Fig 1.
Identification of a genetic program that limits dendrite alignment along epidermal cell-cell junctions.
(A-D) Dual-labeling of epidermal septate junctions (Nrx-IVGFP) and nociceptive C4da neurons (ppk-CD4-tdTomato) in third instar larvae. (A) Maximum projection of a confocal stack showing the distribution of C4da dendrites over epidermal cells in a single dorsal hemisegment. (B) Tracing of epidermal cells depicting distribution of the principal epidermal cell types (epidermal cells, apodemes and histoblasts, pseudocolored as indicated). (C-D) High magnification views of dendrite position over epidermal cells (C) adjacent to posterior segment boundary and (D) dendrite position over apodemes (marked with asterisks) at the posterior segment boundary. Asterisks mark apodemes. (E-I) miR-14 regulates dendrite position over epidermal cells. (E-F) Maximum intensity projections of C4da neurons from (E) wild-type control and (F) miR-14Δ1 mutant larvae at 120 h AEL showing epidermal junction-aligned dendrites pseudocolored in magenta. Insets (E’ and F’) show high magnification views of corresponding regions from control and miR-14Δ1 mutants containing junction-aligned dendrites. Arrows mark dendrite crossing events involving junction-aligned dendrites and dashed lines mark aligned dendrites that are bundled. (E” and F”) Maximum intensity projections show relative positions of C4da neurons (ppk-CD4-tdTomato) and epidermal junctions labeled by the PIP2 marker PLCδ-PH-GFP. (G-K) Quantification of epidermal dendrite alignment phenotypes. Plots depict (G) the fraction of C4da dendrite arbors that are aligned along epidermal junctions at 120 h AEL, (H) the length of aligned stretches of C4da dendrites, (I) the proportion of terminal dendrites that were present in bundles, (J) the proportion of C4da dendrites ensheathed by epidermal cells at 120 h AEL, and (I) the proportion of C4da dendrite arbors that aligned along epidermal junctions over a developmental time course in control and miR-14Δ1 mutant larvae. Box plots here and in subsequent panels depict mean values and 1st/3rd quartile, whiskers mark 1.5x IQR, and individual data points are shown. *P<0.05 compared to pre-EMS control; Mann Whitney test (G-H), Fisher’s exact test (I), unpaired t-test with Welch’s correction (J) or Kruskal-Wallis test with post-hoc BH correction (K).. Sample sizes are indicated in each panel.
Fig 2.
Time-lapse analysis of dendrite-epidermal junction interactions.
C4da neurons (ppk-CD4-tdTomato) were imaged over a 12 h time-lapse (108–120 h AEL) and dynamics were monitored for junction-aligned dendrites identified via co-localization with an epidermal junction marker (A58-GAL4, UAS-PLCδ-PH-GFP). Epidermal junction-aligned dendrites were pseudocolored cyan in the initial time point, and growth (green) and retraction (magenta) were pseudocolored in a composite of the two time points. Representative images are shown for (A) wild-type control and (B) miR-14 mutant larvae. (C) Stacked bar plot shows the fraction of junction-aligned dendrites that were growing, stable, or retracting over the time-lapse. (D) Extent of dynamics. Plot shows the change in length for each aligned dendrite measured (points) as well as mean and standard deviation. (E) Frequency of turnover in junctional alignment. Bars depict mean and standard deviation and data points represent the number of junctional-alignment events gained (green) or lost (magenta) during the time lapse for individual neurons, normalized to the area sampled. (F-H) Time-lapse imaging of new dendrite branch alignment relative to epidermal junctions. C4da neurons were imaged over a 24 h time lapse (96–120 h AEL) and the orientation of dendrite branch growth relative to epidermal junctions was monitored for each new dendrite branch. (F) Bars depict the proportion of newly aligned dendrite stretches (aligned to epidermal junctions at 120 h but not 96 h) that involve new dendrite growth (green) or reorientation of existing dendrites (cyan). Chi-square analysis revealed no significant difference between wild-type controls and miR-14 mutants. (G) A significantly larger proportion of new dendrite branches (present at 120 h but not 96 h) align along epidermal junctions in miR-14 mutants compared to wild-type controls. (H) Comparable portions of unaligned new dendrite branches in miR-14 mutants and wild-type controls orient towards (green) and away from (magenta) the nearest epidermal cell-cell interface. *P<0.05 compared to wild-type control unless otherwise indicated, Chi-square analysis (C, F-H), Kruskal-Wallis test with post-hoc Dunn’s test (D), or one-way ANOVA with post-hoc Sidak’s test (E).
