Developmental Biology 363 (2012) 1–14

Contents lists available at SciVerse ScienceDirect

Developmental Biology

j ourna l homepage: www.e lsev ie r .com/deve lopmenta lb io logy

From the Society for Developmental Biology

Smooth muscle α actin is specifically required for the maintenance of lactation

Nate Weymouth, Zengdun Shi, Don C. Rockey ⁎University of Texas Southwestern Medical Center, Dallas, TX, USA

⁎ Corresponding author at: University of Texas SouthHarry Hines Blvd., Dallas, TX 75390-8887, USA.

E-mail address: don.rockey@utsouthwestern.edu (D

0012-1606/$ – see front matter © 2011 Elsevier Inc. Alldoi:10.1016/j.ydbio.2011.11.002

a b s t r a c t

a r t i c l e i n f o

Article history:Received for publication 16 May 2011Revised 3 November 2011Accepted 4 November 2011Available online 12 November 2011

Keywords:Smooth muscle α-actinMammary glandMyoepithelial cellLactationMouse

Smooth muscle α-actin (Acta2) is one of six highly conserved mammalian actin isoforms that appear to ex-hibit functional redundancy. Nonetheless, we have postulated a specific functional role for the smooth mus-cle specific isoform. Here, we show that Acta2 deficient mice have a remarkable mammary phenotype suchthat dams lacking Acta2 are unable to nurse their offspring effectively. The phenotype was rescued in crossfostering experiments with wild type mice, excluding a developmental defect in Acta2 null pups. The mech-anism for the underlying phenotype is due to myoepithelial dysfunction postpartum resulting in precociousinvolution. Further, we demonstrate a specific defect in myoepithelial cell contractility in Acta2 null mamma-ry glands, despite normal expression of cytoplasmic actins. We conclude that Acta2 specifically mediatesmyoepithelial cell contraction during lactation and that this actin isoform therefore exhibits functionalspecificity.

© 2011 Elsevier Inc. All rights reserved.

Introduction

Actin is a highly abundant structural protein of the cytoskeletonfound in eukaryotic cells, which serves critical roles in cell motilityand muscle contraction. At least six actin isoforms have been identi-fied in eukaryotic cells (Vandekerckhove and Weber, 1978), whichare products of six different genes (Herman, 1993; Miwa et al.,1991; Rubenstein, 1990). Although the isoactins are expressed in spa-tial, temporal, and tissue specific patterns, they exhibit remarkableamino acid identity. Two “nonmuscle” actins, cytoplasmic β and γ-actins, are thought to be present in all cells. The other four actin iso-forms, which include skeletal, cardiac, vascular, and enteric actins,are typically found in specific adult muscle types, although they arealso present in specialized cell types. The muscle specific actins differfrom the nonmuscle cytoplasmic actins at less than 10% of the total375 amino acid residues that make up the primary sequence, whilethe muscle specific isoforms differ from each by only a few residues(McHugh et al., 1991; Miwa et al., 1991).

The remarkable conservation of actin isoforms suggests that actinisoforms are functionally similar. Several cell-based experiments sup-port this conclusion (Gunning et al., 1984; Rubenstein and Spudich,1977; von Arx et al., 1995). For example, when the cytoplasmic β-actin isoform was over expressed in C2 myoblasts, it caused an in-crease in cellular surface area while over expression of γ-isoformcaused cells to roundup (Lloyd et al., 1992). Myogenic cells havebeen shown to regulate the expression of specific actin isoforms in amanner that alters the expression of cytoplasmic actins to muscle

western Medical Center, 5323

.C. Rockey).

rights reserved.

isoforms. Quiescent aortic smooth muscle cells express a significantamount of Acta2. However, when these cultured cells begin to prolif-erate and migrate, they increase the expression of cytoplasmic actins,which suggests a level of functional specificity (Barja et al., 1986).Considerable debate exists about the conclusiveness of such cellculture-based studies and raises questions about the function ofactin isoforms (Bulinski, 2006).

Phenotypic analysis of actin isoform deficient mice provides someevidence supporting the concept that actin isoforms mediate specificfunctions. Mice deficient in skeletal muscle α-actin (Acta1) died earlyin neonatal development due to defects in skeletal muscle growthand/or function (Crawford et al., 2002). Deletion of cardiac muscleα-actin (Actc1) led to defects in myocardial contractility, sarcomericorganization, and an apparent compensatory increase in Acta1mRNA as well as Acta2 protein levels in the hearts of null mice(Kumar et al., 1997). A hypomorphic allele in cytoplasmic β-actin(Actb) proved to be embryonic lethal prior to the beginning of muscledevelopment (Shawlot et al., 1998). A conditional allele for cytoplas-mic γ-actin (Actg1) in which this actin was deleted only from muscletissue was characterized by a progressive myopathy (Sonnemann etal., 2006) without an apparent change in isoactin expression. Thus,all actin isoform mouse knockouts, except smooth muscle γ-actin(Actg2), have been reported and have yielded data suggesting some ele-ment of isoform specificity, as well as overlapping cellular functions(Perrin and Ervasti, 2010).

Here, we have hypothesized Acta2, while exhibiting a high degreeof identity to the other actin isoforms, exhibits functional specificity.We further postulated that such functional specificity was likely tobe observed in a tissue specific fashion and/or under stress condi-tions. Given Acta2's implied role in myofibroblast contraction(Rockey et al., 1993) and the predominance of this isoform in

2 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

myoepithelial cells (Skalli et al., 1986), we propose that Acta2 servesan essential function in myoepithelial contraction and the mainte-nance of lactation. We discovered that Acta2 deficient (−/−) damsfailed to nurse their pups effectively and that the mechanism under-lying this phenotype is linked to hypocontractility of myoepithelialcells due to a lack of Acta2.

Materials and methods

Antibodies

Primary antibodies used were as follows: anti-Acta2 Cy3 (1A4;Part# C6198; Sigma, St Louis, MO), anti-keratin 14 (Part# PRB-155P; Covance, Berkeley, CA), anti-troma1 or anti-keratin 8 (Part#TROMA-1; University of Iowa Hybridoma Bank), anti-vimentin(Part# AB 5733; Chemicon, Temecula, CA), anti-GFAP (6F2; Part#MO761; Dako, Carpinteria, CA), anti-SM22α (Polyclonal antibody #1387-4; a gift from Dr. M. Parmacek), anti-total actin (C4; Part#MAB1501; Chemicon), anti-β-actin ((Actb); AC-15; Part# A1978;Sigma), anti- smooth muscle α-actin ((Acta2); 1A4; Part# A5228;Sigma), anti-smooth muscle γ-actin ((Actg2); B4;(Lessard, 1988)),anti-cytoplasmic γ-actin ((Actg1); monoclonal- (Dugina et al.,2009); polyclonal- (Hanft et al., 2006)), anti-cardiac and skeletal α-actin ((Actc1 and Acta1); 5C5; Part# A2172; Sigma), anti-muscleactins (HUC 1-1; Part# 691391; MP Biomedicals, Solon, OH), anti-myosin sarcomere (MF20; Part# MF20; University of IowaHybridoma Bank), anti-myosin cytoplasmic non-muscle (CMII 23;Part# CMII 23; University of Iowa Hybridoma Bank), anti-myosinsmooth muscle (Part# BT-562; Biomedical Technologies, Stoughton,MA) and anti-GAPDH (1D4; Part# MMS-580S; Covance, Princeton,NJ). Secondary antibodies used were as follows: Donkey anti-mouseCy2, Donkey anti rat Cy5, Donkey anti mouse Cy2 (Molecular Probes,Carlsbad, CA), and Donkey anti mouse IgG HRP (Jackson Immunore-search; West Grove, PA). Milk antibodies: anti-rat β casein antibody(a gift from Dr. C. Kaetzel) and anti-mouse specific milk proteins(anti α-casein, β-casein and WAP; Part# YNRMTM; Accurate Chemi-cal, Westbury, NY). Nuclear stains included BOPRO (MolecularProbes) and DAPI (Vector Labs, Burlingame, CA).

Mice husbandry, maternal care, and genotyping

The targeted disruption of the Acta2 gene by homologous recom-bination has been previously described (Schildmeyer et al., 2000).Using a PCR screen, the genotype of litters from heterozygous matingsrevealed normal Mendelian ratios, as previously reported(Schildmeyer et al., 2000).

In rodents, maternal behavior involves a complex set of activities,which include nest building and repair, sniffing of pups, pup retrieval,licking, grooming as well as blanket nursing. Indeed, several geneshave been shown to be essential for maternal behavior. We carefullymonitored Acta2−/− dams for a full range of maternal behavior aspreviously described (Yeo and Keverne, 1986) and found that thesemice exhibited entirely normal mother–pup interactions.

Acta2−/− mice were maintained on an out bred (BalbC/129/SvJ)genetic background. Acta2 mice were backcrossed to a BalbC strain.All animal care and experiments were carried out in accordancewith NIH guidelines and approved by the local Animal Care and UseCommittee. All mice were housed in cages with filter top cages andsubjected to 12L:12D cycles. For the purposes of breeding and timepregnancies, 10–14 week old females were placed with stud males,checked daily for copulation plugs, and weighted during gestationin order to determine the timing of pregnancy. If a copulation plugwas observed, the female was housed individually and given nestingmaterials. The presence of a copulation plug was designated as day0.5 of pregnancy. Acta2+/+ females (BalbC/129/SvJ and BalbC)birthed between 20 and 21 days after the observation of the

copulation plug. However, 10% of Acta2−/− pregnant females(BalbC/129/SvJ and BalbC) exhibited a failure or delay of parturition(by about 5 days). Thus, monitoring the presence of copulationplugs and recording weights of pregnant females were important todetermine the correct timing of mammary gland development.

Cross fostering and administration of oxytocin

Acta2+/+ and −/− dams that delivered a litter within 24 h ofeach other were used in cross-fostering experiments. Acta2+/+dams were mated with Acta2+/+ males and Acta2−/− dams weremated with Acta2−/− males. The litters born to Acta2−/− and+/+ dams were swapped as described (Chen and Capecchi, 1999).In some experiments, synthetic oxytocin (Sigma, St. Louis, MO) wasinjected into the appropriate mice twice daily 24 h postpartum untilweaning or an experimental endpoint. Administration of daily oxyto-cin rescued several milk letdown phenotypes such as that observed inmf3 null mice (Labosky et al., 1997) and oxytocin (oxt) null mice(Nishimori et al., 1996).

