Ablation of beta1 integrin in mammary epithelium reveals a key role for integrin in glandular morphogenesis and differentiation - PubMed (original) (raw)

. 2005 Nov 21;171(4):717-28.

doi: 10.1083/jcb.200503144.

Na Li, Julia Cheung, Emma T Lowe, Elise Lambert, Rebecca Marlow, Pengbo Wang, Franziska Schatzmann, Timothy Wintermantel, Günther Schüetz, Alan R Clarke, Ulrich Mueller, Nancy E Hynes, Charles H Streuli

Affiliations

Ablation of beta1 integrin in mammary epithelium reveals a key role for integrin in glandular morphogenesis and differentiation

Matthew J Naylor et al. J Cell Biol. 2005.

Abstract

Integrin-mediated adhesion regulates the development and function of a range of tissues; however, little is known about its role in glandular epithelium. To assess the contribution of beta1 integrin, we conditionally deleted its gene in luminal epithelia during different stages of mouse mammary gland development and in cultured primary mammary epithelia. Loss of beta1 integrin in vivo resulted in impaired alveologenesis and lactation. Cultured beta1 integrin-null cells displayed abnormal focal adhesion function and signal transduction and could not form or maintain polarized acini. In vivo, epithelial cells became detached from the extracellular matrix but remained associated with each other and did not undergo overt apoptosis. beta1 integrin-null mammary epithelial cells did not differentiate in response to prolactin stimulation because of defective Stat5 activation. In mice where beta1 integrin was deleted after the initiation of differentiation, fewer defects in alveolar morphology occurred, yet major deficiencies were also observed in milk protein and milk fat production and Stat5 activation, indicating a permissive role for beta1 integrins in prolactin signaling. This study demonstrates that beta1 integrin is critical for the alveolar morphogenesis of a glandular epithelium and for maintenance of its differentiated function. Moreover, it provides genetic evidence for the cooperation between integrin and cytokine signaling pathways.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

Temporal deletion of β1 integrin during mammary gland development. (A) Schematic diagram of β1 integrin allele, Blg-Cre transgene, WapiCre transgene, and Rosa26R alleles. (B–E) X-gal staining of whole mammary glands from Blg-CreTg/+;Rosa26Rfx/fx (B and D) and WapiCreTg/+;Rosa26Rfx/fx mice (C and E). (B) 12-wk nulliparous. (C) 18-wk nulliparous. (D and E) Lactation day 2. Blg-Cre directs Cre expression to mammary epithelium from at least 12 wk of age throughout pregnancy and lactation, whereas WapiCre-induced Cre expression is restricted to differentiated epithelium. (F) PCR analysis of Cre-mediated recombination in tissue from mammary glands of Itgβ1fx/fx;Blg-Cre+/+ (wt) and Itgβ1fx/fx;Blg-CreTg/+ (mt) mice at day 2 of lactation. The 2.1-kb product represents the floxed allele, 1.3-kb the recombined allele, and 0.64-kb the Blg-Cre transgene. Bar, 2 mm.

Figure 2.

Figure 2.

Defects in mammary gland development. Analysis of mammary gland development in Itgβ1fx/fx;Blg-CreTg/+ mice. (A and B) Lactation defects in mutant mice. (A) Some mutant females can lactate enough to support pup survival, but pups are reduced in size (as demonstrated by asterisk; lactation day 13) compared with pups nursed by wild-type females (lactation day 12). (B) Growth-rate analysis of pups from wild-type and mutant females that were able to support some of their litter. Error bars show ± SEM. (C–E) Oil Red O staining of milk fat (C and D) and western analysis of milk protein (E) in mammary glands from wild-type and mutant mice at day 18 of pregnancy. (F–K) Whole mount analysis of wild-type (F and I) and mutant mice with failed lactation (G and J) or reduced lactation capacity (H and K) demonstrates failed lobuloalveolar development in β1 integrin–null mammary glands. (L–Q) Histological analysis of mammary glands from wild-type (L and O) and mutant mice with failed lactation (M and P) or reduced lactation capacity (N and Q) demonstrated reduced alveolar density. Arrows in P and Q demonstrate disruption of alveolar integrity with epithelial cells protruding into the luminal space of alveoli of mutant mice. Bars: (C and D) 25 μm; (F–H) 2 mm; (I–K) 0.5 mm; (L–N) 200 μm; (O–Q) 50 μm.

Figure 3.

Figure 3.

