Structure, development and function of cytoskeletal elements in non-neuronal cells of the human eye

The cytoskeleton, of which the main components in the human eye are actin microfilaments, intermediate filaments and microtubules with their associated proteins, is essential for the normal growth, maturation, differentiation, integrity and function of its cells. These components interact with intra- and extracellular environment and each other, and their profile frequently changes during development, according to physiologic demands, and in various diseases. The ocular cytoskeleton is unique in many ways. A special pair of cytokeratins, CK 3 and 12, has apparently evolved only for the purposes of the corneal epithelium. However, other cytokeratins such as CK 4, 5, 14, and 19 are also important for the normal ocular surface epithelia, and other types may be acquired in keratinizing diseases. The intraocular tissues, which have a relatively simple cytoskeleton consisting mainly of vimentin and simple epithelial CK 8 and 18, differ in many details from extraocular ones. The iris and lens epithelium characteristically lack cytokeratins in adults, and the intraocular muscles all have a cytoskeletal profile of their own. The dilator of the iris contains vimentin, desmin and cytokeratins, being an example of triple intermediate filament expression, but the ciliary muscle lacks cytokeratin and the sphincter of the iris is devoid even of vimentin. Conversion from extraocular-type cytoskeletal profile occurs during fetal life. It seems that posttranslational modification of cytokeratins in the eye may also differ from that of extraocular tissues. So far, it has not been possible to reconcile the cytoskeletal profile of intraocular tissues with their specific functional demands, but many theories have been put forward. Systematic search for cytoskeletal elements has also revealed novel cell populations in the human eye. These include transitional cells of the cornea that may represent stem cells on migration, myofibroblasts of the scleral spur and juxtacanalicular tissue that may modulate aqueous outflow, and subepithelial matrix cells of the ciliary body and myofibroblasts of the choroid that may both participate in accommodation. In contrast to the structure and development of the ocular cytoskeleton, changes that take place in ocular disease have not been analysed systematically. Nevertheless, potentially meaningful changes have already been observed in corneal dystrophies (Meesmann's dystrophy, posterior polymorphous dystrophy and iridocorneal endothelial syndrome), degenerations (pterygium) and inflammatory diseases (Pseudomonas keratitis), in opacification of the lens (anterior subcapsular and secondary cataract), in diseases characterized by proliferation of the retinal pigment epithelium (macular degeneration and proliferative vitreoretinopathy), and in intraocular tumours (uveal melanoma). In particular, upregulation of alpha-smooth muscle actin seems to be a relatively general response typical of spreading and migrating corneal stromal and lens epithelial cells, trabecular cells and retinal pigment epithelial cells.     

Interactions between membrane IgM and the cytoskeleton involve the cytoplasmic domain of the immunoglobulin receptor

Cross-linking induced interactions between the membrane form of immunoglobulin (mIg) and the cytoskeletal matrix have been described by several groups. To date, the function of mIgM association with the cytoskeleton is not yet understood. Delineation of the molecular basis of these interactions will be instrumental in elucidating their function. We have previously shown that the Ig alpha/beta heterodimer is not required for ligand-induced mIgM binding to the cytoskeleton. In this study, we have investigated the role of other B cell-specific proteins in mediating these interactions. For this, we expressed mIgM in the non-hematopoietic human cervical carcinoma cell line HeLa S3 and verified the capacity of the surface-expressed IgM to interact with the cytoskeletal matrix upon cross-linking with anti-mu chain antibodies. We show here that only the mIgM molecule itself and no other B cell-specific protein(s) is required in mediating mIgM interactions with actin filaments. In an attempt to determine the cytoskeleton-binding site of mIgM we investigated the role of the cytoplasmic tail of mIgM (KVK) in binding the receptor to actin-based microfilaments. Using mutated forms of mIgM expressed in J558L cells, we show here that KVK plays a role in mediating these interactions. The absence of KVK did not, however, completely abrogate mIgM-cytoskeletal interactions, suggesting that there are additional molecular requirements for the ligand-induced mIgM binding to the cytoskeletal matrix.     

