Figures and data
Morphology of mouse Merkel-cell afferents in touch domes.
(A–C) Representative axial projections of touch domes from Sarm1+/+; Atoh1A1GFP/A1GFP; TrkCtdTomato/+ mouse. Antibodies for (A) GFP (cyan, visualized with anti-GFP antibody) or (B–C) td-Tomato (red, visualized with anti-DsRed antibody) counterstained with antibodies against K8 (magenta) and NFH (green) to show transgenic label overlap with established markers of Merkel cells and axons. (C) Boxed area from (B) enlarged to show TrkC-tdTomato endings that make contact with a Merkel cell (arrowhead) and do not contact a Merkel cell (arrow). Note that branches that do not contact a Merkel cell are smaller and do not express NFH at detectable levels. (D–E) Representative axial projections of touch domes from skin taken from (D) the crown of the head of a Sarm1+/+; Atoh1A1GFP/A1GFP; TrkCtdTomato/+ female mouse and (E) the lower back of a C57Bl6/J female mouse. K8: magenta, NFH: green. (F–G) Quantification of the number of Merkel cells (F) and terminal branches (G) per touch dome. Crown: N=24 touch domes, 3 mice (2–12 per mouse); P72–P92. Back: N=21 touch domes, 3 mice (7 per mouse); P66. Welch’s t test (two-tail) for Merkel cells (P=0.10) or terminal branches (P=0.16). Data from back skin were previously published in Marshall et al., (2016; Figure 2B–C).
Non-invasive time-lapse imaging reveals that most Merkel cells and axonal terminals show plasticity over 1 month in skin homeostasis.
(A) Diagram of in vivo time-lapse imaging preparation. (B) Timeline of experiments, indicating the hair cycle stages throughout imaging and the age ranges at the beginning and end of imaging as well as at peak anagen. (C) Representative axial projections of a time series of one TrkCtdTomato/+ arbor (red) and associated Atoh1A1GFP/A1GFP Merkel cells over the entire duration of in vivo time-lapse imaging. Day of image relative to first imaging session indicated in bottom left corner. (C’) Legend for symbols used in C. Symbols located in the same positions on adjacent days such that the indicated change takes place between those two sessions. (D) Axial projection of confocal stack of registered axon from (C), labeled with antibodies against K8 (cyan) and NFH (red). (E) Comparison of the percentage of Merkel cells and terminals that showed plasticity or were stable throughout the entirety of imaging. P value: two-tailed Fisher’s exact test. N=309 Merkel cells and 536 terminals from 4 animals. (F) Analysis of the percentage of terminals per touch dome per time point showing plasticity when Merkel cells were stable (cyan outline; N=86) or showed plasticity (cyan filled; N=68). Colored bars indicate median. P value: two-tailed Mann-Whitney test. Each dot represents a single time-lapse session from one touch dome (N=19 touch domes from 4 mice; 154 total time-lapse sessions). (G) Scatter plot of the percentage of Merkel cells and terminal branches showing plasticity in sessions with plasticity in both Merkel cells and terminals. Gray dashed line indicates unity line. Black line is linear regression indicating positive correlation between Merkel-cell and terminal plasticity when both occur. R2=0.37, n=67 time-lapse sessions. (H) Average regression index for each axonal arbor across all time points. N=19 arbors; 4 mice. Bar indicates mean±SD.
Atoh1GFP Merkel cells added during adulthood have short lifetimes compared to existing Merkel cells.
(A) Example axial projections of a touch dome with Atoh1GFP Merkel cells that were added, removed, and showed presumptive movements. The at symbol denotes a Merkel cell (day 0) that was removed in the next session (day 3). Arrow indicates a Merkel cell that was added on day 6 to the empty spot indicated by the arrow on day 3. Asterisk denotes a Merkel cell moving across sessions (day 0–6). (B) Prevalence of each type of plasticity observed in plastic Merkel cells. Areas of overlap indicate Merkel cells that went through multiple types of plasticity. N=195 Merkel cells. (C) Axial projection of the last in vivo image of the touch dome from (A) before the end of time-lapse imaging. (D) Axial projection of confocal registration of in vivo time-lapse touch dome with antibody staining for K8 (cyan) and DsRed to amplify TrkCtdTomato (red). (E) Example axial projection of confocal-registered touch dome that was part of the time-lapse in vivo dataset with antibodies against K8 (cyan) and NFH (red). (F) Example confocal axial projection of neighboring touch dome, which is from the same piece of tissue as E but was not imaged during the time-lapse in vivo study. (G) Comparison of the number of K8-positive Merkel cells in touch domes repeatedly imaged with two-photon imaging (2p) and touch domes that were not subject to repeated irradiation (N). No significant difference in number of Merkel cells per touch dome by Student’s t test (P=0.26; N=12 touch domes in each group from 2–3 animals). Bars indicate mean±SD. (H) Survival plot of Merkel cells that were present on day 0 (Original) and Merkel cells that were added at some point during the in vivo imaging period (Added). Gray dashed line indicated 50% survival. Colored dashed lines indicate median points for each group. Median survival=3 days for added Merkel cells and 21 days for original Merkel cells. P value: Log-rank test to compare survival curves.
