Partial characterization of apoptotic factor in Alzheimer plasma.

We have previously demonstrated that a plasma natriuretic factor is present in Alzheimer's disease (AD), but not in multi-infarct dementia (MID) or normal controls (C). We postulated that the natriuretic factor might induce the increased cytosolic calcium reported in AD by inhibiting the sodium-calcium antiporter, thereby activating the apoptotic pathway. To test for a factor in AD plasma that induces apoptosis, we exposed nonconfluent cultured LLC-PK1 cells to plasma from AD, MID, and C for 2 h and performed a terminal transferase-dUTP-nick-end labeling (TUNEL) assay. The plasma from AD increased apoptosis nearly fourfold compared with MID and C. The effect was dose dependent and the peak effect was attained after a 2-h exposure. Additionally, apoptotic morphology was detected by electron microscopy, and internucleosomal DNA cleavage was found. We inhibited apoptosis by removing calcium from the medium, inhibiting protein synthesis with cycloheximide, alternately boiling or freezing and thawing the plasma, and digesting a partially purified fraction with trypsin. Heating AD plasma to 56 degrees C did not deactivate the apoptotic factor. These results demonstrate the presence of an apoptotic factor in the plasma of patients with AD.

The laminin alpha2-chain short arm mediates cell adhesion through both the alpha1beta1 and alpha2beta1 integrins.

Laminin-2, a heterotrimer composed of alpha2, beta1, and gamma1 subunits, is the primary laminin isoform found in muscle and peripheral nerve and is essential for the development and stability of basement membranes in these tissues. Expression of a domain VI-truncated laminin alpha2-chain results in muscle degeneration and peripheral nerve dysmyelination in the dy2J dystrophic mouse. We have expressed amino-terminal domains VI through IVb of the laminin alpha2-chain, as well as its laminin-1 alpha1-chain counterpart, to identify candidate cell-interactive functions of this critical region. Using integrin-specific antibodies, recognition sites for the alpha1beta1 and alpha2beta1 integrins were identified in the short arms of both laminin alpha1- and alpha2-chain isoforms. Comparisons with a beta-alpha chimeric short arm protein possessing beta1-chain domain VI further localized these activities to alpha-chain domain VI. In addition, we found that the laminin alpha2-chain short arm supported neurite outgrowth independent of other laminin-2 subunits. A heparin/heparan sulfate binding activity was also localized to this region of the laminin alpha2 subunit. These data provide the first evidence that domain VI of the laminin alpha2-chain mediates interactions with cell surface receptors and suggest that these integrin and heparin binding sites, alone or in concert, may play an important role in muscle and peripheral nerve function.

The alpha chain of laminin-1 is independently secreted and drives secretion of its beta- and gamma-chain partners.

A mammalian recombinant strategy was established to dissect rules of basement membrane laminin assembly and secretion. The alpha-, beta-, and gamma-chain subunits of laminin-1 were expressed in all combinations, transiently and/or stably, in a near-null background. In the absence of its normal partners, the alpha chain was secreted as intact protein and protein that had been cleaved in the coiled-coil domain. In contrast, the beta and gamma chains, expressed separately or together, remained intracellular with formation of betabeta or betagamma, but not gammagamma, disulfide-linked dimers. Secretion of the beta and gamma chains required simultaneous expression of all three chains and their assembly into alphabetagamma heterotrimers. Epitope-tagged recombinant alpha subunit and recombinant laminin were affinity-purified from the conditioned medium of alphagamma and alphabetagamma clones. Rotary-shadow electron microscopy revealed that the free alpha subunit is a linear structure containing N-terminal and included globules with a foreshortened long arm, while the trimeric species has the typical four-arm morphology of native laminin. We conclude that the alpha chain can be delivered to the extracellular environment as a single subunit, whereas the beta and gamma chains cannot, and that the alpha chain drives the secretion of the trimeric molecule. Such an alpha-chain-dependent mechanism could allow for the regulation of laminin export into a nascent basement membrane, and might serve an important role in controlling basement membrane formation.

Localization of heparin binding activity in recombinant laminin G domain.

Basement membrane laminin (laminin-1) is a multidomain glycoprotein that interacts with itself, heparin and cells. The interaction with heparin/heparan sulfate proteglycans is thought to be important for the architectural formation of basement membranes and adhesion to cells. The major heparin binding site has been known to reside in the long arm globular domain (G domain). The G domain is in turn subdivided into five subdomains (G1-G5). In order to localize the heparin binding regions further, recombinant G domains (rG and rG5) were expressed in Sf9 insect cells using baculovirus expression vector. By the limited proteolysis of recombinant G domains followed by either heparin affinity HPLC or overlay with radiolabeled heparin, the relative affinity of each subdomain to heparin was assigned as G1>G2 = G4>G5>G3, such that G1 bound strongly and G3 not at all. Since the activity in G1-G3 is cryptic in intact laminin long arm [Sung, U., O'Rear, J. J. & Yurchenco, P. D. (1993) J. Cell Biol. 123, 1255-1268], the active heparin binding site of G domain appears to be located in G4 and proximal G5.

