Rac1 Controls Both the Secretory Function of the Mammary Gland and Its Remodeling for Successive Gestations

Abstract

An important feature of the mammary gland is its ability to undergo repeated morphological changes during each reproductive cycle with profound tissue expansion in pregnancy and regression in involution. However, the mechanisms that determine the tissue's cyclic regenerative capacity remain elusive. We have now discovered that Cre-Lox ablation of Rac1 in mammary epithelia causes gross enlargement of the epithelial tree and defective alveolar regeneration in a second pregnancy. Architectural defects arise because loss of Rac1 disrupts clearance in involution following the first lactation. We show that Rac1 is crucial for mammary alveolar epithelia to switch from secretion to a phagocytic mode and rapidly remove dying neighbors. Moreover, Rac1 restricts the extrusion of dying cells into the lumen, thus promoting their eradication by live phagocytic neighbors while within the epithelium. Without Rac1, residual milk and cell corpses flood the ductal network, causing gross dilation, chronic inflammation, and defective future regeneration.

Keywords: Rac1; ducts; inflammation; involution; lactation; mammary; phagocytosis; regeneration.

Graphical abstract

Figure 1

Loss of Rac1 Leads to Defective Alveolar and Ductal Development in Second Gestation (A) Percentage of litter deaths at day 2 of first and second lactations. (B) Genomic DNA was isolated from WT (Rac1^fx/fx:LSLYFP) and Rac1^−/− (Rac1^fx/fx:LSLYFP:WAPiCre) second-pregnancy mammary glands. PCR shows the presence of the Cre gene in luminal epithelia of transgenic glands and Cre-mediated recombination of the Rac1^fx/fx gene. The remaining full-length floxed allele detected in transgenics represents intact Rac1 in stromal and myoepithelial cells. The 333 bp product represents the full-length floxed allele and the 175 bp product represents the recombined Rac1^−/−allele. n = 3 animals are shown per group. (C) Second-pregnancy day 18 (P18) WT and Rac1^−/− glands, immunostained for YFP reporter gene expression. The presence of YFP in Rac1^−/− glands showed that Cre-mediated recombination occurred in the luminal cells of ducts and alveoli. Bar, 45 μm. (D) Carmine staining of whole-mounted mammary gland of WT and Rac1^−/− mice at pregnancy day 18 of the second gestation. Rac1 loss leads to ductal dilation and severe retardation of alveoli units. Bar, 2.8 mm (insert, 0.6 mm). (E) H&E staining of mammary gland at P18, second gestation. Bar, 80 μm. See also Figure S1.

Figure 2

Second Lactation Cycle Is Severely Defective without Rac1 (A–I and L) Second gestation, P18 glands were used. (A) H&E staining of mammary gland shows the presence of lipid droplets in WT glands (arrow). Note reduced alveolar development and an absence of lipid droplets in Rac1^−/− glands. Bar, 20 μm. (B) Oil red O staining of tissue sections, with dotted lines denoting alveolar edges. In comparison with WT, Rac1^−/− glands do not contain significant quantities of milk fat in alveoli. Bar, 15 μm. (C) Immunofluorescence for lipid envelope protein adipophilin (red) reveals large milk lipid droplets in WT glands but these are significantly reduced in Rac1^−/− glands. Wheat germ agglutinin (WGA-488; green) was used to detect the luminal surface. Bar, 15 μm. (D) Immunofluorescence staining of β-casein shows reduced milk protein in Rac1^−/− glands compared with WT. Bar, 15 μm. (E) qRT-PCR shows defective Csn2 (β-casein) and Csn1s2a (γ-casein) gene expression in Rac1^−/− glands. Error bars ± SEM of n = 4 mice (WT) and n = 5 mice (Rac1^−/−). ^∗∗p < 0.001. (F) qRT-PCR shows reduced Elf5 gene expression in Rac1^−/− glands. Error bars ± SEM of n = 3 mice. ^∗p < 0.05. (G) Immunoblot showing expression and (Y694) phosphorylation of Stat5a. E-cadherin was used to show equal loading. WT, n = 4 mice; Rac1^−/−, n = 5 mice. (H) Quantitation of the Stat5a immunoblot (G) after normalization to loading control (E-cadherin). Error bars ± SEM of n = 4 mice (WT) and n = 5 mice (Rac1^−/−). ^∗p < 0.05. (I) Quantitation of the p-Stat5(Y694) immunoblot (G) after normalization to normalized total Stat5a. Error bars ± SEM of n = 4 mice (WT) and n = 5 mice (Rac1^−/−). ^∗p < 0.05. (J) Immunofluorescence staining of Stat5a at lactation day 2 reveals reduced nuclear translocation in Rac1^−/− alveoli. β-catenin was used to mark cell edges. Bar, 15 μm (insert, 7 μm). (K) Quantitative analysis of Stat5a nuclear translocation. Nine areas/mouse were analyzed. Error bars ± SEM of n = 3 mice per group. ^∗∗p < 0.001. (L) qRT-PCR shows defective prolactin receptor (Prlr) gene expression in Rac1^−/− glands. Error bars ± SEM of n = 3 mice. ^∗p < 0.05. See also Figure S2, Tables S1 and S2.

