Cell Physiology


Airway goblet cells secrete mucin onto mucosal surfaces under the regulation of an apical, phospholipase C/Gq-coupled P2Y2 receptor. We tested whether cortical actin filaments negatively regulate exocytosis in goblet cells by forming a barrier between secretory granules and plasma membrane docking sites as postulated for other secretory cells. Immunostaining of human lung tissues and SPOC1 cells (an epithelial, mucin-secreting cell line) revealed an apical distribution of β- and γ-actin in ciliated and goblet cells. In goblet cells, actin appeared as a prominent subplasmalemmal sheet lying between granules and the apical membrane, and it disappeared from SPOC1 cells activated by purinergic agonist. Disruption of actin filaments with latrunculin A stimulated SPOC1 cell mucin secretion under basal and agonist-activated conditions, whereas stabilization with jasplakinolide or overexpression of β- or γ-actin conjugated to yellow fluorescent protein (YFP) inhibited secretion. Myristoylated alanine-rich C kinase substrate, a PKC-activated actin-plasma membrane tethering protein, was phosphorylated after agonist stimulation, suggesting a translocation to the cytosol. Scinderin (or adseverin), a Ca2+-activated actin filament severing and capping protein was cloned from human airway and SPOC1 cells, and synthetic peptides corresponding to its actin-binding domains inhibited mucin secretion. We conclude that actin filaments negatively regulate mucin secretion basally in airway goblet cells and are dynamically remodeled in agonist-stimulated cells to promote exocytosis.

  • lung
  • mucus
  • exocytosis

in healthy lungs, the viscoelastic mucus gel plays a major role in mucociliary clearance (29), a key component of host defense. Polymeric mucins, the principal component of mucus, are high-molecular-weight glycoconjugates secreted by airway goblet cells and submucosal glands. In airway obstructive diseases such as chronic bronchitis and cystic fibrosis, the airways can become occluded as a result of mucus hypersecretion caused by goblet cell and glandular hyper- or metaplasia, and early phases of these diseases implicate goblet cells in the lower airways selectively (23). While substantial knowledge exists with regard to the regulation of goblet cell secretion by agonists and intracellular messengers (10, 11, 28), very little is known about the downstream effectors activated after cell stimulation. For example, numerous studies have shown that goblet cells respond primarily to purinergic agonist stimulation via a luminal P2Y2 receptor (P2Y2R) that couples through Gq to phospholipase C (PLC) to generate diacylglycerol and inositol 1,4,5-trisphosphate, with a resulting activation of PKC and mobilization of Ca2+ (12, 28). The downstream effector molecules activated by PKC and Ca2+ that direct and regulate mucin granule exocytosis, however, are virtually unknown.

Actin is a structural protein of great importance in eukaryotic cells, playing a major role in muscle contraction, cell motility, and many other cell functions. During the past 20 years, actin filaments have also been shown to regulate membrane trafficking, especially as it relates to endo- and exocytosis (for review, see Refs. 14, 16). Classically, cortical actin filaments are hypothesized to negatively regulate endocrine secretory cell exocytosis by forming a physical barrier between secretory granules and docking sites; in this article, we refer to this hypothesis as “the barrier hypothesis.” In chromaffin cells, for example, several studies have demonstrated that the disruption of actin filaments is essential for exocytosis (18, 59). Disruption of the cortical actin barrier before exocytosis is hypothesized to be caused by two proteins acting in parallel (60). First, the PKC-activated protein myristoylated alanine-rich C kinase substrate (MARCKS; see Ref. 4) tethers actin filaments to the plasma membrane under resting conditions and translocates to the cytosol upon phosphorylation by PKC, thereby disrupting the cytoskeleton. In chromaffin cells (49), platelets (15), and bovine luteal cells (52), the phosphorylation of the membrane-bound MARCKS protein is required for regulated exocytosis. Second, scinderin, a Ca2+-activated, gelsolin-related actin filament-severing and -capping protein (see Refs. 34, 48) has been implicated in cytoskeletal disruption occurring before exocytosis. In stimulated chromaffin cells, scinderin has been shown to control the disassembly of the cortical F-actin network after activation by Ca2+ (59, 62).

Alternatively, more recent data suggest that actin filaments act positively in regulated exocytosis, presumably by transporting or positioning secretory granules relative to plasma membrane docking sites (see, e.g., Refs. 17, 41).

The role of actin filaments in airway goblet cell secretion is generally unknown. In the present study, we investigated the role of actin filaments, MARCKS, and scinderin in regulated mucin secretion, primarily by using SPOC1 cells, a cell line derived from rat trachea that secretes mucin in response to stimulation by P2Y2R agonists (ATP, UTP, and ATPγS), phorbol 12-myristate 13-acetate (PMA), or Ca2+ (2).



DMEM-F-12 culture medium was obtained from GIBCO-BRL (Gaithersburg, MD), and the supplements were purchased from Collaborative Research (Bedford, MA). The following biochemicals were purchased from the following commercial sources: PMA and jasplakinolide, Calbiochem (La Jolla, CA); ATPγS, Roche Applied Science (Indianapolis, IN); ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid (EGTA), Sigma (St. Louis, MO); latrunculin A, Molecular Probes (Eugene, OR); and streptolysin O (SLO), Corgenix (Peterborough, UK). Synthetic peptides representing the actin-binding domains of scinderin and their scrambled controls were synthesized in the University of North Carolina Peptide Core Facility and had the following amino acid sequences (after 66): Sc-ABP1, AAIFTVQMDDYL; scrambled1, YAALMFDIDATQV; Sc-ABP2, RLLHVKGRR; and scrambled2, LVRGKRRLH.

