Preimplantation diagnosis of the beta1 integrin knockout mutation as a model for aneuploid gene testing.

The autosomal beta1 integrin knockout mouse mutation was selected as a model to experimentally determine preimplantation diagnosis test reliability for autosomal gene deletions and duplications. In experiment 1, which analyzed 198 individually disaggregated single blastomeres, the observed test frequencies matched the mathematically predicted frequencies calculated from the independently derived values of 90% normal allele amplification, 92% mutant allele amplification, 4% alternate allele contamination, and 4% failure to transfer amplifiable target DNA into the PCR reaction mix. This experiment correctly predicted a normal embryonic phenotype in 143 (99.3%) of the 144 phenotypically normal autosomal recessive results. Experiment 2 compared single biopsied blastomere test results to test results on the remaining embryonic cells cultured 1 week until trophoblast outgrowth. Single biopsied blastomere analysis correctly predicted a normal autosomal recessive phenotype in 87 (98%) of the 89 embryos that would have been selected for implantation. Experiment 3 compared the PCR results of two biopsied blastomeres tested independently to the PCR result from the remaining cultured blastomeres to improve test reliability. Given that embryos would have been implanted only when two normal results were obtained, 17 of 17 phenotypically normal embryos would have been implanted from among the 44 embryos tested. These experiment 3 results are consistent with the mathematical prediction that about 99.9% of embryos implanted with two unaffected biopsied blastomere results would have had a phenotypically normal genotype.

The human integrin alpha 8 beta 1 functions as a receptor for tenascin, fibronectin, and vitronectin.

The integrin family of adhesion receptors consists of at least 21 heterodimeric transmembrane proteins that differ in their tissue distribution and ligand specificity. The recently identified alpha 8 integrin subunit associates with beta 1 and is predominantly expressed in smooth muscle and other contractile cells in adult tissues, and in mesenchymal and neural cells during development. We now show that alpha 8 beta 1 specifically localizes to focal contacts in cells plated on the extracellular matrix proteins fibronectin or vitronectin. In addition we show that human embryonic kidney cells (293), transfected with alpha 8 cDNA, express alpha 8 beta 1 on their surface and use this receptor for adhesion to fibronectin and vitronectin. Furthermore, alpha 8 beta 1 binds to both fibronectin- and vitronectin-Sepharose and can be specifically eluted from either matrix protein by the arginine-glycine-aspartic acid (RGD)-containing peptide, GRGDSP. Because fibronectin and vitronectin adhesion appeared to be mediated by RGD, we examined additional RGD-containing proteins, including tenascin, fibrinogen, thrombospondin, osteopontin, and denatured collagen type I. We found that only tenascin was able to mediate adhesion of alpha 8-transfected 293 cells. By using recombinant fragments of tenascin in adhesion assays, we were able to localize the alpha 8 beta 1 binding domain of tenascin to the RGD-containing third fibronectin type III repeat. These data strongly suggest that tenascin, fibronectin, and vitronectin are ligands for alpha 8 beta 1 and that this integrin binds to the RGD site in each of these ligands through mechanisms that are distinct and separate from alpha 5- and alpha v-containing integrins.

Deletion of beta 1 integrins in mice results in inner cell mass failure and peri-implantation lethality.

Integrin receptors for extracellular matrix receptors are important effectors of cell adhesion, differentiation, and migration in cultured cells and are believed to be critical effectors of these processes during development. To determine when beta 1 integrins become critical during embryonic development, we generated mutant mice with a targeted disruption of the beta 1 integrin subunit gene. Heterozygous mutant mice were normal. Homozygous loss of beta 1 integrin expression was lethal during early postimplantation development. Homozygous embryos lacking beta 1 integrins formed normal-looking blastocysts and initiated implantation at E4.5. However, the E4.5 beta 1-null embryos in situ had collapsed blastocoeles, and whereas the trophoblast penetrated the uterine epithelium, extensive invasion of the decidua was not observed. Laminin-positive endoderm cells were detected in the inner cell mass area, but endoderm morphogenesis and migration were defective. By E5.5 beta 1-null embryos had degenerated extensively. In vitro analysis showed that trophoblast function in beta 1-null peri-implantation embryos was largely normal, including expression of tissue-specific markers, and outgrowth on fibronectin- and vitronectin-coated, although not on laminin-coated substrates. In contrast, the inner cell mass region of beta 1-null blastocyst outgrowths, and inner cell masses isolated from beta 1-null blastocysts, showed highly retarded growth and defective extraembryonic endoderm morphogenesis and migration. These data suggest that beta 1 integrins are required for normal morphogenesis of the inner cell mass and are essential mediators of growth and survival of cells of the inner cell mass. Failure of continued trophoblast development in beta 1-null embryos after inner cell mass failure could be attributable to either an intrinsic requirement for beta 1 integrins for later stages of trophoblast development, or to the lack of trophic signals from the beta 1-null inner cell mass.

