doi: 10.1242/10.1242/dev.00403
HIF1
is a critical regulator of secretory differentiation and activation, but not vascular expansion, in the mouse mammary gland
Tiffany N. Seagroves1,
Darryl Hadsell2,
Jim McManaman3,
Carol Palmer3,
Debbie Liao1,
Wayne McNulty1,
Bryan Welm4,
Kay-Uwe Wagner5,
Margaret Neville3 and
Randall S. Johnson1,*
1 Division of Biological Sciences, Molecular Biology Section, University of
California San Diego, La Jolla, CA 92093, USA
2 Department of Pediatrics, Baylor College of Medicine, Houston, TX 77030,
USA
3 University of Colorado Health Science Center, Department of Physiology,
Denver, CO 80262, USA
4 Department of Anatomy, University of California San Francisco, San Francisco,
CA 94143, USA
5 Eppley Institute for Research in Cancer and Allied Diseases, University of
Nebraska Medical Center, Omaha, NE 68198, USA
View larger version (47K):
[in a new window]
|
Fig. 1. Mammary epithelial cells contain functional HIF1. (A) Upper panel:
HIF1 , a triplet present at  120 kDa, was detected in purified
mammary epithelial cells (MEC) at normoxia (21% O2) and was induced
dramatically upon exposure to 0.5% oxygen. Lower panel: to demonstrate
equivalent loading, total protein was stained with a reversible protein
detection dye prior to blotting. (B) Purified MEC were cultured at either
normoxia (white bars) or hypoxia (gray bars), harvested, and expression of
target genes was normalized to 18S rRNA. After normalization, the relative
expression of each gene was expressed as a percentage of that observed in
wild-type cells at normoxia (mean±s.e.m.). No significant differences
in gene expression were observed for any gene between Hif1a wild-type
(WT, Hif1af+/f+ only) and null cells
(Hif1af+/f+ cells infected with Adeno-Cre) cultured under
normoxic conditions. At hypoxia, relative to the robust induction of Pgk,
Glut1 and Vegf observed in wild type cells, induction of all of
these mRNAs was decreased by at least 50%. (C) Genomic DNA was prepared from
the mammary glands (mg) and ovaries (ov) of Hif1af+/f+
(Cre-negative) or Hif1af+/f+, MMTV-Cre-positive mice at
either day 15 of gestation (15-P) or at mid-lactation and used for Southern
blotting as described previously (Ryan et
al., 1998 ). As a control for complete excision, DNA was prepared
from primary mouse embryonic fibroblasts (MEF) previously infected with
Adeno-Cre.
|
|
View larger version (108K):
[in a new window]
|
Fig. 2. Defects in secretory differentiation, but not vascular expansion, at day 15
of pregnancy. (A,B) Paraffin wax-embedded sections prepared from wild-type (A)
or Hif1a–/– (B) glands isolated at day 15 of
pregnancy stained with Hematoxylin and Eosin. Scale bar: 50 µm. Note the
striking block in differentiation in the
Hif1a–/– glands (B). (C,D) Patterning of the
vasculature in relationship to the mammary epithelium after lectin (green) and
phalloidin (red) staining in wild-type (C) and
Hif1a–/– glands (D).
|
|
View larger version (81K):
[in a new window]
|
Fig. 4. Block in secretory activation at the transition to lactation. (A,B)
Hematoxylin and Eosin stained sections at day 18 of gestation. Scale bar: 50
µm. Note the uniform size of alveoli and the extent of differentiation in
the wild-type glands (A), versus the mixture of collapsed, non-differentiated
and differentiated alveoli in the Hif1a–/–
glands (B). (C) Expression of Cre (brown, nuclear staining) detected in the
Hif1a–/– gland shown at higher magnification
in D. Areas that expressed Cre contained small, relatively undifferentiated
alveoli (null; white arrows), whereas areas negative for Cre achieved
differentiation (WT; black arrows). (E-G) Multiple antigen labeling
immunostaining was performed to sequentially detect Glut1 (brown stain) and
Cre (purple stain, nuclear localization) using paraffin wax-embedded sections
at day 18 of gestation followed by counterstaining with Methyl Green. The
black arrows indicate the robust, uniform Glut1 staining pattern observed in
wild-type, differentiated alveoli. The white arrows note the relatively weak
Glut1 expression detected in alveoli that also express Cre
(Hif1a–/–). Wild-type (F) and Hif1a
null alveoli (G) were individually imaged at higher power to highlight the
reduced expression of Glut1 in response to deletion of Hif1a. (H)
Average Chalkley score following anti-CD31 immunostaining to determine MVD at
day 18 of gestation, (mean±s.e.m.). (I) DNA, RNA, and protein content
per gram tissue at day 18 of gestation (mean±s.e.m.).
|
|
View larger version (109K):
[in a new window]
|
Fig. 5. Impaired secretory function at lactation. (A,B) Paraffin wax-embedded
sections from mammary glands harvested on the date of birth were stained with
Hematoxylin and Eosin. Scale bar: 50 µm. In comparison with wild-type (A),
glands lacking Hif1 (B) contained fewer alveoli, which were
less differentiated. (C,D) Glands were also harvested at mid-lactation from
weaned dams, allowing milk (indicated by yellow stars) to fill the gland.
Scale bar: 50 µm). In wild-type mice (C), the accumulation of milk fully
distended the alveoli, and a relatively small volume of adipose tissue was
present. By contrast, in the Hif1a–/– glands
(D), accumulation of milk was minimal, large lipid droplets remained trapped
within the epithelial cells (white arrow), and large areas of adipose tissue
were visible. (E) Wet weight (±s.e.m.) of frozen inguinal glands
harvested from lactating dams. (F) Representative growth curve of litters
nursing wild-type (black circles) and Hif1a–/–
dams (gray circles). Pronounced defects in pup weight gain were observed by
day 3 of lactation that persisted until mid-lactation.
|
|
View larger version (20K):
[in a new window]
|
Fig. 6. Changes in milk nutrition and gene expression at mid-lactation. (A)
Analysis of milk volume, water and ion content in milk collected from
wild-type (white bars) and Hif1a–/– glands
(gray bars) at mid-lactation, mean±s.e.m. (B) RTD-PCR was used to
determine the levels of Pgk, Glut1, claudin 7 and claudin 8 mRNA in
mid-lactation tissue, relative to Ck19 expression.
|
|
View larger version (138K):
[in a new window]
|
Fig. 7. The effects of deletion of HIF1 are mammary epithelial cell
autonomous. Primary Hif1af+/f+ mammary epithelial cells
were infected with either Adeno-ßgal or Adeno-Cre and transplanted into
the cleared fat pads of female host mice in order to generate wild-type (A)
and Hif1a–/– (B) epithelial outgrowths,
respectively. Paraffin wax-embedded sections harvested from mice on the date
of birth (without prior weaning of pups) were stained with Mason's Trichrome.
Scale bar: 50 µm (n=3 mice with 100% outgrowths/genotype).
Outgrowths derived from Hif1a–/– mammary
epithelial cells recapitulated the phenotype observed in intact
Hif1a–/– glands. Note the lack of milk
products (indicated by yellow stars) in the
Hif1a–/– outgrowth, and the presence of large,
trapped lipid droplets within the epithelial cells (B, right white arrow). In
addition, there was an abnormal thickening of collagen fibers (stained blue;
B, left white arrow) around the alveoli in these outgrowths compared with wild
type (black arrow). (C,D) Glut1 (brown staining) was detected in paraffin
wax-embedded sections prepared from transplanted outgrowths. Compared to wild
type (C), expression of Glut1 in the Hif1a null outgrowths (D) was
less intense and more patchy.
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2003
