Separable mechanisms drive local and global polarity establishment in the Caenorhabditiselegans intestinal epithelium

Abstract

Apico-basolateral polarization is essential for epithelial cells to function as selective barriers and transporters, and to provide mechanical resilience to organs. Epithelial polarity is established locally, within individual cells to establish distinct apical, junctional and basolateral domains, and globally, within a tissue where cells coordinately orient their apico-basolateral axes. Using live imaging of endogenously tagged proteins and tissue-specific protein depletion in the Caenorhabditiselegans embryonic intestine, we found that local and global polarity establishment are temporally and genetically separable. Local polarity is initiated prior to global polarity and is robust to perturbation. PAR-3 is required for global polarization across the intestine but local polarity can arise in its absence, as small groups of cells eventually established polarized domains in PAR-3-depleted intestines in a HMR-1 (E-cadherin)-dependent manner. Despite the role of PAR-3 in localizing PKC-3 to the apical surface, we additionally found that PAR-3 and PKC-3/aPKC have distinct roles in the establishment and maintenance of local and global polarity. Taken together, our results indicate that different mechanisms are required for local and global polarity establishment in vivo.

Keywords: Caenorhabditiselegans; Apico-basolateral polarity; Intestine; Par3/PAR-3; aPKC/PKC-3.

Fig. 1.

Apical and basolateral proteins localize to the intestinal midline during global polarity establishment. (A-A‴) Schematics of C. elegans embryonic morphogenesis from early pre-bean (stage 1) to comma (stage 4) stages, with minutes post-fertilization (mpf) indicated. Intestinal membranes are indicated in black lines, the intestines are marked with white boxes, and enlarged insets shown in the blue boxes. The body axis directions are indicated for embryos in A-A″. In A‴, the embryo has rotated onto its side, as depicted by the axes. ‘M’ indicates the midline. (B-E‴) Dorsal view of the colocalization of the indicated endogenously tagged proteins (green) and PAR-3 (magenta) in stage 1 (PKC-3, n=12; LGL-1, n=12 ; LET-413, n=11; PAR-1, n=10), stage 2 (PKC-3, n=11; LGL-1, n= 9; LET-413, n=8; PAR-1, n=12), stage 3 (PKC-3, n=5; LGL-1, n=9 ; LET-413, n=13; PAR-1, n=11) and stage 4 (PKC-3, n=16; LGL-1, n=12 ; LET-413, n=15; PAR-1, n=12) intestines. Note that PAR-1 occasionally localized to midbodies (E, arrow) in stage 1 and was highly expressed in germ cells (E‴, asterisk). All images are maximum-intensity projections from live imaging. Intestines are outlined by white dashed lines and the midlines indicated by arrowheads. Enlarged versions of the boxed regions are shown below each panel. Yellow lines in D-D‴ depict the approximate profiles drawn to quantify apical localization of LET-413 and LGL-1 in F,G. Scale bars: 5 µm (panels); 2 µm (magnified views). (F-I) Average line profiles for LGL-1 (F) or LET-413 (H) signal across the intestinal midline from stage 1 to stage 4, and quantification of the apical/lateral LGL-1 (G) or LET-413 (I) signal.

Fig. 2.

Adherens and septate-like junctional proteins localize at different times and places during local and global polarity establishment. (A-C) Dorsal view of colocalization of the indicated endogenously tagged proteins (green) and either PKC-3 (A,B, magenta) or PAR-3 (C, magenta) in live stage 1 (HMR-1, n=11; AFD-1, n=8; DLG-1, n=14), stage 2 (HMR-1, n=13; AFD-1, n=26; DLG-1, n=13), stage 3 (HMR-1, n=11; AFD-1, n=8; DLG-1, n=17) and stage 4 (HMR-1, n=5; AFD-1, n=9; DLG-1, n=12) intestines. (D-H) Dorsolateral views of the indicated endogenously tagged proteins in live stage 5 (1.5-fold) embryos (D, n=15; E, n=12; F, n=15; G, n=12; H, n=14). Asterisks mark germ cells. Magnified views of the blue boxed regions are shown below. All images are maximum-intensity projections from live imaging. Intestines are outlined by white dashed lines and the midlines are indicated by arrowheads. Scale bars: 5 µm (panels); 2 µm (magnified views). (D″-H″) Quantification of the distance between the left and right sides of the HMR-1 and DLG-1 junctional structures, measured with a line profile for signal intensity as shown by the yellow line in D′ (two-tailed paired t-test; n.s., not significant; *P=0.0.0216; **P<0.001).

Fig. 3.

Polarity proteins are differentially required for intestinal structure and function and for larval growth. (A) Percentage of worms at the L1/L2, L3 or L4 larval or adult stages 72 h after egg lay for control [zif-1(gk117);intDeg] or intestine-specific depletion of the indicated proteins (control, n=696; PAR-3^gut(−), n=198; PKC-3^gut(−), n=132; HMR-1^gut(−), n=186; AFD-1^gut(−), n=146; DLG-1^gut(−), n=116; PAR-1^gut(−), n=171; LGL-1^gut(−), n=269). 100% of PAR-3^gut(−) and PKC-3^gut(−) worms arrested as L1 larvae. (B) Percentage of L1 larvae with normal, leaky/hazy or occluded intestinal lumens, or entirely lacking food in the intestine for control or following intestine-specific depletion of the indicated proteins (control, n=13; PAR-3^gut(−), n=16; PKC-3^gut(−), n=17; HMR-1^gut(−), n=18; AFD-1^gut(−), n=12; DLG-1^gut(−), n=11; PAR-1^gut(−), n=15; LGL-1^gut(−), n=10). (C) Representative DIC images of live worms fed blue food coloring showing the indicated phenotypic categories, with higher-magnification views of the boxed regions shown below. Scale bars: 20 µm (panels); 10 µm (magnified views).

