Multiple Guidance Mechanisms Control Axon Growth to Generate Precise T-shaped Bifurcation during Dorsal Funiculus Development in the Spinal Cord

Bridget M Curran
Kelsey R Nickerson
Andrea R Yung
Lisa V Goodrich
Alexander Jaworski
Marc Tessier-Lavigne
Le Ma author has email address

Department of Neuroscience, Jefferson Synaptic Biology Center, Vickie and Jack Farber Institute for Neuroscience, Sydney Kimmel Medical College, Thomas Jefferson University, Philadelphia, PA 19107
Department of Neuroscience, Brown University, Providence, RI 02912
Robert J. and Nancy D. Carney Institute for Brain Science, Providence, RI 02912
Department of Neurobiology, Harvard Medical School, Boston, MA 02115
Department of Biology, Stanford University, Palo Alto, CA 94305

https://doi.org/10.7554/eLife.94109.2

Open access
Copyright information

Figures and data

NF immunostaining reveals DRG axon misprojections in Ntn1^β/β mutants in cleared wholemount embryos and transverse sections.
(A-D) Inverted fluorescent images of whole mount NF staining of CUBIC cleared E10.5 WT (A,C) or Ntn1^β/β (B,D) embryos. The projected images are viewed from the lateral side of the body at the forelimb (A-B) and hindlimb region (C-D). The space above the dorsal funiculus (dotted-line box) is devoid of NF-fibers in the WT embryo (A, C), but filled with misprojecting axons (arrows) from DRGs (*) that grow dorsally in the mutant (B,D). The compass indicates the dorsal (D)-ventral (V) axis and the rostral (R)-caudal (C) axis.
(E) A model depicting DRG sensory axon projections in the spinal cord. Bifurcation is shown in the DREZ (green area) that forms the eventual dorsal funiculus.
(F) Quantification of the number of axonal fibers at proximal, medial, and distal locations from the dorsal funiculus (see purple lines in B and D) in the wholemount staining of WT and Ntn1^β/βembryos (n=4). Two-way ANOVA F(1, 12)=199.5 with Tukey’s HSD Test for multiple comparisons found that the mean value was significantly different between WT and Ntn1^β/β in proximal p<0.0001; medial p<0.0001; and distal, p=0.001 locations.
(G-N) Inverted fluorescent images of NF-stained transverse sections of the brachial region of E10.5 (G-H, K-L) and E11.5 (I-J, M-N) embryos. Misprojections (arrows) extending from the DREZ (green box) are seen in Ntn1^β/β (K-N) but not WT (G-J) animals. Asterisks (*) label the DRG. Images in H, J, L, N are zoomed-in views of the boxed regions around the DREZ in G, I, K, M, respectively. The compass indicates the dorsal (D)-ventral (V) axis and the left (L)-right (R) axis.
**** p<0.0001, *** p<0.001. Bars: 100 μm.

Genetic labeling of DRG neurons confirms DRG sensory axon misprojections and reveals guidance defects during bifurcation in Ntn1^β/β mutants.
(A-C) Sensory axon specific genetic labeling based on tdTomato expression driven by tamoxifen-induced Neurog1:CreER^T2 in E10.5 WT (A) or Ntn1^β/β (B) embryos that were cleared by iDISCO. Fluorescent images of immunostaining using RFP antibodies (for tdTomato) (A-C) or 2H3 antibodies (for NF) (A’-C’) are taken from the lateral side of the spinal cord in the brachial region of cleared wholemount embryos. Color-merged images are shown in A”-C”. Dorsal funiculi are inside the yellow dotted-line boxes and DRGs are marked by asterisks (*). Zoomed-in images (C) from the boxed regions in B (white box) highlights the co-staining of DRG specific tdTomato and NF in abnormal fibers (yellow arrows) that extend from the dorsal funiculus in Ntn1^β/β mutants (n=3). White arrows indicate NF-labeled misprojections that do not co-stain for tdTomato.
(D-H) Sparse sensory axon labeling in WT or Ntn1^β/β animals based on the basal tdTomato fluorescence driven by Neurog1:CreER^T2 without using tamoxifen. Inverted fluorescent images (D,E,G,H,) are viewed along the DREZ from the lateral side of the spinal cord in cleared wholemount embryos. The merged fluorescent image of tdTomato and NF labeling shows the DREZ (dotted-line box) in F. Normal turning is found in WT animals (n =4) for bifurcated (D, n=11/12 axons) and non-bifurcated axons (E, n=14/14 axons). In Ntn1^β/β mutant embryos (n=3), tdTomato labeled axons (purple arrows) can be seen misprojecting from the DREZ in bifurcated (G,H, n=7/14 axons) and non-bifurcated axons (H, n=9/13 axons). Green “V” indicates the location of bifurcation junctions or the turning point of single axons.
(I) Quantification of the defects in sparsely labeled axons above. The percentage of fibers that correctly (black bar) or incorrectly (grey bar) turn at the DREZ is quantified in two groups: bifurcated (+) or non-bifurcated (-). A student’s t-test shows that the mean value of percentage of fibers misprojecting was significantly different between WT and Ntn1^β/βin non-bifurcated axons (t(4)=25, p<0.0001) and bifurcated axons (t(4)=4.533, p=0.0106).
* p<0.05, and **** p<0.0001, Bars: 100 μm.

