Pronephric epithelial proliferation.

(A–C,E,F) Red: anti-BrdU staining, Green: anti-GFP staining. BrdU incorporation was measured for 24 h, between 48 and 72 hpf. (A) Proliferation in the proximal tubule is localized to the neck region (arrow). (B) Proliferation in the anterior part of the distal tubule is low. (C) Proliferation in the ret1-positive pronephric duct segment is high. (D) Cartoon depicting zebrafish pronephros. The proximal tubule (ET33d10 GFP positive domain) and the pronephric duct (ret1:GFP positive domain) are shaded black. The glomerulus and the distal tubule (ET11-9 GFP positive domain) are not shaded. (E) Proliferation in the posterior proximal tubule is low. (F) Proliferation in the posterior distal tubule is high. (G) Quantification of pronephric epithelial proliferation after 24 hour BrdU incorporation. The length of the tubule is plotted on the horizontal axis (measured from the cloaca). The linear density of BrdU positive nuclei (per 100 μm tubule length) is plotted on the vertical axis. Squares: pronephric proliferation between 3 and 4 dpf (n = 3). Circles: proliferation between 1 and 2 dpf (n = 3). See also Figure S1.

Collective epithelial migration stimulates cell proliferation in the distal tubule.

(A, B) Cell proliferation in the distal tubule (ET11-9 GFP domain) after anterior obstruction. Obstruction was induced at 30 hpf, BrdU incorporation was assessed between 2 and 3 dpf. (A) Total number of BrdU+ nuclei in the distal ET11-9 domain. White bar: control (n = 4), black bar: anterior obstruction (n = 4). P = 0.01. (B) Spatial distribution of the BrdU-positive nuclei per 100 µm length of the distal tubule (measured from the posterior border of the ET11-9 domain). Squares: control (n = 4), Circles: anterior obstruction (n = 4). (C,D) Cell proliferation in the distal nephron in mindbomb mutants. mindbomb (mib) heterozygotes were in-crossed and injected with tnnt2 morpholino to control for vascular defects in mib mutants. BrdU incorporation was assessed between 2 and 3 dpf. Homozygous mib mutants were separated from their siblings based on their axis curvature phenotype. (E) Total amount of BrdU incorporation in pronephric duct (black bar: control, n = 8; dark-grey bar: mindbomb, n = 8; P>0.05) and in the posterior distal tubule (light grey bar: control, n = 8; white bar: mindbomb, n = 8; P<0.01). (D) Linear density of the BrdU+ nuclei. Squares: tnnt2MO only control (n = 4); circles: mindbomb +tnnt2MO (n = 4). The underlying bars (B,D) indicate the position of distal tubule/pronephric duct border. (E) 24 hour BrdU incorporation in the posterior proximal tubule (2–3 dpf). (F) BrdU incorporation during 24 hour post-obstruction (2–3 dpf). (G) total number of BrdU positive nuclei in the distal 600 µm of proximal tubule. Black bar: control (n = 4), white bar: obstructed nephrons (n = 8); P<0.01. (H) Proliferation profile in control ET11-9/Tg(atp1a1a.4:GFP) transgenic fish compared to LY294002 treated fish (2–3 dpf). Arrows point to the location of distal tubule/pronephric duct interface. (I) Up-regulation of BrdU incorporation in stretched proximal tubule between 12 and 36 hours post-obstruction. (J) BrdU incorporation was markedly reduced in LY294002 treated, obstructed tubules. (K) Total number of BrdU-positive nuclei in the anterior 500 μm of the proximal tubule. White bar: BrdU incorporation in obstructed nephron/−LY294002 (n = 3), black bar: BrdU incorporation in obstructed nephron/+LY294002 (n = 3); P<0.05. (E,F) Green: GFP (ET33d10 GFP); (I,J) Green – cadherin17; (E,F,I,J) Red: BrdU.

Distal proliferation supports prolonged epithelial migration.

(A–D) 1.5 µm confocal slices of the distal tubule. Green: GFP (ET11-9 GFP transgenic in A–B, Tg(atp1a1a.4:GFP) transgenic in C&D). Red: BrdU. Magenta: DAPI. (A) BrdU incorporation in the distal tubule (2–3 dpf) in control fish. (B) Lack of BrdU incorporation in distal tubule between (2–3 dpf) when embryo is treated with 30 µM LY294002. Cells outside of the kidney continued to incorporate BrdU. (C) DAPI staining of distal tubule in control fish (4 dpf). (D) DAPI staining of distal tubule treated with 30 µM LY294002 (4 dpf). The distal tubule was markedly thinned in LY294002 condition (B, D). (E) estimated cross-sectional tubule area based on measured maximal diameter of the tubule in confocal stacks at 3 dpf. Circles: control (n = 5). Squares: LY294002 treated fish (n = 3). (F) Linear nuclear density (DAPI) in control fish (circles, n = 3) and LY294002 treated fish (squares, n = 3) confirmed linear stretching of the distal tubule in Ly294002 treated fish (4 dpf). (G, H) Kidney segment lengths in (G) control fish and (H) LY294002 treated fish at 4dpf. ET11-9 transgenic fish were used to estimate the segment lengths: ‘a’, ‘b’ and ‘c’ represent proximal tubule, distal tubule and the pronephric duct. (I–N) Individual frames of time lapse movies of the actively migrating pronephric epithelia in the presence of 30 µM LY294002. Arrowheads point to the individual traced cells. Time lapse immediately after addition of LY294002 (I, J), 24 hours after addition of LY294002 (K, L), and 48 hours after addition of LY294002 (M, N). Frame pairs in I and J, K and L, and M and N are separated by 6 hours. (O) Average epithelial migration rates after different durations of LY294002 exposure. Each bar represents average migration rate of 4 different individual cells in the proximal tubule (ET11-9:GFP). (P) Lengths of proximal tubule (designated ‘a’ in G), distal tubule (designated ‘b’ in G) and pronephric duct (designated ‘c’ in G) in control 96 hpf fish and fish treated with 30 μM LY294002 starting at 30 hpf. White bars: control (n = 9), black bars: LY294002 (n = 9). See also Figure S2.

