated with the inability to induce mitogen-activated protein family kinases.

Many cell types, including fibroblasts and primary keratinocytes, increase matrix metalloproteinase 1 (MMP-1) production in response to agonists such as growth factors and phorbol esters. However, the spontaneously transformed human keratinocyte cell line HaCaT, although it increases MMP-1 production in response to epidermal growth factor (EGF), does not respond similarly to stimulation with PMA. This phenomenon occurs even though HaCaT cells remain proliferatively responsive to both agonists, suggesting a HaCaT-specific defect in a PMA-mediated signal transduction pathway. Using an inside-out approach to elucidate the source of this defect, we found that EGF, but not PMA, stimulated MMP-1 promoter activity in transiently transfected HaCaT keratinocytes. In addition, an assessment of fibroblast and HaCaT c-fos and c-jun gene expression after exposure to EGF and PMA showed that although both agonists increased the expression of c-fos and c-jun mRNA in fibroblasts, only EGF did so in HaCaT keratinocytes. Finally, we looked at the activation of mitogen-activated protein (MAP) family kinases after stimulation with EGF or PMA and found that both agonists increased the phosphorylation and activation of fibroblast extracellular signal-regulated protein kinase and c-Jun N-terminal kinase, but only EGF activated the same kinase activities in HaCaT cells. Further, the EGF-mediated increase in MMP-1 gene expression was inhibited by the MAP kinase/ERK kinase (MEK)-specific inhibitor PD98059 and the p38 kinase-specific inhibitor SB203580. Our evidence indicates that although HaCaT MAP kinases are functional, they are not properly regulated in response to the activation of protein kinase C, and that the defect that bars HaCaT MMP-1 expression in response to stimulation with PMA lies before MAP kinase activation.

Differential regulation of the human nidogen gene promoter region by a novel cell-type-specific silencer element.

Transfection analyses of the human nidogen promoter region in nidogen-producing fibroblasts from adult skin revealed multiple positive and negative cis-acting elements controlling nidogen gene expression. Characterization of the positive regulatory domains by gel mobility-shift assays and co-transfection studies in Drosophila SL2 cells unequivocally demonstrated that Sp1-like transcription factors are essential for a high expression of the human nidogen gene. Analysis of the negative regulatory domains identified a novel silencer element between nt -1333 and -1322, which is bound by a distinct nuclear factor, by using extracts from adult but not from embryonal fibroblasts. In embryonal fibroblasts, which express significantly higher amounts of nidogen mRNA as compared with adult fibroblasts, this inhibitory nidogen promoter region did not affect nidogen and SV40 promoter activities. The silencer element seems to be active only in nidogen-producing cells. Therefore this regulatory element might function in vivo to limit nidogen gene expression in response to external stimuli. However, none of the identified regulatory elements, including the silencer, contribute significantly to cell-specific expression of the human nidogen gene. Instead we provide evidence that gene expression in epidermal keratinocytes that are not producing nidogen is repressed by methylation-specific and chromatin-dependent mechanisms.

BM-40 (osteonectin, SPARC) is expressed both in the epidermal and in the dermal compartment of adult human skin.

BM-40 (Osteonectin, SPARC) is the most abundant glycoprotein secreted by human osteoblasts. In situ hybridization studies on the expression of BM-40 mRNA in murine tissues have demonstrated the highest levels of transcripts in bone, but expression was also observed in several other mesenchymal tissues. In contrast, little is known about the expression of BM-40 in human tissues, especially in skin. Total RNA obtained from normal human skin was analyzed by northern blotting and revealed a marked expression of BM-40. To analyze its expression in vivo, in situ hybridization was performed, demonstrating that BM-40 is expressed in fibroblasts, smooth muscle, and endothelial cells in the dermis. Interestingly, BM-40 mRNA was also detected throughout the basal, spinous, and granular layers in the epidermis of adult human skin. Further analysis by immunohistochemistry revealed a marked deposition in the dermis that was most intense directly below the basement membrane in the papillary dermis and around vascular as well as glandular structures. In the epidermis, BM-40 protein could be detected intercellularly in suprabasal layers. This finding is further supported by the intercellular deposition of BM40 detected by immunofluorescence in cultured keratinocytes. This study demonstrates that BM-40 that has previously been thought to be exclusively expressed in extracellular matrix producing cells may in fact play a role in differentiation and maintenance of the epidermis.

Recombinant and tissue-derived mouse BM-40 bind to several collagen types and have increased affinities after proteolytic activation.

The calcium-binding extracellular matrix protein BM-40 was obtained as a mouse cDNA product from a stably transfected kidney cell clone. Electrophoresis and N-terminal sequence analysis demonstrated absence of the proteolytic processing previously observed for a mouse tumour-derived BM-40. Yet the two forms of BM-40 were very similar in their CD spectra, their calcium-dependent change in alpha helix content and their immunological epitopes. In surface plasmon resonance assays, recombinant mouse BM-40 showed distinct binding to the triple-helical domains of collagens I, II, III, IV and V with Kd = 1-4 microM but no binding to collagen VI. These interactions were abolished in the presence of EDTA. Tissue-derived mouse BM-40, however, bound collagens I and IV with Kd = 0.1-0.2 microM. Activation of collagen binding to give a similar Kd could be achieved for recombinant mouse BM-40 by treatment with the matrix metalloproteinase collagenase-3. The major cleavage site was located in helix C of the extracellular calcium-binding module of BM-40 and other less prominent cleavages occurred close to the N-terminus. The sensitive helix C site was just one residue away from that sensitive to endogenous tissue proteolysis, suggesting that cleavage could be a physiological mechanism to modulate collagen binding.

