Defective glomerulogenesis in the absence of laminin alpha5 demonstrates a developmental role for the kidney glomerular basement membrane

Laminins are major components of all basement membranes. They are a diverse group of alpha/beta/gamma heterotrimers formed from five alpha, three beta, and three gamma chains. Laminin alpha5 is a widely expressed chain found in many embryonic and adult basement membranes. During embryogenesis, alpha5 has a role in disparate developmental processes, including neural tube closure, digit septation, and placentation. Here, we analyzed kidney development in Lama5 mutant embryos and found a striking defect in glomerulogenesis associated with an abnormal glomerular basement membrane (GBM). This correlates with failure of the developmental switch in laminin alpha chain deposition in which alpha5 replaces alpha1 in the GBM at the capillary loop stage of glomerulogenesis. In the absence of a normal GBM, glomerular epithelial cells were in disarray, and endothelial and mesangial cells were extruded from within the constricting glomerulus, leading to a complete absence of vascularized glomeruli. In addition, a minority of Lama5 mutant mice lacked one or both kidneys, indicating that laminin alpha5 is also important in earlier kidney development. Our results demonstrate a dual role for laminin alpha5 in kidney development, illustrate a novel defect in glomerulogenesis, and indicate a heretofore unappreciated developmental role for the GBM in influencing the behavior of epithelial and endothelial cells.     

Syndecan binding sites in the laminin alpha1 chain G domain

The laminin alpha1 chain G domain has multiple biological activities. Previously, we identified cell binding sequences in the laminin alpha1 chain G domain by screening 113 synthetic peptide-polystyrene beads for cell attachment activity. Here, we have used a recombinant protein of the laminin alpha1 G domain (rec-alpha1G) and a large set of synthetic peptides to further identify and characterize heparin, cell, and syndecan-4 binding sites in the laminin alpha1 chain G domain. The rec-alpha1G protein promoted both cell attachment and heparin binding (K(D) = 19 nM). Cell attachment to the rec-alpha1G protein was inhibited 60% by heparin and 30% by EDTA. The heparin binding sites were identified by competing heparin binding to the rec-alpha1G protein with 110 synthetic peptides in solution. Only two peptides, AG73 (IC(50) = 147 microM) and AG75 (IC(50) = 206 microM), inhibited heparin binding to rec-alpha1G. When the peptides were compared in a solid-phase heparin binding assay, AG73 showed more heparin binding than AG75. AG73 also inhibited fibroblast attachment to the rec-alpha1G protein, but AG75 did not. Cell attachment to the peptides was studied using peptide-coated plates and peptide-conjugated sepharose beads. AG73 promoted cell attachment in both assays, but AG75 only showed cell attachment activity in the bead assay. Additionally, AG73, but not AG75, inhibited branching morphogenesis of mouse submandibular glands in organ culture. Furthermore, the rec-alpha1G protein bound syndecan-4, and both AG73 and AG75 inhibited this binding. These results suggest that the AG73 and AG75 sites are important for heparin and syndecan-4 binding in the laminin alpha1 chain G domain. These sites may play a critical role in the diverse biological activities involving heparin and syndecan-4 binding.     

Loss of laminin epitopes during glomerular basement membrane assembly in developing mouse kidneys.

Kidney glomerular basement membranes (GMBs) originate in development from fusion of a dual basement membrane between endothelial cells and primitive epithelial podocytes. After fusion, segments of newly synthesized matrix, derived primarily from podocytes, appear as subepithelial outpockets and are spliced into GBMs during glomerular capillary loop expansion. To investigate GBM assembly further, we examined newborn mouse kidneys with monoclonal rat anti-mouse laminin IgGs (MAb) conjugated to horseradish peroxidase (HRP). In adults, these MAb strongly label glomerular mesangial matrices but bind only weakly or not at all to mature GBMs. In contrast, anti-laminin MAb intensely bound newborn mouse GBMs undergoing initial assembly. After intraperitoneal injection of MAb-HRP into neonates, dense binding occurred across both subendothelial and subepithelial pre-fusion GMBs as well as forming mesangial matrices. Considerably less MAb binding was seen, however, in post-fusion GBMs from more mature glomeruli in the same section, although mesangial matrices remained positive. In addition, new subepithelial segments in areas of splicing were negative. These results conflict with those obtained previously with injections of polyclonal anti-laminin IgGs into newborns or adults, which result in complete labeling of all GBMs. Although epitope masking cannot be completely excluded, we believe that decreased MAb binding to developing GBM reflects actual epitope loss. This loss could occur by laminin isoform substitution, conformational change, and/or proteolytic processing during GBM assembly.     