Serum total protein and albumin measurements

Blood chemistry measurements were done as described (Kwon etal., 2001) with several modifications. Blood from pups (n=10–12) ofthree litters for each indicated measurement were collected andplaced in heparinized coated tubes. Serum fraction was isolatedfrom the clotted blood by centrifugation at 4 °C and frozen at−80 °C. The levels of total protein and albumin were determined byUniversity of Texas Southwestern Mouse Metabolic PhenotypingCore using a Vitros 950 Chemistry Analyzer (Johnson and Johnson,Rochester, NY).

Histology and immunohistochemistry

Mammary glands were dissected and embedded in paraffin as de-scribed (Seagroves et al., 1998). Immunohistochemistry was per-formed on formalin fixed specimens, subjected to an antigenretrieval step as previously described (Katoh et al., 1997). Sectionswere incubated with primary antibody for 3 h at room temperature,washed in PBS for 20 min, and incubated with fluorescent labeledsecondary antibody diluted in PBS for 1 h. Sections were washedand mounted (Gel/Mount™ BiØmedia, Fisher Scientific, Ashville,NC). Myoepithelial cells were fixed in a 50:50 mixture of acetoneand methanol, incubated with antibodies, and immunofluorescentsignals were detected using a confocal microscope (LSM 510; CarlZeiss, Inc, Thornwood, NY). Images were imported and arranged inAdobe Illustrator CS2 (Adobe, San Jose, CA).

Analysis of alveolar diameters

Whole-mount preparation was done as previously described(Robinson et al., 1998). The diameter of the duct lumen was mea-sured using a calibrated eyepiece reticule as described (Van Nguyenand Pollard, 2002) in sections from lactation time points. Twentyblinded measurements were taken for each time point in 5 differentmammary glands.

TUNEL assay

Mammary glands were fixed and sectioned at specific time pointsand analyzed for apoptosis by terminal deoxynucleotidyl transferasemediated dUTP nick end labeling (TUNEL) as described (Tepera etal., 2003). The TUNEL terminal transferase was purchased from(Roche, Mannheim, Germany) and utilized per instructions. Foreach time point, 2000–4000 total cells were counted by identificationof DAPI stained nuclei. At least three mammary glands, including

3N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

8 fields for each gland, were examined for each time point. All count-ing was performed in a blinded fashion in random fields.

Myoepithelial cell isolation and contraction

Primary myoepithelial cell isolation was done as previously de-scribed (Nakano et al., 2001), with minor modifications. All myoe-pithelial cells (Acta2+/+ and −/−) used were isolated between 17and 20 dpc. In brief, Acta2−/− mammary gland tissue tended tofloat in the initial wash steps. Therefore, this buoyant material waswashed in suspension, collected and pelleted. Isolated cells weregrown in DMEM medium (Invitrogen) plus 10% fetal bovine serum(FBS) for 7–10 days before further experiments.

Myoepithelial cell contraction was assessed as described (Rockeyet al., 1993), with minor modifications. Briefly, after solidification ofcollagen lattices, myoepithelial cells were trypsinized and layeredon top of the lattice at a density of 0.25×106 cells per well. Cells pas-saged two or fewer times were used for contraction assays. After 24 h,lattices were washed and serum free mediumwas added. Contractionwas induced by an exchange of serum free DMEMmedium containing10% FBS serum. Lattices were immediately released from well wallsby gentle circumferential separation using a sterilized spatula; con-traction was measured electronically over time. Myoepithelial cellsused in rescue experiments were infected with the Acta2 adenovirus(Ad5 Acta2) after 3 days of culture.

Immunoblotting

Mammary glands were carefully dissected to exclude skin or mus-culature and were homogenized in a solution containing 1 ml of10 mM Tris pH 7.5, 0.9% NaCl, and 1.5% SDS. Equal quantities of cell ly-sates (Lowry et al., 1951) (25 μg) were separated using 4–12% Bis TrisNuPAGE (Invitrogen) and 10% SDS PAGE gels, transferred to nitrocel-lulose membranes, placed in blocking solution (5% donkey serum inTBS plus 0.1% Tween). Membranes were then exposed to primary an-tibodies followed by anti mouse secondary antibodies conjugated toHRP (Jackson ImmunoResearch, West Grove, PA). Specific signalswere detected with an enhanced chemiluminescent (ECL) substratekit as per the manufacturer's instructions (Pierce SuperSignal WestDura Extended Duration Substrate, Rockford, IL) and quantified witha Chemigenius2 photodocumentation system (Syngene, Fredrick,MD) and Gene Tools software (Syngene, Fredrick, MD). Loading con-trols included blots probed for GAPDH, α-tubulin, and duplicateCoomassie stained gels.

Acta2 adenovirus

A full-lengthmouse Acta2 cDNA (encoding 377 aa, GenBank acces-sion number: X13297) was cloned from mouse hepatic myofibro-blasts or myoepithelial cells using a PCR approach (sense-Nhe1: 5′ACTAGCTAGCATGTGTGAAGAGGAAGACAGC; antisense-Not1: 5′ATAGCGGCCGCTTAGAAGCATTTGCGGTGGAC). The resulting Acta2cDNA insert was ligated into the pShuttle2 vector (Clontech Adeno-X expression system, Mountain View, CA) and sequenced to verifyits authenticity. The Acta2 cDNA expression cassette was released byCeu1 and Sce1 and then inserted into pAdeno-X. Following the trans-fection of HEK293 cells, Acta2 expression adenovirus was generated,purified, and titered according to manufacturer's instructions (Clon-tech Adeno-X Expression System). The Acta2 adenovirus (Ad5Acta2) has a titer of 7.2×1010pfu/ml and the CMV empty adenovirus(Ad 5 CMV) has a titer of 2.7×1011 pfu/ml. Myoepithelial cells wereinfected with Ad5 CMV or Ad5 Acta2 at the indicated MOI (20–40)for 24 h. Infected cells were then assayed for contraction on collagenmatrices.

cDNA synthesis and real time PCR (qPCR)

Total RNA was extracted from Acta2+/+ and −/− myoepithelialcells using Trizol (Invitrogen). Total RNA (1 μg) was reverse tran-scribed using the High Capacity cDNA Kit (Applied Biosystems, FosterCity, CA). 50–100 ng of cDNA were used for qPCR and amplificationswere performed using the 7900 HT Fast Real Time PCR Machine(Applied Biosystems). Cycling conditions were step 1: 50 °C for2 min, step 2: 95 °C for 10 min, step 3: 95 °C for 15 s, step 4: 60 °Cfor 1 min, cycle back 45 times to step 3. All Taqman primer probesets for all six actin isoforms were purchased from Applied Biosys-tems (Assays on Demand Gene Expression, Applied Biosystems). Toensure specificity, assays were chosen based on amplicons with thehighest degree of sequence identity to specific actin isoformtranscripts.

Taqman primer probe sets used in these experiments were asfollows:

GAPDH Assay ID #MM99999915_g1Cytoplasmic γ-Actin (Actg1) Assay ID #MM01743531_g1Cytoplasmic β-Actin (Actb) Assay ID #MM00607939_s1Smooth Muscle α-Actin (Acta2) Assay ID #MM00725412_s1Smooth Muscle γ-Actin (Actg2) Assay ID #MM00656102_m1Cardiac α-Actin (Actc1) Assay ID #MM0477277_g1Skeletal α-Actin (Acta1) Assay ID #MM0570060_gH

All PCR reactions included 300 nM of each primer, 200 nM probe,and 1X TaqMan Universal PCR Master Mix (Applied Biosystems) con-taining the passive reference ROX. The ΔΔCt method (Livak andSchmittgen, 2001) was used to calculate relative expression levelsand normalized to an endogenous control (GAPDH). Melting curveanalysis confirmed that one product was amplified. All real-time Taq-Man PCR reactions were done in triplicate. In Fig. 8D, relative compar-isons of Acta2+/+ to−/− isoactin transcript levels were determinedwhereby Acta2+/+ levels were set to one.

Statistical analysis

Results are expressed as mean±s.e.m. Significance was estab-lished using the Student's t-test and analysis of variance when appro-priate. Differences were considered significant when pb0.05.

Results

Smooth muscle α-actin−/− dams exhibit a defective nursing phenotype

We discovered that pups nursed by Acta2−/− dams were under-sized and died soon after birth. Further, we noted a lack of milk in thestomachs of pups nursed by Acta2−/− dams postpartum (Fig. 1A,upper row) compared to those pups nursed by Acta2 wild type(+/+) dams (Fig. 1A, lower row), regardless of litter genotype. Inorder to determine the growth characteristics of Acta2−/− pups,we designed a series of matings. We measured postpartum pupweights daily and found Acta2−/− pups nursed by Acta2−/− damswere severely undersized and failed to gain weight compared toActa2+/+ pups nursed by Acta2+/+ dams (Fig. 1B). A majority(67%, n=36/54) of Acta2−/− pups died within 16 days after parturi-tion. Of the remaining pups, the average weight was 3.00 g (n=18)compared to Acta2+/+ pups, which had an average weight of 6.7 g(n=124). By 22 days postpartum, only 4 out of 54 Acta2−/− pupsnursed by Acta2−/− dams lived (average weight of 2.72 g). In con-trast, 122 out of 124 Acta2+/+ pups nursed by Acta2+/+ dams sur-vived (average weight of 8.33 g).