The following antibodies were purchased or received from the following sources: pan-actin monoclonal antibody (Ab-1), Oncogene (Boston, MA); β-actin-specific monoclonal antibody (AC15), Sigma; γ-actin-specific polyclonal antibody (43), J. C. Bulinski (Dept. of Biological Sciences, Columbia University, New York, NY); green fluorescent protein (GFP) polyclonal antibody, Abcam (Cambridge, UK); MUC5AC-specific polyclonal antibody (54), J. K. Sheehan (Dept. of Biochemistry & Biophysics, University of North Carolina, Chapel Hill, NC); phospho-MARCKS-specific polyclonal antibody (sc-12971-R), Santa Cruz Biotechnology (Santa Cruz, CA). A polyclonal anti-scinderin antibody was raised in rat using recombinant mouse scinderin as the immunogen (Laboratory of Hormonology, Marloie, Belgium). The antibody was purified by affinity chromatography over a cyanogen bromide-Sepharose mouse scinderin column, and its specificity was confirmed using ELISA and Western blotting. The anti-scinderin antibody recognized 50 ng of recombinant mouse scinderin and stained only very weakly for 1 μg of human plasma gelsolin.

Secondary antibodies conjugated with either Alexa Fluor 488 or Alexa Fluor 594 were purchased from Molecular Probes for immunostaining. Secondary antibodies for Western blotting were purchased form Jackson ImmunoResearch (West Grove, PA).

For control Western blot analysis, whole cell extracts of human bronchial epithelial cell primary cultures and purified polymorphoneutrophils and red blood cell ghosts were obtained from Scott H. Randell and Hirotoshi Matsui (University of North Carolina, Chapel Hill, NC), respectively, and Peter Mohler and Vann Bennett (Duke University, Durham, NC).

SPOC1 cell culture, mucin secretion experiments, and tracheal xenografts.

SPOC1 cells were seeded as appropriate to the experiment in 6-, 12-, or 48-well cluster plates (Costar, Cambridge, MA) at 8,850 cells/cm2 and were maintained in rat tracheal epithelial cell medium (47) as previously described (3, 50). Except for cells grown solely for passaging, the medium also contained 10 nM retinoic acid. Culture media were changed daily, and the cultures were used for experiments after differentiation of a mucin-secreting phenotype 14–21 days postconfluence. Mature SPOC1 cells were used for mucin secretion experiments after being washed gently three times with warmed DMEM-F-12 separated by 30 min each. The experimental media were then added, and the cells were incubated as indicated in results, after which the media were harvested for mucin assays. For the experiment shown in Fig. 3, SPOC1 cells harvested after routine passaging were seeded into denuded, precannulated rat tracheas at ∼1 × 106 cells/graft in a volume of 50 μl. After the cannulas were heat sealed, the grafts were implanted subcutaneously into the backs of athymic nu/nu BALB/c mice and harvested after a 3-wk incubation using previously described procedures (8). At harvest, the xenografts were opened along the midlines, the mucus plug was gently washed away with DMEM-F-12, and the tissue was incubated for 20 min either without or with 100 μM ATPγS in DMEM-F-12. All animal procedures were performed according to University of North Carolina Institutional Animal Care and Use Committee-approved protocols.

SPOC1 cell permeabilization.

SPOC1 cells were permeabilized using SLO as described previously (50, 53). Briefly, the cells were washed with PBS and then with intracellular buffer (in mM: 130 K-glutamate, 20 PIPES, 1 MgATP and MgCl2, and 3 EGTA, pH 6.8). Free Ca2+ in intracellular buffer was set to 0.1 μM, except in solutions designed to stimulate mucin release, in which it was raised to 10 μM (50, 53). SPOC1 cells were permeabilized with SLO (1 U/ml in intracellular buffer) for 30 s, washed once in intracellular buffer, then incubated with the appropriate experimental solution for 15 min, and samples were collected for mucin analysis. TO-PRO (Molecular Probes, Eugene, OR), a membrane-impermeant, DNA-binding fluorescent dye, was used to confirm cell permeability as previously described (53).

Mucin enzyme-linked lectin assay.

Samples collected from experiments were assessed for mucin content using an enzyme-linked lectin assay described previously (3). Briefly, samples in 96-well high-binding microtiter plates (Costar; Corning Life Sciences, Acton, MA) were incubated overnight at 4°C or 37°C for 2 h, washed with PBS-0.05% Tween 20, and incubated with 2.5 μg/ml of horseradish peroxidase-labeled soybean agglutinin lectin for 1 h at 37°C. The plates were washed, developed with O-phenylenediamine in a citrate phosphate buffer (0.0175 M, 0.01% hydrogen peroxide, pH 5.0), stopped with 4 M H2SO4, and analyzed for optical density at 490 nm in a microtiter plate reader (Dynatech, Alexandria, VA). Standard curves were generated from purified SPOC1 mucins applied to each plate, and the results were expressed in nanograms of mucin released per culture. Importantly, in Western blotting of SPOC1 cell extracts, soybean agglutinin stains only high-molecular-weight glycoconjugates (3).

Western blotting.

After washing and equilibration, SPOC1 cultures were incubated as stated in results, and the cells were washed with ice-cold PBS and then lysed into one of three ice-cold lysis buffers (1): whole cell lysates, 20 mM Tris·HCl, pH 7.4, 2 mM EGTA and EDTA, 1% NP-40, and protease inhibitor cocktail (P-8340, 1:20 dilution; Sigma); cell fraction buffer: the same ingredients, but without NP-40; special lysis buffer for phospho-MARCKS (38) (in mM): 20 Tris·HCl, pH 7.4, 10 EGTA, 2 EDTA, 1 PMSF, 250 sucrose, 2 sodium vanadate, 25 sodium fluoride, 1% NP-40, and protease inhibitor cocktail (P-8340, 1:20 dilution; Sigma). For fractionation, lysed cells were homogenized on ice (15 strokes, Potter-Elvehjem tissue grinder), the extract was centrifuged for 1 h at 100,000 g, and the supernatant was taken as the cytosolic fraction. The pellet was solubilized in buffer supplemented with 1% NP-40, incubated 30–60 min, and centrifuged (14,000 rpm), and the supernatant was taken as the membrane fraction. Importantly, the membrane fraction also contained associated cytoskeleton.