Defects in alveolar structure. Immunofluorescent analysis of alveolar structure from Itgβ1fx/fx;Blg-Cre+/+ (A, C, and E; wild type) and Itgβ1fx/fx;Blg-CreTg/+ (B, D, F, and G; mutant) mice at day 2 of lactation. B and G are from the mouse shown in Fig. 2 P, whereas D and F are from the mouse shown in Fig. 2 Q. (A and B) β1 integrin (red) is lost from luminal cells of mutant mice. Arrows indicate lateral staining. Apical localization of ZO-1 (green) in wild-type and mutant alveoli. (C and D) Laminin (red) is correctly deposited in mutant alveoli. Arrow indicates cells protruding into lumen of mutant alveoli with cell–cell staining of β catenin (green). (E–G) Cell–cell and lateral localization of β catenin (green) in wild-type (E) and mutant alveoli, where cells have detached from the BM and protrude in the lumen (F) or are attached to BM (G). β1 integrin (red) is basal and lateral in alveoli of wild type and basal only in mutant mice. Bars: (A) 13 μm; (B–G) 19 μm.

Figure 4.

Figure 4.

β1 integrin is required for epithelial morphogenesis. Primary mammary epithelial cells from Itgβ1fx/fx mice were infected with either βgal- or Cre-expressing adenovirus; plated on collagen (A–F and M–P), vitronectin (Q–R), or BM matrix (G–L and S–V); and examined by phase-contrast microscopy (A–L) and immunofluorescent staining as shown in M–V. Time after infection was 24 h (A, B, G, and H), 24 h plus trypsinized and replated for a further 24 h (E, F, K, and L), or 48 h (all other panels. (A–F) Cells plated on collagen form a monolayer (A, C, and E), but the cells round up 48 h after integrin deletion (D) or do not spread when replated (F). Cultures were well washed before phase-contrast microscopy. (G–L) Cells plated on the BM matrix form three-dimensional polarized epithelial structures termed acini (G, I, and K), which are not maintained 48 h after integrin deletion (J). When cells already lacking integrin are plated on the BM matrix, they are unable to initiate acinar formation within the first 48 h of culture (L). (M–P) Deletion of β1 integrin results in impaired focal adhesion contacts and a cell-rounding phenotype on collagen. In βgal-infected cells, β1 integrin (inset, green) in focal adhesions colocalizes with ILK (M, red). In Cre-infected cells, smaller focal contacts are present (arrows) as demonstrated by ILK and talin staining (green), but the larger focal adhesions are lost (N). ZO-1 (green) is expressed in mutant cells (P). (Q–R) Focal adhesion complexes are maintained after deletion of β1 integrin when plated on vitronectin. Similar to βgal-infected cells (Q), Cre-infected cells (gray) are well spread and display αv integrin–containing (red) adhesions (R). (S–V) Deletion of β1 integrin results in disruption of epithelial polarity with a failure of acini formation and structure on the BM matrix (lateral cross sections through individual acini). In control acini, ILK (red) is basal (S), β catenin (red) is lateral (U), and ZO-1 (green) localizes to the apical cell surface, where hollow lumens form (S and U). This distribution is disrupted after integrin deletion (T and V). Bars: (A–L) 13 μm; (M and Q–R) 10 μm; (N–P) 6.5 μm; (S–V) 13 μm.

Figure 5.

Figure 5.

Disruption of focal adhesion protein signaling. Western analysis of protein lysates from Itgβ1fx/fx mammary epithelial cells infected with either βgal- or Cre-expressing adenovirus 48 h after infection. Cells cultured on either collagen in growth medium (A) or on BM matrix in differentiation medium with (+) or without (−) Prl (B). Expression of Cre resulted in reduction of β1 integrin levels and decreased amounts of phosphorylated FAK and paxillin regardless of substratum. (C) Small focal regions containing phosphopaxillin are detected in Cre-expressing Itgβ1fx/fx cells (white arrows and inset, showing only phosphopaxillin) cultured on either collagen type IV (Col IV) or I (Col I). Neighboring uninfected cells (suboptimal doses of AdCre virus were used for this immunofluorescence experiment) or control-infected cells are spread and contain substantial phosphopaxillin-containing adhesion complexes (orange arrows). When plated on vitronectin (VN), both Cre- and βgal-infected cells spread and formed focal adhesions containing phosphopaxillin, although the smaller, more centrally located signaling complexes (yellow arrows) are lost from Cre-infected cells. Bar, 10 μm.

Figure 6.

Figure 6.

Erk activation in response to EGF stimulation in the absence of β1 integrin. (A and B) Itgβ1fx/fx primary mammary epithelial cells infected with either βgal- or Cre-expressing adenovirus on collagen. (A) Cells cultured for 36 h after infection in growth medium, starved for 4 h, and stimulated with (+) or without (−) EGF 15 min before harvest. (B) Cells in growth medium for 48 h after infection. Ablation of β1 integrin does not affect the rapid (A) or sustained (B) activation of MAPK pathway as measured by Erk phosphorylation in response to EGF. (C) Primary wild-type epithelial cells were grown on collagen for 24 h and either left on collagen (attached) or maintained in suspension (detached) for times as indicated before 15 min of EGF stimulation. Erk phosphorylation was not affected by the total loss of integrin ligation in mammary epithelial cells.

Figure 7.

Figure 7.