Fig 3.
Junctional-aligned dendrites intercalate between epidermal cells.
(A) Schematic illustrating the axial position of ensheathed, epidermal junction aligned (intercalated) and BM-attached dendrites. (B) Apical dendrite detachment from the BM. Traces depict BM-contacting dendrites in green and BM-detached dendrites in magenta for representative wild-type control and miR-14 mutant C4da neurons. Plots depict (C) the fraction of C4da dendrites apically detached from the BM and (D) the fraction of apical detachment that involves terminal dendrites. *P<0.05 compared to wild-type control, unpaired t-test with Welch’s correction. (E-H) Junction-aligned dendrites apically intercalate between epidermal cells. Maximum intensity projections show distribution of C4da dendrites over individual epidermal cells in wild-type control (E) and miR-14 mutant larvae (F). Orthogonal sections span the width of epidermal cells (cross section; carets mark position of epidermal junctions), the epidermal cell-cell interface (junction section, marked by cyan hatched box) and run adjacent to the epidermal cell-cell interface (junction adjacent section, marked by yellow hatched box). Among these, only junction-aligned dendrites penetrate to the apical surface of epidermal cells. (G-H) Z-projections of confocal stacks depicting axial dendrite position according to a lookup table in wild-type control (G) and miR-14 mutant larvae. White hatched lines outline epidermal junctions, white arrows depict dendrite crossing events involving junction-aligned dendrites, cyan arrows depict apical insertion site of junction-aligned dendrites (I) Schematic depicting approach to measuring dendrite-epidermal junction angles of incidence (left) and scatterplots of axial distance (dendrite to epidermal AJ) versus dendrite-junction angle of incidence (right). Note the inverse linear regression (black lines). (J) Dendrite intercalation is distributed across a range of epidermal cell sizes but preferentially occurs near tricellular junctions. The plot depicts the distribution of edge lengths at epidermal cell-cell interfaces (all junctions, gray) and the length distribution of all epidermal cell-cell interfaces that contain intercalated dendrites (intercalated junctions, cyan). (K) The insertion site for epidermal intercalation is biased towards tricellular junctions. The plot depicts the length from the midpoint of the cell-cell interface to a tri-cellular junction (midpoint) and the distance between the site of apical dendrite intercalation from the nearest tricellular epidermal junction. Measurements were taken from 30 epidermal cell-cell interfaces in miR-14 mutant larvae that contained a single intercalated dendrite. *P<0.05, ns, not significant (P>0.05) compared to wild-type control, Wilcoxon rank sum test.
Fig 4.
miR-14 functions in epidermal cells to control dendrite position.