Studies of the permeation properties of glomerular basement membrane: cross-linking renders glomerular basement membrane permeable to protein

Cross-linking glomerular basement membrane (GBM) has been shown to render it more permeable to protein. Isolated pig GBM was cross-linked with dimethylmalonimidate which reacts selectively with lysine ?-NH₂ groups or with glutaraldehyde, a less selective cross-linking agent. Studies of the ultrafiltration properties of these materials in vitro using cytochrome c, myoglobin, bovine serum albumin and immunoglobulin showed that cross-linking had markedly increased solvent and protein fluxes as compared with native membranes particularly at higher pressures. Filtration studies with serum demonstrated that the cross-linked membranes were more permeable to serum proteins. Thickness measurements under pressure indicated that cross-linked membrane was less compressed than native membrane as pressure was increased. Pore theory did not provide a suitable model for analysis of the results, but analysis of the results using the fibre-matrix hypothesis indicated that cross-linking had the effect of bundling together the fibres (type IV collagen) in the GBM matrix. The effect of cross-linking on filtration could be explained by a combination of contraction of the membrane, fibre bundling and increased rigidity compared with native membrane. Cross-linking of GBM might lead to long-term damage of the glomerular capillary wall in nephritis, so promoting proteinuria.     

Ultrastructural orientation of laminin 5 in the epidermal basement membrane: an updated model for basement membrane organization.

Laminin 5 is a trimeric glycoprotein involved in cell adhesion in the epidermal basement membrane. To determine the precise orientation of laminin 5 in adult human skin, we used plural epitope-specific monoclonal antibodies, a polyclonal antiserum, and postembedding immunogold electron microscopy (IEM). Immunogold labeling distances from the basal keratinocyte plasma membrane (PM) were measured for each gold particle (>200 particles) and the mean distance (nm) calculated. Antibodies included BM165 (recognizing the alpha 3-chain first globular domain) that was measured at 35.40 +/- 2.20 nm from the keratinocyte PM, K140 (recognizing a region adjacent to the beta 3-chain globular domain IV) that measured 45.20 +/- 3.60 nm from the PM, and an anti-laminin 5 polyclonal antiserum that was 43.43 +/- 6.28 nm from the PM. The laminin 5 gamma 2-chain short arm hinge domain was previously localized to the lower lamina densa (LD) at approximately 56.30 +/- 1.65 nm from the keratinocyte PM. Taken together with previous gamma 2-chain data and the distribution of the polyclonal antisera, we estimate that the long axis of laminin 5 is oriented at an angle of approximately 27 degrees from the horizontal lamina lucida (LL)/LD border and propose that the gamma 2-chain lies farthest from the PM. This novel orientation, with the majority of the laminin 5 molecule lying obliquely along the LL/LD border and not perpendicularly, as was first thought, sheds new light on the organization of the basement membrane and likely molecular interactions.     

