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

First published online October 27, 2004
doi: 10.1242/10.1242/dev.01444

This Article Summary Full Text Full Text (PDF) Supplementary Material 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 Crump, J. G. Articles by Kimmel, C. B. Search for Related Content PubMed PubMed Citation Articles by Crump, J. G. Articles by Kimmel, C. B. Social Bookmarking                         What's this?

An essential role for Fgfs in endodermal pouch formation influences later craniofacial skeletal patterning

¹ Institute of Neuroscience, 1254 University of Oregon, Eugene, OR 97403-1254, USA
² Program in Gene Function and Expression, University of Massachusetts Medical School, Worcester, MA 01605, USA
³ Laboratory of Molecular Genetics, NICHD/NIH, Bethesda, MD 20892, USA

View larger version (48K): [in a new window]   Fig. 1. Fgf8 and Fgf3 have redundant functions in the formation of pharyngeal pouches and cartilages. (A-D) Confocal micrographs are merged, lateral views of cranial NCC (green: anti-GFP antibody) and endodermal pouches (red: Zn8 antibody) at 34 hpf; A'-D' show just the red channel. (A,A') In wild-type fli1-GFP animals, the NCC-containing pharyngeal arches are numbered 1-7 (A) and the pouches are numbered p1-p5 (A'). A few hours later, the sixth pouch will form and arches 6 and 7 will separate to form the final arrangement of seven arches. (B,B') fgf8–; fli1-GFP animals have variable defects in pouch structure (arrowhead in B' denotes a misshapen first pouch). Whereas fgf3-MO; fli1-GFP animals have normal pouches (C,C'), fgf8–; fgf3-MO; fli1-GFP animals lack all pouches (D,D'), although pharyngeal endoderm is still present (white line in D'). The Zn8 antibody also recognizes cranial sensory ganglia (dots in A-C, A'-C'). (E-H) Ventral whole-mount views show Alcian-stained pharyngeal cartilages at 4 days. As shown for wild type (E), M and PQ cartilages are derived from the mandibular, or first, arch; CH and HS are hyoid, or second, arch cartilages, and CB1-5 cartilages are formed from the five most posterior branchial arches. fgf8– animals have relatively mild defects in pharyngeal cartilages (F), and in fgf3-MO animals CB cartilages are lost and hyoid cartilages are misshapen (G). However, in fgf8–; fgf3-MO animals, nearly all CB and hyoid cartilages are absent and mandibular cartilages are reduced in size (H). The inset to H is a flat-mount preparation of fgf8–; fgf3-MO cartilages showing that, although reduced in size, M and PQ cartilages retain their distinctive shapes. In E-H, asterisks denote the position of the midline neurocranium that is still present in fgf8–; fgf3-MO animals. Anterior is to the left in all panels. M, Meckel's; PQ, palatoquadrate; CH, ceratohyal; HS, hyosymplectic; CB, ceratobranchial. Scale bars: 50 µm in A-D; 100 µm in E-H.

View larger version (72K): [in a new window]   Fig. 2. Correlated first pouch and hyoid cartilage defects in animals reduced for Fgf8. Confocal projections of fli1-GFP-labeled pharyngeal arches in living wild-type (A) and fgf8-MO (C,E,G) animals at 28 hpf. By this stage of wild-type development, the mandibular (1), hyoid (2), and three branchial arches (3, 4, 5-7) have formed. Pouches are labeled p1-p4 (A), and white arrowheads mark the positions of the first pouch. In fgf8-MO; fli1-GFP animals, variable phenotypes include shape changes in the first pouch (C,E), ectopic pouches (arrow in G), and reductions of more posterior pouches (C,E,G). (I) Confocal micrograph of a fixed fgf8–; fli1-GFP animal with a similar ectopic pouch phenotype (arrow) to the fgf8-MO; fli1-GFP animal in G; Zn8 staining (red) confirms that the non-fli1-GFP-expressing region is probably an ectopic endodermal pouch. (B,D,F,H,J) Flat-mount preparations of Alcian-stained mandibular and hyoid cartilages at 4 days. As labeled in the wild-type example (B), M and PQ are mandibular (1) and CH and HS are hyoid (2) cartilages. (C and D, E and F, G and H) Paired images of individual animals imaged live for fli1-GFP early and subsequently stained for cartilage. Variable hyoid cartilage defects (D,F,H) are correlated with earlier first pouch defects (C,E,G) in individual fgf8-MO; fli1-GFP animals. In H, the black arrowhead marks an apparent ectopic hyoid cartilage that correlates with the ectopic pouch in G. (J) Similar ectopic cartilages (black arrowhead) were seen in some fgf8–; fli1-GFP animals. Anterior is to the left and dorsal is up. M, Meckel's; PQ, palatoquadrate; CH, ceratohyal; HS, hyosymplectic. Scale bar: 50 µm.

