XB-ART-12186Development November 1, 1999; 126 (21): 4715-28.
A two-step mechanism generates the spacing pattern of the ciliated cells in the skin of Xenopus embryos.
The skin of Xenopus embryos contains a population of specialized ciliated cells that are distributed in an evenly spaced pattern. Here we describe two successive steps that govern the differentiation and the generation of the spacing pattern of these ciliated cells. The first step occurs in the inner or sensorial layer of the non-neural ectoderm where a subset of cells are chosen to differentiate into ciliated-cell precursors. This choice is under the control of lateral inhibition mediated by a Suppressor of Hairless-dependent Notch signaling pathway, in which X-Delta-1 is the putative ligand driving the selection process, and a new Enhancer-of-Split-related gene is an epidermal target of Notch signaling. Because nascent ciliated-cell precursors prevent neighboring cells from taking on the same fate, a scattered pattern of these precursors is generated within the deep layer of the non-neural ectoderm. Ciliated-cell precursors then intercalate into the outer layer of cells in the epidermis. We show that the intercalation event acts as a second step to regulate the spacing of the mature ciliated cells. We propose that the differentiation of the ciliated cells is not only regulated by Notch-mediated lateral inhibition, but is also an example where differentiation is coupled to the movement of cells from one cell layer to another.
PubMed ID: 10518489
Article link: Development
Genes referenced: dll1 eno1 esr-5 hes5.1 hes5.10 hes5.2 myh3 notch1 rbpj rpa1 tub tuba4b tubb2b tyro3
Article Images: [+] show captions
|Fig. 1. Ciliated cells are arranged in a regularly spaced pattern and are marked by the expression of an a-tubulin gene. (A,B) Scanning electromicrographs of the skin of a Xenopus embryo (stage 25) taken at a magnification of 200´ (A) and 750´ (B). Note the spacing pattern of cells with tufts of cilia in the outer layer of Xenopus skin (B, Arrow). (C) Xenopus embryo (stage 22) stained in whole mount using HRP immunohistochemistry with a mouse monoclonal antibody directed against acetylated a-tubulin (6-11B1). (D) Xenopus embryo (stage 25) double-labeled by in situ hybridization with the a-tubulin probe (light blue) followed by 6- 11B1 antibody staining (brown). (E) Transverse section of the embryo shown in D, confirming that the same cells stain with both the a-tubulin probe and the 6-11B1 antibody (arrows).|
|Fig. 2. Expression of a-tubulin in the skin of Xenopus embryos at different developmental stages. Expression of a-tubulin detected by whole-mount in situ hybridization on Xenopus embryos at stage 13 (A-C), stage 16 (D-F) and stage 20 (G-I). Dorsal views (A,D,G) and lateral views (B,E,H) show that a- tubulin expression occurs in the non-neural ectoderm (nne) and not in the neural plate (np). Expression is also absent from the cement gland (cg). (C,F,I) Sections of whole-mount-stained embryos show that cells expressing a-tubulin are located in the inner layer (parenthesis or arrow labeled “i”) at stage 13, and in the outer layer (brackets labeled “o”) at stage 16 and older, indicating that the ciliated cells first arise in the inner layer before stage 13 and then subsequently move into the outer layer.|
|Fig. 3. X-Delta-1 is expressed in the ventral ectoderm in a punctate pattern that anticipates a-tubulin expression. (A-C) Xenopus embryo (stage 11) stained in whole mount for the expression of X-Delta-1 by in situ hybridization. View of the ventral side is shown in A, a lateral view in shown in B and a transverse section is shown in C. Note that X-Delta-1 is expressed by scattered cells within the inner layer of the ventral ectoderm (A,C), in addition to cells (B, arrowheads) that are localized to stripes within the neural plate where primary neurons will form (np). (D,E) a-tubulin expression is first detected by in situ hybridization at stage 12 in the ventral ectoderm (D). Lateral view shows that a-tubulin is expressed by scattered cells in the non-neural ectoderm (nne) but not in the neural plate (np). bl, blastopore.|
|Fig. 4. Notch signaling restricts the number of a-tubulin-expressing cells. Albino embryos were injected at 2-cell stage in one blastomere with (A-C) nLacZ RNA alone, (D-F) Notch-ICD and nLacZ RNAs, (G-I) X-Su(H)-DBM and nLacZ RNAs, or (J-L) X-Delta-1Stu and nLacZ RNAs. At stage 17-18 (neurulae), the embryos were fixed, stained in whole mount with X-gal (lightblue reaction product) and then labeled by in situ hybridization for the expression of a-tubulin (dark purple). Left panels show a lateral view of the uninjected side, the middle panels a lateral view of the injected side, and the right panels show a dorsal view. Note that Notch-ICD suppresses a-tubulin expression in 93% of injected embryos (57 embryos in four representative experiments) while the density of a-tubulin-expressing cells increases in embryos injected with X-Su(H)-DBM (55/64 injected embryos) or with XDelta- 1Stu (64/80 injected embryos) in three independent experiments.|