Fig. 4.

PAR-3 and PKC-3 play different roles in AJ and SJ protein localization. (A-L″) Dorsal (stage 3) or dorsolateral (stage 4 and 1.5-fold) live images of the indicated endogenously tagged AJ or SJ proteins in control (A,D,G,J), PKC-3^gut(−) (B,E,H,K) or PAR-3^gut(−) (C,F,I,L) embryos. (B″) Gaps in junctional proteins frequently appeared in 1.5-fold PKC-3^gut(−) intestines (B″, arrow). All images are maximum-intensity projections from live imaging. Intestines are outlined by white dashed lines and the midlines are indicated by arrowheads. Scale bar: 5 µm. Stage 3 HMR-1 (control, n=14; PAR-3^gut(−), n=10; PKC-3^gut(−), n=13); AFD-1 (control, n=4; PAR-3^gut(−), n=3; PKC-3^gut(−), n=6); DLG-1 (control, n=12; PAR-3^gut(−), n=1; PKC-3^gut(−), n=7); PAR-1 (control, n=4; PAR-3^gut(−), n=2; PKC-3^gut(−), n=9). Stage 4 HMR-1 (control, n=8; PAR-3^gut(−), n=18; PKC-3^gut(−), n=12); AFD-1 (control, n=10; PAR-3^gut(−), n=7; PKC-3^gut(−), n=12); DLG-1 (control, n=24; PAR-3^gut(−), n=12; PKC-3^gut(−), n=26); PAR-1 (control, n=13; PAR-3^gut(−), n=16; PKC-3^gut(−), n=16). 1.5-fold HMR-1 (control, n=5; PAR-3^gut(−), n=12; PKC-3^gut(−), n=13); AFD-1 (control, n=6; PAR-3^gut(−), n=8; PKC-3^gut(−), n=10); DLG-1 (control, n=23; PAR-3^gut(−), n=5; PKC-3^gut(−), n=19); PAR-1 (control, n=8; PAR-3^gut(−), n=10; PKC-3^gut(−), n=8).

Fig. 5.

PAR-3 and PKC-3 are not required for apical exclusion of LET-413. (A-C) Dorsal (stage 3) and dorsolateral (stage 4 and 1.5-fold) live images of endogenously tagged LET-413 (green) in control (A-A″), PAR-3^gut(−) (B-B″) or PKC-3^gut(−) (C-C″) embryos. Embryos in A′-C″ co-express endogenously tagged DLG-1 (magenta), with magnified views of the boxed regions shown on the right. All images are maximum-intensity projections from live imaging. Intestines are outlined by white dashed lines and the midlines are indicated by arrowheads. Scale bar: 5 µm (panels); 2 µm (magnified views). Stage 3 (control, n=5; PAR-3^gut(−), n=7; PKC-3^gut(−), n=8); stage 4 (control, n=11; PAR-3^gut(−), n=5; PKC-3^gut(−), n=11); 1.5-fold (control, n=11; PAR-3^gut(−), n=8; PKC-3^gut(−), n=15). (D) Scatter dot plots of the apical and lateral LET-413 signal intensity (signal−cytoplasmic signal) for control (n=11), PKC-3^gut(−) (n=11) and PAR-3^gut(−) (n=7) 1.5-fold embryos. Data show the median with 95% c.i. Two-tailed unpaired Student's t-test was used to determine significance. (E) Quantification of the distance between the LET-413 and DLG-1 junctional structures in control (n=11) and PKC-3^gut(−) (n=9) 1.5-fold embryos. Two-tailed paired Student's t-test was used for LET-413 versus DLG-1 distance in control or PKC-3^gut(−), and two-tailed unpaired Student's t-test was used for control versus PKC-3^gut(−). n.s., not significant; *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001.

Fig. 6.

Local apico-basolateral polarity is established in the absence of PAR-3. (A,B) Localization of ACT-5 (actin, green) and TBA-1 (α-tubulin, magenta) transgenes in live 1.5-fold control (n=1) and PAR-3^gut(−) (n=10) embryos using confocal microscopy. (C-H) Localization of the indicated endogenously tagged proteins in fixed control (n=6) (C) and PAR-3^gut(−) (n=7) (D) 3-fold embryos using confocal microscopy and live control (E,G) and PAR-3^gut(−) (F,H) L1 larvae using a compound microscope. Magnified views of the boxed regions are shown on the right of each panel. Scale bars: 10 µm (panels); 2 µm (magnified views). Control (EPS-8;AFD-1/afadin, n=13; DLG-1;LET-413/Scribble, n=10); PAR-3^gut(−) (EPS-8;AFD-1/afadin, n=26; DLG-1;LET-413/Scribble, n=14).

Fig. 7.

HMR-1 is required for local polarity establishment in the absence of PAR-3. Live images of endogenously tagged AFD-1 (magenta) and DLG-1 (green) localization. (A-D) Dorsolateral images of 1.5-fold intestines from control (n=4), HMR-1^gut(−) (n=9), PAR-3^gut(−) (n=5) and (HMR-1;PAR-3)^gut(−) (n=5) embryos. Magnified views of the boxed regions are shown on the right of each panel. (E-F″) Time courses in control or (HMR-1;PAR-3)^gut(−) L1 larvae for hours post hatching (hph). Magnified views of the boxed regions are shown on the right of each panel. Control (0-1 hph, n=17; 6-7 hph, n=13; 24-25 hph, n=9); (HMR-1;PAR-3)^gut(−) (0-1 hph, n=17; 6-7 hph, n=13; 24-25 hph, n=9). Scale bar: 5 µm (panels in A-D); 10 µm (panels in E-F″); 2 µm (magnified views in A-F″).