Proteins in cell extracts or fractions were resolved by performing 10% SDS-PAGE and transferred electrophoretically to PVDF membranes. The blots were probed with antibodies appropriate to the experiment, and immunoreactive bands were detected using enhanced chemiluminescence (Amersham, Arlington Heights, IL). Imaging of the Western blots was performed with a platform scanner at a resolution of 1,200 × 1,200 dpi × 12 bits and quantitated using MetaMorph image-processing software (Universal Imaging, Downingtown, PA).

Immunolocalization and microscopy.

Paraffin sections (6 μm) of human lung fixed for 10 min in 10% neutral buffered formalin (NBF) were provided by the University of North Carolina Cystic Fibrosis Center histology facility. Cystic fibrosis (CF) and non-CF tissues were used. The tissues were procured from patients with CF and patients without lung disease in accordance with the guidelines of the Committee on the Protection of the Rights of Human Subjects of the University of North Carolina at Chapel Hill. SPOC1 cell xenografts were fixed in 10% NBF and processed for histology, and paraffin sections were cut at 6 μm. Sections were deparaffinized and rehydrated, blocked for 1 h in 3% BSA, and washed. Next, the first antibody was applied with an overnight incubation, followed by washes and a 1-h second antibody incubation. The slides were mounted with an antifade medium (Vectashield; Vector Laboratories, Burlingame, CA).

Immunostained xenograft sections were examined under a Leica DMIRB inverted microscope outfitted for epifluorescence and differential interference contrast using a PlanApo ×100/1.4 oil-immersion objective lens (Leica Microsystems, Wetzlar, Germany) and images were acquired with an Orca-II B water-cooled, charge-coupled device camera (Hamamatsu, Welwyn Garden City, UK). The sections were illuminated with a 100-W HBO mercury lamp and collected after being passed through the appropriate dichroic filter sets: 595DCLP for Alexa Fluor 594 and Q505LP for Alexa Fluor 488 images. Immunostained lung sections were examined with a Zeiss LSM 510-Meta laser scanning confocal microscope using an AchroPlan ×63/1.4 oil-immersion objective lens (Carl Zeiss, Oberkochen, Germany). Images (8-frame averages) were acquired in two channels using the same optical thickness setting for each channel. Alexa Fluor 488 and 594 were excited with, respectively, an Ar laser (488 nm) and a HeNe1 laser (543 nm), and emissions were separated using a 488-/543-nm dichroic filter and passed through a 545-nm beam splitter. Green fluorescence was collected with a 518/25 band-pass filter, and red fluorescence was gathered with a 585 low-pass filter. Lack of cross talk was verified by successively turning off the excitation source adjacent to each channel and confirmed by no visible signal. Photomultiplier tube gain and offset were adjusted to achieve an optimal contrast and avoid saturation in the region of interest.

Molecular cloning.

Total RNA was extracted from SPOC1 cell line and human bronchial epithelial primary cell cultures using the RNeasy extraction kit (catalog no. 74104; Qiagen) and stored at −80°C. SPOC1 cell reverse transcription was achieved with 2 μg of RNA using a first-strand cDNA synthesis kit (Invitrogen) by applying oligo(dT) for mRNA probing and SuperScript II reverse transcriptase for polymerization, stored at −20°C. cDNA (1 μl) was amplified by performing PCR in a 25-μl reaction volume containing 10 mM 2-deoxynucleotide 5′-triphosphate (0.5 μl) and Taq (1 μl). Rat scinderin primers were designed using the known mouse scinderin and gelsolin sequences to be specific in a region of low homology between the two proteins. The scinderin primers were 5′-AACAGTGGTAGGGTCCAGATT-3′ and 5′-AGTGATAGATGCCAGGTTCCTC-3′, with a PCR product of 360 bp. The gelsolin primers were 5′-GGTTCCAACAAGGTGCCCGTG-3′ and 5′-GGTGGCTCCAGAGCTGCTGGC-3′, with a PCR product of 330 bp. The PCR program was predenaturation at 94°C for 3 min and 33 cycles of amplification at 94°C for 45 s, 62°C for 30 s, and 72°C for 45 s, followed by extension at 72°C for 10 min. PCR products were separated by performing electrophoresis in a 1.5% Tris-acetate-EDTA (TAE)-agarose gel run at 110 V for 30 min and visualized using ethidium bromide. Full-length scinderin clones were obtained by performing PCR using Pfu polymerase (GIBCO-BRL) using 5′-ATGGCGCAGGAGCTGCAGCACCCCGAGTT-3′ as forward primer and 5′-TTACCACCTGCTGGAGTCCCAGCCCAGAAA-3′ as reverse primer. The products were cloned using the ZeroBlunt TOPO PCR cloning kit (Invitrogen), and the nucleotide sequence of the products was determined by performing DNA sequencing. Human scinderin was cloned beginning with a blast search, using the mouse scinderin cDNA (GenBank accession no. U04354) that returned the human p-aminoclonidine sequence (GenBank accession no. AC005281). This sequence proved to encode an open reading frame of 1,602 bp with 89% homology to the coding region of the bovine scinderin cDNA (GenBank accession no. D26549) that contained the 5′ end of the gene. The remaining 3′ sequence was isolated by performing 3′ rapid amplification of cDNA ends (RACE) using the commercially available human brain Marathon Ready cDNA library (Clontech, Oxford, UK) according to the manufacturers instructions. Analysis of the RACE products by performing DNA sequencing (ABI Prism 310; PerkinElmer, Wokingham, UK) revealed a further 546 bp, completing the scinderin coding region. A full-length human scinderin cDNA was subsequently isolated from human brain Marathon Ready cDNA using PCR amplification with the gene-specific forward primer 5′-ATGGCGCGGGAGCTATACCACGAAGAGTTC-3′ and reverse primer 5′-TTACCACTTGCTGGAATCCCAGCCCAGGAA-3′. PCR products were obtained and cloned into the TA vector (Invitrogen, Paisley, UK), and DNA sequencing confirmed the full-length coding sequence to be 2,148 bp.

