Figure 1.
Pds5A deficiency results in growth retardation, abnormal skeletal patterning and cleft palate.
(A) The weight of Pds5A−/− P0 mice is significantly lower than that of wildtype and heterozygous littermates (WT, n = 8; Pds5A+/−, n = 30; Pds5A−/−, n = 39). Error bars represent s.e.m. *P<0.001, Student's unpaired, two tailed t-test. (B) Morphology of wildtype and Pds5A mutant mice. Note that newborn Pds5A−/− pups (right) were smaller than their wildtype littermates (left). Note too the absence of milk in the stomach of the Pds5A−/− pup. Scale bar: 0.5 cm. (C, D) Alizarin Red S and Alcian Blue staining of neonatal skulls demonstrates complete cleft palate (arrows) of a Pds5A−/− neonate (D) compared with the proximity of palatal bones in the midline (arrow) in the wildtype control (C). Scale bar: 0.5 mm. (E, F) Palatogenesis defects illustrated using H&E stained coronal sections of E18.5 wildtype (E) and Pds5A−/− (F) embryos. ps, palatal shelf; n, nasal bone; t, tongue. Arrows in F point to the unfused palatal shelves. Scale bar: 1 mm. (G–P) Alizarin Red S and Alcian Blue staining of newborn skeleton. Bone is stained red and cartilage is stained blue. (G–J) Cervical and thoracic vertebrae and ribs. Note the C7-T1 fusions (arrows in H, I) and cervical rib (arrow in J) in Pds5A−/− mice, compared to the normal skeletal patterning in wildtype animals (G). Cervical (C) and Thoracic (T) vertebrae are marked by numbers. (K–N) Sternum morphology. Arrows denote ectopic rib-sternum conjunctions (L–N) in Pds5A−/− mice compared to sternum patterning in wildtype (K). (O, P) Morphology of the first cervical vertebrae. Note the unfused ossification centers at the dorsal tip of the first cervical vertebrae (arrows) in Pds5A−/− animal (P) compared to wildtype (O). Scale bar for G–N: 1 mm; scale bar for O–P: 0.5 mm. (Q, R) Renal abnormalities in neonatal Pds5A-deficient mice. (Q) Bilateral renal agenesis in Pds5A−/− newborn animals. AG, adrenal gland. Scale bar: 1 mm. (R) The total number of glomeruli in kidneys from P0 Pds5A−/− and wildtype mice were counted (n = 4 animals, 8 kidneys). (*, P<0.05, student unpaired t test; mean±s.e.m.).
Figure 2.
Normal development of hippocampal neurons, the superior cervical ganglion, and germ cells in Pds5A−/− mice.
(A, B) Hippocampal neurons from wildtype and Pds5A mutant mice were cultured for 3 days and stained with TuJ1 antibody. The neurons show normal axonal projection and polarity (one axon per neuron). Arrowheads, cell bodies; arrows, axons. (C, D) Tyrosine hydroxylase (TOH) whole-mount staining of sympathetic projections from the superior cervical ganglion in wildtype (C) and Pds5A mutant (D) mice showed that the carotid nerves (indicated by arrows) project normally to the eye. ey, eye; ea, ear. (E, F) H&E staining of germ cells in neonatal testes showed similar numbers of germ cells in wildtype (E) and Pds5A−/− (F) mice. Arrows point to testicular cords surrounding the germ cells. (G, H) H&E staining of 6-week testes transplants of E18.5 wildtype (G) and Pds5A−/− mice (H). Note the presence of elongated spermatids with condensed heads and long tails in both wildtype and Pds5A−/− testis explants (arrows).
Figure 3.
Pds5 gene dosage synergistically affects development but does not cause defects in sister chromatid cohesion or cohesin-chromatin association.
