Mammalian PAR-1 determines epithelial lumen polarity by organizing the microtubule cytoskeleton

Abstract

Epithelial differentiation involves the generation of luminal surfaces and of a noncentrosomal microtubule (MT) network aligned along the polarity axis. Columnar epithelia (e.g., kidney, intestine, and Madin-Darby canine kidney [MDCK] cells) generate apical lumina and orient MT vertically, whereas liver epithelial cells (hepatocytes and WIFB9 cells) generate lumina at cell-cell contact sites (bile canaliculi) and orient MTs horizontally. We report that knockdown or inhibition of the mammalian orthologue of Caenorhabditis elegans Par-1 (EMK1 and MARK2) during polarization of cultured MDCK and WIFB9 cells prevented development of their characteristic lumen and nonradial MT networks. Conversely, EMK1 overexpression induced the appearance of intercellular lumina and horizontal MT arrays in MDCK cells, making EMK1 the first known candidate to regulate the developmental branching decision between hepatic and columnar epithelial cells. Our experiments suggest that EMK1 primarily promotes reorganization of the MT network, consistent with the MT-regulating role of this gene product in other systems, which in turn controls lumen formation and position.

Figure 1.

Endogenous EMK1 in MDCK cells. (A) EMK localizes at the tight junctions: x-y confocal sections at the tight junction region (ap) and at a mid-plane (mid) and a confocal vertical view (x-z) of MDCK cells labeled for EMK (green) and ZO1 (red) in the absence (left) or presence (right) of the COOH-terminal EMK peptide used as an antigen to raise and purify the antibody. Note that the soluble antigen competes with EMK at tight junctions and reduces intracellular labeling. Bars, 10 μm. (B) EMK1 mRNA levels are regulated during the development of polarity RT-PCR from noncontacting (lane 2), just contacting (lane 3), and 3-d confluent (lane 4) MDCK cultures; (lane 1) control PCR without template. Top, amplification with EMK1 primers; right lane, EMK1 cDNA; bottom, amplification with primers for G3PDH, a housekeeping gene.

Figure 2.

EMK1 knockdown inhibits lumen formation in MDCK cells. (A) EMK1 knockdown: left, RT-PCR with EMK1-specific primers of total RNA from control or EMK1-KO cells; EMK1-cDNA was used as a positive control (plasmid); EMK-immunoblot of control and EMK1-KO lysates; asterisk, EMK1 band; right, wide-field image of cells 24 h after transfection with pSUPER (control) or the pSUPER-siEMK1 plasmid (EMK1-KO) taken with the same exposure; cells were immuno-labeled for EMK. (B) Collagen overlay: control (transfected with pSUPER) or EMK1-KO cells were cultured for 24 h on collagen I before being overlaid with additional collagen on the apical surface and analyzed 24 h later. Green, gp135; red, phalloidin; and blue, nucleus. Note the lack of lumen in EMK1-KO cultures. Bars, 10 μm. (C) Ca switch: control or EMK1-KO cells were plated in low Ca medium 24 h upon transfection with pSUPER or pSUPER-KO. After 12 h, cultures were switched to normal medium for 24 h. Top, x-y and x-z confocal projections labeled for gp135 (green), β-catenin (red), and ZO-1 (blue) or for E-cadherin (green). Bars, 10 μm. Bottom, transmission EM of cells sectioned perpendicularly to the substratum; arrow points to microvilli. Bars, 1 μm.

Figure 3.

KN-EMK1 mimics EMK1-KO in MDCK cells. (A) KN-EMK1-myc expression: x-y confocal sections at the tight junction region (ap) and at a mid-plane (mid) and a confocal vertical view (x-z) of MDCK cells labeled for myc (green) and ZO1 (blue) after expression of myc-tagged KN-EMK1 by adenovirus-mediated gene transfer in fully polarized cells. (B) Collagen overlay: cells were transduced with either control (GFP) or KN-EMK1 virus 12 h after plating on collagen and were cultured for an additional 12 h before collagen overlay for 24 h. x-z view of tubule with gp135 (green) and phalloidin (red); nucleus (blue). Note that KN-EMK1 expression leads to loss of apical domain. (C) Ca switch: contact-naive MDCK cells were transduced with the KN-EMK1 virus or a GFP-virus (control) 12 h before Ca switch for 24 h. Red, gp135; blue, ZO-1; and green, E-cadherin. Bars, 10 μm. Note that intracellular GFP fluorescence is negligible under the virus titer used for infection.

Figure 4.

EMK1 promotes lateral lumina in MDCK cells. Wild-type EMK1 was expressed in polarized (A) or in polarizing (B and C) MDCK cells under a tetracycline promotor; control, EMK1 + tet; EMK1-overexpressing cells, EMK1 − tet. (A) Recombinant EMK1-myc expression: x-y-confocal sections at the tight junction region (ap) and at a mid-plane (mid) and a confocal vertical view (x-z) of MDCK cells labeled for myc (green) and ZO1 (blue) after expression of myc-tagged wild-type EMK1 by doxycycline withdrawal from fully polarized cells. Bars, 10 μm. (B) Collagen I overlay: subconfluent monolayers were grown on a collagen I substrate and overlaid with collagen I for 12 (top) or 36 h (bottom); confocal x-z sections depicting the lumen between two cell layers; red, gp135; green, FITC-phalloidin. Bars, 10 μm. (C) Ca switch: filter-grown cells 0 or 24 h after Ca switch. (top) Green, gp135; red, ZO-1, and blue, nucleus. x-y and x-z sections; arrows indicate VACs (0 h) and lateral lumina (24 h); arrowheads indicate the positions of the cell apex and cell base. (middle) E-cadherin (green) and ZO-1 (red) 24 h after Ca switch. Bars, 10 μm. (bottom) Transmission EM of EMK1-cells sectioned parallel to the substratum 24 h after Ca switch; three cells (“nu” indicates nucleus) are arranged around a lumen (asterisks); note microvilli in the lumen that is enlarged in the detailed image on left. Bars: (right) 2 μm; (inset) 1 μm.