In cross fostering experiments in which four different mother pupgenotypes were examined, we reasoned that if stunted postnatalgrowth were due to a defect in the offspring, pup growth would

Avg

. litt

erm

ate

wei

ght (

g)

Days Postpartum

0

2

4

6

8

10

12

2 4 8 12 16 20 22 24

-/- Litter and +/+ Dam

+/+ Litter and -/- Dam

B

D E

Days Postpartum

Avg

. litt

erm

ate

wei

ght (

g)

0

1

2

3

4

5

6

7

8

9

10

2 4 6 8 10 16 18 20 22

+/+ Litter and +/+ Dam-/- Litter and -/- Dam-/- Litter and -/- Dam

(+ Oxytocin 600 mU/kg)

0

2

4

6

8

10

12

0

10

20

30

40

50

60

70

80

90

100

Avg

Pup

Wei

ght (

g)

% S

urvi

val

(n=124) (n=62) (n=79) (n=54)

Dam Genotype

Pup Genotype

+/ + +/+ -/- -/-+/++/+ -/- -/-

% Survival

Avg Pup Weight (g)

0

1

2

3

4

5

6

Tota

l Ser

um P

rote

in o

r A

lbum

in (

g/dl

)

+/+

-/-

Total Protein+/ + +/+-/ - -/-

Albumin(Dams)

(Pups)

F

***

******

***

**

**

***

***1412

A C

Fig. 1. Postnatal growth characteristics of Acta2+/+ and −/− pups. In (A), is shown a representative example of Acta2−/− pups nursed by Acta2−/− dams (upper row) and Acta2+/+pups nursed by Acta2+/+ dams at 4 days postpartum (lower row). Note the absence of milk in the stomach of Acta2−/− pups and the presence of milk in the Acta2+/+ pups nursed byActa2+/+ dams (arrows point to stomachs). In (B), is shown a graphic summary of postpartum growth of Acta2+/+ and −/− mice. In addition, Acta2−/− pups nursed by Acta2−/−damswere given oxytocin daily in an attempt to stimulate milk letdown (*pb0.05, 2–8 days; **pb0.01: (Acta2+/+pups (n=124) vs. Acta2−/− pups (n=54)) or Acta2−/− pups nursedby Acta2−/− dams given oxytocin (n=55), 9–22 days). In (C) and (D), Acta2+/+ and −/− pups were cross fostered to dams of the opposite genotype on day one postpartum. In (C), isshown a representative example; the three pups on the left are Acta2−/− pups cross fostered with an Acta2+/+ dam and the three pups on the right are Acta2+/+ pups cross fosteredwith an Acta2−/− dam (day 13 postpartum). In (D), the weights of pups from multiple cross fostering experiments are depicted graphically (*pb0.05, day 4; **pb0.01: Acta2−/− pups(n=62) vs. Acta2+/+ pups (n=72), 8–24 days). In (E), survival and postnatal growth at day 22 postpartum of all pups fostered is depicted graphically (***pb0.001: pups nursed by Acta2−/− vs. Acta2+/+dams). In (F), levels of circulating proteins and serum albumin in pups nursed by Acta2+/+ or−/− dams (n=10; ** pb0.01: pups nursed by Acta2−/− dams vs. Acta2+/+ dams).

4 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

remain abnormal regardless of dam genotype. However, if the pheno-type was related to the nursing ability of the dam, an Acta2+/+ damnursing Acta2−/− pups should rescue the phenotype. In these exper-iments, all pups were cross-fostered at one day postpartum and mon-itored daily. Fig. 1C shows a representative cross fosteringexperiment. We found that the growth curve of any pup nursed byan Acta2−/− dam was markedly abnormal (Fig. 1D). Acta2−/−pups rescued by cross fostering with an Acta2+/+ dam had an aver-age weight of 9.43 g at 22 days compared to 3.03 g for Acta2+/+pups nursed by an Acta2−/− dam (n=54 for Acta2−/− pups res-cued by Acta2+/+ dams vs. Acta2+/+ pups nursed by Acta2−/−dams, pb0.0001). Further, 87% (54/62) of all Acta2−/− pups crossfostered to an Acta2+/+ dam lived more than 22 days postpartumcompared to only 12% (10/79) of Acta2+/+ pups cross fostered toan Acta2−/− dam (n=79, pb0.0001, Fig. 1E). These findings leadus to conclude that Acta2−/− dams are unable to nurse their pups.To confirm further that pups nursed by Acta2−/− dams are malnour-ished due to a nursing defect, we evaluated total circulating proteinand serum albumin levels in pups (n=10, 5 days postpartum) nursedby Acta2+/+ and −/− dams. There is a significant decrease in totalprotein (Acta2+/+ pup levels 5.22 g/dl vs. Acta2−/− pup levels2.82 g/dl) and serum albumin (Acta2+/+ pups levels 2.42 g/dl vs.Acta2−/− pup levels 1.33 g/dl). These indicators of abnormal bloodchemistry suggest that pups nursed by Acta2−/− dams are severelymalnourished (n=8, pb0.0001, Fig. 1F).

We next queried whether oxytocin mediated maternal abnormal-ities could be involved in the observed phenotype. The rationale forthis is that oxytocin is known to play a role in nursing (Takayanagiet al., 2005) and maternal nurturing behaviors (i.e., pup retrieval,nest building, pup licking, and grooming (Insel and Young, 2001;

Winslow and Insel, 2002)), and defects in oxytocin release or signal-ing often resulting in a phenotype similar to that observed in Fig. 1A(Fahrbach et al., 1985; Insel and Harbaugh, 1989; Pedersen et al.,1985). Oxytocin (600 mU/kg) injected twice daily into Acta2−/−dams from day one postpartum until 22 days postpartum failed toalter the postnatal growth of pups (Fig. 1B, closed circles). Further,oxytocin had no effect on the normal maternal behavior we observedin Acta2−/− dams. Finally, Acta2−/− pups appeared to develop nor-mally when nursed by Acta2+/+ and +/− dams (Fig. 1D). Except fora mild delay or failure in parturition occurring in 10% of Acta2 nulldams, there were no other gross defects besides the described nurs-ing defect in Acta2 null dams.

Pups nursed by smooth muscle α-actin−/− dams die due tomalnutrition

Given the absence of milk and the rescue of postnatal growth inActa2−/− pups by cross fostering with an Acta2+/+ dam, we pre-dicted pups nursed by an Acta2 null dam were unable to thrive dueto malnutrition. Several potential lactation defects might account forthe results in Fig. 1 (Palmer et al., 2006). Thus, we sought to detect his-tological signs of starvation in pups nursed by Acta2 dams. Midthoraxsections of nursed pups 4 days postpartum from each of the differentmother pup combinations revealed smaller midthorax diameters inpups nursed by Acta2−/− dams (Figs. 2A–D), reduced amounts of fatbetween the outer dermis and the musculature (Figs. 2E–H), as wellas the absence of lipid droplets in brown fat regions (Figs. 2I–L). Ofnote, the outer dermis appears to be thicker in Acta2−/− pups nursedby Acta2+/+ dams (Fig. 2F). As a control, an Acta2+/+ pup nursed byan Acta2+/+ dam, which did not nurse for 48 h, displayed a complete

B C D+/+ Dam and +/+ Pup +/+ Dam and -/- Pup -/- Dam and +/+ Pup -/- Dam and -/- Pup

A

I J

E F G H

K L

Dermis

Intradermal Adipose Layer

Dermis

AdiposeLayer

Brown Adipose Muscle

Fig. 2. Severe malnourishment in pups nursed by Acta2−/− dams. Each column contains three pictures from a specific cross fostering pair as indicated on top of each column. In(A–D), pups of the specified genotype and cross fostering dam were sacrificed at day 4 postpartum; midthorax sections are shown (scale bar=1 mm). In (E–H), are representativeviews of the outer dermis (scale bar=200 μm) and in (I–L) are representative images of brown adipose regions as indicated (scale bar=50 μm). Note the smaller midthorax di-ameter (C and D), the reduced layer of adipose fat between the outer dermis and the musculature (G and H) as well as the absence of lipid droplets (K and L) within regions ofbrown adipose tissue of pups nursed by Acta2−/− dams.

5N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

depletion of lipid droplets in brown fat regions and a depletion of gly-cogen stores in the skeletal muscle (N. Weymouth, unpublished). Thedata indicate that the cause for severe malnourishment of pups

+/+

-/-

Virgin 10.5 dpc

18.5 dpc +/+ 18.5 dpc -/-

A B

E F

I J

Fig. 3. Gestational mammary glands in Acta2−/− dams appear structurally normal. In (A–−/− (E–H) dams at different gestational time points (H&E, scale bar=100 μm). In (I–L), aretation (scale bar=1 mm). Abbreviation: dpc = days post-coitus.

nurtured by Acta2−/− dams is that they do not receive adequate nu-trition due to a milk letdown problem, not due to other possible lacta-tion defects (Palmer et al., 2006).

14.5 dpc

20.5 dpc +/+ 20.5 dpc -/-

20.5 dpcC D

G H

K L

H), are shown representative examples of mammary glands from Acta2+/+ (A–D) orshown whole mounts of representative mammary glands at days 18.5 and 20.5 of ges-

+/+ -/-

2.5L

6.5L

10.5L

16.5L

A

C

H

D

E

J

Days Postpartum

B

F

G

I

M

K

02468

1012141618

2.5 4.5 6.5Days Postpartum

2.5 4.5 6.5

Avg

. Dia

met

er (

um) +/+

-/-

L+/+ -/-

0

20

30

40

50

60

Avg

. # D

ucts

/Fie

ld

N

10

**** **

******

***

6 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

7N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

Mammary gland abnormalities in smooth muscle α-actin−/− miceduring gestation and lactation

We considered several possible mechanistic explanations for theinability of newborn pups to develop during lactation. For example,pups could be receiving milk that is poor in nutritional value orthey were not getting enough milk during lactation. Defects in lacta-tion could arise from impaired alveolar differentiation, deficiencies inmilk composition (Zoubiane et al., 2004), secretion (Stinnakre et al.,1994), or abnormalities in milk letdown (Palmer et al., 2006).Mouse milk is composed of 12% protein, 30% lipid, and 5% lactose.The major protein components of milk consist of β-casein, whey acid-ic protein, immunoglobulin A, and α-lactalbumin (Anderson et al.,2007). To address the possibility that Acta2−/− dams had a defectin milk composition or secretion, we analyzed the expression ofmilk proteins by immunoblotting and we found no abnormalities inmilk protein composition in Acta2−/− mammary glands.