Clogging of the glomerular basement membrane

The negative charges of the sulfated glycosaminoglycans (GAGs) of the glomerular basement membrane (GBM) were differentially neutralized by perfusin with high molarity buffers in order to determine whether or not these charges protect the GBM from being clogged by circulating plasma macromolecules. Progressive elimination of the negative charges resulted in clogging of the GBM by perfused native ferritin (NF) and bovine serum albumin as evidenced ultrastructurally by the increase in accumulation of NF in the GBM. In addition, the permeability of the GBM to 125I-insulin, a macromolecule which is normally freely permeable, and the glomerular filtration rate (as determined by [3H]inulin clearance) were markedly reduced after the GBM had been clogged with NF in the presence of high molarity buffer, thereby indicating that clogging severely reduces the ability of the GMB to act as a selective filter. These findings are consistent with the idea that the sulfated GAGs of the GBM serve as anticlogging agents.     

Origin of the glomerular basement membrane visualized after in vivo labeling of laminin in newborn rat kidneys

To examine the origin and assembly of glomerular basement membranes (GBMs), affinity purified anti-laminin IgG was directly coupled to horseradish peroxidase (HRP) and intravenously injected into newborn rats. Kidneys were then processed for peroxidase histochemistry and microscopy. Within 1 h after injection, anti-laminin bound to basement membranes of nephrons in all developmental stages (vesicle, comma, S-shaped, developing capillary loop, and maturing glomeruli). In S-shaped and capillary loop glomeruli, anti-laminin-HRP labeled a double basal lamina between the endothelium and epithelium. Sections incubated with anti-laminin in vitro showed labeling within the rough endoplasmic reticulum of endothelium and epithelium, indicating that both cell types synthesized laminin for the double basement membrane. In maturing glomeruli, injected anti-laminin-HRP bound throughout the GBMs, and double basement membranes were rarely observed. At this stage, however, numerous knobs or outpockets of basement membrane material extending far into the epithelial side of the capillary wall were identified and these were also labeled throughout their full thickness. No such outpockets were found in the endothelial cell layer of newborn rats (and they normally are completely absent in fully mature, adult glomeruli). In contrast with these results, in kidneys fixed 4-6 d after anti-laminin IgG-HRP injection, basement membranes of vesicle, comma, and S-shaped nephrons were unlabeled, indicating that they were assembled after injection. GBM labeling was seen in maturing glomeruli, however. In addition, the outpockets of basement membrane extending into the epithelium were often completely unlabeled whereas GBMs lying immediately beneath them were labeled intensely, which indicates that the outpockets were probably assembled by the epithelium. Injections of sheep anti-laminin IgG followed 8 d later with injections of biotin-rabbit anti-laminin IgG and double-label immunofluorescence microscopy confirmed that GBM formation continued during individual capillary loop expansion. GBM assembly therefore occurs by at least two different processes at separate times in development: (a) fusion of endothelial and epithelial basement membranes followed by (b) addition of new basement membrane from the epithelium into existing GBMs.     

Studies of the permeation properties of glomerular basement membrane: cross-linking renders glomerular basement membrane permeable to protein.

Cross-linking glomerular basement membrane (GBM) has been shown to render it more permeable to protein. Isolated pig GBM was cross-linked with dimethylmalonimidate which reacts selectively with lysine epsilon-NH2 groups or with glutaraldehyde, a less selective cross-linking agent. Studies of the ultrafiltration properties of these materials in vitro using cytochrome c, myoglobin, bovine serum albumin and immunoglobulin showed that cross-linking had markedly increased solvent and protein fluxes as compared with native membranes particularly at higher pressures. Filtration studies with serum demonstrated that the cross-linked membranes were more permeable to serum proteins. Thickness measurements under pressure indicated that cross-linked membrane was less compressed than native membrane as pressure was increased. Pore theory did not provide a suitable model for analysis of the results, but analysis of the results using the fibre-matrix hypothesis indicated that cross-linking had the effect of bundling together the fibres (type IV collagen) in the GBM matrix. The effect of cross-linking on filtration could be explained by a combination of contraction of the membrane, fibre bundling and increased rigidity compared with native membrane. Cross-linking of GBM might lead to long-term damage of the glomerular capillary wall in nephritis, so promoting proteinuria.     