Figure 5.

EMK1 regulates BC formation in WIFB9 cells. (A) EMK1 localization: (left) x-y wide-field images of cells in phase (right) and colabeled for EMK (left) and ZO-1 (middle); (far right) wide-field image labeled for myc after expression of myc-tagged KN-EMK1 by adenovirus-mediated gene transfer. (B) BC formation in control cells: x-y phase and confocal fluorescence images of cells labeled for DPPIV, CE9 (green), and ZO1 (red) at day 5 (top) or day 15 (bottom) after plating. (C) BC formation upon KN-EMK1 expression cells were transduced with a GFP-control virus (GFP) or with KN-EMK1 expressing adenovirus 48 h before analysis. Bars, 10 μm. (fluorescence) DPPIV in control and KN-EMK1 cells at day 9; (histogram) BCs, identified as intercellular translucent holes, were counted in phase images from ∼1,500 cells at day 5, 7, 9, 11, 13, and 15 after plating. Cells were infected at day 5, 7, 9, and 13 with equal pfu of GFP (blue bars) or KN-EMK1 (red bars) expressing adenovirus and analyzed 48 h after infection; black bars, uninfected cells. Data are from three independent experiments. Bars, 10 μm.

Figure 6.

EMK1 promotes a hepatic MT organization. WIFB9 cells at day 11, 2 d after expression of a GFP virus (control) or KN-EMK1 adenovirus (KN-EMK1); EMK1-MDCK cells 24 h after Ca switch in the presence (EMK1) or absence (control) of the recombinant kinase. Confocal sections at the plane of the centrosome; pericentrin, a centrosomal marker (red), MT (green), and nucleus (blue). (scheme) Red, centrosome; green, MT with their +/− end orientation; and brown, luminal markers. Arrows indicate centrosomes. Bars, 10 μm. Note that cells were detergent extracted before fixation, leading to complete GFP extraction in WIFB control cells.

Figure 7.

EMK1 promotes MT assembly at the lateral cortex in MDCK cells. (A) Recombinant EMK1 was induced by doxycycline withdrawal from EMK1-MDCK cells after they had polarized (EMK1); endogenous EMK1 was knocked down (EMK1-KO) by transfection with an EMK1-siRNA vector, and cells were cultured for 48h; controls shown are representative of EMK1-MDCK cells maintained in doxyxycline and of MDCK cells transfected with control siRNA-vector. MT organization, confocal sections of MT at the most apical (ap) and at mid-plane (mid); arrow, MT at lateral cortex; MT stability, apical and lateral MT after 1 h nocodazole treatment at 37°C; and MT nucleation, MT asters (green) that represent the MT nucleation center during MT repolymerization after cold treatment. β-tubulin labeling is in green in all panels. In addition, control and EMK1 cells show β-catenin (red) in the x-y views or the nucleus (blue) in x-z-views. Pericentrin, a centrosomal marker (red), localizes to the β-tubulin aster in the x-z sections. In EMK1-KO cells, E-cadherin and the nucleus are labeled in blue. Bars, 10 μm. Models (right): schematic depiction of the differences in centrosome localization (red dots) and MT arrangements (red lines); black arrows indicate amount of vertical MTs. (B) MT regrowth: serial apical to basal x-y sections from control and EMK1-KO cells labeled for MTs (green), β-catenin (red), and γ-tubulin (blue) 4 min after nocodazole washout. Note MTs released from the MTOC (labeled by γ-tubulin; asterisks) at the lateral cortex in control cells and in the perinuclear region in EMK1-KO cells (arrows). Bars, 10 μm.

Figure 8.

EMK1 regulates MT and lumen polarity in a two-step process. The three model systems used in this work: Ca switch in MDCK cells (top path), collagen overlay of MDCK monolayers (middle path), and long-term cultures of the hepatocyte cell line WIFB9 (bottom path). In all systems, columnar polarity is lost transiently, either upon culture in low Ca²⁺ levels or addition of collagen overlay (MDCK), or by spontaneous loss of a transitional columnar stage (WIFB9). In the nonpolarized state, luminal markers (green) accumulate in intracellular storage compartments (MDCK before Ca switch; WIFB) or are undetectable (MDCK cells under collagen overlay) and the MTs (red lines) have a radial arrangement (MTOC, red dots). Increased EMK1 signaling (1) causes conversion of the nonpolarized stage into a “hepatic polarity stage” with a horizontal MT array and a subluminal centrosome and luminal domains at sites of cell–cell contact (VAC exocytosis in Ca switch, spontaneous intercellular lumen formation in collagen overlay and in WIFB9 cells). In step 2, EMK1-signaling levels regulate the final polarity of the epithelium. High EMK1-signaling levels, as in EMK1-overexpressing MDCK cells or in WIFB9 cells (hypothetical in this case), consolidate lateral lumen polarity and horizontal MT organization (“hepatic-type”). Lower EMK1-signaling levels, such as in control MDCK cells, lead to a relocalization of the luminal domain to the cell apex and vertical MT (columnar organization).