(A) Overall cohesin levels in E14.5 kidney lysates were assessed by Western blotting using anti-SMC3 antibodies. ß-tubulin levels were used to confirm equivalent protein loading. Note that SMC3 levels were equivalent in tissues from wildtype and Pds5A−/− mice. (B) GTG (giemsa-banding using trypsin and giemsa) banded metaphase spreads from mouse embryonic fibroblasts (MEFs) derived from wildtype and Pds5A−/−;Pds5B+/− mice showed no evidence of precocious sister chromatid separation (PSCS). (C) Fluorescent Recovery After Photobleaching (FRAP) using SCC1-EGFP in wildtype and Pds5A−/−;Pds5B+/− MEFs. Half of the nucleus was bleached and fluorescence intensity was measured every 30 sec for 30 min. The images show the recovery dynamics over time. Note no significant difference of cohesin recovery dynamics between wildtype and Pds5A−/−;Pds5B+/− cells after photobleaching. (D) A summary table listing the abnormalities associated with Pds5 deficiencies demonstrates the genetic interactions between Pds5A and Pds5B. The left diagram displays the age when most embryos of the indicated genotype died. The right table lists the phenotypes observed in the indicated embryos. Hom, homozygote; het, heterozygote; wt, wild type; GD, growth delay; CP, cleft palate; SPD, skeletal patterning defects; CHD, congenital heart defects; PNSD, peripheral nervous system defects; LD, lens development defects; PGCD, primordial germ cell defects; CD, cohesion defects; KD, kidney developmental defects. +, presence of the indicated phenotype; - absence of the indicated phenotype; ND, not determined; N/A, not applicable since no embryos were obtained for double homozygous mutants. *, Pds5A homozygotes have delayed ENS precursor migration, but normal SCG projection. [+], highlights functional redundancy and diversification of PDS5A and PDS5B.
Figure 4.
Pds5A mutant mice manifest cardiac abnormalities similar to those observed in CdLS.
(A) A section of a wildtype mouse heart demonstrates the intact ventricular septum and its relationship to the aortic and tricuspid valves.. (B–D) Sections of Pds5A mutant hearts showing: (B) a perimembranous ventricular septal defect (VSD); (C) the pulmonary valve and artery in a heart with double outlet right ventricle; (D) the aorta arising from the right ventricle of the same heart with a muscular conus between the aortic and mitral valve (arrowhead). Arrows indicate the respective defects. Scale bar: 0.5 mm. (E–J) Severe cardiac defects in Pds5A+/−;Pds5B−/− and Pds5A−/−;Pds5B+/− embryos. (E) A section from wildtype E12.5 embryonic heart shows the separation of aorta and pulmonary artery and the developing aortic valves (arrow). (F) A Pds5A−/−;Pds5B+/− E12.5 embryonic heart shows the underdeveloped endocardial cushion of the outflow tract and the non-septated aorta and pulmonary artery (i.e., truncus arteriosus). (G) Normal association of the atrial and ventricular septae with the endocardiac cushion and valve morphogenesis in a wildtype E12.5 heart. (H) A Pds5A+/−;Pds5B−/− E12.5 heart demonstrates the dilated atria and underdeveloped endocardial cushion with no valve formation, and failure of atrioventricular canal septation. (I) Wildtype E12.5 heart with a compact myocardial layer of normal thickness. (J) A Pds5A−/−;Pds5B+/− heart has a thin compact myocardium. Double-headed arrows in (I, J) indicate the compact myocardial layer. Ao, aorta; PA, pulmonary artery; RA, right atrium; LA, left atrium; RV, right ventricle; LV, left ventricle; EC, endocardial cushion; T, trabecular myocardium; C, compact myocardium.
Figure 5.
Synergistic effects of Pds5A and Pds5B deficiencies on enteric nervous system development.
(A–C) Whole mount TuJ1 immunofluorescent antibody staining of E12.5 gut highlights the enteric neurons (A, Pds5A+/−; B, Pds5A−/−; C, Pds5A−/−;Pds5B+/−). White bars define the ileocecal junction and an asterisk denotes the end of the colon. Enteric neurons colonize the bowel to the mid-colon in wildtype mice (A), but migrate only to the proximal colon or ileocecal junction in Pds5A−/− (B) and to near the ileocecal junction in Pds5A−/−;Pds5B+/− mice (C). Scale bars: 10 µm. (D) Schematic representation of the extent of bowel colonization by enteric neurons Pds5A−/− mice at E12.5. [WT (n = 8), Pds5A−/− (n = 6), and Pds5A−/−;Pds5B+/− (n = 3)]. The scale at the top corresponds to the percentage of the respective intestinal segment (small or large intestine) successfully colonized by neurons. (E–G) Whole mount TuJ1 immunofluorescent antibody staining of E11.5 gut highlights the enteric neurons in (E) Pds5A+/−, (F) Pds5B−/− and (G) Pds5A+/−;Pds5B−/− mutant mice. White bars define the ileocecal junction and an asterisk denotes the end of the colon. Note that the enteric neuron migration wavefront in wildtype mice reaches the ileocecal junction at this age while migration was delayed in Pds5B−/− mice. In Pds5A−/−;Pds5B+/− mice only a few enteric neurons in the stomach were detected and none had migrated into the intestine. Scale bars: 10 µm. (H) Schematic representation of ENS defects in Pds5A+/−;Pds5B−/− mice at E11.5 [WT (n = 4), Pds5B−/− (n = 4), Pds5A+/−;Pds5B−/− (n = 6)]. The scale at the top corresponds to the percentage of the respective intestinal segment (small or large intestine) successfully colonized by neurons.