View larger version (68K): [in a new window]   Fig. 3. Fgf8 is required for segmentation of the branchial arches. Bilateral flat-mount dissections of Alcian-stained pharyngeal cartilages at 4 days. As shown for wild type (A), M and PQ are mandibular (1) cartilages, CH and HS are hyoid (2) cartilages, and CB1-5 cartilages are formed from the five most posterior branchial arches (3-7). Note the teeth (*) on the CB5 cartilages. fgf8– animals have variable CB cartilage defects, which include reduced CB number (only 4 CBs per side in B), incompletely formed CB cartilages (arrowhead in C), and fusions between adjacent cartilages (arrow in C). In a representative fgf8–; fli1-GFP clutch (n=172) there was an average of 3.9 CB cartilages per side; 6% of sides had fusions of adjacent cartilages and 2% had incomplete cartilages. (D-K) Time-lapse recordings of wild-type fli1-GFP; H2A.F/Z:GFP (D-G, and see Movie 1 in supplementary material) and fgf8-MO; fli1-GFP (H-K, and see Movie 2 in supplementary material) animals show the cellular basis of branchial arch segmentation. In wild-type animals, branchial arches form as pouches separate the branchial NCC mass into segments in an A–P wave of development. At the beginning of Movie 1 (5-somites, 12 hpf), H2A.F/Z:GFP labels the nuclei of NCC that are migrating ventrolaterally in two streams anterior to, and one stream posterior to, the developing ear. After fli1-GFP initiates in NCC of the pharyngeal arches, selected projections from Movies 1 and 2 show the subdivision of each successive branchial arch (arch 3 at 20 hpf: D,H; arch 4 at 28 hpf: E,I; arch 5 at 34 hpf: F,J; and arches 6 and 7 at 38 hpf: G,K). Pouches are labeled p1-p6, and white arrowheads in D-G indicate the developing vasculature that also expresses fli1-GFP. The white arrows in panels I-K and Movie 2 (in supplementary material) refer to arches 4 and 5, which fail to separate completely in this fgf8-MO animal. Similar cell behaviors were seen in three time-lapse recordings of wild-type animals, and variable defects were observed in four time-lapse recordings of fgf8-MO; fli1-GFP animals. Anterior is to the left in all panels. A-C are ventral views, and dorsal is up and slightly to the right in D-K. M, Meckel's cartilage; PQ, palatoquadrate; CH, ceratohyal; HS, hyosymplectic; CB, ceratobranchial. Scale bar: 50 µm.