(A-B) miR-14-GAL4 is highly expressed in the epidermis with the exception of apodemes, where expression is largely absent. (A) Maximum intensity projection shows miR-14-GAL4, UAS-tdTomato expression in the body wall of third instar larvae additionally expressing NrgGFP to label epidermal junctions. Dashed lines mark apodemes, asterisks mark cells lacking detectable miR-14-GAL4 expression and arrowheads mark lowly expressing cells. (B) Stacked bars depict the proportion of epidermal cells (n = 725 cells) and apodemes (n = 149 cells) with the indicated levels of miR-14-GAL4 expression. *P<0.05, Chi-square test. (C-E) miR-14 activity mirrors miR-14-GAL4 expression patterns. Maximum intensity projections depicting (C) control or (D) miR-14 activity sensor in larvae additionally expressing sr-GAL4, UAS-mCD8-RFP to label apodemes are shown (left) along with lookup tables depicting miRNA sensor intensity (right). Insets show miRNA sensor design and dashed lines outline apodemes. (E) miR-14 sensor expression, which is inversely related to miR-14 activity, is attenuated throughout the epidermis with the exception of apodemes. NS, not significant, *P<0.05, Kruskal-Wallis test followed by Dunn’s multiple comparisons test. (F) miR-14 is dispensable in C4da neurons for dendrite morphogenesis. Total dendrite length and the number of dendrite crossings in miR-14 or drosha mutant C4da MARCM clones are indistinguishable from wild-type controls. NS, not significant, Kruskal-Wallis test followed by Dunn’s multiple comparisons test. (G) Epidermal miR-14 function is required for proper dendrite positioning. Dendrite crossing frequency is shown for larvae expressing control or miR-14 sponge transgenes using the indicated GAL4 drivers. *P<0.05, ANOVA with post-hoc Sidak’s test. (H) Epidermal miR-14 expression is sufficient to support proper dendrite positioning. Dendrite crossing frequency is shown for miR-14 mutant larvae expressing UAS-miR-14 under control of the indicated GAL4 drivers. NS, not significant, *P<0.05, ANOVA with post-hoc Sidak’s test.
Fig 5.
miR-14 regulates sensitivity to noxious mechanical inputs.
(A-B) miR-14 mutants exhibit enhanced nocifensive behavior responses. Plots depict (A) proportion of larvae that exhibit nocifensive rolling responses to von Frey fiber stimulation of the indicated intensities and (B) mean number of nocifensive rolls. (C-D) miR-14 mutation does not affect larval responses to non-noxious mechanical stimuli. Plots depict larval responses to (C) gentle touch and (D) vibration for larvae of the indicated genotypes. (E) Plots depict nocifensive rolling response probability (y-axis) and latency (x-axis) of control and miR-14 mutant larvae to the indicated thermal stimuli. (F-G) miR-14 functions in epidermal cells to control mechanical nociceptive sensitivity. (F) Plot depicts nocifensive rolling responses to 25 mN von Frey fiber stimulation of larvae expressing a miR-14 sponge or a control sponge with scrambled miR-14 binding sites under control of the indicated GAL4 driver (Ubiq, ubiquitous expression via Actin-GAL4; Epi, epidermal expression via A58-GAL4; PNS, md neuron expression via 5-40-GAL4). (G) Plot depicts nocifensive rolling responses to 25 mN von Frey fiber stimulation of miR-14 mutant larvae expressing the indicated GAL4 drivers with or without UAS-miR-14. (H-J) miR-14 acts independent of known pathways for nociceptive sensitization. Plots depict nocifensive rolling responses to 25 mN von Frey fiber stimulation of control or miR-14 mutant larvae (G) 24 h following mock treatment or UV irradiation and (H) following 24 h of feeding vehicle or vinblastine, or (I) of miR-14 mutant larvae carrying loss-of-function mutations in the indicated sensory channels. NS, not significant, *P<0.05, Fisher’s exact test with a BH correction (A, H-J), Kruskal-Wallis test followed by a Dunn’s multiple comparisons test, or Chi-square test (G). (B-D). The number of larvae tested is shown for each condition.
Fig 6.
Apical epidermal intercalation contributes to mechanical hypersensitivity.
(A-E) Neuronal integrin overexpression suppresses miR-14 mutant junctional dendrite alignment defect. Representative composite images show C4da dendrite arbors (white) and epidermal cell-cell junctions (magenta) for (A) wild-type control, (B) miR-14 mutant, (C) control larvae overexpressing UAS-mew and UAS-mys (UAS-Integrins) selectively in C4da neurons (ppk-GAL4), and (D) miR-14 mutant overexpressing UAS-Integrins selectively in C4da neurons. Hatched boxes indicate region of interest shown at high magnification to the right of each image. (E) Plot depicts the proportion of dendrite arbors aligned along epidermal junctions in larvae of the indicated genotypes. *P<0.05, ANOVA with post-hoc Tukey’s test. (F) Neuronal integrin overexpression has no effect on C4da neuron dendrite branch length or number. Plots depict the total dendrite length (left) and terminal dendrite number (right) normalized to segment size in wild-type control and miR-14 mutant larvae. Statistical tests (dendrite length, Fisher’s exact test; dendrite number, unpaired t-test with Welch’s correction) revealed no significant difference in either paramter (n = 7 neurons each). (G) Neuronal integrin overexpression suppresses miR-14 mutant mechanical hypersensitivity. Plot depicts nociceptive rolling responses of larvae of the indicated genotypes to different forces of von Frey stimulation. *P<0.05, Fisher’s exact test with a post-hoc BH correction. The number of larvae tested is shown for each condition.