Figure 6.
Pds5A−/−;Pds5B+/− and Pds5A+/−;Pds5B−/− mice display abnormal lens development.
(A, B) H&E stained sections of eyes from E12.5 wildtype and Pds5A−/−;Pds5B+/− embryos. No lens was formed in the Pds5A−/−;Pds5B+/− embryo (B), whereas the wildtype embryo has a clearly formed lens at this stage (A). (C,D) Whole mount image of the eyes from E11.5 wildtype and Pds5 mutant embryos. Note that the Pds5A+/−;Pds5B−/− eye (arrow, D) is much smaller than wildtype (white dashed circle, C). (E) Immunofluorescent staining of E9.5 eye using PDS5A antibodies demonstrated that PDS5A is expressed at high levels in head surface ectoderm. HSE, head surface ectoderm; M, mesenchyme; OV, optical vesicle; dashed lines, the boundaries among HSE, M, and OV. (F–H) Proliferating cells in the E10.5 eyes of wildtype (G) and Pds5A−/−;Pds5B+/− (H) embryos labeled with BrdU for 1 hr were detected using anti-BrdU antibodies. The eyes were counterstained with hematoxylin. Eye structures are as indicated in G. Note that the lens placode is invaginating in wildtype (G) but not in Pds5A−/−;Pds5B+/− embryos (H), although the percentage of BrdU positive cells in both wildtype and mutant lens placodes was equivalent (F). (I,J) Immunohistochemical analysis using αA-crystallin antibodies of E11.5 wildtype and Pds5A+/−;Pds5B−/− eyes. Note the small size and lack of αA-crystallin-positive cells in a Pds5A+/−;Pds5B−/− lens (J) compared to strong αA-crystallin expression in the majority of cells in a normal size, wildtype lens (I). R, retina; L, lens.
Figure 7.
PDS5B is mutated in a case of familial CdLS.
(A) A schematic diagram of PDS5B showing the location of the R1292Q mutation within the AT-hook domain. Red rectangles, HEAT repeats; grey rectangle, a leucine zipper domain; blue rectangles, AT-hook domains; thin black bars, phosphorylation sites (Ser, Thr, and Tyr) in the C-terminal region of PDS5B. Note: See additional PDS5B structural analysis in supplementary data (See Fig. S6 in Supplementary Data and Materials S1). (B) Sequencing traces displaying the G to A substitution (highlighted with an arrow) in one PDS5B allele in the proband (compare traces of proband and control). This substitution results in an Arg to Gln transition at PDS5B residue 1292. (C) Multispecies alignment of AT-hook domain from PDS5B and PDS5A. Note that the mutated Arg is highly conserved and located within the AT-hook domain. Hs, Homo sapiens; Pa, Pan troglodytes; Ca, Canis familiaris; Mm, Mus musculus; Rn, Rattus norvegicus; Ch, Gallus gallus; Xe, Xenopus tropicalis; Br, Danio rerio; Te, Tetraodon nigroviridis; Fu, Fugu rubripes. (D) The family tree of a proband (arrow) with three affected siblings (II-1,2,3, blackened circle or square), one clinically normal brother (II-4), and two clinically normal parents (I-1 and I-2). One of the affected siblings was deceased and no DNA was available from this individual (II-1). The father, two affected children, and one unaffected child carry the R1292Q mutation (depicted by blackened symbol or small black dot). (E) The PDS5B(R1292Q) mutant has decreased DNA binding. EMSA using the T5-8-T5 AT-rich probe showed decreased DNA binding by mutant PDS5B compared to wildtype PDS5B. The EMSA was performed with the indicated amounts of PDS5B protein (40, 20, 10, 5 ng). The specific nature of the binding was determined by EMSA competition assays (See Fig. S10 in Supplementary Data and Materials S1). Arrows, protein-DNA complexes. C, control.