Quantitative analysis of alpha 1 (I) procollagen transcripts in vivo by competitive polymerase chain reaction.

Due to the limited amount of RNA obtainable from punch biopsies, few data exist on the human alpha 1 (I) procollagen mRNA steady state level in vivo. Therefore, we established a competitive PCR method to quantitate this mRNA in human biopsies. In our approach, the target template and the standard share the same sequence except for a 69 bp deletion, thus competing for the same primer pair and subsequently amplifying at the same rate. Titration of a competitor DNA dilution series against an unknown cDNA sample then allows quantitation of the separated and nonradioactively detected PCR products after densitometrical analysis. Almost identical results were obtained by quantitations of the same total RNA by competitive PCR as well as Northern blot analysis.

A purine-rich sequence in the human BM-40 gene promoter region is a prerequisite for maximum transcription.

BM-40 (osteonectin, SPARC [secreted protein, acidic, rich in cysteine]) is a highly conserved, matrix-associated protein that is found in basement membranes, bones and remodeling tissues throughout vertebrate evolution. We are reporting the characterization of the 5' end of the human BM-40 gene. Sequence comparison of the 5' region revealed significant homologies with the bovine and murine genes, including a purine-rich stretch composed of two boxes, GGA-box 1 and 2, separated by a pyrimidine-rich spacer element. Transfection analyses of the human BM-40 promoter provide strong evidence that this region comprises several distinct regulatory domains, to which different functions can be assigned. GGA-box 1 is thereby absolutely required and sufficient by itself for maximal BM-40 transcriptional activity, whereas the spacer element has a down-regulatory effect. Comparative transfection analyses in human cell lines, positive or negative for BM-40 transcripts, indicate that the GGA-box sequences in the human promoter, in contrast to the bovine promoter, do not significantly contribute to cell-type specific expression in human cells.

Sequence data and chromosomal localization of human type I and type II hair keratin genes.

A cDNA library constructed with poly(A)+ RNA from human scalp was screened with selected fragments of both murine type I and type II hair keratin cDNAs. Two keratin clones, one type I, phKI-2, and one type II, phKII-1, were isolated and sequenced. In Northern blots, cDNA probes containing the 3'-noncoding sequences of the clones specifically hybridized to scalp mRNA species. Based on sequence homology comparisons with the four known murine type I hair keratins mHa1-4, the phKI-2 encoded keratin could be identified as human hair keratin hHa2. Similarly, sequence comparison with the four type II sheep wool keratins K2.9-12 revealed an orthologous relationship between the largest member of the type II wool keratin subfamily, K2.9 (i.e., sHb1) and the phKII-1 encoded human hair keratin (hHb1). The specific 3'-noncoding sequences of hHa2 and hHb1 were also used to isolate genomic fragments for both keratins from human genomic libraries which were than used for fluorescence in situ hybridization to human metaphase chromosomes. The hHa2 gene could be mapped to the long arm of chromosome 17, whereas the hHb1 gene was found on the long arm of chromosome 12. DAPI banding of the chromosomes allowed sublocalization of the hHa2 gene to 17q12-q21 and the hHb1 gene to 12q13, i.e., gene loci that have also been previously determined for human type I and type II epithelial keratins.

Genomic sequences and structural organization of the human nidogen gene (NID).

Nidogen/entactin is a ubiquitous 150-kDa multidomain basement membrane protein. Since in vitro binding studies indicated that nidogen may function as a major mediator in basement membrane organization and assembly, analysis of gene structure and regulation of gene expression will help us to understand many biological processes that involve degradation and reorganization of the basement membrane zone. An approximately 100-kb region of genomic DNA encoding the human nidogen gene (NID) including 5' and 3' flanking sequences has been cloned and characterized by restriction mapping and sequencing. The entire gene is more than 90 kb in length and contains 20 exons. All introns interrupt protein coding sequences. The size of individual introns varies significantly, ranging from 0.6 to 18 kb. Its exon/intron structure revealed that the protein domains of human nidogen are organized in a domain-specific manner with various subdomains being encoded by individual exons, indicating that exon duplication and shuffling have played an important role in determining the present structure of the protein. Comparison of the exon organization with the recently published ascidian nidogen amino acid sequence strongly suggests that vertebrate nidogen might have evolved from a common ancestral precursor resembling ascidian nidogen.

Binding properties and protease stability of recombinant human nidogen.