View larger version (49K): [in a new window]   Fig. 4. Fgf signaling is required during early somite stages for first pouch and hyoid cartilage development. (A-C) Confocal micrographs show Zn8-labeled pharyngeal pouches (red) and GFP-labeled NCC (green) in fli1-GFP animals at 34 hpf. Cranial sensory ganglia (dots) also stain with Zn8. In the wild-type animal (A), an arrowhead marks the first pouch. (B) The first pouch is variably absent (arrowhead) in fli1-GFP animals upon treatment with the Fgf receptor antagonist SU5402 from 10-14 hpf. The absence of the first pouch is selective, as more posterior pouches form normally. (C) Treatment with SU5402 for 1-hour periods starting from 9-13 hpf (a 10.5-11.5 hpf treatment is shown) produce subtler shape changes of the first pouch (arrowhead). (D,E) Flat-mount preparations of Alcian-stained pharyngeal cartilages at 4 days. In wild-type (D), mandibular M and PQ, hyoid CH and HS, and branchial CB1-5 cartilages are labeled. In those animals in which 10-14 hpf SU5402 treatment caused losses of the first pouch early, the HS cartilage was selectively absent later (E). Although M, PQ and CH cartilages are reduced in size, posterior CB cartilages are relatively unaffected. (F) Whereas later treatments with SU5402 (a 20-21 hpf treatment is shown) do not affect first pouch development (arrowhead), they do occasionally disrupt the formation of more posterior pouches (arrow shows an unsegmented branchial NCC mass). (G) Quantitation of first pouch defects after 4-hour treatments with SU5402. The percentages of animals with first pouch losses, in black, and misshapen first pouches, in gray, are plotted. n6-10 hpf=49, n10-14 hpf=99, n14-18 hpf=21. First pouch loss after 10-14 hpf treatment is statistically significant using Tukey HSD test. (H) The percentage of fli1-GFP animals having first pouch defects (primarily shape changes) plotted against the start time of 1 hour treatments with SU5402. n5.5 hpf=17, n7 hpf=14, n8 hpf=24, n9 hpf=14, n10.5 hpf=21, n11 hpf=18, n12 hpf=26, n13 hpf=26, n14 hpf=22, n16 hpf=26, n20 hpf=24, n24 hpf=31. The period of strongest effect is from 9 hpf (90% epiboly) to 13 hpf (8-somites). In order to assess the efficiency of inhibition of Fgf signaling, and the recovery after washout, we fixed sibling controls and examined pea3 expression, a downstream effector of Fgf signaling, at 0 and 4 hours after washout. We know that SU5402 is at least partially being washed out as omission of the washout step leads to severe necrosis of animals. For 4-hour incubation experiments, the levels of pea3 in individual animals, relative to those in similarly staged untreated controls, were as follows: 6-10 hpf, 6/8 reduced at 10 hpf, 11/13 reduced and 2/13 absent at 14 hpf; 10-14 hpf, 5/13 reduced and 8/13 absent at 14 hpf, 6/13 reduced and 7/13 absent at 18 hpf; 14-18 hpf, 5/12 reduced and 7/12 absent at 18 hpf. As pea3 levels were similarly reduced at 18 hpf in 10-14 hpf and 14-18 hpf treatments, yet only 10-14 hpf treatments cause first pouch defects, we conclude that Fgf signaling is required from 10-14 hpf for first pouch development. However, these experiments do not exclude additional requirements for Fgf signaling at later times. Anterior is to the left and dorsal is up. M, Meckel's; PQ, palatoquadrate; CH, ceratohyal; HS, hyosymplectic; SU, SU5402. Scale bar: 50 µm.

View larger version (79K): [in a new window]   Fig. 5. Fgfs are required in neural and mesodermal tissues for first pouch formation. Labeled wild-type tissues (red: A''-E'') were transplanted into fgf8–; fgf3-MO; fli1-GFP animals from 4-6 hpf, and GFP-expressing NCC (green: A'-E') were examined at 34 hpf for rescue of pharyngeal arch structure, a proxy for pouch structure. A'-E' and A''-E'' are confocal projections and are merged in A-E. A'''-E''' are individual confocal sections from A-E and include the Nomarski channel. (A-A''') As transplantations generally contribute donor tissue unilaterally, we used contralateral non-recipient sides of fgf8–; fgf3-MO; fli1-GFP animals as negative controls for rescue (the red staining in A,A'' represents a comparatively small amount of donor tissue that has crossed the midline). In control sides, only mandibular (1) and a few unidentified (?) NCC are evident (A'), and the ear is missing (A'''). (B-B''') Wild-type endoderm (e) fails to rescue pharyngeal arch structure (B') and the ear (B'''). As seen in B'', wild-type endoderm does not segment into pouches in fgf8–; fgf3-MO; fli1-GFP hosts. Wild-type mesoderm (m, C-C''') or wild-type neural tissue (n, D-D''') only partially rescues pharyngeal arch structure in a fraction of animals. In the non-rescued mesodermal example shown (C'), pharyngeal (1 +?) NCC remain unsegmented, revealing a lack of pouches. The neural example shown (D') represents what we scored as partial rescue of arch structure. There is an increase in the amount and organization of NCC, but they are not segmented into ordered pharyngeal arches as in wild-type animals. The lack of rescue of arch structure by wild-type neural tissue is striking, as other structures such as the MHB blood vessel (asterisk in D'), the neural flexure (white arrowhead in D''), and the ear (black arrowhead in D''') are rescued by neural tissue. By contrast, wild-type mesoderm did not rescue the ear (C'''). (E-E''') Both wild-type mesoderm and neural tissue are required together to completely rescue pharyngeal arch structure, and hence pouches, in fgf8–; fgf3-MO; fli1-GFP animals. In this example, a morphologically normal first pouch (p1, arrow) and mandibular (1) and hyoid (2) arches are clearly seen (E'). Some of the more posterior pouches (E', note the segmentation of the branchial (3+) NCC mass) and the ear (black arrowhead in E''') are rescued as well. The identification of mesoderm in the transplants was based on the lack of colocalization with the neural crest marker fli1-GFP in confocal sections, and in this example by the characteristic morphology of the mesodermal cores of the pharyngeal arches (Kimmel et al., 2001) (F) Quantitation of pharyngeal arch rescue by wild-type tissues is plotted as percentage of host sides with complete (black) or partial (gray) rescue of arch structure. nendoderm=34, nmesoderm=11, nneural=30, nneural+mesoderm=6. Complete rescue by neural and mesodermal tissue (neur. + meso.) and partial rescue by mesoderm or neural tissue were statistically significant using Tukey HSD test. In addition, no rescue was seen by wild-type neural crest, a tissue that does not express either Fgf8 or Fgf3 (data not shown). Anterior is to the left and dorsal is up. Scale bar: 50 µm.