Cytoskeleton and cell adhesion molecules: critical targets of toxic agents

Normal development of the nervous system depends upon complex physical interactions between cells and their local environment. These interactions are mediated by several families of cell adhesion molecules (CAMs). Differential expression and function of CAMs are operative in neural tube formation, neuron migration, in post-migratory differentiation, and maintenance of mature neural structure. CAMs also facilitate contact-dependent cell processes, such as formation of cell junctions. Temporal regulation of these molecules during development may provide "windows of vulnerability" to toxicants. In addition to their extracellular binding activities, some CAMs have membrane-spanning domains by which they communicate directly with the cytoskeleton, permitting extracellular signals to be rapidly translated into cell responses via modifications in cytoskeletal organization. These cytologic changes are particularly critical during migration, neurite formation and synaptogenesis. Toxic perturbation of adhesion molecules can have catastrophic effects on morphogenetic processes both directly and via events which depend upon cytoskeletal rearrangement. Toxicants can also act directly upon the cytoskeleton, resulting secondarily in changes of the membrane distribution and function of CAMs. Toxicant-induced changes in CAMs and cytoskeleton may occur contemporaneously. Interference of cell adhesion-cytoskeleton interactions may be a pivotal molecular event dictating developmental consequences of neurotoxicant exposure.     

Mitochondrial distribution and inheritance

Mechanisms mediating the inheritance of mitochondria are poorly understood, but recent studies with the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe have begun to identify components that facilitate this essential process. These components have been identified through the analysis of conditional yeast mutants that display aberrant mitochondrial distribution at restrictive conditions. The analysis of these mutants has uncovered several novel proteins that are localized either to cytoskeletal structures or to the mitochondria themselves. Many mitochondrial inheritance mutants also show altered mitochondrial morphology and defects in maintenance of the mitochondrial genome. Although some inheritance components and mechanisms appear to function specifically in certain types of cells, other conserved proteins are likely to mediate mitochondrial behavior in all eukaryotic cells.     

Role of adhesion molecule cytoplasmic domains in mediating interactions with the cytoskeleton

The past ten years have seen significant progress in cell biology research aimed at understanding how cytoskeletal filaments interact with the plasma membrane. Considerable evidence suggests that both actin microfilaments and intermediate filaments attach to the membrane via the cytoplasmic domains of various membrane proteins including adhesion molecules. Interactions between the cytoskeleton and adhesion molecules appear to be essential for a variety of cellular functions, including cell-cell and cell-extracellular matrix (ECM) interactions, cell motility, receptor-ligand interactions, and receptor internalization. Recently, many of the detailed molecular mechanisms which mediate the associations between actin filaments and adhesion molecules have been identified. Among adhesion molecules that support the attachment of cytoskeletal filaments to their cytoplasmic domains are members of the integrin and cadherin families, the intracellular adhesion molecule-1 (ICAM-1, an immunoglobulin family member), and the glycoprotein Ib/IX complex in platelets. A general conclusion emerging from these studies is that physical associations between cytoskeletal filaments and transmembrane glycoproteins do not occur directly between the filaments and the cytoplasmic tails of adhesion molecules. Instead, these interactions appear to be indirect and involve a complex ensemble of intermediary linker proteins. The severe effects of cytoplasmic domain deletion and mutagenesis on adhesion-dependent functions support the view that receptor cytoplasmic domains play a vital role in regulating receptor function and in mediating communication across the membrane. Transfection studies with mutant and chimeric adhesion molecules, along with protein-binding studies, are clarifying the mechanisms which physically link the cytoskeleton to transmembrane proteins, regulate cytoskeletal organization, mediate signaling across the cell membrane, and regulate the ligand specificity and binding affinity of surface receptors.     

Dystonin is essential for maintaining neuronal cytoskeleton organization

The mouse neurological mutant dystonia musculorum (dt) suffers from a hereditary sensory neuropathy. We have previously described the cloning and characterization of the dt gene, which we named dystonin (Dst). We had shown that dystonin is a neural isoform of bullous pemphigoid antigen 1 (Bpag1) with an N-terminal actin-binding domain. It has been shown previously that dystonin is a cytoskeletal linker protein, forming a bridge between F-actin and intermediate filaments. Here, we have used two different antibody preparations against dystonin and detected a high-molecular-weight protein in immunoblot analysis of spinal cord extracts. We also show that this high-molecular-weight protein was not detectable in the nervous system of all dt alleles tested. Immunohistochemical analysis revealed that dystonin was present in different compartments of neurons--cell bodies, dendrites, and axons, regions which are rich in the three elements of the cytoskeleton (F-actin, neurofilaments, and microtubules). Ultrastructural analysis of dt dorsal root axons revealed disorganization of the neurofilament network and surprisingly also of the microtubule network. In this context it is of interest that we observed altered levels of the microtubule-associated proteins MAP2 and tau in spinal cord neurons of different dt alleles. Finally, dt dorsal root ganglion neurons formed neurites in culture, but the cytoskeleton was disorganized within these neurites. Our results demonstrate that dystonin is essential for maintaining neuronal cytoskeleton integrity but is not required for establishing neuronal morphology.     