Identification of biologically active sequences in the laminin alpha 4 chain G domain

Laminins are a family of trimeric extracellular matrix proteins consisting of alpha, beta, and gamma chains. So far five different laminin alpha chains have been identified. The laminin alpha 4 chain, which is present in laminin-8/9, is expressed in cells of mesenchymal origin, such as endothelial cells and adipocytes. Previously, we identified heparin-binding sites in the C-terminal globular domain (G domain) of the laminin alpha 4 chain. Here we have focused on the biological functions of the laminin alpha 4 chain G domain and screened active sites using a recombinant protein and synthetic peptides. The rec-alpha 4G protein, comprising the entire G domain, promoted cell attachment activity. The cell attachment activity of rec-alpha 4G was completely blocked by heparin and partially inhibited by EDTA. We synthesized 116 overlapping peptides covering the entire G domain and tested their cell attachment activity. Twenty peptides showed cell attachment activity, and 16 bound to heparin. We further tested the effect of the 20 active peptides in competition assays for cell attachment and heparin binding to rec-alpha 4G protein. A4G6 (LAIKNDNLVYVY), A4G20 (DVISLYNFKHIY), A4G82 (TLFLAHGRLVFM), and A4G83 (LVFMFNVGHKKL), which promoted cell attachment and heparin binding, significantly inhibited both cell attachment and heparin binding to rec-alpha 4G. These results suggest that the four active sites are involved in the biological functions of the laminin alpha 4 chain G domain. Furthermore, rec-alpha 4G, A4G6, and A4G20 were found to interact with syndecan-4. These active peptides may be useful for defining of the molecular mechanism laminin-receptor interactions and laminin-mediated cellular signaling pathways.     

Interaction of acetylcholinesterase with the G4 domain of the laminin alpha1-chain

Although the primary function of AChE (acetylcholinesterase) is the synaptic hydrolysis of acetylcholine, it appears that the protein is also able to promote various non-cholinergic activities, including cell adhesion, neurite outgrowth and amyloidosis. We have observed previously that AChE is able to bind to mouse laminin-111 in vitro by an electrostatic mechanism. We have also observed that certain mAbs (monoclonal antibodies) recognizing AChE's PAS (peripheral anionic site) inhibit both laminin binding and cell adhesion in neuroblastoma cells. Here, we investigated the interaction sites of the two molecules, using docking, synthetic peptides, ELISAs and conformational interaction site mapping. Mouse AChE was observed on docking to bind to a discontinuous, largely basic, structure, Val(2718)-Arg-Lys-Arg-Leu(2722), Tyr(2738)-Tyr(2739), Tyr(2789)-Ile-Lys-Arg-Lys(2793) and Val(2817)-Glu-Arg-Lys(2820), on the mouse laminin alpha1 G4 domain. ELISAs using synthetic peptides confirmed the involvement of the AG-73 site (2719-2729). This site overlaps extensively with laminin's heparin-binding site, and AChE was observed to compete with heparan sulfate for laminin binding. Docking showed the major component of the interaction site on AChE to be the acidic sequence Arg(90)-Glu-Leu-Ser-Glu-Asp(95) on the omega loop, and also the involvement of Pro(40)-Pro-Val(42), Arg(46) (linked to Glu(94) by a salt bridge) and the hexapeptide Asp(61)-Ala-Thr-Thr-Phe-Gln(66). Epitope analysis, using CLiPS technology, of seven adhesion-inhibiting mAbs (three anti-human AChE, one anti-Torpedo AChE and three anti-human anti-anti-idiotypic antibodies) showed their major recognition site to be the sequence Pro(40)-Pro-Met-Gly-Pro-Arg-Arg-Phe(48) (AChE human sequence). The antibodies, however, also reacted with the proline-containing sequences Pro(78)-Gly-Phe-Glu-Gly-Thr-Glu(84) and Pro(88)-Asn-Arg-Glu-Leu-Ser-Glu-Asp(95). Antibodies that recognized other features of the PAS area but not the Arg(90)-Gly-Leu-Ser-Glu-Asp(95) motif interfered neither with laminin binding nor with cell adhesion. These results define sites for the interaction of AChE and laminin and suggest that the interaction plays a role in cell adhesion. They also suggest the strong probability of functional redundancy between AChE and other molecules in early development, particularly heparan sulfate proteoglycans, which may explain the survival of the AChE-knockout mouse.     

