
==== Front
Proc Natl Acad Sci U S A
Proc Natl Acad Sci U S A
PNAS
Proceedings of the National Academy of Sciences of the United States of America
0027-8424
1091-6490
National Academy of Sciences

38466856
202401735
10.1073/pnas.2401735121
commCommentarycell-bioCell Biology409
437
Commentary
Biological Sciences
Cell Biology
Too old for hide-and-seek: Cell maturation reveals hidden apical junctional organization
Jensen Corbin C. a
Peifer Mark peifer@unc.edu
a b 1
aDepartment of Biology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599
bLineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599
1To whom correspondence may be addressed. Email: peifer@unc.edu.
11 3 2024
19 3 2024
11 9 2024
121 12 e2401735121Copyright © 2024 the Author(s). Published by PNAS.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).

HHS | National Institutes of Health (NIH) 100000002 NIH R35 GM118096 Mark Peifer
==== Body
pmcEpithelia are a premier animal innovation, separating us from the outside world and demarcating body compartments. To organize individual cells into epithelia, animals evolved specialized machinery that mediates adhesion cell to cell and of cells to the underlying extracellular matrix. These cell junctions link to the cytoskeleton, providing structural support and, in the case of the Actin and myosin cytoskeleton, generating force to drive cell shape change and motility. Pioneering electron microscopy in the 1970s provided an initial picture of cell–cell junction diversity and how they are organized along the cell borders (1) and by the 1990s molecular approaches identified their major protein components. In PNAS, Mangeol et al. substantially update our understanding of cell–cell junctions (2).

Textbook images portray three tiers of junctional complexes (Fig. 1A) (3). In vertebrates, the most apical are tight junctions (TJs; zonula occludens), where transmembrane claudins and their transmembrane and cytoplasmic partners assemble into TJ strands to regulate paracellular movement of molecules, sealing the sheet to passive diffusion. Just below them are adherens junctions (AJs; zonula adherens), where transmembrane E-cadherin (Ecad) and its cytoplasmic catenin partners mediate cell adhesion. Both junctions link to Actin, via proteins like ZO-1 in TJs and alpha-catenin in AJs. More basally, vertebrates have desmosomes, where desmosomal cadherins reinforce adhesion and link to intermediate filaments. Insect cells share apical AJs but their claudin-based barrier junctions, called septate junction, are basal to AJs, and desmosomes are absent (Fig. 1A).

Fig. 1. An emerging model of apical junctions. (A) The classical view of cell–cell junctions. (B) A revised model of cell-cell junctions depicting emergence during maturation of a two-layered structure, with Nectins/Afadin apical to E-Cadherin/catenins. Adapted from ref. 2.

Over time, new discoveries revised this simple picture. Work in Drosophila and C. elegans identified cell polarity regulators that had to fit into this picture. The “Par complex” (Par3, Par6, and aPKC) and “Crumbs complex” (Crumbs, Stardust/PALS, and PATJ) regulate different aspects of apical polarity. Both localize apical to AJs, with the Par complex localized to vertebrate TJs and the Crumbs complex even more apical (4). Other data revealed temporal complexity. For example, in mammalian cultured cells, the TJ protein ZO-1 localizes to AJs early in epithelial junction formation in some cell types (5), before segregating apically, while in Drosophila Bazooka/Par3 initially co-localizes with Ecad as AJs assemble and later localizes between AJs and its apical Par complex “partners” (6). The Drosophila Crumbs complex assembles later in embryogenesis than AJs or the Par complex and fully formed septate junctions assemble even later. TJ formation also follows AJ assembly in mammalian embryos.

In PNAS Mangeol et al. substantially update our understanding of cell-cell junctions.

The role of AJs in cell shape change is well established. It was initially suggested that Actin linkage is simple and direct: Ecad links cells together by homophilic adhesion, and cadherin cytoplasmic tails bind beta-catenin (βcat) which binds alpha-catenin which binds Actin. Work in the mid-2000s challenged this picture (7), and it was revised in two important ways (8). A network of proteins helps link AJs to Actin, and this network is mechanosensitive, tightening linkage when force is exerted on it. Afadin and its Drosophila homolog Canoe are prominent proteins in the network. Afadin, identified as an Actin-binding protein localized to AJs (9), is a cytoplasmic binding partner of Ig-superfamily nectins (10) that mediate homophyllic and heterophyllic adhesion. The Takai lab initially suspected that nectins would be key for cell–cell adhesion, like Ecad, but later work revealed a role in cell sorting (e.g. ref. 11). Canoe/Afadin is also not essential for adhesion but plays a key role in linking AJs to the cytoskeleton during diverse morphogenetic events in flies and mammals. In its absence, the cytoskeleton separates from AJs, disrupting morphogenesis (12, 13). With these revisions, the current view is that AJs are the major force-transducing lateral junction, with cadherin–catenin complexes mediating this linkage.