Dystonin is an essential component of the Schwann cell cytoskeleton at the time of myelination

A central role for the Schwann cell cytoskeleton in the process of peripheral nerve myelination has long been suggested. However, there is no genetic or biological evidence as yet to support this assumption. Here we show that dystonia musculorum (dt) mice, which carry mutations in dystonin, a cytoskeletal crosslinker protein, have hypo/amyelinated peripheral nerves. In neonatal dt mice, Schwann cells were arrested at the promyelinating stage and had multiple myelinating lips. Nerve graft experiments and primary cultures of Schwann cells demonstrated that the myelination abnormality in dt mice was autonomous to Schwann cells. In culture, dt Schwann cells showed abnormal polarization and matrix attachment, and had a disorganized cytoskeleton. Finally, we show that the dt mutation was semi-dominant, heterozygous animals presenting hypo- and hyper-myelinated peripheral nerves. Altogether, our results suggest that dt Schwann cells are deficient for basement membrane interaction and demonstrate that dystonin is an essential component of the Schwann cell cytoskeleton at the time of myelination.     

Loss of response of carnitine palmitoyltransferase I to okadaic acid in transformed hepatic cells

The specific activity of carnitine palmitoyltransferase I (CPT-I) was similar in mitochondria isolated from rat Fao and human HepG2 hepatoma cells and from rat hepatocytes, but almost twofold higher in permeabilized hepatoma cells than in permeabilized hepatocytes. Short-term exposure to okadaic acid induced a ca. 80% stimulation of CPT-I in hepatocytes, whereas no significant response of the enzyme from hepatoma cells was evident. Thus, the high CPT-I activity displayed by hepatoma cells may be reached by hepatocytes upon challenge to okadaic acid. Reconstitution experiments with purified mitochondrial and cytoskeletal fractions showed that the cytoskeleton of hepatocytes produced a more remarkable inhibition of CPT-I than the cytoskeleton of Fao cells. The present data may be explained by a disruption of interactions between CPT-I and cytoskeletal components in tumor cells that may be involved in the okadaic acid-induced activation of hepatic CPT-I as previously suggested.     

Study of regulation of mitochondrial respiration in vivo. An analysis of influence of ADP diffusion and possible role of cytoskeleton

The purpose of this work was to investigate the mechanism of regulation of mitochondrial respiration in vivo in different muscles of normal rat and mice, and in transgenic mice deficient in desmin. Skinned fiber technique was used to study the mitochondrial respiration in the cells in vivo in the heart, soleus and white gastrocnemius skeletal muscles of these animals. Also, cardiomyocytes were isolated from the normal rat heart, permeabilized by saponin and the "ghost" (phantom) cardiomyocytes were produced by extraction of myosin with 800 mM KCl. Use of confocal immunofluorescent microscopy and anti-desmin antibodies showed good preservation of mitochondria and cytoskeletal system in these phantom cells. Kinetics of respiration regulation by ADP was also studied in these cells in detail before and after binding of anti-desmine antibodies with intermediate filaments. In skinned cardiac or soleus skeletal muscle fibers but not in fibers from fast twitch skeletal muscle the kinetics of mitochondrial respiration regulation by ADP was characterized by very high apparent Km (low affinity) equal to 300-400 microM, exceeding that for isolated mitochondria by factor of 25. In skinned fibers from m. soleus, partial inhibition of respiration by NaN3 did not decrease the apparent Km for ADP significantly, this excluding the possible explanation of low apparent affinity of mitochondria to ADP in these cells by its rapid consumption due to high oxidative activity and by intracellular diffusion problems. However, short treatment of fibers with trypsin decreased this constant value to 40-70 microM, confirming the earlier proposition that mitochondrial sensitivity to ADP in vivo is controlled by some cytoplasmic protein. Phantom cardiomyocytes which contain mostly mitochondria and cytoskeleton and retain the normal shape, showed also high apparent Km values for ADP. Therefore, they are probably the most suitable system for studies of cellular factors which control mitochondrial function in the cells in vivo. In these phantom cells anti-desmin antibodies did not change the kinetics of respiration regulation by ADP. However, in skinned fibers from the heart and m. soleus of transgenic desmin-deficient mice some changes in kinetics of respiration regulation by ADP were observed: in these fibers two populations of mitochondria were observed, one with usually high apparent Km for ADP and the second one with very low apparent Km for ADP. Morphological observations by electron microscopy confirmed the existence of two distinct cellular populations in the muscle cells of desmin-deficient mice. The results conform to the conclusion that the reason for observed high apparent Km for ADP in regulation of oxidative phosphorylation in heart and slow twitch skeletal muscle cells in vivo is low permeability of mitochondrial outer membrane porins but not diffusion problems of ADP into and inside the cells. Most probably, in these cells there is a protein associated with cytoskeleton, which controls the permeability of the outer mitochondrial porin pores (VDAC) for ADP. Desmin itself does not display this type of control of mitochondrial porin pores, but its absence results in appearance of cells with disorganised structure and of altered mitochondrial population probably lacking this unknown VDAC controlling protein. Thus, there may be functional connection between mitochondria, cellular structural organisation and cytoskeleton in the cells in vivo due to the existence of still unidentified protein factor(s).     