β- and γ-actin overexpression.

The retroviral expression plasmid pLXSNb was derived from the Moloney murine leukemia virus (42) with a long-terminal repeat (LTR) promoter to drive expression. This vector contains a cloning site and uses an internal SV40 promoter to regulate the neomycin resistance (NeoR) gene. To improve selection efficiency, the SV40 promoter was removed and replaced with a poliovirus internal ribosomal entry site (IRES). To obtain a higher level of expression, a second vector, QXCIN (Clontech), was used. This vector uses an internal cytomegalovirus (CMV) promoter to drive expression of the insert/IRES/NeoR cassette. The cDNA encoding for yellow fluorescent protein (YFP)-β-actin (pEYFP-actin, catalog no. 6902-1; Clontech) was inserted into the multiple cloning site for both vectors using NotI and BamHI digestion. Because γ-actin differs from the β-isoform by only four highly conserved amino acid residue substitutions at the NH2 terminus (as detailed in results), we generated YFP-γ-actin constructs using a four-point site-directed mutagenesis scheme within the YFP-β-actin sequence (QuickChange kit, catalog no. 200519-5; Stratagene). The mutagenic primers were 5′-CGAGCTATGGAAGAAGAAATGGCCGCGCTCGTCATCGAC-3′ and its complement, with the mutated sites underlined.

Retroviral particles were obtained by transfection of human embryonic kidney-293T packaging cells using a Ca2+-phosphate precipitation method (25); titers were determined to be between 5 × 105 and 5 × 106 viral cells/ml. Subsequently, SPOC1 cells in passage 6 were exposed to 1 ml of virus-containing solution for 4 h in the presence of 8 μg/ml of polybrene (Aldrich Chemical). The selection for gene transfer started the next day and continued for ∼1 wk with medium changed daily using G418 at a concentration of 125 μg/ml. Actin-overexpressing cells were used for experiments between passages 8 and 13.

Statistical analysis.

Data are presented as means ± SE for a specified number of SPOC1 cell cultures, with each culture in a given experiment originating from a different passage. ANOVA and Student's t-test were used as appropriate to determine statistical significance between data sets. Data were deemed statistically significant at P < 0.05.


Localization of actin filaments in polarized human airway epithelial cells.

To determine whether actin is positioned in well-differentiated goblet cells to play a role in exocytosis, human lung sections were immunostained and observed using confocal microscopy. For these experiments, CF tissues were used because of their higher incidence of goblet cells; however, lung tissues from patients free of lung disease were also stained as controls, and their more sporadic goblet cells showed similar staining patterns (data not shown). Paraffin sections were stained with a monoclonal pan-actin antibody and, to label goblet cells in the superficial epithelium, a polyclonal MUC5AC antibody (54). The luminal margin of the airway epithelium exhibited a distinctly apical distribution of actin, but one that was specific to cell type (Fig. 1). In ciliated cells, actin localized to the apical membrane and microvilli, which gave it a “bushy” appearance. In goblet cells, however, actin took on the appearance of a prominent subplasmalemmal sheet lying between the apical membrane and mucin granules. In both cell types, actin signals were weak in the cytoplasm and along the basolateral surface. Notably, actin was not detected within the secretory granule region of goblet cells. Stress fibers in the basal pole of the cells were not observed because of the orientation of the sections and our focus on the luminal aspect of the airways. Interestingly, actin in basal cells was distributed in the cytoplasm as well as in the cell cortex (data not shown).

Fig. 1.

Actin is preferentially localized to the apical membrane domain of human airway goblet and ciliated cells. Thin lung tissue sections from patients with cystic fibrosis (CF) were double stained with polyclonal anti-MUC5AC and a monoclonal pan-actin antibodies and scanned using confocal microscopy (magnification, ×63; 8-frame averages). Top, left: differential interference contrast (DIC) image showing goblet and ciliated cells; right: red channel, pan-actin antibody fluorescence (Alexa Fluor 594 secondary antibody) showing actin distribution in the epithelium. Bottom, left: green channel mucin-specific-antibody (MUC5AC) fluorescence (Alexa Fluor 488 secondary antibody) identified mucin granules in goblet cells; right, merged red and green fluorescence images.

Studies of nonpolarized, nonsecretory cells have shown that β- and γ-actin have different cellular distributions, with γ-actin present in stress fibers and in the cell cortex and β-actin localized to the cortex and in other specialized compartments (see Ref. 27). In epithelial cells, β-actin has been reported to localize preferentially to the apical membrane, whereas γ-actin associates with both apical and basolateral membranes (64, 65). We used actin isoform-specific antibodies to determine the distribution of these isoforms in airway epithelial cells. Figure 2 shows that β- and γ-actin have similar distributions in both ciliated and goblet cells, with both isoforms localizing uniformly to the apical membrane. As a control for specificity of the antibodies, note the bright neutrophil infiltrating the epithelium that stained strongly for β-actin but faintly positive for γ-actin (Fig. 2, asterisks). This difference in expression was confirmed by performing Western blotting on isolated neutrophil and red blood cell ghost extracts, which showed a signal 3.7 times stronger for β-actin than for γ-actin (Fig. 2).

Fig. 2.