Fig 7.
miR-14 regulation of epidermal gap junction assembly controls dendrite position and nociceptive sensitivity.
(A-B) miR-14 mutation affects epidermal barrier function. Maximum intensity projections show C4da arbors (ppk-CD4-tdGFP) and rhodamine-conjugated dextran labeling in cross-section of wild-type control and miR-14 mutant larvae. Dashed lines indicate the position of the orthogonal xz sections (middle), and bottom images show xz maximum intensity projections. Arrows indicate apically-shifted dendrite branches and carets mark apical dextran infiltration at cell-cell junctions. (C) Schematic depicting position of epidermal junctional complexes (left), alleles used for genetic interaction studies (center), and markers used for analysis of junctional assembly (right). (D-F) ogre and Inx2 genetically interact with miR-14 to regulate dendrite position. Representative images show 120 h AEL C4da neurons from (D) miR-14Δ1/+ heterozygous mutant and (E) miR-14Δ1/+, Inx2G0118/+ double heterozygous mutant larvae. (F) Morphometric analysis of C4da dendrites in larvae heterozygous for miR-14 and the indicated epidermal junction genes showing the mean number of dendrite-dendrite crossing events per neuron normalized to dendrite length. *P<0.05, ANOVA with post-hoc Dunnett’s test. (G) ogre and Inx2 genetically interact with miR-14 to regulate mechanical nociceptive sensitivity. Plots depict the rolling probability and frequency of multiple roll responses evoked by 25 mN von Frey fiber stimulation in larvae of the indicated genotypes. *P<0.05 compared to miR-14 heterozygous controls, Fisher’s exact test with a post-hoc BH correction (G). (H-O) miR-14 regulates GJ assembly. Maximum projection images show C4da dendrites (green) and ogre immunoreactivity in the epidermis of a wild-type control larva (H). (I) Zoomed images corresponding to hatched box in (H) show the relative position of C4da dendrites and ogre at epidermal cell-cell junctions. Arrows mark sites of disconiuties in junctional ogre immunoreactivity, which most frequenly occurs at tricellular junctions (I’).Orthogonal sections show ogre distribution at a representative bicellular junction (outlined with hatched lines) in xy (J) or xz projections (K). C4da dendrites are confined to the basal face of a continuous belt of ogre immunoreactivity in control larvae. (L-O) miR-14 mutation disrupts organization of ogre immunoreactivity at epidermal cell-cell junctions. (L-M) The belt of ogre immunoreactivity exhibits irregularity in width, signal intensity, and frequent discontinuities (arrows). (N, O). C4da dendrites intercalate into gaps in ogre and immunoreactivity and penetrate apically into the GJ domain. (P-R) Selective epidermal overexpression of Inx genes suppresses miR-14 mutant dendrite alignment and mechanonociception defects. (P) ogre immunoreactivity at GJs is reduced in miR-14 mutants. Plot depicts intensity of ogre immunoreactivity signal normalized to arm immunoreactivity at epidermal cell-cell interfaces in control and miR-14 mutant larvae. P<0.05, unpaired t-test with Welch’s correction. (Q) GJ belt width is reduced in miR-14 mutants. Violin plot depicts the distribution of GJ belt widths at epidermal cell-cell interfaces in control and miR-14 mutant larvae Each data point represents the average GJ belt width measured across the full length of a cell-cell interface. *P<0.05, Kolmogorov-Smirnov test. (R) Composite images show C4da dendrites pseudocolored green to label epidermal junctional alignment in a miR-14 mutant (left) and a mir-14 mutant expressing UAS-ogre selectively in epidermal cells (right). (R) Plot depicts the fraction of C4da dendrite arbors aligned along epidermal junctions at 120 h AEL for the indicated genotypes. NS, not significant, *P<0.05, Kruskal-Wallis test followed by Wilcoxon rank sum test with BH correction. (S) Fraction of larvae of the indicated genotypes that exhibit nocifensive rolling responses to 25 mN von Frey stimulation. NS, not significant, *P<0.05, Fisher’s exact test with a BH correction. (T) Epidermis-specific Inx gene knockdown increases epidermal junctional alignment of C4da dendrites. The plot depicts the proportion of C4da dendrite arbors aligned along epidermal junctions in larvae expressing the indicated RNAi transgenes. Quantitative analysis of Inx protein knockdown and representative images of dendrite phenotypes are shown in S9 Fig. *P<0.05, Kruskal-Wallis test followed Dunn’s multiple comparisons test.