We next considered whether defects in alveolar differentiation, ashas been previously described (Liu et al., 1997), could account for theobserved nursing phenotype. Thus, we explored the formation oflobuloalveolar structures capable of producing milk in Acta2−/−mice. Histological and whole mount analysis of Acta2+/+ and −/−mammary glands at various stages of pregnancy revealed that duringearly to mid pregnancy, mammary glands appeared normal (Figs. 3A–C, E–G). There were no differences in ductal elongation and formationof lobuloalevolar structures. However, during the late stages of gesta-tion (18–20 days post-coitus (dpc)), we observed alveolar distensionof mammary glands from Acta2−/−mice (Figs. 3D,H). This differencewas even more pronounced in whole mounts, where mammaryglands from Acta2−/− mice appeared to have a greater alveolar den-sity (Figs. 3 I–L). Mammary glands from Acta2−/− mice (Fig. 3H)were isolated 20 days after a copulation plug was noted. A potentialproblem with parturition in Acta2−/− dams might explain the alve-olar distension and milk stasis noted in (Fig. 3H). The data suggestthat the deficiency of Acta2 does not affect ductal morphogenesis oralveolar differentiation.

We noted that postpartummammary glands from Acta2−/−micewere fully engorged with milk, raising the possibility that milk stasisand precocious involution occur even in the presence of a normalsuckling stimulus. This led us to postulate that the observed pheno-type in Acta2−/− mice is due to an abnormality in milk letdown.We also considered whether the neuroendocrine milk reflex arc, me-diated by oxytocin, and essential for milk letdown could be abnormalin Acta2−/− mice (Palmer et al., 2006). Proper sucking and milk re-moval during lactation are needed for the maintenance of lactation(Wagner et al., 1997). Further, oxytocin-mediated contraction ofmyoepithelial cells (Gimpl and Fahrenholz, 2001) also appears to beimportant (e.g., ablation of oxytocin or the oxytocin receptor leadsto a milk letdown phenotype with premature involution in null mam-mary glands (Nishimori et al., 1996; Takayanagi et al., 2005)). Nota-bly, involution, occurring normally after lactation and triggered bypup weaning, is characterized by two distinct phases: an earlyphase of milk engorgement, followed by epithelial apoptosis, andthen a tissue remodeling phase characterized by an increased expres-sion of proteinases (Lund et al., 1996). Acta2−/− mammary glandsexhibited signs of milk stasis and involution shortly after parturition(Figs. 4 B, D, F, H, J, and L). Acta2+/+ mammary glands displayed a

Fig. 4. Mammary glands of Acta2−/− mice exhibit milk stasis, alveolar distension, and preindicated times during lactation (H&E, scale bar=50 μm). Indicated days precede either dsame mammary tissue shown in (E) and (F), respectively (H&E, scale bar=200 μm). InActa2+/+ dam (K) and Acta2−/− dam (L) (scale bar=1 mm). In (M), the alveolar diamdata depicted graphically at the indicated times after parturition (n=5 fields and 6 mice(N), ducts were counted as described in the Materials and methods and the average numActa2+/+ dams vs. Acta2−/− dams at all time points).

typical acinar like morphology of milk producing epithelial cells lin-ing the ducts, which in turn are surrounded by a layer of contractilemyoepithelial cells. These ducts are surrounded by an extensiveframework of connective tissue, known as the stroma, consisting ofextracellular matrix, adipose tissue, lymphatic tissues, and fibroblasts,which often lie in the intervening vasculature of the gland. Duringlactation in Acta2+/+ dams, there were no significant histologicalchanges until weaning (Figs. 4A, C, E, G, I, and K). However,Acta2−/− dams exhibited morphological changes during lactation,consistent with premature involution (Figs. 4B, D, F, H, and J). Ofnote, mammary ducts are generally surrounded by extensive connec-tive tissue. However, this connective tissue stroma was much less ap-parent surrounding the alveoli in the normal lactating gland as shownin Fig. 4, panels A, C, E, G, and I. The adipose stroma has reappeared inthe sections from Acta2−/− mice, characteristic of dams with lacta-tion failure where adipose lipid is not mobilized for milk synthesis.Thus, histological changes, shown in Fig. 4, panels D, F, and H, wereapparent. During early lactation (0–6 days postpartum), there weresigns that the first phase of involution had already occurred. Therewas extensive distension of the alveoli due to milk engorgement(Fig. 4M). Furthermore, the surrounding stroma had been dramatical-ly altered or even remodeled at these early stages. At both 10 and16 days of lactation, there were a large number of apoptotic cells inthe lumen of Acta2−/− glands (Figs. 4F and J) and a collapse of alve-olar structures (Fig. 4H). Throughout lactation, mammary glandsfrom Acta2−/− mice had a greater average alveolar diameter(Fig. 4M) and a decrease in the number of alveoli (Fig. 4N) than didthose from Acta2+/+ mice.

Mammary glands from smooth muscle α-actin−/− mice exhibitprecocious involution highlighted by myoepithelial cell apoptosis

Given that lactating mammary glands from Acta2−/− mice dis-played signs of precocious involution and apoptosis, we measuredthe number of apoptotic cells at various stages of lactation. InActa2+/+ mice, TUNEL positive cells were not detected in early ormid-stage lactation (Figs. 5A–D). In contrast to Acta2−/− mice,TUNEL analysis revealed significantly more apoptotic cells liningducts and cells sloughed into the lumen (Figs. 5E–H, Fig. S1). Quanti-tative changes in the number of apoptotic cells in Acta2−/− micewere quite remarkable during lactation (Fig. 5I). These data indicatethat mammary glands from smooth muscle α-actin−/− mice exhibitprecocious involution highlighted by apoptosis.

In several samples from Acta2−/− mice, we observed the pres-ence of apoptotic cells on the outside of the ducts, especially in lategestation and shortly after parturition (Figs. 5E–G, and K, Fig. S2, ar-rows in Figs. 5E and K). This finding led us to query whether theremay be abnormalities in myoepithelial cells, which are found sur-rounding epithelial cell acini, during lactation. To test this possibility,we examined myoepithelial and epithelial cells at time points duringgestation and postpartum. We used an anti-keratin 14 antibody tolabel myoepithelial cells and an anti-keratin 8 antibody to label epi-thelial cells. Additionally, in Acta2+/+ mice, we labeled myoepithe-lial cells with an anti-Acta2 antibody. In mammary glands fromActa2+/+ mice, branched hollow ducts of epithelial cells were sur-rounded by keratin 14 and smooth muscle α-actin positive myoe-pithelial cells (Figs. 6A–D; Note — Fig. 6 contains only merged

cocious involution during lactation. In (A–H), mammary glands were harvested at theays post-coitus (dpc) or lactation (L). (I and J) are higher magnification views of the(K and L), are images of representative mammary glands 12 h after parturition: aneter from multiple sections was quantified as in the Materials and methods and theper time point; **pb0.01: Acta2+/+ dams vs. Acta2−/− dams for all time points). Inber of ducts per field is depicted (n=5 fields and 6 mice per time point; ***pb0.001

+/+

-/-

18.5 dpc 4.5L 10.5L6.5L

A B C

E HF G

I

D

0

5

10

15

20

25

30

18.5 dpc 4.5L 6.5L 10.5L

TU

NE

L P

ositi

ve C

ells

(% to

tal)

+/+-/-

J K***

***

***

Fig. 5. Enhanced apoptosis in Acta2−/− mammary glands. In (A–H), Acta2+/+ and −/− mammary glands were harvested at the indicated time points. Tissues were subjected toTUNEL as in the Materials and methods. Merged images of DAPI stained nuclei and TUNEL positive cells are shown (DAPI/blue, TUNEL positive cells/pink, scale bar=50 μm). Arrowspoint to apoptotic cells which lie outside of ducts, surrounding the acini, during late gestation. Indicated days precede either days post-coitus (dpc) or lactation (L). In (I), the ratio ofTUNEL positive per DAPI positive cells, per field, are depicted (*pb0.05 Acta2+/+ dams vs. Acta2−/− dams at 18 dpc; **pb0.01 Acta2+/+ dams vs. Acta2−/− dams at 4 L;***pb0.001 Acta2+/+ dams vs. Acta2−/− dams at 6 L and 10 L; n=6 mice per group and 10 random fields per mouse) In (J and K), are shown higher magnification images ofA and E respectively (J — Acta2+/+, K — Acta2−/−, scale bar=100 μm).

8 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

images of keratin 8 (pink), Acta2 (red), and keratin14 (green); seeFigs. S2–S3 for complete immunofluorescence labeling data), whichhave been described as basket-like cells that surround the acini(Franke et al., 1980). Interestingly, mammary glands fromActa2−/− mice appeared to contain fewer myoepithelial cellsexpressing keratin 14 (Figs. 6E–I, Fig. S2–3) as well as other myoe-pithelial or smooth muscle markers (such as keratin 5, smooth mus-cle calponin, GFAP, vimentin, smooth myosin heavy chain, andP-cadherin; data not shown) compared to Acta2+/+ mice. The de-crease expression in myoepithelial specific markers, such as keratin14, in Acta2−/− mammary glands (Figs. 6E–I, Fig. S2–3, Fig. S5) andthe quantitative changes in the number of apoptotic cells inActa2−/− mammary glands (Fig. 5I) suggest a reduction of myoe-pithelial cells in Acta2−/− mammary glands at all stages of gestationand lactation as well as in the virgin state.

Isolated myoepithelial cells express smooth muscle and myoepithelialspecific markers and contract

On the basis of our findings and the fact that Acta2 may be impor-tant for cellular contraction in vivo, we determined whether Acta2 isrequired for myoepithelial cell contraction. First, we confirmed thepurity of isolated cells (Figs. 7A–P). We examined the following pro-teins in primary myoepithelial cells: SM22α (Figs. 7B and D), Acta2

(Figs. 7C, D, G, H, K, and L), keratin 14 (Figs. 7F, H, J, and L), keratin5 (Fig. S4), vimentin (Fig. S4), desmin (Fig. S4), and smooth musclecalponin (Fig. S4). Interestingly, SM22α, a calponin family actin bind-ing protein, and Acta2, were co-localized along stress fibers. Interme-diate filaments were identified in a delicate network of threadlikestructures surrounding the nucleus and extending into the cytoplasm(Fig. S4). Also, epithelial PtK2 control cells were immunolabeled todetect keratin 8, an epithelial intermediate filament (Fig. 7N and P,Fig. S4); this was identified in only 2% of cells isolated suggestingminimal contamination by epithelial cells. In order to assay the con-tractile properties of myoepithelial cells, we needed to confirm thatthe only change in contractile proteins concerns Acta2 and notchanges in myosin levels. Using Western blot analysis, non-muscleand smooth myosin heavy chains are expressed in myoepithelialcells and levels do not change in null cells. Skeletal muscle myosinwas not identified in myoepithelial cells. Collectively, these data indi-cate the isolation of a homogenous population of contractile myoe-pithelial cells (Fig. S4).