β- and γ-actin are localized to the apical membrane of goblet and ciliated cells in human airway epithelium. Sections double stained with nonmuscle actin isoform-specific antibodies as in Fig. 1 revealed the distribution of respective actin isoforms in airway epithelium. Left, top: DIC image showing goblet and ciliated cells; left, middle: red channel monoclonal anti-β-actin-specific antibody fluorescence (Alexa Fluor 594 secondary antibody). Right, middle: green channel polyclonal anti-γ-actin-specific-antibody fluorescence (Alexa Fluor 488 secondary antibody); right, top: merged red and green fluorescence images. *Infiltrating neutrophil. Bottom: Western blot analysis was performed to test the specificity of β- and γ-actin antibodies in cell lysates, SPOC1 cells, polymorphoneutrophils (PMN), and red blood cell ghosts (RBC, which do not express γ-actin). The amount of lysate applied to the gels was identical for each cell type but was adjusted by cell type to avoid saturation.

Distribution of actin isoforms in SPOC1 cells before and after stimulation.

SPOC1 cells adopt a multilayered morphology when grown in culture, with the mucin-containing cells in the top layer being rather flat. To gain a columnar phenotype favorable for immunolocalization, we instead grew cells in denuded rat tracheas as xenografts in nude mice. Under these conditions, as shown in Fig. 3, A and B, SPOC1 cells grew as a pseudostratified epithelium in which all of the columnar cells had a goblet cell-like phenotype (see also Refs. 3, 47). In these cells, β- and γ-actin assumed the same cellular distribution observed in human airway goblet cells (Fig. 3B, left; compare with Fig. 2). To test whether apical actin filaments are altered in their distribution during regulated mucin secretion, we stained SPOC1 cell xenografts after 20 min of maximal stimulation by ATPγS. As shown in Fig. 3B, both β- and γ-actin virtually disappeared from the apical pole, a period during which the mucin pools in the cells were largely depleted. These results suggest that the actin filaments function as a barrier to secretion in goblet cells under resting conditions and show that they undergo dramatic rearrangement and disruption during purinergic stimulation.

Fig. 3.

Cortical actin network disappears after agonist stimulation of SPOC1 cells grown in xenografts. SPOC1 cells, grown in xenografts for 3 wk, were exposed either to vehicle (left) or to a purinergic agonist, 100 μM ATPγS (right), for 20 min. A: thin sections of xenografts in which Alcian blue-periodic acid-Schiff staining (AB-PAS) was used to identify mucin granules. B: sections stained for β-actin (middle, red channel) and γ-actin (bottom, green channel); top: DIC images corresponding to the fluorescence images at bottom. Arrows indicate positions of apical membrane as identified by DIC. The sections shown in A and B are from the same xenografts.

Actin dynamics in SPOC1 cells: involvement of MARCKS and scinderin.

Because MARCKS and scinderin are hypothesized to effect actin filament disruption before exocytosis in other secretory cells (60), we tested whether these proteins are expressed in SPOC1 cells and/or in human airways and whether they might also participate in regulated mucin secretion. MARCKS expression in SPOC1 cells was tested by performing Western blotting of whole cell extracts obtained from cultures incubated for 40 min with DMEM-F-12 vehicle, a maximal dose of ATPγS (100 μM), or a dose of PMA (30 nM) sufficient to fully activate PKC (1, 3). The blots were probed with a phospho-MARCKS-specific antibody as an index of MARCKS translocation to the cytosol (38). The immunoblots (Fig. 4) indicate that MARCKS was phosphorylated progressively after stimulation by both ATPγS and PMA, reaching a peak at 10–20 min and declining slightly to a stable plateau. Hence, MARCKS protein appears to be expressed in SPOC1 cells and is phosphorylated after agonist stimulation, indicating a translocation from the membrane in a manner consistent with its postulated role in regulated secretion (60).

Fig. 4.

Time course of myristoylated alanine-rich C kinase substrate (MARCKS) phosphorylation induced by ATPγS (100 μM) or phorbol 12-myristate 13-acetate (PMA; 30 nM) stimulation. A: Western blotting of SPOC1 cell extracts, probed with a phospho-MARCKS antibody, harvested at the times indicated and run on SDS-PAGE gel with equal amounts of protein. B: bar graph of the means ± SE (n = 5) of the corresponding integrated intensities of the phospho-MARCKS bands normalized to t = 0 control.

Human airway epithelial and SPOC1 cells were found to express scinderin mRNA by performing RT-PCR (Fig. 5A, left). Interestingly, gelsolin mRNA was expressed in human bronchial epithelial cells, but apparently not in SPOC1 cells, a result consistent with the notion that gelsolin and scinderin have mutually exclusive expression patterns in different cell types (33). The scinderin cDNAs were cloned and sequenced [GenBank accession nos. AF276507 (human) and NM_198748 (rat)], and an alignment of the bovine, rat, mouse, and human nucleotides sequences showed an identity of 79.3%. A scinderin-specific antibody used to probe a Western blot of SPOC1 cell extract showed that the protein was also expressed (Fig. 5A, right). In further analysis, scinderin was found in both the cytosol and membrane fractions, and its distribution did not appear to change after exposure of SPOC1 cells to purinergic agonist (data not shown).

Fig. 5.

Scinderin is expressed in SPOC1 cells, and peptides corresponding to the scinderin actin-binding domains inhibit mucin secretion. A, left: RT-PCR for gelsolin and scinderin expression in duplicate pools of human bronchial epithelial primary cells (lanes 1 and 2) and SPOC1 cells (lanes 3 and 4); right: Western blot probing SPOC1 cell lysate with a scinderin-specific antibody. B: effect of synthetic peptides corresponding to the actin-binding domains of scinderin and the respective scrambled peptides on regulated mucin release in SPOC1 cells. Data represent means ± SE (n = 5). Peptide concentration is given as the amount of each of the 2 actin-binding domain peptides or their scrambled controls; i.e., the total peptide concentration would be twice the amount shown. *P < 0.05.