Mapping of network-forming, heparin-binding, and alpha 1 beta 1 integrin-recognition sites within the alpha-chain short arm of laminin-1.

Cell-interactive and architecture-forming functions are associated with the short arms of basement membrane laminin-1. To map and characterize these functions, we expressed recombinant mouse laminin-1 alpha-chain extending from the N terminus through one third of domain IIIb. This dumbbell-shaped glycoprotein (r alpha 1(VI-IVb)'), secreted by mammalian cells, was found to possess three activities. 1) Laminin polymerization was quantitatively inhibited by recombinant protein, supporting an alpha-chain role for a three-short arm interaction model of laminin self-assembly. 2) r alpha 1(VI-IVb)' bound to heparin, and the activity was localized to a subfragment corresponding to domain VI by 125I-heparin blotting. 3) PC12 rat pheochromocytoma cells adhered to, and rapidly extended branching neurites on, r alpha 1(VI-IVb)', with adhesion inhibited by alpha 1 and beta 1 integrin chain-specific antibodies. The ability of anti-laminin antibody to block PC12 cell adhesion to laminin was selectively prevented by absorption with r alpha 1(VI-IVb)' or alpha-chain domain VI fragment. This active integrin-recognition site could furthermore be distinguished from a second cryptic alpha 1 beta 1-binding site exposed by heat treatment of fragment P1', a short arm fragment lacking globules. Thus, a polymer-forming, a heparin-binding, and the active alpha 1 beta 1 integrin-recognition site are all clustered at the end of the alpha-chain short arm, the latter two resident solely in domain VI.

Basal lamina assembly.

From studies of the 'classical' components, models for the assembly and structure of an idealized basal lamina have been developed. In particular, the evidence supports the concept of enmeshed collagen and laminin polymers, in which nidogen/entactin acts as a bridge between these molecules and provides anchorage for diverse matrix components. Different basement membranes, however, possess different members of the basic basal lamina families, such as the newly described alpha 6 (IV) collagen, alpha 2 (merosin) laminin, and beta 3 laminin (in kalinin/nicein) chains. Even though these members share homologous domains and sequences, and are likely to share certain functions, they also possess unique characteristics that are expected to provide for basal lamina heterogeneity. A combination of genetic, recombinant and biochemical approaches are now being applied to elucidate the special roles of both old and new components.

Domain-specific activation of neuronal migration and neurite outgrowth-promoting activities of laminin.

The ECM glycoprotein laminin has profound and varied actions on neurons in vitro. Little is known about how laminin's multiple domains and receptor-binding sites interact in determining its overall effects. Here, it is shown that laminin's ability to promote migration of olfactory epithelium neuronal cells maps to distal long arm domain E8 and is mediated by alpha 6 beta 1 integrin. Surprisingly, treatment of laminin with antibodies against its short arms (domains E1' or P1') uncovered a new neuronal migration-promoting activity, mediated by a beta 1 integrin other than alpha 6 beta 1. Laminin treated with anti-short arm antibodies also promoted beta 1 integrin-dependent neurite outgrowth from late embryonic retinal neurons, which are normally unresponsive to laminin. These "antibody-induced" migration and neurite outgrowth activities mapped to laminin's distal long arm, far from the site(s) of antibody binding. Evidence is presented that the induced activities are not actually cryptic in laminin, but are suppressed by an activity that is located in laminin's P1' domain and that may be lacking in the laminin homolog merosin.

A short sequence upstream of the 5' major splice site is important for encapsidation of HIV-1 genomic RNA.

A series of linker scanning and deletion mutations has been constructed in the 5' leader sequence of HIV-1. One virus with a 13-base-linker substitution upstream of the 5' major splice site was as impaired in its ability to replicate as a virus with a large deletion, which included these 13 bases, and was less efficient in packaging its genomic RNA than viruses carrying mutations between the 5' major splice site and the gag translation initiation site. These observations have led to the identification of a conserved pattern of repeated sequence elements associated with sequences experimentally defined as necessary for encapsidation of Moloney murine leukemia virus, spleen necrosis virus, avian leukosis-sarcoma viruses, and human immunodeficiency virus type 1.

