Altered processing of integrin receptors during keratinocyte activation

We used monoclonal antibodies against specific integrin subunits to examine the role of integrin receptors in keratinocyte activation. We found that before activation, beta 1 subunits in keratinocytes showed a diffuse distribution, whereas after activation, keratinocytes organized beta 1 receptors into marginal adhesion plaques. In immunoprecipitation experiments with antibodies against beta 1 integrin subunits, we found mostly immature subunits synthesized in keratinocytes freshly harvested from skin. Moreover, integrin receptor complexes immunoprecipitated from these cells by monoclonal antibodies against alpha 2, alpha 3, or alpha 5 subunits contained only immature beta 1 subunits. With keratinocytes cultured 4-7 days, anti-beta 1 antibodies immunoprecipitated mostly mature beta 1 subunits, and integrin complexes immunoprecipitated from cultured cells by anti-alpha subunit antibodies contained mostly mature beta 1 subunits. Antibodies directed against beta 1 subunits also inhibited keratinocyte migration. Based on these results, we suggest that up-regulation of migration by activated keratinocytes depends on changes in processing of pre-beta 1 subunits to mature beta 1 subunits. We also studied the distribution of integrin subunits in skin and on keratinocytes migrating out of skin explants. Whereas beta 1, alpha 2, and alpha 3 subunits were detected in keratinocytes in skin and migrating out of explants, alpha 5 subunits were observed only in migrating cells.     

Nested collagen matrices: a new model to study migration of human fibroblast populations in three dimensions

Fibroblast-3D collagen matrix culture provides a model system to analyze cell physiology under conditions that more closely resemble tissue than conventional 2D cell culture. Previous work has focused primarily on remodeling and contraction of collagen matrices by fibroblasts, and there has been little research on migration of cell populations within the matrix. Here, we introduce a nested collagen matrix model to analyze migration of fibroblasts in 3D collagen matrices. Nested collagen matrices were prepared by embedding contracted cell-containing matrices (also called dermal equivalents) inside cell-free matrices; migration occurred from the former to the latter. Control experiments with human dermal fragments in place of dermal equivalents confirmed the reliability of the model. Human fibroblast migration in nested collagen matrices occurred after a lag phase of 8-16 h, and cells migrating out of the inner matrices were bipolar with leading dendritic extensions. Migration was myosin II, Rho kinase and metalloproteinase-dependent but did not require plasma fibronectin. Platelet-derived growth factor but not lysophosphatidic acid or serum stimulated cell migration, although all three of these physiological agonists promote matrix remodeling and contraction. The nested collagen matrix model is a relatively easy, rapid and quantitative method to measure migration of cell populations. Our studies using this model demonstrate important differences between regulation of fibroblast migration and remodeling in collagen matrices.     

Interaction of fibronectin-coated beads with attached and spread fibroblasts : Binding, phagocytosis, and cytoskeletal reorganization

After 15 min incubations, binding of 0.8-, 6-, and 16-microns fibronectin-coated latex beads occurred primarily at the margins of chick embryo fibroblasts that previously were attached and spread on fibronectin-coated glass coverslips. Extensive phagocytosis of the smallest beads and some phagocytosis of the larger beads occurred within 2 h. Following binding of the 16-micron beads, there were no changes in overall cell shape or in the distribution of several cytoskeletal proteins. There was, however, a local accumulation of actin and alpha-actinin patches adjacent to the sites where the beads were bound. The formation of alpha-actinin patches could be detected with 6- or 16-microns beads shortly after initial bead binding to the cells, but a similar reorganization of alpha-actinin in response to the binding of 0.8-micron beads was not detected. The patches of alpha-actinin appeared to be associated with membrane ruffles, since such structures were observed by scanning electron microscopy (SEM) to be sites of cell interaction with 6- but not 0.8-micron beads. Also, two other cytoskeletal proteins normally absent from membrane ruffles, tropomyosin and vinculin, were not detected at the sites of cell-bead interaction. No reorganization of vinculin at the cell-bead interaction sites was observed even when the 16-microns beads remained bound at the cell surfaces for up to 6 h. Nevertheless, prominent vinculin plaques were observed at the marginal attachment sites on the ventral cell surfaces. Consequently, formation of mature focal adhesions may be restricted to linear regions of cell-substratum interaction.     