Finally, we tested the hypothesis that Acta2 is required for myoe-pithelial cell contraction by measuring contraction of isolatedActa2+/+ and −/− cells using a collagen lattice system. Initially,we assessed contraction using silicone rubber substrates (Harris etal., 1980); however, the preparation of uniform substrates was prob-lematic for quantitative measurements, and this led us to utilize

A B C D

E F H I

Virgin

+/+

-/-

12.5 dpc 18.5 dpc 4.5L

Fig. 6. Mammary myoepithelial cell numbers in Acta2−/− dams are reduced during gestation and lactation. Mammary gland tissue sections from Acta2+/+ and −/− dams wereimmunolabeled with anti-cytokeratin 14 (Alexa Fluor 488/green), anti-Acta2 (Cy3/red), and anti-cytokeratin 8 (Cy5/pink) antibodies as in the Methods. In (A–D), are shown rep-resentative merged images of stained Acta2+/+ mammary glands at the indicated time points (scale bar=20 μm). In (E–I), are shown representative merged images of stainedActa2−/− mammary glands at the indicated time points (scale bar=20 μm). Figures S2 and S3 contain a complete representation of individual staining: (cytokeratin 14(green), Acta2 (red), and cytokeratin 8 (pink)). Indicated times are symbolized: the days postpartum proceed the letter “L” which stands for lactation. Abbreviation: dpc = dayspost-coitus.

9N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

collagen matrices, which appear to be more physiological substratesthan polyacrylamide substrates and moreover allow the precisequantification of cellular contraction (Miron-Mendoza et al., 2008).We found that Acta2+/+ myoepithelial cells were significantlymore contractile than Acta2−/− cells were (Fig. 7Q) at all time pointsafter the initiation of contraction.

In order to explore at a molecular level the role of smooth muscleα-actin, we expressed Acta2 in isolated Acta2 null cells using an ade-noviral construct as described in the Materials and methods. Smoothmuscle α-actin was highly expressed when Acta2−/− cells wereinfected at a MOI of 10–80 (Fig. S4). Further, Acta2 was found inactin stress fibers at the same MOIs. Forced expression of Acta2 inActa2−/− cells rescued contraction at MOI 20 or 40 (Fig. 7Q, Fig. S4).

Expression of actin isoforms in smooth muscle α-actin−/− mammaryglands and myoepithelial cells

Given previous data suggesting redundancy in actin function, weconsidered the possibility that actin isoforms could be differentiallyregulated in response to the loss of Acta2. Therefore, we investigatedexpression profiles of different actin isoforms (Figs. 8A–C) inActa2+/+ and −/− mammary gland tissues at various stages of de-velopment. Acta2 expression was upregulated between 0 and10 days of gestation (Fig. 8A) (the increase in Acta2 expressionnoted in Fig. 8A probably represents an increasing proportion ofmyoepithelial cells in the whole gland). There were no noticeablegestational changes in total actin using the C4 monoclonal antibody,which recognizes all six actin isoforms (Fig. 8B). Quantitative analysisrevealed no statistically significant differences in total actin levelsafter multiple samples were analyzed (Fig. 8B lower graph) betweenActa2+/+ and −/− tissues. Cytoplasmic β-actin levels remainedunchanged and no compensating expression was noted (Fig. 8C). Col-lectively, these results demonstrate no compensatory response ofother actins due to the loss of Acta2 expression in mammary glandtissues.

We next determined which isoactins were expressed in myoe-pithelial cells and whether actin isoform expression changed inActa2−/− cells. Specifically, we assessed both transcript and proteinlevels for all six actin isoforms (Figs. 8D and E). As expected, cytoplas-mic actins (β/γ) mRNAs were present in myoepithelial cells, but in-terestingly, there were no quantitative changes noted in Acta2−/−cells. The levels of other muscle actins mRNAs were not significantlydifferent than background levels but presented for completeness.Acta2 mRNA was abundantly expressed in Acta2+/+ cells but absentin Acta2−/− cells (Fig. 8D). Actin isoform protein levels were mea-sured via immunoblotting using five different monoclonal antibodiesspecific for different actin isoforms as indicated on the right hand sideof both panels (Fig. 8E). First, we tested the specificity of each anti-body using purified actins from different tissues (Fig. 8E, left panel,1 μg/lane). Smooth muscle α and γ-actin were abundant in gizzardand not found in platelets (Fig. 8E, upper left panels). Cytoplasmic ac-tins (β/γ) were abundant in platelets and surprisingly quite low inheart (Fig. 8E, middle left panels). Using the 5C5 monoclonal anti-body, which recognizes both cardiac and skeletal α-actins, either orboth muscle actins were found in skeletal and cardiac muscles butnot expressed in platelets (Fig. 8E, lower left panels). Next, wedetermined the actin isoform composition in myoepithelial cells andwhether there were changes in actin isoform levels due to the lossof Acta2 expression (Fig. 8E right panel; Fig. S5 K–Q fordensitometry quantification (n=5–8)). As expected, Acta2 wasabundantly expressed in Acta2+/+ cells but not in Acta2−/− cells(Fig. 8E right panel). Smooth muscle γ-actin was not present inActa2+/+ or −/− cells (Fig. 8E, right panel). Of note, smooth muscleγ-actin, was found in gizzard but was not present in platelets (Fig. 8Eupper left panel) using the same anti-smooth muscle γ-actin monoclo-nal antibody (B4) used for analysis of myoepithelial lysates (Fig. 8Eright panel). Further, cardiac and skeletal α-actins were not presentin Acta2+/+ or −/− cells (Fig. 8E, right panel). As expected, bothcytoplasmic actins (γ/β) were present in Acta2+/+ and −/− cells(Fig. 8E, right panel), but no quantitative changes in the cytoplasmicactins (β/γ) were found in Acta2 null cells (Fig. S5, O–P). Total actin

+/+

A B C D

E F G H

I J K L

M N

O P

Q

0

20

40

60

80

100

0 60 120 180 240

Latti

ce C

ontr

actio

n (

% in

itial

latti

ce a

rea)

Time (min)

+/+

-/-

1020

+/+

-/-

***

*********

***

-/- Ad5 CMV

-/- Ad5 Acta2 MOI 20

-/-Ad5Acta2 MOI 40

Fig. 7. Phenotypic characterization and rescue of impaired myoepithelial contractility in Acta2−/− cells. Myoepithelial cells from Acta2+/+ and−/− dams were isolated as in theMaterials and methods. In (A–L), after two days in culture, myoepithelial cells were subjected to immunolabeling to detect the following: sm22α (B, Alexa Fluor 488/green), cyto-keratin 14 (F and J; Alexa Fluor 488/green), and smooth muscle α-actin (C, G, and K; Cy3/red). In (D, H, and L) are shown merged images of sm22α or cytokeratin 14 and smoothmuscle α-actin immunolabeling (scale bar=50 μm). Control PtK2 epithelial cells were immunolabeled with Acta2 (M, Cy3/red) or cytokeratin 8 (N, Cy5/pink). Control smoothmuscle A7r5 cells in (O) were dual immunolabeled with smooth muscle α-actin (Cy3/red) and cytokeratin 14 (Alexa Fluor 488/green). In (P), A7r5 cells were dual immunolabeledwith cytokeratin 8 (Cy5/pink) and cytokeratin 14 (Alexa Fluor 488/green) (scale bar=20 μm). Merged images for (O and P) are shown. In (A, E, I, D, H, L, and M-P), BOPRO1/bluewas used to label nuclei. In (Q), isolated myoepithelial cells were subjected to collagen lattice contraction assays as described in Materials and methods. In some experiments, iso-lated Acta2−/− myoepithelial cells were infected with an adenovirus expressing smooth muscle α-actin (Ad5 Acta2) at the indicted MOIs as in Materials and methods (2nd: n=4for each time point, ***pb0.001 for Ad5 Acta2 treated cells (MOI 20 or 40) vs. Acta2−/− cells; 1st: n=6 for each time point, ***pb0.001 for Acta2+/+ cells vs. Acta2−/− cells).

10 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

protein levels were decreased in Acta2−/− cells (Fig. S5 Q, Fig. 8E,right panel). In aggregate, these results led us to conclude thatmyoepithelial cells contain the following three actin isoforms:

cytoplasmic β-actin, cytoplasmic γ-actin, and smooth muscle α-actin.Further, there was no compensatory upregulation of any actin isoformin Acta2−/− cells.

+/+ -/- +/+ -/- +/+ -/- +/+ -/-Aor

ta Virg

in

6.5

dpc

10.5

dpc

18.5

dpc

30

25

20

15

10

5

0Virgin 6.5 dpc 10.5 dpc 18.5 dpc

SM

α-a

ctin

(rel

ativ

e ex

pres

ion)

0.00.20.40.60.81.01.21.41.61.82.0

+/+ +/+ +/+ -/-Aor

ta Virg

in

Virgin 6.5 dpc 10.5 dpc

+/+ -/-

6.5

dpc

18.5

dpc

β-actin -/-+/+-/-

6.5

dpc

-/-

10.5

dpc

Total actin

Coomassie

α-tubulin

+/+ -/-Gizz

ard

Platele

ts

Skelet

al

Heart

Platele

ts

Platele

ts

Heart

SM α-actin

SM γ-actin

Cytoplasmic γ-actin

Cytoplasmic β-actin

SM α-actin

SM γ-actin

Cytoplasmic γ-actin

Cytoplasmic β-actin

Cardiac α-actinSkeletal α-actin

2.5

2.0

1.5

1.0

0.5

0.06.5 dpc 18.5 dpc

+/+-/- +/+

-/-+/+-/-

A B C

D E

Total Actin42 kDa42 kDa

Cardiac α-actinSkeletal α-actin

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

42 kDa

55 kDa

SM α-actin 42 kDaTotal Actin

SM α-actin Cytoplasmicβ-actin

Cytoplasmic γ-actin

SM γ-actin Cardiac α-actin

Skeletalα-actin

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Act

in m

RN

A(r

elat

ive

expr

essi

on)

