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

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 Ahn, K. Articles by Crenshaw, E. B. Search for Related Content PubMed PubMed Citation Articles by Ahn, K. Articles by Crenshaw, E. B., III

BMPR-IA signaling is required for the formation of the apical ectodermal ridge and dorsal-ventral patterning of the limb

¹ Department of Neuroscience, University of Pennsylvania Medical School, Philadelphia, PA, USA
² Department of Molecular Genetics, University of Texas MD Anderson Cancer Center, Houston, TX, USA
³ Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709, USA
⁴ Procter & Gamble Pharmaceuticals, Mason, OH, USA

View larger version (42K): [in a new window]   Fig. 1. Generation of a Cre-mediated knockout of the Bmpr gene. (A) The loxP/Cre system requires, at the minimum, two pedigrees of animals: an activator strain that directs the tissue-specific expression of the Cre recombinase gene (Transgenic Pedigree #1) and a responder strain, in which loxP sites (large arrowheads) have been introduced flanking critical regions, such as exon 2, of the Bmpr target gene (Transgenic Pedigree#2). Intercross matings of the two strains induces an intramolecular recombination event between the two loxP sites that excises the intervening sequences (exon 2 in this case). (B) The mating scheme used to generate mutant and normal littermates used in this study (the term ‘flox’ is used when either the floxP or floxP-neo allele could be used (see D), as both appear to function equivalently). (C) Southern blot analyses of tissues from an 18.5 dpc embryo demonstrate that the Brn4-Cre transgene efficiently mediates the rearrangement of the floxP-neo allele of the Bmpr gene (see D for the structure of this allele). In tissues derived from the neural tube, such as spinal cord, hindbrain and forebrain, efficient rearrangement of the floxP-neo allele results in a conversion of the 6.3 kb fragment to a 2.2 kb fragment; the null allele yields a 4.0 kb fragment. The small amount of Cre-mediated rearrangement in the limbs results from the ectodermally derived cells in the limbs that express the Brn4-Cre transgene (see Fig. 2); the majority of the limb derives from mesenchymal tissue that does not express the Brn4-Cre transgene. Labeled size standard (lane 1) is 1 kb DNA Ladder (Life Technologies); sizes of hybridized bands (kb) are given (left). (D) To introduce loxP sites (arrowheads) into the first and second intron, a targeting vector was engineered to contain one loxP site in the first intron, and a neor gene flanked by loxP sites in the second intron; successful targeting of this construct resulted in the floxP-neo allele depicted in this panel. The residual neor gene in the second intron of the floxP-neo allele apparently did not interfere with the function of the gene, as no discernible phenotype was detected in mice homozygous for this allele. An allele in which the neor gene was specifically removed (floxP) was generated by partial excision of the locus with Cre recombinase (Y. M., M. C. H. and R. B., unpublished). No differences in phenotype were observed whether we used the floxP-neo or the floxP allele. Box (red) above exon 3 of modified alleles corresponds to H23 probe used to genotype Bmpr pedigrees. Nh, NheI; S, SacI.

View larger version (81K): [in a new window]   Fig. 2. ROSA reporter analyses demonstrated that Cre-mediated recombination occurred in the limb ectoderm prior to hindlimb bud formation, but after the forelimb bud had begun to form. Cre-mediated rearrangement of the ROSA reporter results in activation of lacZ expression and elucidates the temporal and spatial domain of ectopic Brn4-Cre gene expression in the embryonic limb. (A) This panel depicts a 9.75 dpc embryo that is doubly transgenic for the ROSA reporter and the Brn4-Cre transgene. The initial expression of the ROSA reporter was detected in ventrolateral ectoderm at a time in which the forelimb bud had begun to form (white arrowhead), but before initial hindlimb bud formation, which occurred several hours later. Arrows designate major axes: D, dorsal; V, ventral; R, rostral; C, caudal. (B) Vibratome section (100 µm) of a 10.5 dpc hindlimb demonstrated that Cre-mediated lacZ expression was restricted to the ectoderm of the limb. Cre-mediated lacZ expression was highest in the AER, although significant levels of expression were found in the ventral ectoderm. Limited induction of expression occurred in the dorsal limb ectoderm. Arrows designate major axes: D, dorsal; V, ventral; Di, distal; P, proximal (C-E). These panels demonstrate Cre-mediated expression of lacZ from the earliest stages of hindlimb formation in embryos sacrificed at 10.0 dpc. As the hindlimb bud grows, expression becomes particularly high in the pre-AER region at the distal tip of the limb. The hindlimbs depicted represent the typical variability in age of embryos sacrificed at 10.0 dpc; hindlimbs were ordered progressively based upon limb bud size. (F) By 12.0 dpc, the hindlimb expresses the activated lacZ throughout much of the dorsal ectoderm, as well as the ventral ectoderm and AER. (G) Both the extent and timing of ROSA reporter activation are different in the forelimb. At 10.25 dpc, the forelimb attains a more advanced limb bud stage than the hindlimb, but the degree of ROSA reporter activation is reduced when compared with earlier limb bud stages of the hindlimb (compare with E). (H) At 12.0 dpc, the degree of ROSA reporter is reduced compared to the hindlimb (compare with F). However, there is considerable induction of the ROSA reporter gene in the AER at this stage of embryogenesis.