In an exciting new paper, Mangeol et al challenge this view (2). Using Stimulated Emission/Depletion (STED) superresolution microscopy both in vivo and in vitro, the authors obtained a view of apical junction organization in intestinal epithelia with unprecedented detail. STED bypasses the diffraction limit of light by using two lasers; one to excite the fluorophore and the second to deplete the resulting emission. The second laser creates a donut shape-dampening effect that allows only emission of signal from the very center of the donut to reach the detector, increasing resolution. The authors further increased their spatial resolution via cryo-sectioning, whereby the z-axis effectively becomes the x/y-axis. These approaches increased resolution threefold in the junctional plane and sevenfold in the apicobasal plane. The authors then sectioned biopsies of the human small intestine, immunolabelled junctional proteins and used STED to determine the spatial organization of apical junctions in vivo and their association with junctional Actin.

In these cells, apical microvilli transition into an apicolateral Actin belt. AJs and TJs were clearly distinguished. ZO-1 associated with the most apical region of the 300 nm wide Actin belt. Nectin-2 and Afadin occupied the more basal part of the lateral Actin belt, and as expected, there was no overlap between Afadin signal and that of ZO-1. Unexpectedly, however, Ecad and βcat did not localize with the Actin band. Quantification revealed that the peak Ecad/catenin signal was shifted 180 nm basal of the peak Actin signal. Furthermore, Ecad had very little overlap with Afadin. These data suggest a two-layered AJ with distinct Ecad/catenin and Nectin/Afadin/Actin zones (Fig. 1B). This segregation in vivo prompted the authors to extend these experiments in vitro using cultured Caco-2 intestinal cells seeded on filters, grown to confluence and allowed to fully polarize for 14 d. Strikingly, they observed similar results to those in vivo. Ecad and the catenins clustered together in a region below Nectin-3 and Afadin, with ZO-1 even more apical. On par with their in vivo results, Ecad localized 150 nm basal of the peak Actin belt. It is worth noting that there was some overlap between Ecad and Actin, but the Actin signal overlapping Ecad was 6-times weaker than that overlapping Afadin. Consistent with their in vivo data, Nectin and Afadin co-localized with the peak Actin signal and localized apical to βcat and basal to ZO-1.

These results were striking, suggesting new AJ complexity. The authors next asked how this was missed before. The use of STED is one reason, but they also discovered a more interesting reason—things change over time, as cells fully polarize. At day 6, when Caco cells just reached confluence, Ecad and Afadin signals overlapped, while both were already basal to TJs (Fig. 1B). By day 9, Ecad and βcat localization became largely distinct from Afadin. Interestingly, at days 6 and 9, the peak Actin signal localized with ZO-1 and TJs, before relocating to the Nectin/Afadin layer at day 14. Thus, the dual-band AJ only became apparent after cells had time to mature, and as such, the authors termed this phenomenon the zonula adherens matura.

The authors then set out to define mechanisms driving segregation of Nectin/Afadin/Actin from Ecad/catenins. They generated Afadin-Knockout (KO) cells using CRISPR and repeated their experiments in mature cells at day 14. Afadin-KO profoundly changed apical junctions. Nectin-3 and ZO-1 now colocalized in a band apical to Ecad, whereas Actin localization was more diffuse and no longer tightly associated with the Nectin band. Changes in Actin localization were also apparent when they visualized the sheet from above. Afadin-KO cells had a gap in the Actin belts where cells meet that was not seen in wild-type cells, suggesting Actin belts move from lateral junctions to the terminal web.

These changes in Actin localization in Afadin-KO cells prompted the question of what effect this has on mechanical tension. Cells in a monolayer experience three sources of tension. The first is through adhesion to extracellular matrix, the second is tension at cell–cell junctions as cells push and pull on neighbors, and the third is tension at the apical surface. All are inter-related as changing one can affect the others. Afadin’s localization at apical junctions suggested tension differences lay there. The authors addressed Afadin’s role in apical tension and force transmission between cells two ways. First, they co-cultured wild-type (WT) and Afadin-KO cells and examined WT–KO hetero-junctions. At most WT–KO interfaces (75.7%), the lateral membrane curved in toward the WT cells, consistent with higher apical tension in WT cells. To directly test whether WT–KO junctions had higher interfacial tension, the authors used laser ablation to sever cell–cell junctions and measured recoil velocity. Surprisingly, the recoil velocity was not significantly different between WT–WT and WT–KO borders, indicating the two junction types have similar tension. They then tested whether the curvature difference was a result of Afadin-KO cells having a lower apical surface tension. They used laser ablation to disrupt the terminal web, an Actin meshwork just below the apical surface that works with junctional Actin belts to form the apical contraction apparatus, and measured the cellular response. WT cells spread 30% faster than Afadin-KO cells immediately following laser ablation, and WT cell surface area was 50% larger 50 s post-ablation. Both results indicate that Afadin-KO cells have significantly lower apical surface tension than WT cells. Thus, the curvature results indicate that higher apical tension in WT cells causes them to “pull” on more “squishy” Afadin-KO cells. Together, these data suggest that Afadin is important for cells to generate appropriate apical tension.