Wholemount NF immunostaining reveals the loss of the dorsal funiculus in Slit1^-/-;Slit2^-/-;Ntn1^β/β triple mutants with distinct misprojections.
(A-B) Inverted fluorescent images of NF staining in CUBIC-cleared E10.5 embryos in the forelimb region. The dorsal funiculus (between dotted-lines) is evident in the control Slit1^-/-;Slit2^+/+;Ntn1^+/+ embryo and the space above it has few labeled fibers (arrows, A), but in the Slit1^-/-;Slit2^-/-; Ntn1^β/β mutant (B), extensive NF-labeled axons (arrows) wander into the dorsomedial region of the spinal cord while the dorsal funiculus (between dotted-lines) is reduced and diminished. DRGs are marked by green asterisks (*).
(C-N) Inverted fluorescent images of NF staining in wholemount embryos of E10.5 embryos with various genotypes are viewed from the lateral side (D-G) or the dorsal surface (J-M). Arrows denote sensory misprojections found in Slit1^-/-;Slit2^+/+;Ntn1^+/+ (D,J, n=5), Slit1^-/-;Slit2^-/-;Ntn1^+/+ (E,K, n=6), Slit1^-/-;Slit2^+/+; Ntn1^β/β (F,L, n=8) and Slit1^-/-;Slit2^-/-;Ntn1^β/β(G,M, n=3) animals. In Slit1^-/-;Slit2^-/-;Ntn1^β/β triple mutants (G,M), the dorsal funiculus is nearly absent with misprojecting fibers. The number of axonal fibers at proximal, medial, and distal locations between the DREZ and the dorsal midline is quantified in the bar graphs from the lateral (H) or the top-down (N) view. The orientation and measurement location are shown in the corresponding cartoons (C and I) and DRGs are marked by green asterisks (*). Statistics from two-way ANOVA analysis in H: for proximal misprojections: Slit1^-/- vs. Slit1^-/-;Slit2^-/-p=0.0122; Slit1^-/- vs. Ntn1^β/β p=0.0011; Slit1^-/- vs. Slit1^-/-;Slit2^-/-;Ntn1^β/βp<0.001; for medial misprojections: Slit1^-/-vs. Ntn1^β/βp=0.0122; Slit1^-/- vs. Slit1^-/-;Slit2^-/-;Ntn1^β/βp<0.0001; and for distal misprojections in Slit1^-/- vs. Slit1^-/-;Slit2^-/-;Ntn1^β/β p<0.0001. Statistics from two-way ANOVA analysis in I: for proximal misprojections: Slit1^-/- vs. Slit1^-/-;Slit2^-/-p=0.0388; Slit1^-/-vs. Ntn1^β/β p=0.0002; Slit1^-/- vs. Slit1^-/-;Slit2^-/-;Ntn1^β/β p<0.0002; for medial misprojections: Slit1^-/-vs. Ntn1^β/β p=0.0388; Slit1^-/- vs. Slit1^-/-;Slit2^-/-;Ntn1^β/βp<0.0001; and for distal misprojections in Slit1^-/-vs. Ntn1^β/β p=0.0175; Slit1^-/- vs. Slit1^-/-;Slit2^-/-;Ntn1^β/β p=0.0004.
(O-T) Inverted fluorescent images of NF staining in cross sections of E10.5 embryos with various genotypes (P-S, n=3, except n=4 for Ntn1^β/β). Orange arrows indicate axonal misprojections found in Slit1^-/-;Slit2^-/- mutants inside the spinal cord and blue arrows indicate misprojections associated with Ntn1^β/βmutants at the pial surface. Blue lines with arrow at both ends indicate the distance from dorsal projections to the midline. Quantification of defects (T) is based on the distance of the dorsal fibers from the midline (blue bars) and the length of misprojection fibers inside the spinal cord (orange bars). The cartoon (O) illustrates the two types of misprojections (brown lines) and the quantification of dorsal distance (blue lines). One-way ANOVA: for horizontal, F= 18.34, df=3, and p=0.0006; for dorsal, F=44.16, df=3, and p<0.0001.
* p<0.05, ** p<0.01, *** p<0.001 and **** p<0.0001, ns, not significant. Bars: 100 μm.