Simulations of the erythrocyte cytoskeleton at large deformation. I. Microscopic models

Three variations of a polymer chain model for the human erythrocyte cytoskeleton are used in large deformation simulations of microscopic membrane patches. Each model satisfies an experimental observation that the contour length of the spectrin tetramers making up the erythrocyte cytoskeleton is roughly square root of 7 times the end-to-end distance of the tetramer in vivo. Up to modest stress, each brushy cytoskeletal network behaves, consistently, like a low-temperature, planar network of Hookean springs, with a model-dependent effective spring constant, keff, in the range of 20-40 kBT/s(o)2, where T is the temperature and s(o) is the force-free spring length. However, several features observed at large deformation distinguish these models from spring networks: 1) Network dimensions do not expand without bound in approaching a critical isotropic tension (square root of 3 keff) that is a characteristic limit of Hookean spring nets. 2) In surface compression, steric interactions among the chain elements prevent a network collapse that is otherwise observed in compression of planar triangulated networks of springs. 3) Under uniaxial surface tension, isotropy of the network disappears only as the network is stretched by more than 50% of its equilibrium dimensions. Also found are definitively non-Hookean regimes in the stress dependence of the elastic moduli. Lastly, determinations of elastic moduli from both fluctuations and stress/strain relations prove to be consistent, implying that consistency should be expected among experimental determinations of these quantities.     

The bacterial actin nucleator protein ActA of Listeria monocytogenes contains multiple binding sites for host microfilament proteins

BACKGROUND: Several intracellular pathogens, including Listeria monocytogenes, use components of the host actin-based cytoskeleton for intracellular movement and for cell-to-cell spread. These bacterial systems provide relatively simple model systems with which to study actin-based motility. Genetic analysis of L. monocytogenes led to the identification of the 90 kD surface-bound ActA polypeptide as the sole bacterial factor required for the initiation of recruitment of host actin filaments. Numerous host actin-binding proteins have been localized within the actin-based cytoskeleton that surrounds Listeria once it is inside a mammalian cell, including alpha-actinin, fimbrin, filamin, villin, ezrin/radixin, profilin and the vasodilator-stimulated phosphoprotein, VASP. Only VASP is known to bind directly to ActA. We sought to determine which regions of the ActA molecule interact with VASP and other components of the host microfilament system. RESULTS: We used the previously developed mitochondrial targeting assay to determine regions of the ActA protein that are involved in the recruitment of the host actin-based cytoskeleton. By examining amino-terminally truncated ActA derivatives for their ability to recruit cytoskeletal proteins, an essential element for actin filament nucleation was identified between amino acids 128 and 151 of ActA. An ActA derivative from which the central proline-rich repeats were deleted retained its ability to recruit filamentous actin, albeit poorly, but was unable to bind VASP. CONCLUSIONS: Our studies reveal the initial interactions that take place between invading Listeria and host microfilament proteins. The listerial ActA polypeptide contains at least two essential sites that are required for efficient microfilament assembly: an amino-terminal 23 amino-acid region for actin filament nucleation, and VASP-binding proline-rich repeats. Hence, ActA represents a prototype actin filament nucleator. We suggest that host cell analogues of ActA exist and are important components of structures involved in cell motility.     