Expression of the beta 6 integrin subunit in development, neoplasia and tissue repair suggests a role in epithelial remodeling.

The alpha v beta 6 integrin was identified in cultured epithelial cells and functions as a fibronectin receptor. We have now used monoclonal antibodies to determine in vivo expression patterns of the beta 6 subunit in normal and pathological human or primate tissues, and during experimental wound healing or induced lung injury. The results indicate that beta 6 expression is restricted to epithelia and is up-regulated in parallel with morphogenetic events, tumorigenesis, and epithelial repair. During development of the kidney, lung, and skin, we found that beta 6 is expressed by specific types of epithelial cells, whereas it is mostly undetectable in normal adult kidney, lung and skin. In contrast, we detected high-level expression in several types of carcinoma. For example, beta 6 is almost invariably neo-expressed in squamous cell carcinomas derived from the oral mucosa, often focally localized at the infiltrating edges of tumor islands. Expression of beta 6 is also upregulated in migrating keratinocytes at the wound edge during experimental epidermal wound healing. Similarly, beta 6 expression is induced in type II alveolar epithelial cells during lung injury caused by injection of live bacteria. We also observed beta 6 expression in adult lungs and kidneys at focal sites of subclinical inflammation, as well as in a variety of clinical specimens from patients with chronic or acute inflammation of the lungs or kidneys. From these findings and earlier results, we hypothesize that alpha v beta 6 affects cell spreading, migration and growth during reorganization of epithelia in development, tissue repair, and neoplasia.

A novel phosphoglucomutase-related protein is concentrated in adherens junctions of muscle and nonmuscle cells.

Using five monoclonal antibodies raised against a human uterine smooth muscle extract, we have identified a novel antigen which runs as a closely spaced doublet in SDS-gels. The proteins (60/63 kDa) co-purify, are present in a 1:1 ratio as judged by Coomassie Blue staining, and are immunologically closely related, if not identical. No N-terminal sequence could be obtained from a mixture of the 60/63 kDa proteins, but the sequence of four polypeptides liberated by V8 protease or cyanogen bromide cleavage showed that the proteins are closely related to the glycolytic enzyme phosphoglucomutase type 1. Affinity-purified polyclonal antibodies and three different monoclonal antibodies to the 60/63 kDa proteins cross-reacted with rabbit skeletal muscle phosphoglucomutase type 1, whilst two additional monoclonal antibodies were specific for the 60/63 kDa proteins. Peptide maps of the 60/63 kDa proteins and phosphoglucomutase 1 are markedly different, and the purified proteins have no detectable phosphoglucomutase activity. Staining of cultured smooth muscle cells and fibroblasts with antibodies to 60/63 kDa proteins showed that the antigen is concentrated in focal contacts at the ends of actin bundles and is also associated with actin filaments. About 60% of the cellular 60/63 kDa proteins were found in the detergent-insoluble fraction, suggesting a physical association with the cytoskeleton. The highest levels of protein immunoreactivity were found in muscles. The antigen is concentrated in muscle adherens junctions, including smooth muscle dense plaques, cardiomyocyte intercalated disks, and striated muscle myotendinous junctions. Among epithelial cells, the 63 kDa isoform of the protein was found only in cultured keratinocytes where immunofluorescent staining was localized in cell-to-cell adherens junctions. Expression of the 60/63 kDa proteins in vascular smooth muscle cells is developmentally regulated and correlates with the differentiated contractile phenotype of these cells.