Merkel cells are contacted by two morphologically distinct types of afferent terminals.
(A–A’) In vivo image of touch dome with (A) and without (A’) Merkel-cell overlay displaying all three contact types. Symbols indicate examples of each ending type: asterisks, uninnervated; hashtags, kylikes; arrowheads, boutons. (B–B’) Confocal image of same touch dome as A, labeled with antibodies against dsRed (TrkCtdTomato, red) with (B) and without (B’) K8 (cyan) overlay. Symbols as in A. (C–D) Single z planes of zoomed in regions from A’, used to measure length and width of individual endings. Areas for each ending indicated below image. (E) Histogram of the maximum ellipse area for each identified terminal, so that each terminal is represented once. Red line: best fit bimodal gaussian. (F) Gaussian components underlying best fit bimodal gaussian in E. Threshold (gray dashed line) defined as the crossing point between the two component gaussians. (G) Ancient Greek kylix. Note the shallow cup and thin stem of the kylix, which closely resembles the previously described Merkel-cell contact. Image from the Metropolitan Museum of Art, reproduced under Creative Commons license CC0 1.0. (H) Survival plot comparing boutons and kylikes. P value: Log-rank test. N=339 kylikes and 120 boutons. (I) Schematic indicating the probability of morphological changes in terminal branches that contact Merkel cells. N=1,120 kylikes, 153 boutons, and 168 empty terminals. The most probable event for each ending type was for the morphology to remain the same (kylikes: 85%, boutons: 39%, empty/removed endings: 87%). (J) Comparison of stable and added Merkel cells with the first known contact of each type. P value: two-tailed Fisher’s exact test. (K) Comparison of the percentage of stable and removed Merkel cells with the last known contact of each type. P value: two-tailed Fisher’s exact test.
The plasticity of Merkel cells but not axon terminals is synchronized across arbors by the hair cycle.
(A–B) Plots indicating the Merkel-cell (A) and terminal branch (B) plasticity indices for each afferent on each session. Color scale in B applies to both A and B. White lines mark afferents belonging to each animal. Black squares indicate sessions with missing or unanalyzable data. (C–E) Schematic of analysis pipeline to test similarities in temporal patterns of plasticity across afferents. (C) Time series data from animal F1 are aligned to external factors, e.g. hair growth. Then, a correlation matrix is generated to calculate a correlation coefficient for each pair of afferents. (D) To test whether correlations are higher than chance, data are shuffled 1,000 times and a correlation matrix is generated for each set of shuffled data. (E) Correlation coefficients from experimental (C) and shuffled (D) datasets are compared using cumulative histograms. Gray dashed line indicates 50%, colored dashed lines indicate median correlation coefficients for each group. (F–K) Analysis of Merkel-cell plasticity index (F–H) and terminal branch plasticity index (I–K) for every afferent used in analysis following the same organization as C–E. (H,K) Cumulative histograms of correlation coefficients from experimental data and mean±SD cumulative distribution of shuffled permutations. P values: permutation test.
TrkCtdTomato afferent terminals are more stable if they contact Merkel cells.
(A–A’) Axial projections of TrkCtdTomato terminals of one identified afferent over time without (A) and with (A’) Merkel-cell overlay. Days elapsed since day 0 indicated in top left corner. (B) Axial projection of confocal registration with antibodies against K8 (cyan) and NFH (red). Arrowhead, stable occupied branch; arrow, occupied branch that persists after Merkel-cell removal; blue dashed circle, location of removed Merkel cell. Symbols located in the same position on adjacent days indicate change occurring between those two sessions. All changes were noted in three-dimensional stacks. (C) Comparison of the percentage of Merkel-cell contacting (occupied) and empty terminals that were stable for the duration of imaging or showed plasticity. P value: two-tailed Fisher’s exact test. (D) Survival curve comparing empty and occupied terminal neurites. Gray dashed line indicates 50% survival. Colored dashed lines indicate median values for each group (9 d for empty terminals and 30 d for occupied terminals). P value: log-rank test. (E) Trajectories of individual branches associated with removed Merkel cells aligned to the day of Merkel-cell removal. Bar on right indicates how many branches were either not removed (persisted through imaging) or removed before, with, or after Merkel-cell removal.
Merkel cells limit branch number and promote branch structural maturation.