View larger version (130K): [in a new window]   Fig. 3. Phospho-SMAD1 immunolabeling and Msx2 gene expression demonstrate loss of BMPR-IA signaling between 9.75 and 10.0 dpc in limb ectoderm. (A) Transverse section of a 9.75 dpc normal embryo demonstrating phospho-SMAD1 immunolabeling in the lateral mesoderm (lm) and the overlying ectoderm in the hindlimb field. Strong immunolabeling is detected on the ventral side of the coelom (co), and an overall gradient of labeling is detected with the highest labeling in the most ventral region of the embryo. (B) A higher magnification view of A. (C) Phospho-SMAD1 labeling is detected preferentially in the dorsal neural tube of normal embryos adjacent to the roofplate, which is a rich source of BMP factors. (D) Mutant embryo section corresponding to the normal embryos shown in A. (E) A higher magnification view of D. At this stage, only an occasional ectodermal cell shows reduced levels of phospho-SMAD1 immunolabeling, as indicated by the arrowhead. (F) Phospho-SMAD1 labeling is detected preferentially in the dorsal neural tube of mutant embryos, suggesting functional redundancy of BMP receptor function in the neural tube of these embryos. (G) Transverse section through hindlimb bud of a 10.0 dpc normal embryo demonstrating phospho-SMAD1 immunolabeling in the lateral mesoderm and the overlying ectoderm. Inset demonstrates that phospho-SMAD1 immunolabeling is low or undetectable in the dorsal ectoderm (see arrowhead). (H) Higher magnification view of G, showing the predominantly nuclear immunolabeling. (I) Phospho-SMAD1 immunolabeling in dorsal neural tube of 10.0 dpc normal embryo. (J) Phospho-SMAD1 labeling in a section of mutant embryo demonstrates that immunolabeling is not detected in most ectoderm cells, but robust immunolabeling is detected in the underlying mesoderm. Section comparable with that shown in G. (K) Higher magnification view of J demonstrates that only an occasional cell demonstrates phospho-SMAD1 immunolabeling in the mutant ectoderm (arrow), whereas the vast majority of ectodermal cells are not immunolabeled at 10.0 dpc (arrowhead). (L)Phospho-SMAD1 immunolabeling in the dorsal neural tube of mutant 10.0 dpc embryo. (M) Phospho-SMAD1 immunolabeling is detected in neural retina of 9.75 dpc eye. (N) Phospho-SMAD1 immunolabeling in dorsal hindbrain of 9.75 dpc embryo. (O) In situ hybridization demonstrates that expression of Msx2 at 10.0 dpc is detected in ventral hindlimb (bracket, transverse plane of section). Arrow indicates the border between lateral and paraxial mesoderm; arrowhead indicates Msx2 expression in the dorsal midline. (P) Expression of Msx2 in the mutant is abolished in the ventral hindlimb indicating that BMP signaling has been abrogated (bracket corresponds to region that normally expresses Msx2; arrowhead indicates Msx2 expression in the dorsal midline).