Cell and heparin binding in the distal long arm of laminin: identification of active and cryptic sites with recombinant and hybrid glycoprotein.

The long arm of laminin, which binds heparin and cells, consists of three polypeptides (A, B1, and B2) joined in a coiled-coil rod attached to a terminal A chain globule (G). Previously, we found that recombinant globular domain (rG) supported heparin and myoblast binding (Yurchenco, P. D., U. Sung, M. D. Ward, Y. Yamada, and J. J. O'Rear. 1993. J. Biol. Chem. 268:8356-8365). To further analyze long arm functions, we expressed the distal moiety of the mouse laminin A chain extending from the middle of the rod to the carboxyl terminus (rAiG). This larger glycoprotein, secreted by Sf9 insect cells infected with recombinant baculovirus, was intercalated in vitro into the corresponding disulfide-linked B chain segments of laminin fragment E8 (distal long arm rod and proximal globule). The hybrid molecule (B-rAiG) possessed a structure similar to laminin long arm as judged by electron microscopy and limited proteolysis. By joining rAiG with E8-B chains, the affinity of G domain for heparin decreased from that observed with rAiG and rG to one similar to native protein. HT1080 cells adhered to E8, rAiG, and B-rAiG, less well to rG, and not to denatured E8/B-rAiG, the A and B chain moieties of E8, or to a mixture of rG and E8-B chains. Cell adhesion to E8 and B-rAiG, in contrast to rAiG, was inhibited with antibodies specific for alpha 6 and beta 1 integrin chains. Since intercalation (a) restored a conformationally dependent alpha 6 beta 1 integrin recognition site present in native protein, (b) inactivated a cryptic cell binding activity in the A chain, and (c) inhibited a heparin binding site present in proximal G domain, we conclude that biological activities of laminin are different from that of its isolated subunits.

Recombinant laminin G domain mediates myoblast adhesion and heparin binding.

A recombinant mouse cDNA fragment encoding the G domain of the basement membrane laminin A chain was inserted into the eukaryotic baculovirus expression vector pVL1392 modified to produce fusion proteins carrying the rat fibronectin signal. G domain, expressed and secreted as a soluble glycoprotein (rG), was purified to near homogeneity without denaturing conditions. By electron microscopy rG possessed the same globular morphology as found in laminin. rG was cleaved with elastase into two fragments, rG70 and rG50. The latter fragment possessed the identical N terminus as laminin fragment E3 and both shared the same secondary structure by circular dichroism. rG, and rG containing a deletion (residues 2980-3028) rich in basic residues bound to heparin with similar avidity. rG also promoted mouse C2 myoblast cell adhesion and spreading, and evaluation of myoblasts on rG70 and rG50 further revealed that cell spreading was an activity confined to the more proximal sequence of rG70. Antibody specific for rG70 completely blocked cell adhesion to intact laminin in contrast to antibody specific for E3. Finally rG did not inhibit laminin polymerization. These data support the role of G domain in cell and heparin binding, but not laminin self-assembly, and the approach provides a means to further characterize these functions.

A novel laminin B1 chain variant in avian eye.

Two distinct recombinant cDNA clones having homology to mouse laminin B1 have been isolated from an adult chicken eye library by cross-species nucleic acid hybridization. DNA sequence analysis identified one cDNA as the chicken homologue of the prototypic EHS laminin B1 chain. The second recombinant cDNA encodes a portion of a laminin B1-like protein, which is neither the chicken homologue of laminin B1 nor s-laminin, and thus represents a new laminin B1 variant.

Spontaneous changes in nucleotide sequence in proviruses of spleen necrosis virus, an avian retrovirus.

We determined the nucleotide sequence of about 1 kilobase of DNA 3' to the 5' long terminal repeat of three noninfectious ad one infectious proviral DNA clones of spleen necrosis virus, an avian retrovirus, to determine if the types of nucleic acid changes involved in retrovirus mutation shed light on special features of retrovirus replication. An open reading frame was found starting 411 base pairs from the end of the long terminal repeat. It contained sequences coding for the 36 amino acids at the amino terminus of the p30 of a related reticuloendotheliosis virus [Oroszlan, S., Barbacid, M., Copeland T., Aaronson, S. A. & Gilden, R. V. (1981) J. Virol. 39, 845-854]. Therefore, the open reading frame represents the 5' end of the gag gene. A mutation in one noninfectious provirus changed the initiation codon for the gag polypeptide; a mutation in another noninfectious provirus caused premature termination of gag polypeptide synthesis; and a nontandem duplication into gag resulting from a mistake in initial (+) strand DNA synthesis changed amino acids and the reading frame in a third noninfectious provirus. These mutations appear to be responsible for the lack of infectivity of these provirus clones and indicate a higher relative frequency of mutation in this region of the genome. In addition, all four clones have multiple other mutations. These mutations are mostly base pair substitutions and many are clustered for any one clone, reflecting certain special features of retrovirus replication.