(A) Axial projections of touch dome where there was an increase in the number of branches added after Merkel-cell removal. (B–C) Schematic of Merkel cells from Day 9 image in A, color coded to indicate which Merkel cells are removed on each day. Color code in C applies to B and C. (D) Comparison of the number of branches added in the session after Merkel cells were (≥1) or were not (0) removed. P value: Mann-Whitney rank test. (E–F) Example z projections (left) and rotated three-dimensional projections (right) from the same arbor, in a session without overshoot (E) and with overshoot (F). Day of imaging is indicated in upper right corner. Scale bars indicate 10 microns. Stack from F is thicker than E, thus rotated image is thicker. (G) Comparison of the percentage of afferents from each animal displaying overshoot. P value is result from Fisher’s exact test (N=4–6 touch domes per animal). (H) Analysis of timing of branch overshoot relative to the nearest event of Merkel-cell removal. Positive values indicate days after Merkel cell is removed; negative values indicate days before Merkel-cell removal. (I–J) Axial projections of touch domes from WT (K14+/+; Atoh1fl/+; TrkCtdTomato/+) and Atoh1CKO (K14Cre/+; Atoh1fl/lacz; TrkCtdTomato/+) 8 w animals stained for K8 (cyan), NFH (green) and DsRed to amplify TrkC-tdTomato (red). (K–N) Comparison across genotypes of branching parameters: number of terminal branches (K), number of branch points (L), highest branch order (M), and complexity index (N). P values: Welch’s t test (K–L), Student’s t test (M), or Mann-Whitney rank test (N). N=16–18 touch domes from 2 mice; 7–9 touch domes per mouse. Bars indicate mean±SD (K‒M) or median (N).
A model of how sensory axons respond to Merkel-cell plasticity.
(A) Three Merkel cells are contacted by either a kylix or a bouton. Merkel cell-kylix connections inhibit surrounding branches. (B) One Merkel cell is removed. The bouton grows into a kylix and inhibits surrounding branches. (C) Branches are added and lengths increase such that branches overshoot the Merkel-cell level after a Merkel cell-kylix complex is removed. Legend indicated at the bottom.
Hair follicles are unique landmarks to register afferent locations for several weeks.
Adult mice go through hair growth cycles after post-natal hair morphogenesis. The first cycle usually begins between P28 and P35 in C57Bl6/J mice, but differs with genetic background, sex, body size, and stress level. Telogen, the resting phase of hair growth, has small hair follicles with hairs that are not actively growing. Anagen, the phase of active hair follicle growth, is a period of massive cellular proliferation. The skin thickens to accommodate larger hair follicles that are actively growing hair and appears back from the long lengths of actively growing hair that are present under the skin. Cells throughout the skin turnover to expand the thickness of the skin to accommodate enlarged hair follicles. The final phase, catagen, is a short phase in which follicles actively regress back to their resting size. Epifluorescence z-stacks were used to register the locations of hair follicles surrounding touch domes. Hair follicle locations were used to confirm the identity of each afferent that was part of the time-lapse imaging study in an un-biased manner. A–B) Single z planes of epifluorescence z-stacks of the same areas taken throughout time-lapse imaging. Yellow circles indicate guard hair follicle immediately adjacent to touch dome. Black circles indicate surrounding hair follicles used as landmarks. A and B are different touch domes from the same animal; afferent ID indicated on left of each row. Imaging day indicated in top left corner; hair-cycle stage indicated in bottom left corner. Note that the relative positions of hair follicles surrounding individual touch domes are unique and persist throughout hair cycle stages for ∼1 month. Cyan: Atoh1-GFP; red: TrkC-tdTomato.
An afferent that had complete loss of the Merkel-cell cluster showed complex branching structure that could not be quantified with existing methods.
In vivo time series of touch dome no. 1 from mouse M2. Individual branches could not be registered with confidence after day 3. Note hyper-branched morphology on day 6 that started to refine on day 9. Cyan: Atoh1-GFP; red: TrkC-tdTomato.
The percentage of Merkel cells and axon terminals that showed plasticity, as well as regression index, is heterogeneous across touch domes within individual mice but does not differ with sex or age.
(A–C) Comparison of the range of plasticity in Merkel cells (A) and terminal neurites (B), as well as regression index (C) between imaging sessions for each touch dome analyzed. Note the variance in the amount of plasticity across afferents in the same animal. Bars indicate means, error bars indicate SDs, and dots show all time points. P values are results from Kruskal-Wallis nonparametric ANOVA (F2 Merkel cells) or ordinary one-way ANOVA (M1 terminals). (D–F) Comparison of average plasticity in Merkel cells (D) and terminals (E), and regression index (F) separated by age and sex. No significant interaction of either age or sex by two-way ANOVA for any parameter.
Two-photon laser exposure does not synchronize Merkel-cell or axon plasticity across arbors.
Analysis of Merkel-cell plasticity index (A–C) and terminal branch plasticity index (D–F) for every afferent used in analysis following the same organization as Figure 5. (C,F) Cumulative histogram of correlation coefficients from exposure-aligned experimental data and mean±SD cumulative histogram from correlation matrices of shuffled permutations. P values: permutation test. Gray dashed line indicates 50%, colored dashed lines indicate median correlation coefficients for each group. White lines in A, B, D, E indicate the arbors from each animal; black squares indicate days with unanalyzable or missing data.
Kylix formation requires Merkel cells. (A–B) Axial projections from substacks of wildtype (A) and Atoh1CKO (B) touch domes from Figure 7. Projections are of 3 z planes to show the tangled branches of the Atoh1CKO animal and absence of mature kylikes. K8 (cyan), NFH (green) and DsRed to amplify TrkC-tdTomato (red).
Information for individual mice.