To test whether scinderin possesses the capacity to sever actin filaments relevant to mucin granule exocytosis, synthetic peptides corresponding to the two actin-binding domains of scinderin (66) were presented to SLO-permeabilized SPOC1 cells activated to release mucin by elevating Ca2+ to 10 μM (53, 50). As shown in Fig. 5C, the presence of these peptides inhibited Ca2+-activated mucin secretion in a concentration-dependent manner, with a maximum of −41.7 ± 11.1% at 30 μM. Scrambled peptides used as a control under the same conditions had no effect on secretion. We interpreted these results to indicate that the synthetic peptides competed with the action of the endogenous scinderin in activated SPOC1 cells.

Actin disruption and stabilization effects in SPOC1 cells.

Disruption of the actin cortical network by cytochalasin D or the sponge-derived G actin-sequestering agents latrunculins A and B (57b) has been shown to induce secretion in chromaffin and other cells (9, 14). Consistent with these findings, latrunculin A treatment of SPOC1 cells resulted in a concentration-dependent increase in mucin secretion under both basal (nonagonist control) and agonist-stimulated conditions, with maximal effects at 10 μM of 77 and 85%, respectively (Fig. 6). These results suggest that actin filament disruption in mucin-secreting cells activates granule release and that it does so in a manner that maintains the responsiveness of the cells to agonist.

Fig. 6.

Effect of actin filament disruption by latrunculin A on basal and regulated mucin secretion in SPOC1 cells. SPOC1 cells grown in 12-well plates were incubated with increasing concentrations of latrunculin A up to 10 μM in the absence or presence of ATPγS (100 μM). Data were normalized to control wells (no latrunculin) in both cases. Data are means ± SE; n = 6. *P < 0.05.

Conversely, stabilization of actin filaments with phalloidin or jasplakinolide (6) has been shown to inhibit regulated exocytosis (41, 51). In SLO-permeabilized SPOC1 cells activated by an elevation of Ca2+ concentration to 10 μM, jasplakinolide inhibited mucin release in a concentration-dependent manner, with maximal effect at 10 μM of −51.4 ± 22.3% (Fig. 7). Together, the latrunculin A and jasplakinolide data indicate that actin filament disassembly in SPOC1 cells is not only permissive but also necessary to enable mucin granule exocytosis upon stimulation by agonist.

Fig. 7.

Effect of actin filament stabilization by jasplakinolide on mucin secretion in streptolysin O (SLO)-permeabilized, Ca2+-activated SPOC1 cells. SPOC1 cells were grown on 48-well plates and permeabilized using SLO for 30 s and washed once in intracellular buffer, with Ca2+ = 0.1 μM. They were then incubated with intracellular buffer Ca2+ = 10 μM in the presence of increasing concentrations of jasplakinolide for 15 min. Data were normalized to the control well (no jasplakinolide), which represents Ca2+-activated mucin release. Data are means ± SE; n = 4. *P < 0.05.

Overexpression β- and γ-actin in SPOC1 cells.

As an independent test of the barrier hypothesis regarding regulated mucin secretion, β- and γ-actin were overexpressed in SPOC1 cells. Because SPOC1 cells require 2–3 wk to differentiate, a stable expression system was necessary; therefore, we used a retroviral expression system, repeating infections on individual lots of passage 6 cells at least four times to ensure that the results obtained were not due to inopportune insertion of the provirus. The retroviruses generated expressed either YFP-β-actin (22) or YFP-γ-actin. Because nonmuscle β-actin differs from nonmuscle γ-actin by only four amino acid residues in the NH2-terminal segment, Asp2-Asp3-Asp4- … -Val10 and Glu2-Glu3-Glu4- … -Ile10, respectively (20, 38), γ-actin vectors were obtained by site-directed mutagenesis of the β-actin vectors. The expression of each isoform was driven by either the LTR or the CMV promoter, leading, respectively, to weak or strong expression in the resulting cell lines. As a control, SPOC1 cells were infected using an empty retroviral vector. After selection, infected SPOC1 cells were grown for 15–18 days and then were either assessed for actin isoform expression levels or challenged with secretagogues for 30 min, and then mucin release was quantified. Data from the retrovirally infected SPOC1 cells were normalized to basal mucin release determined for each infected cell type.

YFP-actin expression was determined by performing Western blotting (Fig. 8, A and B, left) using a GFP antibody for YFP-actin and β- or γ-actin-specific antibodies for endogenous actin. Quantification of the blots revealed that YFP-actin expression driven by CMV was >2.2 times stronger than the expression driven by LTR. Estimation of overexpressed YFP-actin relative to endogenous actin was not possible, because the conjugated forms of β- or γ-actin were recognized only partially by pan-, β-, and γ-actin antibodies, likely because of camouflage of the epitope by the YFP protein.

Fig. 8.

Retroviral overexpression of yellow fluorescent protein (YFP)-β- or YFP-γ-actin inhibits regulated mucin secretion in an expression level-dependent manner. SPOC1 cells were infected, selected, and grown on 12-well plates. YFP-β-actin (A) and YFP-γ-actin (B) were overexpressed in SPOC1 cells using retroviruses with weak (W; long-terminal repeat) or strong (S; cytomegalovirus) promoters. Western blots (bottom left) depict the levels of YFP-actin expression driven by the two promoters. Western blot anti-green fluorescent protein (GFP) band intensities are graphed above the respective blots. Bottom bands in the blots represent endogenous β- and γ-actin, respectively. Mucin release (right) from non-overexpressing (C) and actin-overexpressing (W, S) cells, with stimulation by ATPγS (100 μM) or PMA (100 nM). Data are means ± SE (n = 4 or 5), relative to stimulated control cells (empty virus). *P < 0.05. **P < 0.005.

Strikingly, overexpression of YFP-β- and YFP-γ-actin led to a decrease in mucin secretion in an expression level-dependent manner for cells stimulated with either 100 μM ATPγS or 100 nM PMA. SPOC1 cells expressing YFP-β- or YFP-γ-actin behind the weak LTR promoter exhibited inhibition ranging from 41 to 52% of control; for cells expressing actin behind the stronger CMV promoter, the inhibition ranged from 67 to 79% (Fig. 8, A and B, right).