Characterization of reticuloendotheliosis virus strain T DNA and isolation of a novel variant of reticuloendotheliosis virus strain T by molecular cloning.

Reticuloendotheliosis virus strain T (REV-T) is a highly oncogenic avian retrovirus which causes a rapid neoplastic disease of the lymphoreticular system. Upon infection, this virus gives rise to two species of unintegrated linear viral DNA, which are 8.3 and 5.5 kilobase pairs long and represent the helper virus (REV-A) and the oncogenic component (REV-T), respectively. Restriction endonuclease cleavage maps of these two DNA components indicate that REV-T DNA has a large portion of the genome deleted with respect to REV-A DNA and a substitution about 0.8 to 1.5 kilobase pairs long that is unrelated to REV-A DNA. These additional sequences comprise the putative transforming region of REV-T (rel). A chicken spleen cell line transformed by REV-T produced virus which upon infection gives rise to three species of unintegrated linear viral DNA (8.3, 5.5, and 3,3 kilobase pairs). We isolated the proviruses of the 8.3- and 3.3-kilobase pair species from this cell line by cloning in the phage vector Charon 4A. Restriction enzyme mapping showed that the two proviral clones are proviruses of REV-A and a variant of REV-T, respectively. A subclone of the variant REV-T provirus specific for the rel sequences of REV-T was used as a hybridization probe to demonstrate that the rel sequences are different from the putative transforming sequences of Schmidt-Ruppin Rous sarcoma virus strain A, avain myelocytomatosis virus, avian myeloblastosis virus, avian erythroblastosis virus, Abelson murine leukemia virus, and Friend erythroleukemia virus. In addition, the rel-specific hybridization probe was used to identify a specific set of sequences which are present in uninfected avian DNAs digested with several restriction enzymes. The corresponding cell sequences are not arranged like rel in REV-T.

Mapping of alterations in noninfectious proviruses of spleen necrosis virus.

Ten recombinant lambda phage containing proviruses of spleen necrosis virus (SNV) were previously obtained. Six of the proviruses are infectious and four are not infectious in infectious DNA assays. In this paper, we show that these noninfectious proviruses are not infectious because of alterations in the viral DNA. We constructed recombinants between infectious and noninfectious proviruses and tested these recombinants in an infectious DNA assay. In addition, we carried out cotransfection of a noninfectious provirus with a restriction endonuclease-generated fragment of viral DNA. The alterations in the viral DNA resulting in lack of infectivity were mapped to regions of viral DNA of 1 to 2 kilobase pairs. These results and other biochemical data indicate that alterations in retrovirus proviruses occur at a high frequency.

Infectious and noninfectious recombinant clones of the provirus of SNV differ in cellular DNA and are apparently the same in viral DNA.

Ten clones of Charon 4A containing proviruses of spleen necrosis virus, an avian retrovirus, and flanking chicken DNA sequences were isolated and characterized. Some clones gave rise to progeny with viral DNA sequences deleted or duplicated, probably as a result of crossing-over in the 600 bp terminal redundancy in viral DNA. The cellular sequences are different in each clone, indicating that all the proviruses are integrated in different sites in cellular DNA. Six clones are infectious and four are not. All the infectious molecules containing a provirus are of a similar size and are smaller than the noninfectious molecules containing a provirus. The viral DNA is not apparently different in eight clones, but two clones, one infectious and one noninfectious, lack two restriction sites each. Large changes in proviral DNA therefore do not seem responsible for the lack of infectivity of some clones. These results are consistent with the hypothesis that neighboring cellular DNA sequences control proviral expression (infectivity).

DNA of noninfectious and infectious integrated spleen necrosis virus (SNV) is colinear with unintegrated SNV DNA and not grossly abnormal.

The cleavage sites of eight restriction endonucleases in linear spleen necrosis virus (SNV) DNA were mapped, and the map was oriented with respect to viral RNA. With the aid of this map, several structural features of the viral DNA were elucidated: unintegrated linear SNV DNA is terminally redundant; the majority of SNV DNA molecules integrated in chicken DNA, which were previously shown to be present in many sites in cellular DNA, are colinear with unintegrated viral DNA; no tandem integration of proviral molecules is detectable; and the majority of integrated SNV DNA molecules, including integrated SNV DNA molecules previously shown to be noninfectious, do not have an altered restriction enzyme digestion pattern.