Overall, the exciting data presented by the authors revise the picture of AJs, revealing that, in at least some cell types, a two-layered AJ is found, with apical nectins/Afadin and more basal Ecad/catenins. This fits with the observation that in many cell types, the Ecad–catenin complex localizes all along the lateral membrane, while Afadin is restricted to AJs. Using biopsies from human small intestines provided compelling evidence that this phenomenon is not merely a by-product of cell lines or culture conditions, thus challenging the classical view of AJs in intestinal epithelia and potentially beyond. It will be exciting to see how widespread this is, especially in places where epithelia are being rapidly remodeled. For example, all of Drosophila embryonic morphogenesis occurs in ~14 h, relative to the multi-day maturation here, and during these events Canoe and βcat localization overlap, with only a slight apical shift of Canoe (14). Even in MDCK cells maturation after calcium switch takes hours rather than days (15). The key role of Afadin seen here echoes the critical roles played by Canoe and Afadin in vivo, where Canoe/Afadin loss detached the actomyosin cytoskeleton from cell junctions (12, 13). However, the importance of nectins in this remains an open question—for example, loss of the Drosophila nectin relative Echinoid disrupts only a subset of the morphogenetic events requiring Canoe, and Echinoid’s PDZ-binding motif, which interacts with Canoe/Afadin, is not essential for its function (16).

Author contributions

C.C.J. and M.P. wrote the paper.

Competing interests

The authors declare no competing interest.

See companion article, “The zonula adherens matura redefines the apical junction of intestinal epithelia,” 10.1073/pnas.2316722121.
==== Refs
1 M. G. Farquhar, G. E. Palade, Junctional complexes in various epithelia. J. Cell Biol. 17 , 375–412 (1963).13944428
2 P. Mangeol , The zonula adherens matura redefines the apical junction of intestinal epithelia. Proc. Nat. Acad. Sci. U.S.A., 121 , e2316722121 (2024).
3 B. Vanslembrouck, J. H. Chen, C. Larabell, J. van Hengel, Microscopic visualization of cell-cell adhesion complexes at micro and nanoscale. Front Cell Dev. Biol. 10 , 819534 (2022).35517500
4 E. Martin, R. Girardello, G. Dittmar, A. Ludwig, New insights into the organization and regulation of the apical polarity network in mammalian epithelial cells. FEBS J. 288 , 7073–7095 (2021).33448150
5 J. Ikenouchi, K. Umeda, S. Tsukita, M. Furuse, S. Tsukita, Requirement of ZO-1 for the formation of belt-like adherens junctions during epithelial cell polarization. J. Cell Biol. 176 , 779–786 (2007).17353356
6 T. J. Harris, M. Peifer, The positioning and segregation of apical cues during epithelial polarity establishment in Drosophila. J. Cell Biol. 170 , 813–823 (2005).16129788
7 S. Yamada, S. Pokutta, F. Drees, W. I. Weis, W. J. Nelson, Deconstructing the cadherin–catenin–actin complex. Cell 123 , 889–901 (2005).16325582
8 A. S. Yap, K. Duszyc, V. Viasnoff, Mechanosensing and mechanotransduction at cell-cell junctions. Cold Spring Harb. Perspect. Biol. 10 , a028761 (2018).28778874
9 K. Mandai , Afadin: A novel actin filament-binding protein with one PDZ domain localized at cadherin-based cell-to-cell adherens junction. J. Cell Biol. 139 , 517–528 (1997).9334353
10 K. Takahashi , Nectin/PRR: An immunoglobulin-like cell adhesion molecule recruited to cadherin-based adherens junctions through interaction with Afadin, a PDZ domain-containing protein. J. Cell Biol. 145 , 539–549 (1999).10225955
11 H. Togashi , Nectins establish a checkerboard-like cellular pattern in the auditory epithelium. Science 333 , 1144–1147 (2011).21798896
12 J. K. Sawyer, N. J. Harris, K. C. Slep, U. Gaul, M. Peifer, The Drosophila afadin homologue Canoe regulates linkage of the actin cytoskeleton to adherens junctions during apical constriction. J. Cell Biol. 186 , 57–73 (2009).19596848
13 W. Choi , Remodeling the zonula adherens in response to tension and the role of afadin in this response. J. Cell Biol. 213 , 243–260 (2016).27114502
14 A. Schmidt , Polychaetoid/ZO-1 strengthens cell junctions under tension while localizing differently than core adherens junction proteins. Mol. Biol. Cell. 34 , ar81.37163320
15 H. C. Yu-Kemp , Micron-scale supramolecular myosin arrays help mediate cytoskeletal assembly at mature adherens junctions. J. Cell Biol. 221 , e202103074 (2022).34812842
16 C. Laplante, L. A. Nilson, Asymmetric distribution of Echinoid defines the epidermal leading edge during Drosophila dorsal closure. J. Cell Biol. 192 , 335–348 (2011).21263031