View larger version (60K): [in a new window]   Fig. 6. Pharyngeal endoderm and cranial NCC defects in animals lacking Fgf8 and Fgf3. nkx2.7 (A-D) and axial (E-H) label pharyngeal endoderm during early pouch morphogenesis stages (18 hpf). nkx2.7 and axial are in blue, and, in E-H, krox-20 in red labels R3 and R5. In wild-type animals (A,E), the first pouch (p1: arrows) has formed anterior to R3, and a more posterior endodermal mass that will give rise to the remaining pouches (black lines) is situated adjacent to R4-R6 territory. The first pouch is variably lost in fgf8– animals (asterisk in B, question mark in F). Whereas pharyngeal endoderm develops normally in fgf3-MO animals (C,G), in fgf8–; fgf3-MO animals (D,H) pharyngeal endoderm is present as a single anterior mass (black line) and no pouches are evident. (I-P) dlx2, in blue, labels cranial NCC; in red (I-L), pax2a labels the MHB and krox-20 labels R3 and R5. (I) In 18 hpf wild-type animals, mandibular (1), hyoid (2), and branchial (3) NCC streams give rise to seven pharyngeal arches. (M) At 33 hpf, the third branchial stream has generated arches 3-5 and arches 6 and 7 have yet to separate. In fgf8– (J,N) and fgf3-MO (K,O) animals, the migration and coalescence of NCC to form the pharyngeal arches is largely normal. In fgf8–; fgf3-MO animals, the mandibular (1) stream is disorganized and hyoid and branchial streams are fused together (2/3) at 18 hpf (L). By 33 hpf (P), nearly all hyoid and branchial NCC are absent, and mandibular (1) NCC are present but reduced. Anterior is to the left in all panels. A-D are dorsal views, and E-P are lateral views. R3 and R5, rhombomeres 3 and 5; MHB, midbrain-hindbrain boundary; p1, first pouch. Scale bar: 50 µm.

View larger version (112K): [in a new window]   Fig. 7. Pharyngeal pouches form by an Fgf-dependent lateral migration of endodermal cells. (A-E) Representative still images from a time-lapse confocal recording of pharyngeal pouch and arch development in wild-type animals (see Movies 3, 4 in supplementary material). Pharyngeal endoderm has been labeled in red (A'-E'; merged with GFP in A-E) by transplanting TAR* endoderm into a fli1-GFP; H2A.F/Z:GFP host at 4 hpf; this technique leads to mosaic labeling of endoderm in the host animal. In green, H2A.F/Z:GFP allows the nuclei of every cell to be seen, and fli1-GFP marks NCC of the pharyngeal arches. In wild-type development, endodermal cells are spread out in a monolayer over the surface of the yolk at 10 hpf (A,A'). Concomitant with medial ectodermal movements to form the neural keel, endodermal cells migrate medially and begin to aggregate (14 hpf: B,B'). Shortly after medial migration, individual endodermal cells destined to become the first pouch (arrow in C') then migrate back out laterally (C: 18 hpf). As seen in Movie 4, pouch endodermal cells extend cytoplasmic processes laterally during migration. At the same time, cranial NCC begin to condense and express fli1-GFP. By 22 hpf, the first two pouches (D': p1,p2) have formed and interdigitate three NCC-containing pharyngeal arches (D: 1-3). Although in this example most second pouch cells are not labeled in red, their development can still be observed based on transient expression of fli1-GFP (asterisks in C,D). Also, the characteristic flexure of the neural keel near the MHB is visible by H2A.F/Z:GFP (arrowhead in D). At 30 hpf, three pouches (E': p1-p3) and four arches (E: 1-4) are well developed. (F-J) Representative still images from a time-lapse recording of pharyngeal development in an animal reduced for Fgf8 and Fgf3 (see Movies 5, 6 in supplementary material). Similar cell behaviors were seen in two separate recordings. Labeled endoderm (red) was generated by transplantation of wild-type TAR* endoderm into fgf8–; fgf3-MO; H2A.F/Z:GFP animals (see text for discussion of experimental rationale, including how wild-type and mutant endoderm probably behave similarly in a mutant host). In fgf8–; fgf3-MO animals, the generation (F,F') and medial migration (G,G') of pharyngeal endoderm is normal. However, by 18 hpf, the lateral migration of endodermal cells is delayed (H,H'). Also, the migration of endodermal cells is disorganized, with cytoplasmic processes not being oriented laterally as in wild-type animals (arrows in Movie 6 in supplementary material). By 22 hpf lateral endodermal cells form an extended anterior mass (white line in I'). A confirmation of the fgf8–; fgf3-MO phenotype is the lack of a neural flexure (arrowhead in I), increased cell death at the MHB, and the lack of an ear (data not shown). By 30 hpf, pharyngeal endoderm has not segmented into discrete pouches and remains a single anterior mass (white line in J'). Although animals did not carry the fli1-GFP transgene, reduced mandibular (1) and possibly hyoid (2?) arches are visible as condensations of H2A.F/Z:GFP-expressing cells. The views are dorsolateral with anterior to the left. Scale bar: 50 µm.