To determine whether YFP actin enhanced the normal barrier function of the apical actin filaments or whether the inhibition observed was an artifact of overexpression, we treated YFP-β- and YFP-γ-actin-overexpressing cells with latrunculin A. As shown in Fig. 9, mucin secreted in the latrunculin A-treated cells attained the same level for non-, weakly, and strongly YFP-actin-overexpressing cells. First, the results of this experiment imply that actin-overexpressing and wild-type SPOC1 cells contain the same total amount of mucin. Second, the results show that inhibition of mucin secretion by actin isoform overexpression is specific to actin filaments. In control (empty vector) cells, latrunculin A increased the response to ATPγS, similar to its effect in noninfected cells (Fig. 6). Notably, latrunculin A stimulated the same nominal degree of response from all infected cells, except for those strongly overexpressing YFP-γ-actin. In the case of YFP-γ-actin driven by the CMV promoter, there may have been some nonspecific effects; however, even in this case, breakdown of the actin filament barrier resulted in strong stimulation of mucin secretion. Collectively, these findings are novel and reveal that actin overexpression inhibits mucin secretion in a manner that can be overcome by an actin filament-disrupting agent, offering additional support for the barrier hypothesis.

Fig. 9.

Latrunculin A (Lat) treatment of YFP-β- and YFP-γ-actin-expressing cells restores mucin secretion. SPOC1 cells described in Fig. 8 were incubated with either ATPγS (100 μM) in the absence or presence latrunculin A (3 μM) as indicated. A: SPOC1 cells overexpressing β-actin. B: SPOC1 cells overexpressing γ-actin. Data are means ± SE, n = 3. *P < 0.05 and **P < 0.005, actin overexpression cells relative to control cells (empty vector); †P < 0.05, latrunculin-treated cells relative to ATPγS-challenged cells.


Actin is a ubiquitous protein that in nonmuscle cells is responsible for functions ranging from controlling cell shape and motility to serving as a major, membrane-associated protein scaffold. β- and γ-actin, the two predominant isoforms expressed in nonmuscle cells, differ by only four amino acids at the NH2 terminus (for review, see Refs. 20, 27) and, despite this similarity, are differentially distributed in many cells. In motile cells, for instance, γ-actin is distributed throughout the cell and in stress fibers, whereas β-actin localizes primarily to cortex, especially the leading edge (21, 24). In neurons, β-actin is enriched in immature growing axons relative to the cell body and γ-actin is uniformly distributed throughout the cell (19, 63). In the gastric epithelium, β-actin localizes preferentially at the apical membrane of parietal cells, whereas γ-actin is found at both membranes and predominates at the basolateral membrane (64, 65). In ciliated and goblet cells of human airways and in SPOC1 cells, β- and γ-actin appeared to colocalize strongly to the apical membrane (Figs. 2 and 3). Although it is tempting to speculate that the differential distribution of β- and γ-actin in many cells may mean that they subserve different functions, in fact such discrimination between isoforms has yet to be demonstrated. At least in experimental situations, the two isoforms appear to function indiscriminately in homogenous or mixed populations in actin filaments (27).

It is widely accepted that the basic engine for cell motility is the actin cytoskeleton, an engine composed of closely coordinated actin filament polymerization and degradation, cell adhesion, and membrane remodeling. Intense research efforts during the past two decades have revealed the formation of protrusive structures involved in motility to be due to actin polymerization after the activation of the Wiscott-Aldrich syndrome (WASp/SCAR) family by Rho GTPases (Cdc42, Rac), phosphatidylinositol 4,5-bisphosphate, or several of the Src homology domain 3 proteins such as Grb2, followed by filament branching, nucleation, and extension under the influence of WASp/SCAR and actin-related protein 2/3 complex (45, 46). The polymerizing actin filaments, anchored within the leading edge, provide a motive force to extend the plasma membrane outward from the cell (see Ref. 36). Interestingly, this same actin polymerization machinery has been implicated in intracellular organelle transport (26, 35a), endocytosis (4), and various aspects of exocytosis (see below).

The role of actin filaments in exocytosis is controversial, which is to say that actin potentially has multiple functions in secretion and/or that the precise mechanisms may be cell-type dependent. Classically, the actin cortex has been viewed as a barrier that impedes the movement of secretory granules to the plasma membrane (7, 30a, 57a, 62); however, in other cells such as PC12 and insulin-secreting cells, dynamic actin remodeling may be required for the exocytotic process (17; 30b; 35b). Recent biophysical studies within the near membrane environment of secretory cells suggest that actin polymerization is required for exocytosis as though the secretory granules were being guided actively to plasma membrane docking sites (40). A key distinguishing feature between these barrier and transport roles, experimentally, is the effect of cytochalasin D and/or latrunculin on secretion. In studies favoring a positive role for actin filaments in secretion, disruption was inhibitory (44), whereas disruption was stimulatory in studies suggesting a negative barrier role (37). In addition to barrier and transport roles, actin filaments have also been shown in Xenopus oocytes and pancreatic acinar cells to polymerize around secretory granules after exocytotic fusion. These new filaments function to provide the motive force for expulsion of granule contents and/or to maintain the “Ω-figure” as a prelude to endocytotic membrane retrieval (39, 56). In both cases, latrunculin inhibited actin filament coat formation around the granules and promoted fusion of adjacent granules. Latrunculin, however, either had no effect on exocytosis or was stimulatory in these cells. Hence, actin filaments may function variously to regulate exocytosis negatively in resting cells, to guide granules to docking sites after cell stimulation, and/or to stabilize Ω-figures after exocytosis.