|Fig. 5. Lateral inhibition in the non-neural ectoderm feeds back on X-Delta-1 expression. (A-C) The ventral ectoderm of albinos embryos (stage 11) that were uninjected (A) or injected in one blastomere at the 2-cell stage with RNA encoding Notch-ICD (B) or X-Su(H)-DBM (C) along with nLacZ RNA as a tracer. Embryos were stained with X-Gal (sky blue reaction product) or for the expression of X-Delta-1 using whole-mount in situ hybridization (dark purple). Note that the density of X-Delta-1 expression increases when Notch signaling is blocked (C) and decreases when Notch signaling is increased (B). (D,E) Isolated ectoderm was assayed for the levels of a-tubulin (a-tub) and X-Delta-1 RNAs by RPA. Ectoderm was isolated at stage 9/10 from embryos that were uninjected (uninj.) or injected at 2-cell stage in both blastomeres with RNA encoding Notch-ICD or X-Su(H)-DBM. RNA was extracted at stage 18 or 12 and assayed simultaneously for the levels of a-tubulin and EF1-a RNA, or for X-Delta-1 and EF1-a RNA, respectively. Note that increased Notch signaling (Notch-ICD injection) decreases, while decreased Notch signaling (X-Su(H)-DBM injection) increases the levels of X-Delta-1 and a-tubulin when normalized to recovery of RNA using the levels of EF1-a RNA. RNA from whole uninjected embryos (w. emb.) was used as positive control.|
|Fig. 6. ESR-6e encodes a WRPW-bHLH protein expressed epidermally and activated by the Notch pathway. (A) ESR-6e encodes a WRPW-bHLH protein related to other Xenopus ESR genes that are activated by the Notch pathway. ESR-1 and ESR-7 are expressed in the developing nervous system, while ESR-5 is expressed in the paraxial mesoderm during segmentation. Percentages of similarity between the bHLH domains of ESR-6e and other ESR factors identified in Xenopus are given as well as the overall similarity for the full-length proteins. (B) Expression of ESR-6e RNA in stage 14 (neural plate) Xenopus embryos as detected by whole-mount in situ hybridization. Note that ESR-6e expression is mostly localised to the non-neural ectoderm as shown by the white dashed line marking the border of the neural plate. A high level of expression is also observed in the cement gland (white arrowhead) and a low level in the central neural plate (black arrowhead). (C) Total RNA extracted from embryos at the indicated stages was assayed by RPA for the levels of ESR-6e RNA and EF1-a RNA as a loading control. The levels of ESR-6e at different stages are plotted in a bar graph after normalizing to the levels of EF1-a RNA. EF1-a RNA is first transcribed at the mid-blastula transition, reaching steady levels at stage 12. Thus, the levels of ESR-6e RNA are likely to be overestimated after normalizing to the levels of EF1-a RNA at stage 8 and 10. (D) Notch-ICD RNA was coinjected with nLacZ RNA in one blastomere of 2-cell-stage embryos. At stage 14, the embryos were stained with X-gal (light blue) and for ESR-6e expression using in situ hybridization (dark blue). Note that Notch-ICD RNA induces high levels of ESR-6e expression but only in the non-neural ectoderm in 48 injected embryos on a total of 54. In this example, the Notch-ICD-injected area (circled by the black dashed line) extends into the neural plate but ESR-6e overexpression stays confined to the prospective epidermis (white dashed line) as in B. Note also that the embryo shown in D has been stained for less time than the one in B, as indicated by the reduced signal from the basal expression of ESR-6e. (E,F) Transverse sections of a Notch-ICD-injected embryo as in D. (G) RNase protection assay to measure the levels of ESR-6e RNA in ectoderm isolated from embryos injected with RNA encoding Notch-ICD or X-Su(H)-DBM as described in the legend to Figure 5. RNA was extracted at stage 11 from intact ectoderm (w. caps) or from the inner layer (inner) or outer layer (outer) after manual separation. Note that Notch-ICD induces higher levels of ESR-6e expression in both layers. X-Su(H)-DBM also increases the levels of ESR-6e in the inner layer but not in the outer layer.|
|Fig. 7. XESR-6e restricts the number of ciliated cell precursors. (A-D) One blastomere of 2-cell-stage albino embryos was injected either with 0.5 ng of ESR-6e RNA or 2 ng of ESR-6eÆb RNA along with nLacZ RNA as a tracer. At stage 19-20, the injected embryos were fixed and stained with X-gal (light blue) and a-tubulin expression by in situ hybridization (dark blue). Embryos are oriented with anterior to the left with injected sides shown in B and D and the uninjected sides shown in A and C. (A,B) Injection of ESR-6e RNA suppressed the formation of a- tubulin-expressing cells in 90% of the embryos that survived the injection and developed normally (23 embryos in two independent experiments) whereas (C,D) injecting RNA encoding a DNA-binding mutant, ESR-6eÆb, produced an increase in a-tubulin-expressing cells in 90% of 48 injected embryos. (E) RPA on ectoderm isolated from embryos injected with RNA encoding ESR-6e and ESR-6eÆb (0.5 and 2 ng respectively) as described in Fig. 5. RNA was extracted from stage 11 animal caps for assessing X-Delta-1 expression and at stage 16-19 for a-tubulin. Note that ESR-6e decreases the levels of both X-Delta-1 and a- tubulin RNA while ESR-6eÆb increases their levels. (F) RPA of RNA isolated from ectodermal caps of embryos injected with RNA encoding wild-type (0.5 ng) and basic domain mutant forms of ESR-6e and ESR-7 (2 ng) as described above. Note that injection of RNA encoding ESR- 6e or ESR-6eÆb leads to a decrease or increase, respectively, in the levels of a-tubulin RNA. Conversely, RNA encoding ESR-7 causes a decrease in the levels of a-tubulin RNA while RNA encoding ESR-7Æb has no apparent effect.|