+/+-/-

Tot

al A

ctin

(rel

ativ

e ex

pres

ion)

b-ac

tin(r

elat

ive

expr

esio

n)

***

* * *

Fig. 8. Actin isoform composition in Acta2−/− mammary glands and isolated myoepithelial cells. In (A–C), Acta2+/+ and −/− mammary glands were harvested at the indicatedtimes and lysates were subjected to immunoblotting as in Materials and methods. In (A), is shown a representative immunoblot (n=6) of Acta2 expression at different dpc. In(B), is shown a representative immunoblot of total actin (n=6) levels at the indicated dpc. In (C), is shown a representative immunoblot of β-actin expression (upper panel,n=4) and total actin expression (lower panel, n=4). For (A–C), data were scanned, quantified, and depicted graphically below each immunoblot as in Materials and methods(*pb0.05 for Acta2+/+ vs. Acta2−/− tissues). In (D), mRNA levels of all six actin isoforms were measured in myoepithelial cells by qPCR using isoform specific probes. Expressionlevels were normalized to GAPDH and calculated by the ΔΔCt method. Ratios of fold changes are presented graphically where Acta2+/+ levels were set arbitrarily at 1 (n=10,**pb0.001 for Acta2+/+ vs. Acta2−/− transcript). In (E), actin isoform protein levels were measured in myoepithelial cells by immunoblotting with actin isoform specific mono-clonal antibodies (n=5–10). In the left panel, purified actins (1 μg per lane) from the indicated tissues were used to test the specificity of each actin isoform antibody (n=5). In theright panel, lysates of Acta2+/+ and Acta2−/−myoepithelial cells were subjected to immunoblotting to detect specific actin isoforms (n=10). Blots were scanned, quantified, andpresented graphically in Fig. S5. α-tubulin and coomassie stained gel serve as loading controls. Abbreviation: dpc = days post-coitus.

11N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

Discussion

We have discovered a remarkable mammary gland phenotype inActa2 deficient mice. Acta2−/− mammary glands have normal levelsof cytoplasmic (β/γ) actins (Actb, Actg1, respectively) and there is nocompensatory change in total actin (Figs. 8A–C), yet milk ejectionduring nursing was defective (Fig. 1). This was presumably due toan inability of myoepithelial cells to contract (Fig. 7Q). BecauseActa2−/− myoepithelial cells are hypocontractile, mammary glandsexhibit postpartummilk stasis, alveolar dilation, and precocious invo-lution (Fig. 4), which ultimately leads to the starvation of nursingpups (Fig. 2). We provide a comprehensive analysis of Acta2's rolein mammary gland development, which indicates significant struc-tural changes during lactation, and precocious involution in Acta2null glands due to myoepithelial dysfunction (Figs. 3–6). Also, wehave shown that myoepithelial cells contain only three actin iso-forms: cytoplasmic γ-actin (Actg1), cytoplasmic β-actin (Actb), andActa2. Most importantly, smooth muscle γ-actin (Actg2) is not pre-sent in myoepithelial cells nor is there any compensatory upregula-tion of actin isoform levels in Acta2−/− cells (Figs. 8D–E). Acta2serves a specific function in myoepithelial contraction without com-pensation by a similar actin isoform, (i.e., such as Actg2). Thus, Acta2is required for the maintenance of lactation.

Our findings address a fundamental question in actin biology as itrelates to the issue of actin isoform specificity. Molecular geneticstudies of actin isoform mutants in model organisms and mouseactin knockouts have generated valuable data, which might help ex-plain the functional significance of amino acid differences among iso-actins (Hennessey et al., 1993; Kato-Minoura et al., 1997; Mahaffey etal., 1985; Perrin and Ervasti, 2010; Willis et al., 2006). On one hand, ithas been proposed that minor amino acid differences in actin iso-forms are functionally not important and that multiple actin genesmay actually only reflect differences in total actin requirements fora particular cell type (Nowak et al., 2009; Wagner et al., 2002). S. cer-evisiae has one actin isoform (ACT1) and it is essential for viability.Despite only 90% identity, human Actb can substitute for mutants ofACT1 without any loss of viability (Karlsson et al., 1991). Data fromseveral actin knockouts including Actb, Acta1 (skeletal α-actin),Actc1 (cardiac α-actin), and Acta2 null mice have revealed a paucityof phenotypes and compensatory expression of closely identical iso-actins which would suggest that actin isoforms are not tied to a spe-cific function (Crawford et al., 2002; Kumar et al., 1997; Perrin andErvasti, 2010; Schildmeyer et al., 2000; Shawlot et al., 1998). Recent-ly, it was shown that Actc1 expression is sufficient to replace skeletalmuscle functional deficiencies in Acta1 null mice (Nowak et al., 2009).These studies conclude by suggesting these actin isoforms are

12 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

functionally equivalent and thus redundant. On the other hand, it ispossible that small amino acid differences are functionally important.For example, the Actc1 phenotype could only be partially rescuedwith cardiac expression of Actg2 (Kumar et al., 1997). Likewise, over-expression of Actg1 is unable to rescue the postnatal lethality of Acta1deficient mice even when total actin was restored to normal levels(Jaeger et al., 2009). In a conditional allele for Actg1 in skeletal muscle,mice exhibited overt symptoms of skeletal myopathy: decrease mo-bility, limb weakness, and joint contractures. This phenotype suggeststhat Actg1 serves a specific function even in skeletal myocytes(Sonnemann et al., 2006). Previously, it has been shown that Acta2null mice have a vascular phenotype in which null mice have a de-creased in blood pressure and an increase in blood flow, yet compen-satory expression of isoactins was noted in Acta2 null aortas(Schildmeyer et al., 2000). Collectively, these studies support the no-tion actin isoactins can mediate specific cellular processes and pointout the inherent difficulties in proving that an isoactin is requiredfor a specific function.

We performed a strict test of our hypothesis that Acta2 specificallymediates myoepithelial contraction by exogenous expression of Acta2in Acta2 deficient myoepithelial cells, which rescued contraction(Fig. 7Q). We considered an alternative approach — transgenic ex-pression of Acta2 specifically in null mammary glands using a myoe-pithelial specific promoter or by the exogenous expression of Acta2in isolated null myoepithelial cells. We chose the latter approach be-cause it allows us to rule out any potential non-autonomous effects(e.g. vascular defects in Acta2 mice (Schildmeyer et al., 2000) or po-tential problems associated with oxytocin mediated milk secretion(Lollivier et al., 2006)) which might play a role in the observed phe-notype. This cell based approach would preclude us from extensivemammary gland transplantation experiments or the creation of an-other transgenic mouse to demonstrate that Acta2 is specifically re-quired for myoepithelial contraction and the maintenance oflactation. Clearly, we have utilized this approach to measure and res-cue compromised myoepithelial contraction in null cells (Fig. 7Q andS4).

In order to draw a distinction between whether a defect in myoe-pithelial total actin levels or Acta2 functional specificity is responsiblefor the observed phenotype, we specifically evaluated compensatoryexpression of actin isoforms in Acta2 null cells. It was previously dem-onstrated that Acta2−/− aortas had an increase in Acta1 expression,which led to the suggestion that Acta1 might substitute for Acta2'sfunctionality (Schildmeyer et al., 2000). Interestingly, transgenic ap-proaches have been used in an attempt to rescue phenotypes ob-served in mice deficient for Acta1 (Jaeger et al., 2009; Nowak et al.,2009) and Actc1 (Kumar et al., 1997) to address the question ofactin isoform specificity. The upregulation in expression of anotherisoactin in response to the loss of Acta2 in myoepithelial cells wouldsuggest that another isoactin contributes to myoepithelial contractil-ity. In Fig. 8, we show that there was no upregulation in expression ofanother isoactin in Acta2 null cells, suggesting that myoepithelial con-tractility is mediated by Acta2. Cytoplasmic β-actin (Actb)−/− micehave an embryonic lethal phenotype; however, Actb+/− dams areable to nurse their litters normally (Shawlot et al., 1998). Littersnursed by Actg1−/− dams exhibited normal postnatal growth(Belyantseva et al., 2009). These studies argue against the functionalcontribution of cytoplasmic (β/γ) actins in this phenotype. Thus,our data confirms that Acta2, without functional contribution by cyto-plasmic (β/γ) actins, is essential for myoepithelial contraction.

We speculate that there may be a requirement for a particularcomposition of actin in myoepithelial cells to mediate contractilityand the integrity of the actin cytoskeleton needed for cell survival. Itis possible that a reduction in total actin levels due to the deletionof Acta2 causes cell specific apoptosis given the composition of isoac-tins in a given cell type. Our data emphasize that Acta2 is highlyexpressed in myoepithelial cells; in this case, the ratio of Acta2 to

other actin isoforms may be critical and in Acta2's absence, the cellis likely unable to compensate. This is consistent with previous obser-vations. Pharmacological disruption of actin dynamics (CelesteMorley et al., 2003; Gourlay and Ayscough, 2005; Gourlay et al.,2004; Posey and Bierer, 1999; Zhang et al., 1997), ablation of Actc1in hearts (Abdelwahid et al., 2004), or depletion of cytoplasmic actins(β/γ) isoforms triggers apoptosis (Dugina et al., 2009; Harborth et al.,2001). It is possible that cell specific apoptosis contributes to an ob-served phenotype in mice carrying a null allele for an actin isoform.Given the increase in apoptosis found during lactation in Acta2 nullmammary glands (Fig. 5) and the possible loss of myoepithelial cells(Fig. 6), we have shown that apoptosis most likely contributes tothe observations of milk stasis and precocious involution (Fig. 4).