View larger version (24K): [in a new window]   Fig. 8. Fgf8 and Fgf3 as positional determinants of pharyngeal segmentation. Model of pharyngeal segmentation in wild type (A,B) and animals lacking Fgf8 and Fgf3 (C,D). This model is based on fgf8 expression (Reifers et al., 2000) and fgf3 expression (Maves et al., 2002) at 13 hpf (8-somites). A and C are lateral views with anterior to the left and dorsal up, and B and D are cross-sectional views at the level of R2. (A,B) In wild type, Fgf8 protein, dark blue, is produced in neural MHB-R2 and R4 domains and in the lateral mesoderm. Fgf3 protein, light blue, is co-produced with Fgf8 in the MHB and R4 (striped domains represent overlap). Mandibular (1) NCC (Hox negative: light green) are generated adjacent to MHB-R2 territory, whereas hyoid (2) and branchial (3) NCC (Hox positive: dark green) have their origins at R4 and R6-R7 axial levels, respectively. Likewise, the first (p1) endodermal pouch (Hox negative: red) develops ventrolateral to R2, and the second (p2) and more posterior (p3+) pouches (Hox positive: wine) form ventrolateral to R4 and R6-R7, respectively. The ear (black circle with two dots) develops adjacent to R5. (B) A cross-sectional view shows that during lateral migration pouch endodermal cells are in close proximity to Fgf-expressing ventral neural keel and lateral mesoderm. Pharyngeal pouches would form where Fgf expression in the hindbrain coincides with Fgf8 expression in the underlying lateral mesoderm. (C,D) In fgf8–; fgf3-MO animals, NCC and pharyngeal endoderm are generated but subsequent morphogenesis is defective. Pharyngeal endoderm remains unsegmented and hyoid and branchial NCC streams fuse. The structure of the hindbrain is also defective in animals lacking Fgf8 and Fgf3. MHB, R1, R5 and R6 regions fail to develop, and R2 and R3 are reduced in size (Brand et al., 1996; Walshe et al., 2002; Maves et al., 2002; Reifers et al., 1998). As Fgfs are required in both neural and mesodermal tissues to promote the formation of pouches, Fgf signaling may link early neural and mesodermal patterning to segmentation of the pharyngeal endoderm. In a direct model, Fgfs from the lateral mesoderm and ventral hindbrain act as chemoattractants to promote the lateral migration of pouch endodermal cells (B). In animals lacking Fgf8 and Fgf3, pouch endodermal cells would fail to get the appropriate cues to migrate laterally (D). Alternatively, in an indirect model, Fgfs function to regulate the structure of, and gene expression in, the hindbrain and lateral mesoderm. In the absence of Fgf signaling, guidance cues for pouch endodermal migration would be reduced or absent. LM, lateral mesoderm; MHB, midbrain–hindbrain boundary; R, rhombomere.

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

Twitter