|
Figure 1.
fgfrs are expressed adjacent to the developing optic tract. A, Schematic illustrating where along the optic tract the first RGC axons have reached at each stage of embryonic development. B–E, Lateral whole-mount views of stage 35/36 X. laevis brains in which expression of four fgfrs (xfgfr1–xgfgr4) was determined by in situ hybridization (blue staining). HRP followed by a DAB reaction was used to anterogradely label RGC axons from the contralateral eye (brown fibers). Asterisks indicate the location where RGC axons make a caudal turn in the mid-diencephalon, and dashed lines in B–E delineate the approximate rostral boundary of the optic tectum. chi, Optic chiasm; di, diencephalon; mhb, midbrain–hindbrain border; tec, optic tectum; tel, telencephalon. Scale bar, 100 μm.
|
|
Figure 2.
RGC axons terminate in the mid-diencephalon when FGF signaling is inhibited. A–D, Representative examples of whole-mount stage 40 brains in which RGC axons have been labeled with HRP. Stage 33/34 embryos were exposed to control DMSO (A), 100 μM SU5402 solution (B, C), or 20 μg/ml recombinant mouse sFGFR3/IIIc (D). With global inhibition of FGF signaling, most axons stop at the mid-diencephalic turn, and only some grow past en route to the optic tectum. Note that brain in C is mounted slightly ventrally compared with brains in A and B. E, Schematic illustrating optic tract width ratio analysis: 10 evenly spaced concentric rings were overlaid on normalized brains, originating at the chiasm and terminating at the midbrain–hindbrain border (2 easily identifiable morphological landmarks, 5 rings shown here). The width of the optic tract at the third (y) and fifth (x) ring was measured, and the ratio of x/y was calculated for each embryo as an estimate of the number of axons that navigated beyond the turn. F, Average optic tract width ratio for embryos exposed to control DMSO or SU5402 solutions for 22 or 29 h. A significant decrease was observed in the optic tract width ratio after exposure to SU5402 in both the 22 and 29 h exposure groups (Newman–Keuls post hoc test). There was no significant impact from increased exposure time. Asterisks in A–D indicate the point where RGC axons make caudal turn in mid-diencephalon, and the dotted lines indicate approximate rostral boundary of the optic tectum. A↔ P, Anterior/posterior axis; chi, optic chiasm; di, diencephalon; mhb, midbrain–hindbrain border; tec, optic tectum; tel, telencephalon. Scale bar, 50 μm. Graph depicts mean ± SEM. *p < 0.05, **p < 0.01.
|
|
Figure 3.
FGF signaling maintains xsema3A and xslit1 expression in the forebrain. A–F, Stage 33/34 embryos were exposed to a control DMSO (A, C, E) or 100 μM FGFR inhibitor (SU5402) (B, D, F) solution for 10 h and then processed for guidance cue expression by in situ hybridization. To best observe expression patterns, embryos were viewed from a lateral perspective for xsema3A (A, B) and xslit1 (C, D) and a dorsal perspective for xslit2 (E, F). Embryos were scored in a blinded manner for intensity of staining from 1 (indicating light staining) to 3 (indicating dark staining). G, Bar graph illustrates the average intensity scores for each riboprobe. Graphs indicate a significant decrease in xsema3A and xslit1 but not xslit2 expression after SU5402 treatment (Mann–Whitney rank sum test). H, Representative image of the change in xsema3A mRNA levels after exposure to SU5402 as assessed by RT-PCR. I, Stage 33/34 embryos were exposed to a control DMSO or 100 μM SU5402 solution for 2, 6, or 10 h and then processed for xsema3A expression by in situ hybridization and analyzed as in G. A significant decrease in xsema3A expression was observed after 6 and 10 h of exposure to the inhibitor (Dunn's post hoc method). A↔ P, Anterior/posterior axis; pi, pineal gland; tel, telencephalon. Scale bar, 100 μm. Graphs depict mean ± SEM. **p < 0.01.
|
|
Figure 4.
FGF signaling is sufficient to induce xsema3A and xslit1 expression. A–D, Stage 28 embryos were electroporated with pCS2–gfp (gfp) (A, B) or pCS2–gfp and pCS2–xfgf8 (xfgf8; C, D). Twenty-four hours later, embryos were processed for xsema3A (A, C) or xslit1 (B, D) expression by in situ hybridization. Regions of expanded or premature expression are indicated with arrows. E, F, Representative images of the change in xsema3A (E) and xslit1 (F) mRNA levels after gfp and xfgf8 electroporation as assessed by RT-PCR. dd, Dorsal diencephalon; hy, hypothalamus; mb, midbrain; pi, pineal gland. Scale bar, 100 μm.
|
|
Figure 5.
Forebrain map generally maintained after late-stage FGFR inhibition. A–H, Stage 33/34 embryos were exposed to a control DMSO (A, C, E, G) or 100 μM SU5402 (B, D, F, H) solution for 10 h and then processed for xlhx1 (A, B), xlhx5 (C, D), xdll3 (E, F), and xlhx2 (G, H) expression by in situ hybridization. I, Each embryo was photographed and scored in a blinded manner: a score of 1 indicates light staining, and a score of 3 indicates dark staining. The average score for control- and SU5402-treated groups is summarized in the bar graph. No change in xlhx1, xlhx5, or xdll3 expression was observed after exposure to SU5402, but xlhx2 expression was diminished. Scale bar, 100 μm. Graphs show mean ± SEM. *p < 0.01, Mann–Whitney rank sum test.
|
|
Figure 6.
RGC axons fail to navigate beyond the mid-diencephalon when XSema3A and XSlit1 levels are diminished. Fluorescently tagged antisense oligonucleotides targeted to xsema3A and xslit1 mRNA were introduced into stage 28 embryos via electroporation, and, at stage 39, optic tracts were anterogradely labeled with HRP. To maintain fluorescence, brains were not dehydrated and cleared as they were in Figure 2, so tracts appear less distinct. Differences in the thickness of the optic projections is at least in part accounted for by small variations in effectiveness of the HRP-labeling method. A–D, Shown here are embryos electroporated with 500 μM xsema3A-S control (A), 250 μM xslit1-AS plus 250 μM xsema3A-AS (B, C), or 500 μM xslit1-AS (D) oligonucleotides. Insets show distribution of oligonucleotides. A severe arrested-at-turn phenotype is obvious in C, whereas milder arrested-at-turn phenotypes are shown in B and D: at least half of the RGC axons stop at the mid-diencephalic with the remaining axons extending on toward the optic tectum. E, Optic tract widths were measured for pre-turn and post-turn sections of the optic tracts, as described in Figure 2E. A significant decrease in the optic tract width ratio was observed after electroporation with 500 μM xslit1-AS or 250 μM xslit1-AS plus 250 μM xsema3A-AS (S3A-AS) when compared with xsema3A-S (S3A-Sense) electroporated control embryos. F, G, Schematic illustrating model: F depicts FGFs signaling to neuroepithelial cells and promoting expression of xsema3A and xslit1, which go on to guide RGC axons. G, At stage 33/34, XSema3A and XSlit1 are present in advance of incoming RGC axons, and, by stage 35/36, RGC axons can sense these two guidance cues. The repulsive gradient generated by XSema3A and XSlit1 drives RGC axons caudally toward the optic tectum (stage 37/38). Asterisks indicate site of mid-diencephalic turn, and dotted lines indicate approximate rostral boundary of optic tectum. Graph indicates mean ± SEM. **p < 0.01, Newman–Keuls post hoc test.
|