Inhibition of laminin alpha 1-chain expression leads to alteration of basement membrane assembly and cell differentiation

The expression of the constituent alpha 1 chain of laminin-1, a major component of basement membranes, is markedly regulated during development and differentiation. We have designed an antisense RNA strategy to analyze the direct involvement of the alpha 1 chain in laminin assembly, basement membrane formation, and cell differentiation. We report that the absence of alpha 1-chain expression, resulting from the stable transfection of the human colonic cancer Caco2 cells with an eukaryotic expression vector comprising a cDNA fragment of the alpha 1 chain inserted in an antisense orientation, led to (a) an incorrect secretion of the two other constituent chains of laminin-1, the beta 1/gamma 1 chains, (b) the lack of basement membrane assembly when Caco2-deficient cells were cultured on top of fibroblasts, assessed by the absence of collagen IV and nidogen deposition, and (c) changes in the structural polarity of cells accompanied by the inhibition of an apical digestive enzyme, sucrase-isomaltase. The results demonstrate that the alpha 1 chain is required for secretion of laminin-1 and for the assembly of basement membrane network. Furthermore, expression of the laminin alpha 1-chain gene may be a regulatory element in determining cell differentiation.     

Degradation of glomerular basement membrane by human neutrophils in vitro

The glomerular basement membrane is susceptible to immunologic injury when immune complexes or anti-basement-membrane antibodies become lodged in its network. We have studied the digestion of glomerular basement membrane prepared from normal human kidney by isolated neutrophils. In the absence of immunoglobulin aggregates or immune complexes, there was little evidence of neutrophil adherence to the membrane, of release of lysosomal enzymes, or of digestion. However, when the basement membrane contained immunoglobin G (IgG) aggregates generated in situ by heating the membrane impregnated with IgG to 63 degrees C, electron micrographs showed neutrophils adherent to the basement-membrane surface and phagocytosis of smaller fragments. Lysosomal enzymes were detectable in the extracellular medium, and measurements of either total protein or hydroxyproline solubilized showed digestion of 80 micrograms basement membrane/h per 10(7) cells. Hydroxyproline solubilization was almost totally inhibited by phenylmethylsulphonyl fluoride, indicating that the neutrophil serine proteinases, elastase and cathepsin G are responsible for degradation. These findings provide direct evidence for the digestion of extracellular matrix by neutrophils stimulated in situ by deposited immune complexes as a contributor to inflammatory tissue damage.     

Identification of neurite outgrowth active sites on the laminin alpha4 chain G domain

The laminin alpha4 chain is widely distributed in various mesodermal tissues, including the perineurium of peripheral nerves, dorsal root ganglion (DRG), skeletal muscle, and capillaries, and plays important roles in synaptic specialization at the neuromuscular junction and in microvascular formation. The C-terminal globular domain (G domain) of the laminin alpha4 chain was previously found to be critical for heparin binding and cell attachment activity. Here, we focused on neurite outgrowth activity of the laminin alpha4 chain G domain. We found that the recombinant alpha4 chain G domain protein (rec-alpha4G) promoted neurite outgrowth of rat pheochromocytoma PC12 cells. When 114 overlapping synthetic peptides that covered the entire G domain were tested for neurite outgrowth activity, nine peptides were active, but the 105 remaining peptides did not exhibit activity. Three of the nine active peptides, A4G6 (LAIKNDNLVYVY), A4G20 (DVISLYNFKHIY), and A4G107 (VIRDSNVVQLDV), strongly promoted neurite outgrowth of PC12 cells. A4G107 was found to form amyloid-like fibrils in Congo red, X-ray, and electron microscopy analyses. We also synthesized cyclic peptides to evaluate their conformational requirements. Cyclic peptide A4G82X (cyc-A4G82X;TLFLAHGRLVFX, where X is norleucine) significantly enhanced neurite outgrowth activity, but the rest of the cyclic peptides eliminated the activity. The A4G82 sequence is located on the loop region, suggesting that the activity of A4G82 is required for a loop conformation. These peptides also exhibited neurite outgrowth activity with dorsal root ganglion (DRG) explants and with DRG cells from E14.5 mouse embryos, indicating that they are active in both neuronal cell lines and native neuronal cells. Taken together, the data suggest that the peptides from the laminin alpha4 chain G domain promote neurite outgrowth activity via a specific conformation.     