We have shown that there are significant structural and develop-mental changes in Acta2−/− mammary glands, which are indicativeof milk stasis and precocious involution (Figs. 4–6). Given that myoe-pithelial contractility is compromised, involution occurs shortly afterparturition within 2–4 days postpartum. Recently, a descriptive ac-count of Acta2's role in myoepithelial contractility reported no struc-tural changes in Acta2 null glands and no developmentalabnormalities due to Acta2's absence (Haaksma et al., 2011). Theseconclusions are not consistent with both previous observations (Liet al., 1997; Palmer et al., 2006;Wagner et al., 1997) and the data pre-sented here (Figs. 4–6). Here, we have provided a complete histolog-ical and immunofluorescence analysis of Acta2's role in postpartumdevelopment of the mammary gland at multiple time points duringlactation; as shown in Figs. 4–6, there are multiple signs that preco-cious involution has occurred shortly after parturition. Specifically,connective tissue stroma is significantly less prevalent surroundingthe alveoli in Acta2+/+ lactating glands (Figs. 4 A, C, E, G, and I),whereas the adipose stroma has reappeared in the sections of Acta2null glands (Figs. 4 A B, D, F, H, and J), characteristic of dams with lac-tation failure where the adipose lipid is not mobilized for milk syn-thesis, even at 2 days postpartum. Further, previous work hasshown that disruption of the neuroendocrine reflex arc by litter re-moval (Wilde et al., 1999), teat sealing (Li et al., 1997), ablation ofoxytocin (Li et al., 1997; Wagner et al., 1997) or the oxytocin receptor(Takayanagi et al., 2005), leads to build-up of milk in the mammarygland triggering involution within 24–48 h postpartum, fundamen-tally similar to the morphological phenotype we have reported.Thus, it appears that milk ejection and milk removal facilitated bymyoepithelial contraction is needed for normal postpartum develop-ment of the mammary gland (Bruckmaier and Wellnitz, 2008; Wildeet al., 1999).

Mechanistically, there appear to be two contributing factors forthe disruption of lactation in Acta2 null mammary glands. First, wehave shown that Acta2 appears to mediate myoepithelial contractionin an isoform specific manner and secondly, we have demonstrated aprofound decrease in the number of myoepithelial cells in lactatingActa2 null glands. These two features combine to cause a severemilk stasis phenotype that includes increased apoptosis and preco-cious involution in Acta2 null glands shortly after parturition, whichinhibits the ability of nursing pups to thrive. Thus, we conclude thatActa2, which differs little in molecular structure from other isoactins,has a specific function in the mammary myoepithelium and that sub-tle amino acid differences in actin isoforms are functionallyimportant.

Conflict of interest statement

The authors declare no competing financial interests.

Acknowledgments

We thank the following investigators for providing specific re-agents: Dr. Robert Schwartz: SMA null mouse, Dr. Ted Salamon:

13N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

PtK2 cells, Dr. Micheal Parmacek: SM22α antibody, Dr. WernerFranke: keratin antibodies, Dr. James Ervasti: cytoplasmic γ-actin an-tibody, Dr. Christine Chaponnier: cytoplasmic γ-actin antibody, andDr. Elaine Fuchs: keratin 5 antibody. We thank Ron Ordas, VenkateshVaradan, Tim Oliver, Romona Rodriguiz, and Richard Premont fortheir technical assistance. Also, we are thankful to Dr. James Ervastifor sharing unpublished observations on the Actg1−/− mouse.

This work was supported by the NIH (Grants R01 DK 50574 andR01 DK 60338 to DCR) and the Burroughs Welcome Fund (DCR: re-cipient of a BWF Translational Scientist Award).

Appendix A. Supplementary data

Supplementary data to this article can be found online at doi:10.1016/j.ydbio.2011.11.002.

References

Abdelwahid, E., Pelliniemi, L.J., Szucsik, J.C., Lessard, J.L., Jokinen, E., 2004. Cellular dis-organization and extensive apoptosis in the developing heart of mice that lack car-diac muscle alpha-actin: apparent cause of perinatal death. Pediatr. Res. 55,197–204.

Anderson, S.M., Rudolph, M.C., McManaman, J.L., Neville, M.C., 2007. Key stages inmammary gland development. Secretory activation in the mammary gland: it'snot just about milk protein synthesis! Breast Cancer Res. 9, 204.

Barja, F., Coughlin, C., Belin, D., Gabbiani, G., 1986. Actin isoform synthesis and mRNAlevels in quiescent and proliferating rat aortic smooth muscle cells in vivo and invitro. Lab. Invest. 55, 226–233.

Belyantseva, I.A., Perrin, B.J., Sonnemann, K.J., Zhu, M., Stepanyan, R., McGee, J., Frolen-kov, G.I., Walsh, E.J., Friderici, K.H., Friedman, T.B., Ervasti, J.M., 2009. Gamma-actinis required for cytoskeletal maintenance but not development. Proc. Natl. Acad. Sci.U. S. A. 106, 9703–9708.

Bruckmaier, R.M., Wellnitz, O., 2008. Induction of milk ejection and milk removal indifferent production systems. J. Anim. Sci. 86, 15–20.

Bulinski, J.C., 2006. Cell biology. Actin discrimination. Science 313, 180–181.Celeste Morley, S., Sun, G.P., Bierer, B.E., 2003. Inhibition of actin polymerization en-

hances commitment to and execution of apoptosis induced by withdrawal of tro-phic support. J. Cell. Biochem. 88, 1066–1076.

Chen, F., Capecchi, M.R., 1999. Paralogous mouse Hox genes, Hoxa9, Hoxb9, and Hoxd9,function together to control development of the mammary gland in response topregnancy. Proc. Natl. Acad. Sci. U. S. A. 96, 541–546.

Crawford, K., Flick, R., Close, L., Shelly, D., Paul, R., Bove, K., Kumar, A., Lessard, J., 2002.Mice lacking skeletal muscle actin show reduced muscle strength and growth def-icits and die during the neonatal period. Mol. Cell. Biol. 22, 5887–5896.

Dugina, V., Zwaenepoel, I., Gabbiani, G., Clement, S., Chaponnier, C., 2009. Beta andgamma-cytoplasmic actins display distinct distribution and functional diversity. J.Cell Sci. 122, 2980–2988.

Fahrbach, S.E., Morrell, J.I., Pfaff, D.W., 1985. Possible role for endogenous oxytocin inestrogen-facilitated maternal behavior in rats. Neuroendocrinology 40, 526–532.

Franke, W.W., Schmid, E., Freudenstein, C., Appelhans, B., Osborn, M., Weber, K.,Keenan, T.W., 1980. Intermediate-sized filaments of the prekeratin type in myoe-pithelial cells. J. Cell Biol. 84, 633–654.

Gimpl, G., Fahrenholz, F., 2001. The oxytocin receptor system: structure, function, andregulation. Physiol. Rev. 81, 629–683.

Gourlay, C.W., Ayscough, K.R., 2005. A role for actin in aging and apoptosis. Biochem.Soc. Trans. 33, 1260–1264.

Gourlay, C.W., Carpp, L.N., Timpson, P., Winder, S.J., Ayscough, K.R., 2004. A role for theactin cytoskeleton in cell death and aging in yeast. J. Cell Biol. 164, 803–809.

Gunning, P., Ponte, P., Kedes, L., Hickey, R.J., Skoultchi, A.I., 1984. Expression of humancardiac actin in mouse L cells: a sarcomeric actin associates with a nonmuscle cy-toskeleton. Cell 36, 709–715.

Haaksma, C.J., Schwartz, R.J., Tomasek, J.J., 2011. Myoepithelial cell contraction andmilk ejection are impaired in mammary glands of mice lacking smooth musclealpha-actin. Biol. Reprod. 85, 13–21.

Hanft, L.M., Rybakova, I.N., Patel, J.R., Rafael-Fortney, J.A., Ervasti, J.M., 2006. Cytoplas-mic gamma-actin contributes to a compensatory remodeling response indystrophin-deficient muscle. Proc. Natl. Acad. Sci. U. S. A. 103, 5385–5390.

Harborth, J., Elbashir, S.M., Bechert, K., Tuschl, T., Weber, K., 2001. Identification of es-sential genes in cultured mammalian cells using small interfering RNAs. J. Cell Sci.114, 4557–4565.

Harris, A.K., Wild, P., Stopak, D., 1980. Silicone rubber substrata: a new wrinkle in thestudy of cell locomotion. Science 208, 177–179.

Hennessey, E.S., Drummond, D.R., Sparrow, J.C., 1993. Molecular genetics of actin func-tion. Biochem. J. 291 (Pt 3), 657–671.

Herman, I.M., 1993. Actin isoforms. Curr. Opin. Cell Biol. 5, 48–55.Insel, T.R., Harbaugh, C.R., 1989. Lesions of the hypothalamic paraventricular nucleus

disrupt the initiation of maternal behavior. Physiol. Behav. 45, 1033–1041.Insel, T.R., Young, L.J., 2001. The neurobiology of attachment. Nat. Rev. Neurosci. 2,

129–136.

Jaeger, M.A., Sonnemann, K.J., Fitzsimons, D.P., Prins, K.W., Ervasti, J.M., 2009. Context-dependent functional substitution of alpha-skeletal actin by gamma-cytoplasmicactin. FASEB J. 23, 2205–2214.

Karlsson, R., Aspenstrom, P., Bystrom, A.S., 1991. A chicken beta-actin gene can comple-ment a disruption of the Saccharomyces cerevisiae ACT1 gene. Mol. Cell. Biol. 11,213–217.

Kato-Minoura, T., Hirono, M., Kamiya, R., 1997. Chlamydomonas inner-arm dyneinmutant, ida5, has a mutation in an actin-encoding gene. J. Cell Biol. 137,649–656.

Katoh, T., Numata, T., Konishi, K., Furuya, K., Yazawa, M., 1997. Essential light chain ex-change in smooth muscle myosin. Anal. Biochem. 253, 78–84.

Kumar, A., Crawford, K., Close, L., Madison, M., Lorenz, J., Doetschman, T., Pawlowski, S.,Duffy, J., Neumann, J., Robbins, J., Boivin, G.P., O'Toole, B.A., Lessard, J.L., 1997. Res-cue of cardiac alpha-actin-deficient mice by enteric smooth muscle gamma-actin.Proc. Natl. Acad. Sci. U. S. A. 94, 4406–4411.

Kwon, Y.T., Xia, Z., Davydov, I.V., Lecker, S.H., Varshavsky, A., 2001. Construction andanalysis of mouse strains lacking the ubiquitin ligase UBR1 (E3alpha) of the N-end rule pathway. Mol. Cell Biol. 21, 8007–8021.

Labosky, P.A., Winnier, G.E., Jetton, T.L., Hargett, L., Ryan, A.K., Rosenfeld, M.G., Parlow,A.F., Hogan, B.L., 1997. The winged helix gene, Mf3, is required for normal develop-ment of the diencephalon and midbrain, postnatal growth and the milk-ejectionreflex. Development 124, 1263–1274.