Modeling pronephric migration and proliferation.

(A) Initial arrangement of cells in the simulation. Squares: cell positions. All cells in the chain are presumed to be responsive to fluid flow. The difference between the tubule segment (long bar) and the duct segment (short bar) is that cells in the tubule segment migrate faster than cells in the duct segment. (B) Collective migration results in piling up of epithelial cells in the proximal nephron (bracket) and an increase in total number of epithelial cells (gray squares indicate newly formed cells) due to stretch-dependent distal tubule proliferation. The arrowhead points to the zone of proliferation. The number of iterations = 700. (C) When the threshold for stretch-induced cell proliferation was increased, migration came to a premature halt and, as a result, proximal convolution was significantly reduced (bracket) and distal tubule became overstretched (bracketed arrow). The number of iterations = 700. (D) When only the distal half of the kidney is subjected to the directional migration bias, the ectopic convolution develops (bracket). The number of iterations = 100.The starting number of cells in simulations A–D was 40. (E,F) The model predicted that if at any time during active migration (such as in E (100 iterations)) the cue for directional migration (fluid flow) was eliminated, the direction of migration would temporarily reverse (F, arrow in E). (E, F) Arrowheads point to the same cell. The additional number of iterations in F = 100. The starting number of cells in simulations E–F was 20. (G) Simulations predicted that inhibiting distal proliferation should result in premature arrest of migration (lower trace) while in control condition the migration rate remained relatively constant (upper trace). The migration rate was measured over the five cells (20th–24th cells counting from the back end of the chain). (H) Inhibiting distal proliferation in live embryos using LY294002 resulted in linear stretch of the distal kidney epithelial cells as evidenced by the increase in the internuclear distance (inset). The same effect is predicted in our simulations (upper trace – no proliferation condition vs. lower trace – control condition). The total number of cells in this simulation = 40. The distal half is displayed. (I,J) Reversal of the direction of migration can be observed during stochastic transient tubule obstruction. (I) Transiently obstructed tubule (above) and unobstructed tubule (below) in the ET33d10 transgenic fish at time 0. (J) the same two tubules 1 hour later. Arrowheads point to the individual traced cell in each tubule. Arrows indicate the direction of migration. See also Movie S10. (K) Diagram showing the portion of the zebrafish kidney that was imaged in I and J.

mindbomb mutants show robust proliferative response to cell stretch, related to Figure 2. Mindbomb mutants were obstructed at 36 hpf and incubated in the presence of BrdU between 48 and 72 hpf. The embryos were then fixed and stained with anti- Cadherin17 (green) and anti- BrdU (red) antibody. Robust BrdU incorporation was observed. Single 1.5 µm confocal slice is shown.

Obstruction-induced cell stretch does not depend on cell proliferation, related to Figure 2. 48 hpf ET33d10 fish were subjected to distal obstruction for 8hours, fixed and stained with anti- pospho histone H3 (green) and anti alpha6F (red) antibodies (A,B). 60 μM Camptothecin was used to inhibit cell proliferation (monitored by the amount of phospho histone H3 staining). Maximal luminal diameter of the proximal kidney was used as a measure of radial cell stretch and was identical in the experimental and the control condition (C). (A,B) show flattened confocal stacks. The apparent kidney staining with phospho histone H3 in (B) is actually in a different focal plane (in the CNS). No significant kidney phospho histone H3 staining is observed after 8hours of obstruction.

Inhibiting CDK4/CyclinD1 signaling phenocopies the effects of inhibiting PI3K, related to Figure 3. Inhibiting CDK4/CyclinD1 using 20 μM NSC625987 results in exaggerated linear stretching of the distal tubule (A), ET11-9 transgenic. This chemical treatment also resulted in significant reduction in distal kidney BrdU incorporation (B), P<0.05. (C): the segment lengths were affected by NSC625987 treatment in a manner very similar to LY294002. The proximal tubule becomes significantly longer (P<0.01), the collecting duct becomes significantly shorter (P<0.01). Unlike the LY294002 result, the distal tubule is slightly shorter in NSC625987 treatment condition. This small, but statistically significant effect (P<0.05) could be due to small direct effect of NSC625987 on the migration rate (distal tubule elongation is migration dependent).

Partial inhibition of proximal convolution by LY294002, related to Figure 4. Tg(atp1a1a.4:GFP) transgenics were analysed for the presence and the extent of proximal convolution at 4 dpf in control embryos (A) vs. the LY294002 treated embryos (B). The estimation of the degree of convolution by a convolution index shows partial but significant reduction in the degree of proximal convolution (C, P<0.01).

Acknowledgments
This image is the copyrighted work of the attributed author or publisher, and ZFIN has permission only to display this image to its users. Additional permissions should be obtained from the applicable author or publisher of the image. Full text @ PLoS One