Figure 3

Loss of Rac1 from Differentiated Epithelium Results in Defective Lactation (A) Carmine staining of whole-mounted mammary gland of WT and Rac1^−/− mice at lactation day 2 of the first gestation. Note the smaller alveoli in Rac1^−/− glands. Bar, 2.8 mm (insert, 0.33 mm). (B) H&E staining of above mice. (C) Immunostaining of β-casein shows reduced milk protein in P18 Rac1^−/− glands compared with WT. Bar, 30 μm. (D) Reduced milk lipid droplets revealed by adipophilin staining in P18 Rac1^−/− glands. Wheat germ agglutinin (WGA488) was used to demark apical lumens. Bar, 50 μm. (E) Immunoblot showing reduced β-casein in P18 Rac1^−/− glands. (F) β-casein immunoblot was quantified using the Odyssey imaging system (LICOR Biosciences). Black diamonds, WT glands; red squares, Rac1^−/− glands. (G) qRT-PCR shows defective Csn2 and Csn1s2a gene expression in P18 Rac1^−/− glands. Error bars ± SEM of n = 3 mice, ^∗p < 0.05. (H) Immunofluorescence staining of β-casein and YFP in WT and Rac1-KO primary cultures. These are confocal images through the center of 3D acini cultured on a basement membrane matrix. Bar, 10 μm. (I) Immunoblotting shows reduced β-casein levels in Rac1-KO primary cultures. (J) qRT-PCR shows defective Csn2 gene synthesis in response to lactogenic hormones in Rac1-KO primary cultures. Error bars ± SEM of n = 3 samples, ^∗p < 0.05. See also Figure S4.

Figure 4

Baobab Ducts Arise in Post-lactational Involution, and Glandular Lumens Are Full of Dead Cells (A) Carmine staining of whole-mounted mammary glands at 4 weeks post-lactational involution. This shows baobab ducts in Rac1^−/− tissue. Bar, 2.8 mm (insert, 0.7 mm). (B) H&E staining of WT and Rac1^−/− mammary glands at involution day 2, day 4, and 4 weeks. Note Rac1^−/− baobab ducts are present at involution day 2 and persist throughout involution. Bar, 200 μm. (C) Immunofluorescence staining of YFP shows that Rac1 ablation in ducts coincides with the baobab phenotype. Bar, 45 μm. (D) H&E stain shows accumulation of dead cells in Rac1^−/− alveolar and ductal lumens. (E) Cleaved caspase-3 staining showing dead cell accumulation in lumens of Rac1^−/− glands. Arrows point to dead cells in alveoli (top panel) and ducts (bottom panel). Bar, 45 μm (top), 200 μm (bottom). (F) Cleaved caspase-3 staining in WT and Rac1-KO primary cultures induced to die for 5 hr in suspension. (G) Immunoblot of (F) showing no difference in caspase-3 activation in WT and Rac1-KO suspension cultures. See also Figure S5.