Mid-term surgical results after valve replacement with the CarboMedics valve prosthesis

Located at the level of the Z-line, the transverse cytoskeletal network of insect-flight muscle interconnects adjacent myofibrils with one another, and interconnects peripheral myofibrils with the cell membrane. This network has been presumed to keep myofibrils in register, or to distribute tension laterally among myofibrils. In this study, we used scanning-electron microscopy to reveal details of the three-dimensional arrangement of this network. The network is seen to interconnect longitudinal elements of the cytoskeletal network which surround each myofibril. The arrangement is not unlike that seen in vertebrate skeletal muscle. Interestingly, the transverse network makes contact with cell components such as dense bodies and mitochondria. Such contacts imply potential roles over and above those noted above. The network may be involved not only in mechanical function, but possibly also in intracellular communication.     

Role of the actin cytoskeleton on epithelial Na+ channel regulation

The regulatory role of actin filament organization on epithelial Na+ channel activity is reviewed in this report. The actin cytoskeleton, consisting of actin filaments and associated actin-binding proteins, is essential to various cellular events including the maintenance of cell shape, the onset of cell motility, and the distribution and stability of integral membrane proteins. Functional interactions between the actin cytoskeleton and specific membrane transport proteins are, however, not as well understood. Recent studies from our laboratory have determined that dynamic changes in the actin cytoskeletal organization may represent a novel signaling mechanism in the regulation of ion transport in epithelia. This report summarizes work conducted in our laboratory leading to an understanding of the molecular steps associated with the regulatory role of the actin-based cytoskeleton on epithelial Na+ channel function. The basis of this interaction lies on the regulation by actin-binding proteins and adjacent structures, of actin filament organization which in turn, modulates ion channel activity. The scope of this interaction may extend to such relevant cellular events as the vasopressin response in the kidney.     

Structure and dynamics of the fenestrae-associated cytoskeleton of rat liver sinusoidal endothelial cells

This article describes the cytoskeleton associated with fenestrae and sieve plates of rat liver sinusoidal endothelial cells. Fenestrae control the exchange between the blood and parenchymal cells. We present evidence indicating that several agents that change the fenestrae and sieve plates also cause changes in the cytoskeleton. Cultured liver endothelial cells (LECs) were slightly fixed and treated with cytoskeleton extraction buffer. Detergent-extracted whole mounts of cultured cells were prepared for either scanning electron microscopy (SEM) or transmission electron microscopy (TEM). Extracted cells show an integral intricate cytoskeleton; sieve plates and fenestrae are delineated by cytoskeleton elements. Fenestrae are surrounded by a filamentous, fenestrae-associated cytoskeleton with a mean filament thickness of 16 nm. Sieve plates are surrounded and delineated by microtubuli, which form a network together with additional branching cytoskeletal elements. The addition of ethanol to cultured cells enlarged the diameter for these fenestrae-associated cytoskeleton rings by 5%, whereas serotonin treatment reduced the diameter by 20%. These observations indicate that the fenestrae-associated cytoskeleton probably changes the size of fenestrae after different treatments. After treatment with cytochalasin B the number of fenestrae increased. However, cytochalasin B did not change the structure of the fenestrae-associated cytoskeleton ring, but disperses the microtubuli. In conclusion, LECs have a cytoskeleton that defines and supports sieve plates and fenestrae. Fenestrae-associated cytoskeleton is a dynamic structure and plays a role in maintaining and regulating the size of fenestrae after different treatments. Therefore, the fenestrae-associated cytoskeleton controls the important hepatic function of endothelial filtration.     

Cytosolic free calcium and the cytoskeleton in the control of leukocyte chemotaxis

In response to a chemotactic gradient, leukocytes extravasate and chemotax toward the site of pathogen invasion. Although fundamental in the control of many leukocyte functions, the role of cytosolic free Ca2+ in chemotaxis is unclear and has been the subject of debate. Before becoming motile, the cell assumes a polarized morphology, as a result of modulation of the cytoskeleton by G protein and kinase activation. This morphology may be reinforced during chemotaxis by the intracellular redistribution of Ca2+ stores, cytoskeletal constituents, and chemoattractant receptors. Restricted subcellular distributions of signaling molecules, such as Ca2+, Ca2+/calmodulin, diacylglycerol, and protein kinase C, may also play a role in some types of leukocyte. Chemotaxis is an essential function of most cells at some stage during their development, and a deeper understanding of the molecular signaling and structural components involved will enable rational design of therapeutic strategies in a wide variety of diseases.     