Identification of neurite outgrowth promoting sites on the laminin alpha 3 chain G domain

Laminins are expressed in specific tissues and are involved in various biological activities including promoting cell adhesion, growth, migration, neurite outgrowth, and differentiation. The laminin alpha3 chain is mainly located in the skin and is also expressed in the floor plate of the developing neural tube. Previously, we showed that the human laminin alpha3 chain LG4 module binds to syndecan-2/4, a membrane-associated proteoglycan, and promotes human fibroblast adhesion. Here, we have evaluated the neurite outgrowth activity of the laminin alpha3 chain LG4 and LG5 modules. Three overlapping recombinant proteins, which contained LG4 and/or LG5 modules of the human laminin alpha3 chain, were prepared using a mammalian cell expression system. Two proteins, rec-alpha3LG4-5 and rec-alpha3LG4, promoted cell attachment and neurite outgrowth of rat pheochromocytoma PC12 cells, but rec-alpha3LG5 was inactive. Twenty-two peptides covering the entire LG4 module were synthesized and tested for cell attachment and neurite outgrowth activity to identify active sites of the LG4 module. A3G75 (KNSFMALYLSKG, alpha3 chain 1411-1422) and A3G83 (GNSTISIRAPVY, alpha3 chain 1476-1487) promoted PC12 cell attachment and neurite outgrowth. Additionally, A3G75 and A3G83 inhibited PC12 cell attachment to rec-alpha3LG4. These results suggest that the A3G75 and A3G83 sites are important for PC12 cell attachment and neurite outgrowth in the laminin alpha3 chain LG4 module. We also conjugated the A3G75 and A3G83 peptides on chitosan membranes to test their potential as bio-materials. These peptide-conjugated chitosan membranes were more active for neurite outgrowth than the peptide-coated plates. These results suggest that the A3G75- and A3G83-conjugated chitosan membranes are applicable as bio-medical materials for neural tissue repair and engineering.     

Metal-binding properties of the isolated glomerular basement membrane

The nature of binding of metal cations to the glomerular basement membrane has been investigated using isolated bovine glomerular basement membrane. Highest-affinity binding for a number of ions is attributable to the glycosaminoglycans (mostly heparan sulfate) of the membrane. Some ions, such as divalent Mn, Ca and Ni, have specific binding sites on these polymers, while for others the ion-polyelectrolyte interaction is of a non-specific nature. Both structural and binding data indicate a linear charge density of close to unity for the heparan sulfate of the glomerular basement membrane, which at the ionic composition of the plasma filtrate corresponds to a polymer surface potential of about -45 mV. Several independent observations are better explained by a model of counter-ion condensation about the glycosaminoglycans than by conventional double layer theories. These include the valence dependence of ion binding, the sharp ejection of divalent ions at a critical concentration of La3+, and the relative insensitivity of 63Ni2+ binding to NaCl concentration in the neighbourhood of physiological ionic strength. In its interactions with metal ions, the glomerular basement membrane behaves like a dilute solution of polyelectrolytes. This conclusion has important consequences for the extent of charge reduction of the filtration barrier of the kidney, bathed as it is in an electrolyte solution of mainly monovalent salts.