View larger version (84K): [in a new window]   Fig. 4. Cre-mediated inactivation of the Bmpr gene in the limb resulted in severe abnormalities in the hindlimb and more subtle defects in the forelimb. Although variability existed in the hindlimb phenotype, it was typically quite severe. (A) The most severe phenotype resulted in complete agenesis of the hindlimb in approximately one-fifth of the mutants (8/42 hindlimbs), as depicted in this P0 neonate (arrow). (B) Forelimb development was comparatively normal, although malformations, including partial polydactyly (arrow) and dysplastic digits (data not shown), were detected. (C) A representative malformation of the hindlimb, including a partial duplication of the distal region of the digit (arrowhead) and hematoma at the tip of the digits (arrow). The proximal skeletal elements of the digit (metatarsals and proximal phalanges) were typically reduced in number with the majority of the hindlimbs containing two (11/42) or three (13/42) digits. (D) A cleared whole-mount preparation of the forelimb of a Bmpr mutant was stained with Alizarin Red and Alcian Blue to visualize the bone structure. This panel demonstrates a partial polydactylous digit composed of a distal phalange, including the nail. (E) Forelimb from opposite side of animal depicted in D. (F) A mutant skeletal preparation demonstrates a reduction in digit number associated with syndactyly and a partial duplication (arrowhead) of the digits more distally. (G) Skeletal preparation of a rare polydactylous hindlimb demonstrating four proximal metatarsals, but seven distal phalanges. (H) A transverse section through a normal hindlimb at the level of the digits reveals the typical dorsal/ventral organization of the musculature and tendons. Arrows indicate the flexor digitorum profundus tendon in two of the digits. (I) A similar section through a mutant hindlimb demonstrated a double dorsal phenotype. The flexor digitorum profundus tendon did not display the prominent phenotype observed in the normal animal (asterisks), and the musculature formed an overall mirror-image symmetry unlike the obviously polarized structure of the normal animal. The hindlimb depicted in this panel is comparatively normal when compared to the distribution of mutant hindlimb phenotypes observed in the mutants. However, this mutant hindlimb still demonstrated a partial duplication of the digit, which is displaced ventrally (arrowhead) compared with the other digits. (J) Lateral view of normal hindlimb skeletal prep from a normal P10 mouse, Arrow indicates the sesamoid process, which is a ventral structure. The inset is a higher magnification view of the sesamoid process. (K) Lateral view of mutant P10 hindlimb demonstrating lack of sesamoid process. (L) A higher magnification view of normal digit transverse section indicated by right arrow in H. (M) A higher magnification view of the mutant digit for comparison to L (area indicated by right asterisk in I).

View larger version (86K): [in a new window]   Fig. 5. BMPR-IA is required for apical ectodermal ridge formation. (A-D) Fgf8 expression was detected in the pre-AER region before overt hindlimb bud development in normal embryos (A; 10.0 dpc), and is a molecular marker for the apical ectodermal ridge at later stages of hindlimb formation (C; 11.5 dpc). Fgf8 expression was detected in very few cells at 10.0 dpc in most mutant embryos examined (B). Presumably incomplete penetrance of the Bmpr conditional knockout allows the expression of Fgf8 in a small number of cells (arrow). Expression was detected later in hindlimb embryogenesis, but the levels of expression are variable (D-F). (C-F) Hindlimbs (11.5 dpc) double labeled for Fgf8 expression in the AER (arrowhead) and Shh expression in the ZPA (arrow). (C) The normal pattern of Fgf8 and Shh expression. Expression of Fgf8 and Shh varies in the mutant from no expression detected (D; Category ‘0’ in Table 1) to a majority of the pattern detected in normal embryos (F; Category ‘3’ in Table 1). A focal patch of Fgf8 expression is depicted in F (small arrow). Focal patches were detected that do not lie at the distal tip of the hindlimb, but they were not consistently deflected in either a dorsal or ventral direction. Interestingly, we did not see a ventral extension of AER gene expression, as seen with the En1 knockouts (Loomis et al., 1998). (G) At 10.5 dpc, Bmp4 expression is detected in both the ZPA (arrow) and the AER (arrowhead) of normal embryos. (H) Bmp4 gene expression is variable in the AER of mutant embryos and absent in the embryo depicted. Bmp4 expression is consistently detected in the ZPA of 10.5 dpc mutants.

View larger version (85K): [in a new window]   Fig. 6. BMPR-IA is required for dorsal/ventral patterning of the hindlimb. (A,C,E,G) Wild type; (B,D,F,H) mutant. (A-D) Lmx1b expression was restricted to dorsal mesoderm in normal 11.5 dpc hindlimb (A). However, the mutant hindlimbs demonstrated a double dorsal phenotype with Lmx1b expression being detected in ventral mesoderm (B; arrowhead). Lmx1b expression in the mutant forelimb (D) was indistinguishable from the control forelimb (C). (E) Wnt7a expression was normally restricted to the dorsal ectoderm of the limb field/bud in a normal 10.25 dpc embryo. (F) Wnt7a expression was expanded into the ventral region of the limb field/bud in the Bmpr conditional mutants. (G) En1 expression was restricted to the ventral region of the limb field/bud in a normal 10.25 dpc embryo, which is necessary to suppress Wnt7a expression in this region. (H) En1 expression was almost completely lost in the Bmpr conditional mutant. Arrow indicates region where En1 gene expression should normally exist.