|Fig. 8. Spacing of ciliated cells is imposed by intercalation. Notch signaling was blocked in the ectoderm of Xenopus embryos by injecting 2 ng of X-Su(H)-DBM RNA into two blastomeres of 4-cell-stage embryos, along with nLacZ RNA as a tracer. (A-C) At stage 17, a sample of embryos were fixed and the distribution and density of the a-tubulin-expressing cells was assessed by in situ hybridization with a-tubulin probe (dark blue) after detection of the tracer by X-gal staining (light blue). (C) A section through the embryo shown in A,B illustrates that blocking Notch signaling can result in several layers of a-tubulin-expressing cells. (B) The white dashed line indicates an area of the uninjected side of the embryo that has been populated by cells descending from the injected blastomeres as a result of cell mixing. (D-N) The remaining embryos were fixed at stage 25. In these embryos, the distribution of injected RNA was measured by staining with Magenta-gal (magenta), the density of a-tubulin-expressing cells was measured by in situ hybridization using just BCIP for detection (light blue), and the status of differentiated ciliated cells was measured by staining with the 6-11B1 antibody using HRP immunohistochemistry (brown). Note that BCIP detection gives rise to a higher level of background (diffuse sky blue staining) compared to NBT-BCIP detection. (D-H) Example of an embryo with a relatively mild increase in the number of a-tubulin-expressing cells. Lateral views are shown with anterior to the left. (F,G) Magnified views of the area framed in D and E, respectively, showing that all the a-tubulin-expressing cells have differentiated since they are also stained with 6- 11B1. The density of the differentiated ciliated cells approximately doubled on the injected side without affecting the morphology of the skin. (H) Transverse section through the embryo shown in D and E showing that the extra a-tubulin-expressing cells have reached the surface, resulting in a spacing pattern twice as dense on the injected side: seventeen ciliated cells can be counted on the injected side, versus nine on the uninjected side. nc, notochord; sm, somitic mesoderm; lm, lateral mesoderm; epi, epiderm. (I-N) An embryo with a severe overproduction of a- tubulin-expressing cells. (I,J,L,M) Lateral views are shown anterior to the left. (L,M) Magnified views of the area framed in I and J. Note the continuous pattern of a-tubulin staining on the injected side (compare I to J and L to M). The arrow points to a region where 6-11B1 Ab staining can be distinguished despite the high level of blue staining due to a-tubulin overexpression. (K,N) Sections through the embryo shown in I and J allow the detection of the 6-11B1 staining much better than the whole-mount view. (K) The injected area, marked by purple-stained nuclei, lies between the two arrowheads. Note that overproduction of a-tubulin-expressing cells in this area leads to a thickening and altered morphology of the epidermis. (N) A high-power view of an adjacent section shows that most of the a-tubulin-expressing cells have differentiated since they are also stained by 6-11B1 antibody whichever position they have in the ectoderm. The differentiated ciliated cells (blue and brown) that have reached the surface of the skin are surrounded by non-ciliated cells (asterisks). A fraction of the double-labeled cells were retained in the inner layer of the epidermis (arrowheads). Note that some of them are likely to be migrating to the surface (arrows) indicating that intercalation is an ongoing process.|
|Fig. 9. Schematic representation of the two-step mechanism spacing the ciliated cells. (1) Lateral inhibition takes place in the inner ectodermal layer at late gastrula stages and drives a subset of cells to express higher levels of X-Delta-1 (green cells). These cells inhibit the differentiation of their neighbours (orange cells) and take on the ciliated-cell fate as marked by expression of a-tubulin (blue cells). (2) At neurula stages, the inner layer cells organize into a monolayer and the a-tubulin-expressing cells intercalate into epithelial junctions of the outer layer (columnar red cells). Differentiation of the ciliated cells occurs at early tadpole stages while they reach their definitive position in the epithelium. At the same time, division of the surface layer cells produces new interstitial locations available for intercalation of additional ciliated cells.|