Figure 1.
Schematic diagram of the barrier between the maternal and embryonic vasculatures within the placental labyrinth.
The placental endothelial basement membrane, which normally contains LMα5, lies between the fetal endothelium and the trilaminar trophoblast cellular structure. Mononuclear trophoblasts line the maternal blood spaces, whereas the other two trophoblast layers are syncytial due to cell-cell fusion. Maternal red blood cells (RBCs) lack nuclei, whereas fetal RBCs retain nuclei until late gestation.
Figure 2.
LMα5 is expressed in both endothelial cells and trophoblasts in the normal placenta.
(A) Fluorescence-activated cell sorting was performed on dissociated E18.5 wild-type labyrinth cells after staining with a phycoerythrin (PE)-conjugated CD31/PECAM antibody. CD31+ (endothelial cell) and CD31− (trophoblast; indicated as baseline) populations were collected. (B) RT-PCR using RNA prepared from the two cell types showed that LMα5 was expressed in both: Lane 1, DNA marker; 2 and 3, LMα5 in CD31(−) and (+) cells, respectively; 4, negative control; 5 and 6, GAPDH in CD31(−) and (+) cells, respectively. (C) RNA was subjected to real time RT-PCR to quantitate the levels of laminin α1 (lama1), α5 (lama5), β1 (lamb1), and β2 (lamb2) mRNAs. Error bars represent standard deviations.
Figure 3.
Mosaic placental labyrinths containing wild-type trophoblasts and LMα5−/− endothelial cells show LMα5 deposition and normal vascularization.
(A, B) Schematic diagrams of the strategy for conditional mouse LMα5 mutation. Using the Cre/loxP system, we generated LMα5flox/ko; Sox2Cre embryos. Sox2cre, when inherited from a male, is active in epiblast, but not in trophectoderm. Thus, epiblast-derived cells (A), which include the embryo proper as well as extraembryonic endothelial cells, are not able to synthesize LMα5, but trophoblasts, which derive from trophectoderm (B), can. (C–H; C′–H′) Analysis of LMα5 expression and tissue architecture in control (top rows) and LMα5flox/ko; Sox2cre mutant (bottom rows) embryos. LMα5 was not expressed in the kidney of LMα5flox/ko; Sox2cre embryos (C′; counterstained with anti-nidogen in D′; compare with control, C and D), which show developmental abnormalities typical of LMα5−/− embryos (E′; arrows indicate exencephaly and syndactyly) not seen in control (E). In contrast, LMα5 was present in the placental labyrinth of LMα5flox/ko; Sox2cre embryos (F′) and of control (F), and placental LM-111 and PECAM expression and localization were similar to those observed in control LMα5+/− placenta (G–H, G′–H′). Cytokeratin 8 (CK8) was used to identify trophoblasts (G, G′).
Figure 4.
Mosaic placental labyrinths containing LMα5−/− trophoblasts and hLMα5-expressing endothelial cells show hLMα5 deposition and normal vascularization.
(A) Schematic diagram of the strategy for forcing expression of hLMα5 in endothelial cells on the LMα5−/− background. Cre recombinase driven by the Tie2 promoter removes a floxed STOP located between the Rosa26 promoter and the reverse tetracycline transactivator (rtTA). rtTA binds and activates the tetracycline-inducible TetO7 promoter in the presence of doxycycline, thereby driving transcription of the hLMα5 cDNA in endothelial cells. (B–G) LMα5−/−;ROSA26TA;hLMα5;Tie2cre embryos (top panels) were compared with LMα5−/− embryos (bottom panels). Mouse LMα5 was undetectable in kidney (B, B′) or placenta (E, E′). Human LMα5 was detected in both kidney and placental vasculatures of LMα5−/−;ROSA26TA;hLMα5;Tie2cre embryos (C, F) but not of LMα5−/− embryos (C′,F′), both of which show the typical LMα5 null phenotype (D, D′). Expression of hLMα5 in endothelial cells was associated with a normalized placental labyrinth architecture, demonstrated by the LM-111 antibody staining pattern (compare G and G′).
Figure 5.
Analysis of placental labyrinth vasculature at E14.5.
Frozen sections of placenta were stained with antibodies to LM-111 to label all basement membranes, to cytokeratin 8 (CK8) to label trophoblasts (green in A–C, A′–C′), and to PECAM to label endothelial cells (green in D–F, D′–F′). The reduced vascular complexity in the LMα5−/− labyrinth (B, E) was rescued and made similar to normal (A, D) by hLMα5 secretion from LMα5−/−;ROSA26TA;hLMα5;Tie2cre endothelial cells (C, F) exposed to doxycycline.
Figure 6.
Phenotype of various LMα5 mutant and control embryos at E18.5.
Genotypes are indicated. The mother was fed doxycycline beginning at E0.5. Endothelial expression of hLMα5 in LMα5−/−;RO26Ta;hLMα5;Tie2cre embryos resulted in a larger (though still not quite normal) embryo size compared to the nontransgenic LMα5−/− embryo.
Figure 7.
Expression of hLMα5 in glomerular endothelial cells rescues glomerular vascularization in LMα5−/− kidney.
(A,B) Analysis of human (A) and mouse (B) LMα5 expression (green) relative to WT1 (red), which stains podocyte nuclei at E18.5. Human LMα5 is visible primarily in the glomeruli (open arrows) and in arterioles in embryos carrying the transgenes (A,A′) but is absent from the mutant lacking the transgenes (A″; open arrowheads indicate glomeruli). (C) Status of glomerular vascularization was revealed by PECAM (green) and WT1 (red) double staining. PECAM-positive endothelial cells were properly localized inside glomeruli when either mouse or hLMα5 or both were present in the GBM (C, C′), but glomerulogenesis failed in the absence of LMα5 (C″). (D,E) Transmission electron microscopic analysis of glomeruli in control (D) and rescued mutant (E) kidney reveals that both have an intact GBM (arrows), capillaries containing red blood cells (RBCs), and podocytes with foot processes (arrowheads). Bars in D and E are 500 nm.