View larger version (110K): [in a new window]   Fig. 7. Expression pattern of gene products that are upstream and downstream of BMPR-IA signaling during early hindlimb development. (A-C) Expression of BMPR-IA ligands, Bmp4 and Bmp7, and their relationship to the domain of Msx2 expression in normal 9.75 dpc hindlimbs, which is just before limb bud formation. (A) High levels of Bmp4 expression are detected in tissues of the ventral part of the embryo, including the lateral mesoderm from which the limb mesenchyme develops. (B) Bmp7 expression is highest in the most ventral region of the embryo in the outermost cell layers. Bmp7 expression demonstrates a diminishing gradient of expression from ventral to dorsal, and its expression domain ends at the junction of lateral and intermediate mesoderm (indicated by an arrow in C). (C) The Msx2 expression pattern is similar to that of Bmp7, except the dorsal border of expression (arrowhead) does not extend as far dorsomedially (difference between arrow and arrowhead). (D-G) Expression of BMPR-IA ligands, Bmp4 (D) and Bmp7 (F), and their relationship to the domain of Fgf8 (E) and Msx2 (G) expression in normal 10.0 dpc hindlimb buds. (D,F) The expression of both Bmp7 and Bmp4 have been significantly downregulated when compared with their 9.75 dpc expression domain. Furthermore, the Bmp4 expression domain has been restricted to the pre-AER region as indicated by the expression of pre-AER marker, Fgf8, in a serial section (E).

View larger version (25K): [in a new window]   Fig. 8. The hypothesized roles of regulatory genes during limb development. (A) The expression patterns of genes that regulate DV patterning of the limb ectoderm. This schematic depicts a transverse section of the limb bud with proximal to the left, distal to the right ending in the AER, and dorsal towards the top. (B) Model for the genetic pathway regulating dorsal/ventral patterning of the limbs. As shown in this study, Bmpr signaling in the limb ectoderm is required for En1 gene expression (red in A) in the ventral ectoderm. Previous studies have demonstrated a role for En1 in suppressing the expression of Wnt7a (yellow) in ventral ectoderm (Cygan et al., 1997; Loomis et al., 1996; Loomis et al., 1998), thereby restricting its expression to the dorsal ectoderm. Wnt7a has been shown to induce the expression of Lmx1b (blue) in the mesoderm underlying the dorsal ectoderm (Cygan et al., 1997; Loomis et al., 1998; Riddle et al., 1995). Figure adapted and modified from Johnson and Tabin (Johnson and Tabin, 1997). (C) Hypothesized molecular mechanisms that dictate DV patterning in the presumptive limb region before limb bud formation and the presumptive fate maps of limb ectoderm. Ventral embryonic regions that express Bmp4 are depicted in various shades of blue with lateral mesoderm (lm) depicted as the darkest shade of blue, and intermediate mesoderm (im) with lighter blue shading. Somitic mesoderm (so) is depicted as red, in those regions that express noggin, and as orange in the remainder of the somite. In the ectoderm, purple denotes regions that correspond to chick ectoderm previously fate mapped to form dorsal limb ectoderm (Altabef et al., 1997; Michaud et al., 1997). Green denotes regions fated to contribute to ventral ectoderm. The arrowhead denotes the boundary between the dorsal and ventral limb compartments (Altabef et al., 1997; Michaud et al., 1997). Our model hypothesizes that the expression of BMPs in the ventral and lateral mesoderm (blue) induce the ventral identity of the overlying ectoderm (green). This induction is blocked by noggin expression in the somite, which previously has been shown to express noggin (Capdevila and Johnson, 1998; Hirsinger et al., 1997; Marcelle et al., 1997; McMahon et al., 1998; Reshef et al., 1998; Tonegawa and Takahashi, 1998). In addition, this region has been demonstrated to induce the overlying ectoderm to form dorsal limb ectoderm (Michaud et al., 1997). Our model proposes that the inhibition of BMP induction in ectoderm overlying the most medial parts of the lateral mesoderm (where purple ectoderm overlies blue mesoderm) is accomplished by noggin. The region of ventral ectoderm hypothesized in this model corresponds to the domain of Msx2 gene expression, which is lost in the Bmpr mutant. (D) Schematic illustration of morphogenetic movements of ectoderm to form dorsal limb ectoderm (purple), AER (yellow) and ventral ectoderm (green).