Keith R. Porter Lecture, 1996. Of mice and men: genetic disorders of the cytoskeleton

Since the time when I was a postdoctoral fellow under the supervision of Dr. Howard Green, then at the Massachusetts Institute of Technology, I have been interested in understanding the molecular mechanisms underlying growth, differentiation, and development in the mammalian ectoderm. The ectoderm gives rise to epidermal keratinocytes and to neurons, which are the only two cell types of the body that devote most of their protein-synthesizing machinery to developing an elaborate cytoskeletal architecture composed of 10-nm intermediate filaments (IFs). Our interest is in understanding the architecture of the cytoskeleton in keratinocytes and in neurons, and in elucidating how perturbations in this architecture can lead to degenerative diseases of the skin and the nervous system. I will concentrate on the intermediate filament network of the skin and its associated genetic disorders, since this has been a long-standing interest of my laboratory at the University of Chicago.     

Inhibition of phospholipase D activity by fodrin. An active role for the cytoskeleton

Phospholipase D (PLD) is a major enzyme implicated in important cellular processes such as secretion and proliferation. The knowledge of its regulation is essential to understand the control of these phenomena. Several proteins activating PLD have been described in the last years. In this report, we chromatographed bovine brain cytosolic proteins to identify fodrin, the non-erythroid spectrin, as the first described inhibitor of PLD. A cytosolic fraction with an inhibitory effect on PLD activity loses its capacity after immunoprecipitation of fodrin. Moreover, at 1 nM, purified fodrin blocks fully and quickly PLD activity, whatever the stimuli used. In contrast, fodrin has no effect on adenylate cyclase activity. Fodrin-analogous proteins like dimeric or tetrameric erythroid spectrin have the same inhibitory effect on PLD, at higher concentrations. Other cytoskeletal proteins, actin and vimentin, are inefficient on PLD inhibition. The mechanisms implicated in PLD modulation such as post-translational modifications of fodrin and the role of small G-proteins on the cytoskeleton regulation are discussed. In conclusion, this study reveals that fodrin is involved in the control of PLD activity, suggesting that the cytoskeleton could have an active role in control of secretion and proliferation.     

Antagonistic roles of neurofilament subunits NF-H and NF-M against NF-L in shaping dendritic arborization in spinal motor neurons

Dendrites play important roles in neuronal function. However, the cellular mechanism for the growth and maintenance of dendritic arborization is unclear. Neurofilaments (NFs), a major component of the neuronal cytoskeleton, are composed of three polypeptide subunits, NF-H, NF-M, and NF-L, and are abundant in large dendritic trees. By overexpressing each of the three NF subunits in transgenic mice, we altered subunit composition and found that increasing NF-H and/or NF-M inhibited dendritic arborization, whereas increasing NF-L alleviated this inhibition. Examination of cytoskeletal organization revealed that increasing NF-H and/or NF-M caused NF aggregation and dissociation of the NF network from the microtubule (MT) network. Increasing NF-H or NF-H together with NF-M further reduced NFs from dendrites. However, these changes were reversed by elevating the level of NF-L with either NF-H or NF-M. Thus, NF-L antagonizes NF-H and NF-M in organizing the NF network and maintaining a lower ratio of NF-H and NF-M to NF-L is critical for the growth of complex dendritic trees in motor neurons.     

Neurexin IV, caspr and paranodin--novel members of the neurexin family: encounters of axons and glia

Axonal insulation is of key importance for the proper propagation of action potentials. In Drosophila and other invertebrates, it has recently been demonstrated that septate junctions play an essential role in axonal insulation or blood-brain-barrier formation. Neurexin IV, a molecular component of Drosophila septate junctions, has been shown to be essential for axonal insulation in the PNS in embryos and larvae. Interestingly, a vertebrate homolog of Neurexin IV, caspr--also named paranodin--has been shown to localize to septate-like junctional structures. These vertebrate junctions are localized to the paranodal region of the nodes of Ranvier, between axons and Schwann cells. Caspr/paranodin might play an important role in barrier formation, and link neuronal membrane components with the axonal cytoskeletal network.