Recombinant human nidogen was obtained from transfected kidney cell clones as a 150-kDa protein with a three-globule structure. It was modified by sulfation and O-glycosylation and a lower level of N-glycosylation than mouse nidogen. Recombinant nidogens of both species were, however, indistinguishable in their affinities for laminin-1 and a recombinant laminin gamma 1 chain fragment and showed a similar binding to collagen IV and the heparan sulfate proteoglycan perlecan. The two nidogens were also equivalent in the promotion of ternary complex formation between these ligands, indicating that this function has been conserved during mammalian evolution. Fewer zinc-binding sites could be identified in human nidogen and correlated with a lower capacity of zinc to prevent binding to laminin and collagen IV. Most remarkable was the greater sensitivity of human nidogen to endogenous proteolysis in cell culture, yielding fragments of 90-145 kDa. Studies with several exogenous proteases, including thrombin and leucocyte elastase, showed lack of stability of the N-terminal globular domain G1 in contrast to what was found for mouse nidogen. Since such degradation could be important for basement membrane remodelling, this difference between human and mouse may be biologically significant.

Recombinant analysis of human alpha 1 (XVI) collagen. Evidence for processing of the N-terminal globular domain.

The N-terminal non-collagenous domain NC11 of the human collagen alpha 1 (XVI) chain was obtained as a recombinant 35-kDa protein from stably transfected kidney cell clones. This form had undergone proteolytic trimming at a basic cleavage motif indicating a similar release in vivo. Domain NC11 showed a globular shape after rotary shadowing and was resistant to neutral proteases. Specific antibodies could be raised against recombinant NC11 and were used for the analysis of other cell clones transfected with the full-length alpha 1 (XVI) chain. Immunoprecipitation of detergent extracts of metabolically labelled cells demonstrated the presence of disulfide-bonded 200-kDa polypeptides possessing NC11 epitopes. This material was partially resistant to pepsin, indicating the formation of alpha 1 (XVI) chain homotrimers with a triple-helical conformation. Yet a substantial proportion of these homotrimers was degraded to fragments of variable size (35-150 kDa) when secreted into the culture medium. Several of these fragments could be obtained on a semi-preparative scale from cells grown in hollow fiber cassettes and showed substantial hydroxylation of proline, consistent with triple-helix formation. Edman degradation demonstrated the origin of some from the N-terminal and of one from a more C-terminal position of collagen XVI. This extensive degradation may be explained by the release of NC11 and by further cleavages within some of the nine interruptions of the triple-helical domain of the alpha 1(XVI) chain. Whether this process also occurs in situ remains to be shown.

Recombinant expression and properties of the Kunitz-type protease-inhibitor module from human type VI collagen alpha 3(VI) chain.

The Kunitz-type inhibitor motif (domain C5) present at the C-terminus of the human collagen alpha 3(VI) chain was prepared in a recombinant form from the culture medium of stably transfected kidney cell clones. The 76-residue protein was disulfide bonded and showed a high stability against protease treatment. The recombinant protein lacked, however, any inhibitory activity for trypsin, thrombin, kallikrein and several other proteases, which could be due to a few unusual substitutions in the region crucial for inhibitor binding. A sensitive radioimmunoassay detected low concentrations of C5 epitopes in normal human serum and fibroblast culture medium and showed a lack of cross-reaction with aprotinin. Antibodies against C5 immunoprecipitated collagen VI obtained from fibroblast medium. The C5 epitopes could not be detected on intact collagen VI purified from guanidine extracts of human placenta. Collagen VI was shown to possess several alpha 3(VI) chain bands (approximately 200 kDa) and reacted strongly with antibodies to an N-terminal recombinant fragment. Immunofluorescence with anti-C5 antibodies failed to stain several human tissues but produced a distinct intracellular staining of cultured fibroblasts. The data indicate the rapid loss of the C5 domain after biosynthesis of collagen VI.

Changes in calcium and collagen IV binding caused by mutations in the EF hand and other domains of extracellular matrix protein BM-40 (SPARC, osteonectin).

Recombinant expression in a human kidney cell-line was used to prepare mutant human BM-40 with deletions including the N-terminal acidic domain, a central alpha-helical domain and the C-terminal EF hand domain. Two putative EF hand motifs were altered by point mutations. Only elimination of the whole EF hand domain or its single disulfide bond decreased production and secretion indicating that the C-terminal region of BM-40 is essential for correct folding. Deletions in the alpha-helical domain changed a single disulfide bond in this domain and caused oligomerization. Several mutations resulted in significant conformational changes as shown by CD spectroscopy and epitope analysis. Fluorescence titration demonstrated a single high-affinity (Kd = 0.08 microM) calcium-binding site with a low dissociation rate constant. A Glu to Lys mutation in the -Z position of the C-terminal EF hand motif and lack of a stabilizing disulfide bridge caused a 30 to 100-fold reduction in calcium affinity, while an Asp to Lys mutation in the X position had only a small effect. Deletions in the alpha-helical domain caused an even more dramatic reduction in calcium binding and abolished calcium-dependent binding of BM-40 to collagen IV. Both binding properties are critically dependent on a conformational interaction between the EF hand and the alpha-helical domain, which contains the collagen-binding site.