The alpha(2) integrin subunit-deficient mouse: a multifaceted phenotype including defects of branching morphogenesis and hemostasis

Abstract

The alpha(2)beta(1) integrin is a collagen/laminin receptor expressed on platelets, endothelial cells, fibroblasts, and epithelial cells. To define the role of the alpha(2)beta(1) integrin in vivo, we created a genetically engineered mouse in which expression of the alpha(2)beta(1) integrin was completely eliminated. Mice deficient in the alpha(2)beta(1) integrin are viable, fertile, and develop normally with no excess lethality of homozygotes. Both alpha(2)beta(1)-integrin protein and alpha(2) mRNA were undetectable in the alpha(2)-null mice. Gross and histological evaluation of the heart, lungs, kidneys, gastrointestinal tract, pancreas, skin, and reproductive tracts revealed no abnormalities. However, quantitative analysis of mammary gland branching morphogenesis demonstrated that branching complexity is markedly diminished in the alpha(2)-deficient animals. Studies in the alpha(2)-deficient animals do not support the proposed roles for the alpha(2)beta(1) integrin on fibroblasts and keratinocytes in wound healing. When compared to platelets from wild-type littermates, platelets from alpha(2)-null mice failed to adhere to type I collagen under either static or shear-stress conditions. Although platelets from alpha(2)-deficient animals aggregated in response to collagen, they did so with prolonged lag time and lessened intensity. The alpha(2)beta(1) integrin-null mouse thus exhibits diverse, sometimes subtle, phenotypes consistent with the widespread pattern of alpha(2)beta(1) integrin expression.

Figure 1.

Targeting of the α₂ integrin gene. A: The targeting construct was prepared from an 8.0-kb EcoRI genomic clone containing exon 1 (black box). The targeting vector contained a 2.0-kb BamHI fragment that included 5′ UTR of the α₂ integrin gene upstream of the 5′ loxP (triangle) site. A 600-bp BamHI fragment containing exon 1 and flanking sequence was placed between the two downstream loxP sites. A 2.4-kb Aval/EcoRI fragment containing intronic sequence was placed between the 3′ loxP site and the PGK-TK cassette. The fragment used for Southern blot hybridization (probe a), primers (1 and 2) for PCR genotyping, restriction enzyme sites, the 5′ UTR, and intron 1 are indicated. After homologous recombination, transient transfection with a plasmid encoding Cre-recombinase resulted in removal of both the neo-cassette and exon 1. A single loxP site remained. Restriction enzyme sites: B, BamHI; A, Aval; R, EcoRI; X, Xho. B: Southern blot analysis of wild-type, heterozygous, and α₂-deficient animals. Tail DNA from mice obtained from the heterozygous intercrosses was digested with EcoRI and Xhol, electrophoresed, and hybridized with probe a. The wild-type allele generates a 6.5-kb fragment and the targeted allele generates a 4.5-kb fragment. C: PCR genotyping of tail DNA. The genotype was confirmed by PCR analysis of tail DNA using primer 1, directed against the 5′ UTR of the α₂ integrin gene and primer 2 against intron 1. The wild-type allele generated an 870-bp product and the mutant allele generated a 320-bp product.

Figure 2.

Expression of the α₂ integrin subunit protein and mRNA. A: Flow cytometric analysis to detect extracellular expression of the α₂β₁ integrin, CD31, and CD41/61 was performed on purified platelets from wild-type, heterozygous, and α₂-deficient mice using fluorescein isothiocyanate-conjugated monoclonal anti-CD49b, CD31, or CD41/61 antibodies or isotype control. B: Immunoblot analysis of purified protein lysates from platelets or mammary glands derived from wild-type, heterozygous, and α₂-deficient animals was performed with polyclonal rabbit antiserum directed against the carboxy-terminal 20 amino acids of the murine α₂ integrin cytoplasmic domain or against actin. C: RT-PCR was performed on total RNA isolated from the mammary gland of wild-type, heterozygous, and α₂-deficient mice. mRNA from wild-type and heterozygous animals generated an 840-bp product.

Figure 3.

Branching morphogenesis of the mammary gland. A: Histological sections of mammary glands of wild-type (+/+) and α₂-deficient (−/−) animals 8 to 12 weeks of age were evaluated (top). Whole-mount preparations of 8- to 10-week-old virgin wild-type (+/+) and α₂-deficient (−/−) mammary glands were stained with carmine solution. Terminal ducts were photographed at ×100 (middle) and ×1000 (bottom). B: The number of branch points per terminal duct was quantitated morphologically. The mean and SEM are shown.

Figure 4.

Platelet adhesion and aggregation. A: Platelets purified from wild-type (+/+), heterozygous (+/−), and α₂-deficient (−/−) animals were assayed for adhesion to type I collagen. The assays were performed for 1 hour in phosphate-buffered saline with 2 mmol/L of Mg²⁺ or ethylenediaminetetraacetic acid. B: Platelet adhesion to type I collagen at shear rates of 150 seconds⁻¹ was assayed in a parallel-plate flow chamber apparatus. Micrographs of the collagen substrate and the deposition of platelets from wild-type (+/+), heterozygous (+/−), and α₂-deficient (−/−) mice are depicted. C: The surface area covered by adherent wild-type, heterozygous, or α₂-deficient platelets was quantitated and expressed as a percentage of wild-type platelet adhesion. D: Platelet aggregation of heterozygous (+/−) or α₂-deficient (−/−) platelets was initiated by either the addition of 2.5 or 5 μg/ml of soluble equine tendon collagen or 0.5 U/ml of thrombin.

Figure 5.

α₂β₁ integrin and wound healing. A: Photomicrographs of full-thickness skin biopsies from wild-type or α₂-null mice revealed no differences in skin development. The dermis of both animals contained comparable numbers of fibroblasts and organized collagen bundles. B: The size of a 100-mm full-thickness wound was digitally quantitated from day 0 to day 18, when all wounds were healed. Both wild-type and α₂-null animals exhibit similar rates of wound healing.