Post-translational modifications of proteins: some problems left to solve

Three major questions regarding the post-translational modification of amino acid side chains in proteins are briefly considered: (1) What are the biological functions of the reactions, (2) what is the specificity of the processing reactions in selecting only a few or sometimes even only one residue for modification, and (3) how do we solve the uniqueness of the processing steps in the production of recombinant proteins? The answers to these questions are not obvious at this time.     

Quantitative freeze-fracture analysis of the frog neuromuscular junction synapse – I. Naturally occurring variability in active zone structure

The orderly arrays of intramembranous particles (IMPs) found in the p-face of freeze-fracture replicas of the frog neuromuscular junction (‘active zones’) are believed to be involved in transmitter release. Some or all of the particles represent voltage-dependent Ca2+ channels. Since there is a great heterogeneity in the amount of transmitter released by different frog motor nerve terminals we sought to determine whether active zone (AZ) structure displayed a similar heterogeneity by using a novel freeze-fracture procedure providing large, intact replicas containing significant portions of motor nerve terminals from the cutaneous pectoris muscle of the frog, Rana pipiens. Using only junctions in which more than 50 AZs or more than 50 μm of nerve terminal were included in the fractures, we measured AZ length, AZ intramembranous particle density, terminal width at each AZ, space between AZs, the angle of AZ orientation with respect to the longitudinal axis of the nerve terminal, exposed pre-synaptic nerve terminal surface area and a calculated value for mean AZ length per unit terminal length. The analysis led to the following conclusions. There is an approximate 5-fold range in mean AZ length/micrometre terminal length. Terminal width is a good predictor of AZ length. Particle density does not vary significantly within a given AZ, nor between AZs from the same or different junctions. The distance between AZs is not related to AZ length, i.e. shorter AZs are no more or less likely to be closer to the adjacent AZ. The probability of release from any AZ on action potential invasion is small. If most of the IMPs are Ca2+ channels, either the probability of channel opening or the efficacy of triggering release is very low or both. That the variability in release efficacy in different terminals is much greater than ultrastructural variability in terminals suggests that regulation of release is dominated by physiological processes that do not have obvious ultrastructural correlates. On the other hand, the apparent excess of AZ relative to the number of vesicles released indicates that the amount and variability in amount of AZ is important in ways that need to be elucidated.     

5' sequence of vesicular stomatitis virus N-gene confers selective translation of mRNA.

The infection of cells by vesicular stomatitis virus results in the rapid inhibition of host-cell protein synthesis, but not of viral protein synthesis. To determine if this translational selectivity might be conferred by the viral mRNA, we constructed a plasmid (pUCLN beta-4) containing the 5' end of the viral nucleocapsid (N)-gene, including the ribosome binding site, fused in frame with the gene encoding beta-galactosidase, and compared it to a control plasmid (pMC1924) containing the cellular rabbit beta-globin gene 5' end fused with the beta-galactosidase encoding gene. Both plasmids contained identical promoter and 3' nontranslated regions and expressed similar levels of beta-galactosidase in the indicator cell line 293. In cells transfected with either plasmid, viral infection resulted in a approximately 70% decrease in protein synthesis by five hours. The level of beta-galactosidase from cells transfected with pMC1924 also decreased concomitantly with the decrease in total protein synthesis. However, the level of beta-galactosidase from cells transfected with pUCLN beta-4 was not affected by viral infection. Our data suggest that sequences in the 5' end of the viral mRNA allow for the selective translation of the viral message in the presence of an inhibited translational machinery.