Figure 5

Reduced Cell-ECM Adhesion in Rac1^−/− Epithelia Leads to Increased Cell Shedding (A) H&E staining of WT and Rac1^−/− alveoli at involution day 4. In WT glands, dying cells are contained in vacuoles within viable neighbors. In Rac1^−/− glands, apoptotic cells are extruded out of the epithelium. Bar, 35 μm. (B) Incidence of dead cell extrusion from epithelium. The number of cell corpses shed into lumens were counted in 50 alveoli per gland (involution day 2). Error bars ± SEM of n = 3 mice. (C) E-cadherin staining shows cell-cell junctions are not disrupted following deletion of Rac1 in monolayer culture. YFP was used to detect Cre recombination. (D) Paxillin staining shows marked reduction in focal adhesions following Rac1 deletion in culture. (E) Vinculin and actin staining show reduced focal adhesion and distinct actin rearrangement, following Rac1 deletion. (F) Rac1-KO cells fail to attach when re-plated onto a collagen or laminin-111 matrix. Bar, 10 μm (C–E). See also Figure S6.

Figure 6

Rac1 Is Required for MEC Phagocytosis in Involution (A) Electron micrographs of involution day 2 glands. Left, arrow points to a dead cell inside a live MEC in WT, also note the presence of milk fat globules. Right, arrow points to a necrotic cell in the lumen of Rac1^−/− alveolus. Note, there is no engulfment in transgenics. Bar, 1 μm (left), 2 μm (right). (B) WT apoptotic MECs labeled with cell tracker red were added to WT and Rac1-KO MEC cultures, labeled with cell tracker green. Note, there is cell corpse engulfment by WT but not by Rac1-KO MECs. Bar, 3 μm. Histogram shows the percentage of cells containing engulfed apoptotic bodies. (C) Electron micrographs of engulfment by MECs in culture as in (B). Left, arrow points to an internalized apoptotic body in live WT MEC. Right, arrow points to a necrotic cell, with no evidence of engulfment in Rac1-deficient MECs. Bar, 0.5 μm (left), 1 μm (right). (D–F) Electron micrographs of WT and Rac1^−/− involuting glands. Left panels: (D) Arrow points to macropinocytic engulfment at the apical membrane of WT MECs. Bar, 1 μm (left), 0.5 μm (right). (E) Arrow shows milk within WT MECs. (F) Arrow shows WT MEC forming a phagocytic cup around a milk fat globule. Bar, 2 μm. Right panels: Loss of Rac1 prevents all phagocytic and macropinocytic activity. (G) Immunofluorescence staining of β-casein in WT and Rac1^−/− involution day 2 tissue sections. Arrow points to cell-associated milk protein in WT glands. In Rac1^−/− glands, β-casein was only detected in the lumen of alveoli and not within cells. β1-integrin and WGA-488 were used to demark cell surfaces. Bar, 7 μm. (H) Electron micrograph showing extensive macropinocytosis in WT MEC cultures but not in Rac1 depleted cells. Bar, 1 μm. (I) H&E staining of involution day 4 glands. Note, the alveoli have collapsed in WT glands but Rac1^−/− alveoli remain engorged with milk and dead cells. Bar, 20 μm. (J) Immunofluorescence staining of β-casein shows loss of milk from the lumen in WT glands but not Rac1^−/− at involution day 4 tissues. Bar, 7 μm. See also Figure S7.

Figure 7

Loss of Rac1 Induces Chronic Inflammation (A) Involution day 2 glands show early recruitment of F4:80 positive macrophages in Rac1^−/− tissue. These were only detected in WT glands at day 4. Bar, 30 μm. (B) Macrophages in lumen of Rac1^−/− alveoli. Bar, 10 μm. (C and D) qRT-PCR shows elevated chemokines (C) CCL2 and (D) CCL7 in Rac1^−/− glands at involution day 2 (gray), day 4 (red), and 4 weeks (blue). n = 3 mice were used per group. Error bars ± SEM of triplicates taken from the same RNA sample. (E) H&E staining of involution week 4 glands showing persistent inflammation without Rac1. Boxed areas show large foamy macrophages in duct lumens. Bar, 20 μm. (F) Foamy macrophages in Rac1^−/− duct lumens stain positive for F4:80. Bar, 20 μm. (G) Annexin V-647⁺ cells quantified by flow sorting show no differences in phosphatidylserine exposure in WT or Rac1-deficient dying MECs. R6 = Annexin V-647⁺ fraction; WT = 83.87, Rac1-KO = 82.95. Inset shows representative micrographs of positively marked cells. See also Table S3.