Single axon DiI labeling of DRG central afferents shows altered turning of bifurcated axons in various Slit and Ntn1 mutants.
(A-C) Single cell DiI labeling of DRG axons in open book preparations (A) of spinal cords isolated from E12.5 (B) WT and (C) Slit1^-/-;Slit2^-/-;Ntn1^β/βmutants. Maximum z-projection are depth color coded (∼45 μm represented by the color bar) with multiple bifurcating axons visible in field of view. Arrows point to misprojecting fibers and green “V” indicates the location of bifurcation junctions.
(D-G) Skeletonized tracings of DiI labeled axons from embryos with various genotypes. WT axons show stereotyped T-bifurcation along the DREZ (D, n=4 animals). Axons in Slit1^-/-;Slit2^-/-mutants overshoot the DREZ (E, n=3 animals). In Ntn1^β/β mutant embryos, axons stray from the DREZ and have disrupted turns along the DREZ (F, n=5 animals). Axon in Slit1^-/-;Slit2^-/-;Ntn1^β/βtriple mutants are completely disorganized around the DREZ (G, n=2 animals). Dashed lines in D show the angle (θ) measurement.
(H-J) Quantification of the turning angle of bifurcated branches along the DREZ (H) from single axon labeling above showed the mean angles are similar for all genotypes (one way ANOVA [F(3,114) = 0.8229, p=n.s.). However, the deviation of the angle from the average was significantly different among different genotypes (I). One way ANOVA with Tukey’s Post-hoc for multiple comparisons ([F(3,114) = 15.97, p<0.0001]): WT and Slit1^-/-;Slit2^-/-;Ntn1^β/β, p<0.0001; Slit1^-/-;Slit2^-/-and Slit1^-/-;Slit2^-/-;Ntn1^β/β, p=0.0001; and Ntn1^β/β and Slit1^-/-;Slit2^-/-;Ntn1^β/β, p<0.0001. Additionally, the angle difference between the two branches showed significantly difference among genotypes (J). One way ANOVA with Tukey’s Post-hoc for multiple comparisons ([F(3,114) = 12.99, p<0.0001]): WT and Slit1^-/-;Slit2^-/-;Ntn1^β/β, p<0.0001; Slit1^-/-;Slit2^-/-and Slit1^-/-;Slit2^-/-;Ntn1^β/β, p=0.0004; and Ntn1^β/β and Slit1^-/-;Slit2^-/-;Ntn1^β/β, p=0.0001. Only the comparison between the triple mutants and other genotypes are shown here.
*** p<0.001, **** p<0.0001, ns: not significant. Bars: 100 μm.

NF immunostaining reveals the loss of the dorsal funiculus in Robo1Robo2^--/--;DCC^-/- triple mutants.
(A-D) Inverted fluorescent images of NF staining in CUBIC-cleared E10.5 embryos viewed from the lateral side in the forelimb region. WT (A, n=3) have relatively few axons straying from the dorsal funiculus (boxed region). Purple arrows denote sensory misprojections found in Robo1Robo2^++/--;DCC^+/+ embryos (C, n=2) or DCC^-/- (B, n=6). In Robo1Robo2^--/--;DCC^-/-mutants (D, n=3), the dorsal funiculus is completely lost and disorganized with axons wandering into the dorsomedial region of the spinal cord.
(E-H) Inverted fluorescent images of NF staining in cross sections of E11.5 embryos with various genotypes (n=3, except n=4 for WT). Orange arrows indicate horizontal misprojections inside the spinal cord found in Robo1Robo2^--/-- mutants and blue arrows indicate dorsal misprojections along the pial surface associated with DCC^-/-mutants. Lines with arrow at both ends (blue) indicate the distance between dorsal misprojections to the midline.
(I-J) Quantification of dorsal and horizontal misprojections based on the distance from the roof plate to the DREZ (I, blue bars) and the total length of horizontal fibers (J, orange bars). One-way ANOVA: for horizontal, F= 28.93, df=3, and p=0.0001; for dorsal, F=31.4, df=3, and p<0.0001.
** p<0.01, *** p<0.001, **** p<0.0001, ns: not significant. Bars: 100 μm.

Sign up for email alerts