Lessard, J.L., 1988. Two monoclonal antibodies to actin: one muscle selective and onegenerally reactive. Cell Motil. Cytoskeleton 10, 349–362.

Li, M., Liu, X., Robinson, G., Bar-Peled, U., Wagner, K.U., Young, W.S., Hennighausen, L.,Furth, P.A., 1997. Mammary-derived signals activate programmed cell death dur-ing the first stage of mammary gland involution. Proc. Natl. Acad. Sci. U. S. A. 94,3425–3430.

Liu, X., Robinson, G.W., Wagner, K.U., Garrett, L., Wynshaw-Boris, A., Hennighausen, L.,1997. Stat5a is mandatory for adult mammary gland development and lactogen-esis. Genes Dev. 11, 179–186.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) method. Methods 25,402–408.

Lloyd, C., Schevzov, G., Gunning, P., 1992. Transfection of nonmuscle beta- and gamma-actin genes into myoblasts elicits different feedback regulatory responses from en-dogenous actin genes. J. Cell Biol. 117, 787–797.

Lollivier, V., Marnet, P.G., Delpal, S., Rainteau, D., Achard, C., Rabot, A., Ollivier-Bous-quet, M., 2006. Oxytocin stimulates secretory processes in lactating rabbit mam-mary epithelial cells. J. Physiol. 570, 125–140.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement withthe Folin phenol reagent. J. Biol. Chem. 193, 265–275.

Lund, L.R., Romer, J., Thomasset, N., Solberg, H., Pyke, C., Bissell, M.J., Dano, K., Werb, Z.,1996. Two distinct phases of apoptosis in mammary gland involution: proteinase-independent and -dependent pathways. Development 122, 181–193.

Mahaffey, J.W., Coutu, M.D., Fyrberg, E.A., Inwood, W., 1985. The flightless Drosophilamutant raised has two distinct genetic lesions affecting accumulation of myofibril-lar proteins in flight muscles. Cell 40, 101–110.

McHugh, K.M., Crawford, K., Lessard, J.L., 1991. A comprehensive analysis of the devel-opmental and tissue-specific expression of the isoactin multigene family in the rat.Dev. Biol. 148, 442–458.

Miron-Mendoza, M., Seemann, J., Grinnell, F., 2008. Collagen fibril flow and tissuetranslocation coupled to fibroblast migration in 3D collagen matrices. Mol. Biol.Cell 19, 2051–2058.

Miwa, T., Manabe, Y., Kurokawa, K., Kamada, S., Kanda, N., Bruns, G., Ueyama, H., Kaku-naga, T., 1991. Structure, chromosome location, and expression of the humansmooth muscle (enteric type) gamma-actin gene: evolution of six human actingenes. Mol. Cell. Biol. 11, 3296–3306.

Nakano, H., Furuya, K., Yamagishi, S., 2001. Synergistic effects of ATP on oxytocin-induced intracellular Ca2+ response in mouse mammary myoepithelial cells. Pflu-gers Arch. 442, 57–63.

Nishimori, K., Young, L.J., Guo, Q., Wang, Z., Insel, T.R., Matzuk, M.M., 1996. Oxytocin isrequired for nursing but is not essential for parturition or reproductive behavior.Proc. Natl. Acad. Sci. U. S. A. 93, 11699–11704.

Nowak, K.J., Ravenscroft, G., Jackaman, C., Filipovska, A., Davies, S.M., Lim, E.M., Squire,S.E., Potter, A.C., Baker, E., Clement, S., Sewry, C.A., Fabian, V., Crawford, K., Lessard,J.L., Griffiths, L.M., Papadimitriou, J.M., Shen, Y., Morahan, G., Bakker, A.J., Davies,K.E., Laing, N.G., 2009. Rescue of skeletal muscle alpha-actin-null mice by cardiac(fetal) alpha-actin. J. Cell Biol. 185, 903–915.

Palmer, C.A., Neville, M.C., Anderson, S.M., McManaman, J.L., 2006. Analysis of lac-tation defects in transgenic mice. J. Mammary Gland Biol. Neoplasia 11,269–282.

Pedersen, C.A., Caldwell, J.D., Johnson, M.F., Fort, S.A., Prange Jr., A.J., 1985. Oxytocin an-tiserum delays onset of ovarian steroid-induced maternal behavior. Neuropeptides6, 175–182.

Perrin, B.J., Ervasti, J.M., 2010. The actin gene family: function follows isoform. Cyto-skeleton (Hoboken) 67, 630–634.

Posey, S.C., Bierer, B.E., 1999. Actin stabilization by jasplakinolide enhances apoptosisinduced by cytokine deprivation. J. Biol. Chem. 274, 4259–4265.

Robinson, G.W., Johnson, P.F., Hennighausen, L., Sterneck, E., 1998. The C/EBPbeta tran-scription factor regulates epithelial cell proliferation and differentiation in themammary gland. Genes Dev. 12, 1907–1916.

Rockey, D.C., Housset, C.N., Friedman, S.L., 1993. Activation-dependent contractilityof rat hepatic lipocytes in culture and in vivo. J. Clin. Invest. 92, 1795–1804.

Rubenstein, P.A., 1990. The functional importance of multiple actin isoforms. Bioessays12, 309–315.

14 N. Weymouth et al. / Developmental Biology 363 (2012) 1–14

Rubenstein, P.A., Spudich, J.A., 1977. Actin microheterogeneity in chick embryo fibro-blasts. Proc. Natl. Acad. Sci. U. S. A. 74, 120–123.

Schildmeyer, L.A., Braun, R., Taffet, G., Debiasi, M., Burns, A.E., Bradley, A., Schwartz, R.J.,2000. Impaired vascular contractility and blood pressure homeostasis in thesmooth muscle alpha-actin null mouse. FASEB J. 14, 2213–2220.

Seagroves, T.N., Krnacik, S., Raught, B., Gay, J., Burgess-Beusse, B., Darlington, G.J.,Rosen, J.M., 1998. C/EBPbeta, but not C/EBPalpha, is essential for ductal morpho-genesis, lobuloalveolar proliferation, and functional differentiation in the mousemammary gland. Genes Dev. 12, 1917–1928.

Shawlot, W., Deng, J.M., Fohn, L.E., Behringer, R.R., 1998. Restricted beta-galactosidaseexpression of a hygromycin-lacZ gene targeted to the beta-actin locus and embry-onic lethality of beta-actin mutant mice. Transgenic Res. 7, 95–103.

Skalli, O., Ropraz, P., Trzeciak, A., Benzonana, G., Gillessen, D., Gabbiani, G., 1986. Amonoclonal antibody against alpha-smooth muscle actin: a new probe for smoothmuscle differentiation. J. Cell Biol. 103, 2787–2796.

Sonnemann, K.J., Fitzsimons, D.P., Patel, J.R., Liu, Y., Schneider, M.F., Moss, R.L., Ervasti,J.M., 2006. Cytoplasmic gamma-actin is not required for skeletal muscle develop-ment but its absence leads to a progressive myopathy. Dev. Cell 11, 387–397.

Stinnakre, M.G., Vilotte, J.L., Soulier, S., Mercier, J.C., 1994. Creation and phenotypicanalysis of alpha-lactalbumin-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 91,6544–6548.

Takayanagi, Y., Yoshida, M., Bielsky, I.F., Ross, H.E., Kawamata, M., Onaka, T., Yanagi-sawa, T., Kimura, T., Matzuk, M.M., Young, L.J., Nishimori, K., 2005. Pervasive socialdeficits, but normal parturition, in oxytocin receptor-deficient mice. Proc. Natl.Acad. Sci. U. S. A. 102, 16096–16101.

Tepera, S.B., McCrea, P.D., Rosen, J.M., 2003. A beta-catenin survival signal is requiredfor normal lobular development in the mammary gland. J. Cell Sci. 116,1137–1149.

Van Nguyen, A., Pollard, J.W., 2002. Colony stimulating factor-1 is required to recruitmacrophages into the mammary gland to facilitate mammary ductal outgrowth.Dev. Biol. 247, 11–25.

Vandekerckhove, J., Weber, K., 1978. At least six different actins are expressed in ahigher mammal: an analysis based on the amino acid sequence of the amino-terminal tryptic peptide. J. Mol. Biol. 126, 783–802.

von Arx, P., Bantle, S., Soldati, T., Perriard, J.C., 1995. Dominant negative effect of cyto-plasmic actin isoproteins on cardiomyocyte cytoarchitecture and function. J. CellBiol. 131, 1759–1773.

Wagner, C.R., Mahowald, A.P., Miller, K.G., 2002. One of the two cytoplasmic actin iso-forms in Drosophila is essential. Proc. Natl. Acad. Sci. U. S. A. 99, 8037–8042.

Wagner, K.U., Young III, W.S., Liu, X., Ginns, E.I., Li, M., Furth, P.A., Hennighausen, L.,1997. Oxytocin and milk removal are required for post-partum mammary-glanddevelopment. Genes Funct. 1, 233–244.

Wilde, C.J., Knight, C.H., Flint, D.J., 1999. Control of milk secretion and apoptosis duringmammary involution. J. Mammary Gland Biol. Neoplasia 4, 129–136.

Willis, J.H., Munro, E., Lyczak, R., Bowerman, B., 2006. Conditional dominant mutationsin the Caenorhabditis elegans gene act-2 identify cytoplasmic and muscle roles for aredundant actin isoform. Mol. Biol. Cell 17, 1051–1064.

Winslow, J.T., Insel, T.R., 2002. The social deficits of the oxytocin knockout mouse. Neu-ropeptides 36, 221–229.

Yeo, J.A., Keverne, E.B., 1986. The importance of vaginal–cervical stimulation for mater-nal behaviour in the rat. Physiol. Behav. 37, 23–26.

Zhang, X., Minale, L., Zampella, A., Smith, C.D., 1997. Microfilament depletion and circum-vention of multiple drug resistance by sphinxolides. Cancer Res. 57, 3751–3758.

Zoubiane, G.S., Valentijn, A., Lowe, E.T., Akhtar, N., Bagley, S., Gilmore, A.P., Streuli, C.H.,2004. A role for the cytoskeleton in prolactin-dependent mammary epithelial celldifferentiation. J Cell Sci. 117, 271–280.