In goblet cells and SPOC1 cells, neither β- nor γ-actin could be identified around the mucin granules near the apical membrane or anywhere within the mucin granule region of the apical pole; rather, actin was positioned prominently as a thick sheet between the granules and the apical membrane (Figs. 2 and 3). After stimulation, both isoforms disappeared from the SPOC1 cell apical membrane domain (Fig. 3B, right). Latrunculin A increased basal secretion (Fig. 6), suggesting that disruption of actin filaments alone is sufficient to induce secretion. Furthermore, agonist-stimulated SPOC1 cell mucin secretion was enhanced, indicating that even in stimulated cells, actin filaments act to regulate granule release negatively (Fig. 6). Additional evidence of a barrier role for actin in SPOC1 cells includes the inhibition of mucin secretion in activated cells by filament stabilization using jasplakinolide (Fig. 7) and by overexpression of actin isoforms (Fig. 8). In addition, the two proteins hypothesized to effect disruption of the actin barrier in secretory cells, MARCKS and scinderin (58, 61), are not only expressed in SPOC1 cells but also appear to act in accordance with the barrier hypothesis after cell activation. MARCKS is phosphorylated, indicating a translocation to the cytosol, and scinderin appears to sever actin filaments (Figs. 4 and 5, respectively). Hence, several lines of independent evidence strongly support a barrier role for actin filaments in regulated mucin secretion from SPOC1 cells. Two aspects of the data suggest that it is the predominant function. First, the apparent lack of actin within the mucin granule region makes an actin-based granule transport mechanism difficult to envision. Second, the stimulation of secretion in resting cells by latrunculin A suggests that actin filament disruption alone may be sufficient to trigger the exocytotic process. Consequently, if actin-based mucin granule transport does occur in goblet cells, it may not be a prerequisite for secretion as it is in other cells (17, 41).

The role of MARCKS in regulated exocytosis in goblet cells is presently controversial. Recent data showing inhibition of mucin secretion by a myristoylated NH2-terminal MARCKS peptide led to the novel hypothesis that the protein plays a role in a transport of granules by actin filament from the cytosol to the apical membrane (32, 55). Our data regarding MARCKS phosphorylation suggest that translocation from the membrane to cytosolic fraction upon agonist stimulation (Fig. 4) and stimulation of mucin secretion by latrunculin A (Fig. 6) are more consistent with the barrier hypothesis (60). The current data do not rule out a granule transport role for MARCKS; however, our data do suggest that if MARCKS plays a role in mucin granule transport, it is likely to be subservient to the barrier function.

Interesting possibilities potentially reconciling the apparent contradictory roles of actin filaments in secretion are that their barrier and transport roles are interwoven and that a factor determining which role dominates in a particular cell type is granule size. Secretory granules generally range in size from ∼50 nm to >1 μm in diameter, and mucin granules at 0.8–1.5 μm (13, 31) are among the largest known. Larger-diameter granules may have an increased likelihood of interaction with plasma membrane docking sites because of the greater surface area and smaller angle of curvature of the limiting membrane. For the smallest granules, positioning by actin filaments to such plasma membrane sites may be necessary to ensure a dynamic secretory response. From a regulatory point of view, a barrier to secretion formed by actin filaments is a very attractive notion: secretory granules are retained in the cells at low energetic expense by the actin barrier as they mature and the cell awaits activation, and they gain rapid access to exocytotic docking sites while the actin filaments are remodeled dynamically upon receipt of an appropriate signal.

The present data do not allow us to draw conclusions regarding the possibility of a postexocytotic role for actin filaments in supporting Ω-figures in mucin-secreting cells as revealed recently in other cells (5, 39). It is interesting to note, however, that because mucin granules in airway goblet cells are exocytosed completely in <100 ms (13), such a role may be unnecessary.

To conclude, β- and γ-actin participate in the formation in airway goblet cells of an apical actin cap that acts as a physical barrier to mucin secretion under resting conditions. Upon stimulation, the actin filaments rearrange under the actions of MARCKS and scinderin to allow granule access to the plasma membrane. However, it is likely an oversimplification of what really happens in regulated mucin secretion to think that actin filaments simply require disassembly to provide access to docking sites for mucin granule exocytosis. Further studies are required to investigate whether actin filament disassembly in the apical region is directed or random, whether β- and γ-actin form independent isoform-specific filaments that depend on different regulatory systems, whether local actin polymerization operates secondary to disruption to guide granules to exocytotic sites, and the specific roles of PKC and MARCKS as well as Ca2+ and scinderin in regulating and/or effecting actin filament remodeling.


K. De Pestel acknowledges support for the development of the scinderin-specific antibody by Grant FWO-G.0007.03 (to C. Ampe, Department of Biochemistry) and Grant GOA 12051401 of the Concerted Research Actions of the Flemish Community (to J. Vandekerckhove and C. Ampe). Primary support for these studies was derived from grants to C. Ehre from La Foundation Vaincre la Mucoviscidose, Paris, France, from National Heart, Lung, and Blood Institute Grant HL-63756 (to C. W. Davis), and from the North American Cystic Fibrosis Foundation.


We are grateful to Mireille Ladislas, Novartis Horsham Research Centre, for expert technical assistance in the cloning and analysis of human scinderin and to Drs. Michael Chua and Silvia Kreda for expert assistance in microscopy.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30a.
  31. 30b.
  32. 31.
  33. 32.
  34. 33.
  35. 34.
  36. 35a.
  37. 35b.
  38. 36.
  39. 37.
  40. 38.
  41. 39.
  42. 40.
  43. 41.
  44. 42.
  45. 43.
  46. 44.
  47. 45.
  48. 46.
  49. 47.
  50. 48.
  51. 49.
  52. 50.
  53. 51.
  54. 52.
  55. 53.
  56. 54.
  57. 55.
  58. 56.
  59. 57a.
  60. 57b.
  61. 58.
  62. 59.
  63. 60.
  64. 61.
  65. 62.
  66. 63.
  67. 64.
  68. 65.
  69. 66.
View Abstract