Impaired epithelial function in Itgβ1fx/fx;WapiCreTg/ + mice. (A and B) Impaired lactogenesis was observed in Itgβ1fx/fx;WapiCreTg/+ mice. (A) Pup growth-rate analysis of wild-type and mutant mice from first pregnancy. Error bars indicate ± SEM. (B) Reduced size of pups from mutant females (asterisk) is maintained after second pregnancy. (C–F) Whole mount analysis shows no significant difference in lobuloalveolar development between wild-type and mutant mice (E and F are higher power views). (G–J) Histological analysis demonstrates a slight reduction in alveolar density from the mammary glands of mutant mice at lactation day 2. Alveolar morphology is essentially normal, although some epithelial cells detach and shed into the lumen of alveoli (J, arrow). (K–N) As with the Blg-Cre deletions, β1 integrin remains in the myoepithelial population (orange arrows) but is absent in luminal cells. This is particularly noticeable on lateral cell surfaces, where the integrin normally present (K and M, white arrows) has disappeared (L and N, white arrows). ZO-1 is located on the apical surface of luminal cells in both wild-type and mutant alveoli (M and N). (O–T) Oil Red O staining of milk fat (O and P) and β casein staining (Q–T) in mammary glands from wild-type and mutant mice at day 18 of pregnancy. Milk protein (casein) and lipid (Oil Red O) is dramatically diminished in β1 integrin–null mammary glands. Bars: (C and D) 2 mm; (E and F) 0.5 mm; (G and H) 200 μm; (I and J) 50 μm; (K–L) 47 μm; (M and N) 15 μm; (O, P, S, and T) 25 μm; (Q and R) 100 μm.

Figure 8.

Figure 8.

β1 integrin regulates Prl-mediated differentiation and Stat5 activation. (A) Itgβ1fx/fx primary epithelial cells plated on the BM matrix and infected with AdCre resulted in the down-regulation of β1 integrin and failed differentiation as demonstrated by impaired β casein synthesis after Prl stimulation for 24 h. Note that wild-type cells infected with AdCre virus still differentiated in response to Prl (not depicted). (B) Immunofluorescent staining of acini with antibodies to β1 integrin (green) and β casein (red) formed from Cre- and βgal-infected Itgβ1fx/fx cells demonstrated impaired differentiation. (C) Western analysis of whole mammary gland lysates from Itgβ1fx/fx;WapiCreTg/+ (M) mice show reduced phosphorylation levels of Stat5 compared with Itgβ1fx/fx;WapiCre+/+ (C) controls at days 3 and 14 of lactation in their first pregnancy. (D–F) Prl-induced Stat5 nuclear translocation is disrupted after loss of β1 integrin. (D) Stat5 (red) nuclear staining is prominent in wild type but is severely diminished in mutant alveoli from Blg-Cre and WapiCre mammary glands. β catenin (green) is localized on the lateral surfaces of luminal epithelial cells of wild-type, Blg-Cre, and WapiCre alveoli. (E) Triple immunofluorescence of Cre-infected Itgβ1fx/fx primary epithelial cells stained for Cre (gray), Stat5 (red), and Dapi (blue). The infected cells fail to translocate Stat5 to the nucleus; two examples are depicted within white box. In contrast, noninfected cells (which do not stain for Cre) do show Stat5 translocation after 15 min of Prl stimulation (arrows). (F) Quantification of the number of cells undergoing Stat5 activation after 15 min of Prl stimulation. Error bars indicate ± SEM. Bars: (B and E) 13 μm; (D) 40 μm.

References

    1. Aggeler, J., J. Ward, L.M. Blackie, M.H. Barcellos-Hoff, C.H. Streuli, and M.J. Bissell. 1991. Cytodifferentiation of mouse mammary epithelial cells cultured on a reconstituted basement membrane reveals striking similarities to development in vivo. J. Cell Sci. 99:407–417. - PubMed
    1. Akiyama, S.K., S.S. Yamada, K.M. Yamada, and S.E. LaFlamme. 1994. Transmembrane signal transduction by integrin cytoplasmic domains expressed in single-subunit chimeras. J. Biol. Chem. 269:15961–15964. - PubMed
    1. Aszodi, A., E.B. Hunziker, C. Brakebusch, and R. Fassler. 2003. Beta1 integrins regulate chondrocyte rotation, G1 progression, and cytokinesis. Genes Dev. 17:2465–2479. - PMC - PubMed
    1. Barcellos-Hoff, M.H., J. Aggeler, T.G. Ram, and M.J. Bissell. 1989. Functional differentiation and alveolar morphogenesis of primary mammary cultures on reconstituted basement membrane. Development. 105:223–235. - PMC - PubMed
    1. Bill, H.M., B. Knudsen, S.L. Moores, S.K. Muthuswamy, V.R. Rao, J.S. Brugge, and C.K. Miranti. 2004. Epidermal growth factor receptor-dependent regulation of integrin-mediated signaling and cell cycle entry in epithelial cells. Mol. Cell. Biol. 24:8586–8599. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources