QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 8 April 2004
doi: 10.1242/dev.01090

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mericskay, M. Articles by Sassoon, D. Search for Related Content PubMed PubMed Citation Articles by Mericskay, M. Articles by Sassoon, D. Social Bookmarking                         What's this?

Wnt5a is required for proper epithelial-mesenchymal interactions in the uterus

Mathias Mericskay^1,*, Jan Kitajewski² and David Sassoon^1,

¹ Brookdale Department Molecular, Cell and Developmental Biology, Mount Sinai Medical School, 1 G Levy Place, New York, NY 10029, USA
² Department of Pathology and OB/GYN, Columbia University, 630 West 168th Street, New York, NY 10032, USA

View larger version (70K): [in a new window]   Fig. 1. Wnt5a mutant FRTs lack posterior structures. Hoxa10 whole-mount in situ hybridization of the FRT from wild-type (A,C) and Wnt5a-/- (B,D) at P0. The Wnt5a mutant FRT lacks cervical and vaginal structures and the uterine horns are short and convoluted. Arrows in C and D indicate the anterior limit of Hoxa10 expression at the uterotubal junction. Scale bars: 1 mm in A,B; 0.4 mm in C,D. Msx1 (E,F), Wnt7a (G,H), Wnt4 (I,J) and Wnt5a (K,L) 35S in situ hybridization of paraffin wax embedded sections from wild-type (E,G,I,K) and Wnt5a mutant (–/–) (F,H,J,L) P0 FRTs. Silver grains are superimposed as red upon a phase contrast image. Ao, aorta; ce, cervix; ova, ovary; ovi, oviduct; ur, ureter; ut, uterus; va, vagina. Scale bar: 200 µm in E-L.

View larger version (78K): [in a new window]   Fig. 2. Postnatal development of the Wnt5a mutant FRT grafts. Wild-type and Wnt5a mutant P0 FRT were cut into fragments along the anteroposterior axis (as indicated) and grafted under the renal capsule of an intact adult host. Grafts were harvested 3 weeks after grafting into the host. Paraffin sections from wild-type (A-D) and Wnt5a mutant (E-G) were stained with Haematoxylin and Eosin. (H-M) High magnification of corresponding boxed area in A-G. All structures formed in the Wnt5a mutant displayed normal characteristics of each compartment, including postnatal smooth muscle (smc) differentiation, stromal compartment (stro) and ciliated epithelium (ci) in the oviduct [inset in K (scale bar: 10 µm)]. Scale bar: 250 µm for low magnification; 50 µm for high magnification. ost, ostium; amp, ampulla.

View larger version (126K): [in a new window]   Fig. 3. Uterocervical junction forms in the Wnt5a mutant despite the lack of an identifiable cervix at birth. (A) Haematoxylin-Eosin staining of a Wnt5a mutant posterior graft at low magnification. Scale bar: 400 µm. (B,C) High magnification of boxed areas in A showing the transition from simple columnar epithelium to stratified epithelium (arrowhead; scale bar: 40 µm). The transition is accompanied by a correct formation of thick smooth muscle layers in the uterine area and sparse smooth muscle bundles in the vaginal region as shown by smooth muscle myosin heavy chain in situ hybridization (D, SMHC). (E) Wnt7a also shows a normal and sharp boundary of expression at the level of the uterocervical transition and Wnt4 (F) expression shows the correct pattern of epithelial (ep) and stromal (st) expression in the uterus (proestrus stage) and stratified epithelium (se) in the vagina (see insets).

View larger version (140K): [in a new window]   Fig. 4. Wnt5a and Wnt7a are required for gland formation in grafts grown into ovariectomized hosts. (A-F) Haematoxylin-Eosin staining of grafts grown for 3 weeks in ovariectomized hosts (A-E) and cycling host (F). Note the presence of glands (gl) in the control Wnt5a+/– graft (A) and Lef1-/- (D) but not in the Wnt5a-/- (B), Wnt7a-/- (C) and Wnt7a/Lef1-/- (E), nor the wild-type graft grown in a cycling host (F). Scale bar: 50 µm. (G-J) Pattern of Wnt gene expression during gland formation in the wild-type uterus at P15 (G-I, adjacent sections) and in a 3 week-old wild-type graft (J). Lumen (lum) and smooth muscle (sm) are indicated. (G) Wnt7a is expressed in the `invaginated' luminal epithelium but not in the glandular epithelium (gl). In a section that passes through the site of invagination (J), we observe a boundary of Wnt7a expression between the luminal epithelium and the glandular epithelium. Wnt5a is expressed in the stroma surrounding the glands (H). Low levels of Wnt5a expression are also detected in the epithelium. Wnt4 expression is most abundant in the subepithelial stroma (I). Scale bar: 20 µm.

View larger version (91K): [in a new window]   Fig. 5. Stromal Wnt5a expression is required for gland formation in the uterus. (A) Schema of the recombinant graft procedure. The mesenchymal sheath is separated from the epithelial tube by mild trypsin digestion and gentle mechanical manipulation. The mesenchyme (mes) from either wild-type or Wnt5a-/- is recombined with wild-type or Wnt5a-/- epithelium (epi) and grafted under the renal capsule of an adult host. (B-E) Haematoxylin-eosin staining of the recombinants. Glands (gl) form in the wild-type mes/wild-type epi (B) and wild-type mes/Wnt5a-/- epi (D) but not in the Wnt5a-/- mes/wild-type epi (C), or in the Wnt5a-/- mes/Wnt5a-/- epi (E). (F-H) Haematoxylin-Eosin staining of grafts (frozen sections) derived from the same Wnt5a-/- individual infected at birth by the retroviral backbone (control, F), Wnt5a expressing retrovirus (G) and a Wnt4 expressing retrovirus (H). We observe that Wnt5a rescues the formation of glands whereas Wnt4 does not. Scale bar in D: 50 µm.

View larger version (110K): [in a new window]   Fig. 6. Wnt5a is required for the uterotrophic response and for DES-mediated repression of Wnt7a. Neonate (P0) uterine horns from control and Wnt5a-/- individuals were separated into two pools of grafts that were grown in two ovariectomized hosts for 3 weeks. Each host received control (two left-hand columns) and mutant (two right-hand columns) grafts. For each experiment, one host was injected intraperitoneally daily from day 18 to day 20 with DES resuspended in saline and one host was injected with saline alone, as indicated. Hosts were sacrificed on day 21 and the grafts harvested for analyses. Results are shown for a Wnt5a+/– individual and a Wnt5a-/- individual. (A-H) Haematoxylin-Eosin staining at low magnification (A-D; scale bar: 250 µm) and high magnification (E-H; scale bar: 20 µm). Note the aberrant uterotrophic response in the Wnt5a-/- graft (D) showing enlarged lumen and thin uterine walls when compared with the Wnt5a+/– graft (B). The Wnt5a-/- epithelium does show an increase in height and thickness in response to DES (H). (I-T) In situ hybridization for Wnt7a (I-L), Hoxa10 (M-P) and Hoxa11 (Q-T). Wnt7a is repressed by DES in the Wnt5a+/– control graft (J) but not in the mutant (L). Hoxa10 and Hoxa11 are strongly repressed by DES in the subepithelial stroma of the Wnt5a+/– control graft (N,R) but not in the Wnt5a-/- graft (P,T). Lines delineate the limits between the luminal epithelium, the stroma and the myometrium.

View larger version (157K): [in a new window]   Fig. 7. Regulation of Wnt7a, Hoxa10 and Hoxa11 genes is deficient in the Wnt5a mutant during estrous. Genotypes are indicated at the top and in situ probes on the left. (A) Control and Wnt5a-/- grafts were grown for 3 weeks in a cycling host that was sacrificed at diestrus when the level of circulating estrogen is low. Wnt7a and Hoxa genes expression is normal in the Wnt5a-/- graft. Note the expected lack of expression of Hoxa10 and Hoxa11 genes in the anterior mutant FRT, i.e. tubular and ostium (infundibulum) region of the oviduct. Lines in the Hoxa10 photographs delineate the limit between the epithelium and the stroma. (B) As in A but the host was sacrificed at late proestrus when the level of circulating estrogen is high. Wnt7a and Hoxa10 and Hoxa11 are downregulated in the control wild-type and Wnt5a+/– grafts but not in the Wnt5a-/- grafts.

View larger version (180K): [in a new window]   Fig. 8. Estrogen signaling is intact in Wnt5a mutant grafts. In situ hybridization for Esr1 (A-F), Wnt5a (G-J), Pgr (K-P) and Msx1 (Q-T) from host uteri and grafts grown in saline or DES conditions, as indicated at the bottom. Genotypes are indicated on the left. Esr1 expression increases to very high levels in adult mice after ovariectomy (A) but is repressed after prolonged exposure to DES, except in the epithelium and the smooth muscle layer (B). Although lower in 3-week-old grafts, Esr1 expression is similarly regulated in both control Wnt5a+/– (C,D) and Wnt5a-/- grafts (E,F). Wnt5a is downregulated in the stroma and activated in the epithelium and smooth muscle, and Wnt5a mutant transcript is correctly regulated even in absence of Wnt5a product (H,J). Pgr gene regulation is also identical in control and mutant grafts although Pgr is not downregulated in the epithelium from grafts (N,P) as in the host (L), probably because of stage difference between immature 3-week-old grafts and sexually mature host uterus. Msx1 is repressed by DES in both control (R) and Wnt5a mutant grafts (T).

View larger version (51K): [in a new window]   Fig. 9. Role and regulation of Wnt genes during FRT morphogenesis and estrogenic response. (A) Schematic comparison of Hoxa code and Hoxa mutant phenotypes with Wnt5a and Wnt7a mutant phenotypes in the FRT. During fetal development, all the Hoxa genes, Wnt7a and Wnt5a are expressed all along the anteroposterior axis of the FRT (not shown). At birth, domains of Hoxa genes expression start to regionalize along the anteroposterior axis of the FRT (see left diagram). The regionalization of Hoxa10 to the uterine horn slightly precedes regionalization of Wn7a and Wnt5a also to the uterine horns that occur a few days after birth. The Hoxa10 mutant phenotype presents a bona fide homeotic transformation of the anterior 25% of the uterine horn into an oviduct-like structure. Loss of Hoxa11, or one allele of each Hoxa10 and Hoxa11 genes, or loss of Wnt7a affects primarily the uterine horns; however, Wnt7a phenotype can also affect the oviduct and the vagina. Loss of Hoxa13 or Wnt5a affects the caudal growth of the Müllerian ducts and the growth of the genital tubercle (not shown). (B) Postnatal uterine morphogenesis. Wnt7a is required for correct epithelial organization, the radial growth and patterning of the adjacent mesenchymal cells, and the organization of the smooth muscle layers. Wnt7a is required for maintenance (dotted arrows) of high levels of Wnt5a, Wnt4, Hoxa10 and Hoxa11 genes. Wnt5a signals cooperate with an unknown factor X to allow Wnt7a downregulation during gland formation (this study). (C) Wnt5a-dependant and Wnt5a-independent uterotrophic response to DES. DES binding to stromal Esr1, downregulates Wnt7a in the epithelium through a factor X that is functional or present only when Wnt5a is expressed. The factor X could be the same or different to the factor X required for Wnt7a repression during glandulargenesis. DES, through factor X, represses the levels of Hoxa10 and Hoxa11 in the stroma either directly or through repression of Wnt7a. Correct Wnt5a dependant downregulation of Wnt7a and Hoxa genes by prolonged estrogenic signal may be involved in the stimulation of glandulargenesis, fluid retention by the stroma and possibly preparation of the uterine wall for embryo implantation.

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

Twitter