Fig 8.
Epidermal dendrite intercalation tunes C4da neuron calcium levels and nociceptive sensitivity.
(A-F) Epidermal dendrite intercalation regulates baseline calcium levels in C4da neurons. (A) Larval preparation for imaging mechanically-evoked calcium responses in C4da neurons. (B) Control and miR-14 mutant larvae exhibit comparable amplitudes of GCaMP6s responses (ΔF/F0) in C4da neurons to mechanical stimulus (n = 15 larvae each, p = 0.217, Wilcoxan rank sum test). (C) GCaMP6s fluorescence intensity is significantly elevated in miR-14 mutant C4da axons prior to mechanical stimulus and 5 min after mechanical stimulus (n = 15 larvae each, p = 0.0003 pre-stimulus, p = 0.002 post-stimulus, Wilcoxon rank sum test). (E-F) Ratiometric calcium imaging using a GCaMP6s-Cherry fusion protein expressed selectively in C4da neurons (ppk-GAL4, UAS-Gerry). (E) Representative images depict fluorescence intensity of Cherry and GCaMP6s for wild-type control and miR-14 mutant larvae. (F) miR-14 mutants exhibit elevated GCaMP/mCherry fluorescence ratios in C4da axons. Points represent measurements from individual abdominal segments (A2-A8) from 10 larvae of each genotype (n = 64 data points in control larvae, 56 in miR-14 mutants, 66 in miR-14 + ppk>Integrin larvae). *P<0.05, Kruskal-Wallis test followed by Dunn’s post-hoc test. (G) C4da dendrites are confined to territory between muscle adhesion sites at apodemes. (H-J) Mutation of the PS2 integrin ligand Tig prevents dendrite intercalation between apodemes. Representative maximum intensity projections of larvae expressing ppk-CD4-tdTomato to label C4da dendrites and NrgGFP to label epidermal junctions are shown for (H) TigA1/+ heterozygote control and (I) TigA1/X mutant larvae. C4da dendrites are depicted in magenta and apodemes in green using an apodeme mask to subtract NrgGFP signal in other cells. (H’ and I’) C4da dendrites are pseudocolored according to their orientation at apodemes (magenta, aligned along apodeme junctional domains; cyan, invading apodeme territory). (J) TigA1/X mutant larvae exhibit a significant reduction in junctional dendrite alignment at apodemes and an increase in dendrite invasion into apodeme territory. *P<0.05, ANOVA with a post-hoc Sidak’s test. (K) TigA1/X mutant larvae exhibit no significant alteration in dendrite arbor length outside of apodeme domains. No significant difference was detected between the two groups using an unpaired t-test. (L-M) Dendrite intercalation at apodemes tunes nociceptive sensitivity. (L) TigA1/X mutant larvae exhibit a significant reduction in rolling responses to noxious mechanical stimulus. (M) TigA1/X mutant larvae exhibit a significant reduction in GCaMP/mCherry fluorescence ratios in C4da axons Points represent measurements from individual neurons. *P<0.05, Wilcoxon rank sum test in (L-M).