| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Physiology and Pharmacology, University of Queensland (A.S.Y., V.C., S.W.M.),,St. Lucia, Brisbane, Australia 4072; and the Department of Anatomy and Cell Biology, University of Alberta (B.R.S.), Edmonton, Canada T6G 2H7
Address all correspondence and requests for reprints to: Dr. Alpha Yap, Department of Physiology and Pharmacology, The University of Queensland, St. Lucia, Brisbane, Australia 4072. E-mail: yap{at}plpk.uq.oz.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Primary porcine thyroid cell cultures provide a particularly useful model system to study the cellular mechanisms responsible for follicular morphogenesis. When seeded at high density in the presence of TSH, freshly isolated porcine thyroid cells adhere to one another and reorganize into follicles within three-dimensional cell aggregates. In the absence of TSH, cells instead form nonfollicular two-dimensional epithelial monolayers (13, 14, 15, 16, 17). The morphogenetic influence of TSH is mediated by intracellular cAMP, as it can be exactly mimicked by cAMP analogs and other agents that stimulate cAMP production (e.g. forskolin and PGE2). Taken together, these observations identify TSH, acting through the cAMP-protein kinase A signaling pathway, as a principal determinant of thyroid follicular morphology in vitro (13, 15, 16, 18). Recent studies, moreover, presented evidence that TSH regulates thyroid cell adhesion (4, 5, 7) and cytoskeletal organization (19), and inhibits cell locomotility (20) in a manner predicted to favor follicle formation.
Importantly, thyroid cell patterning in culture is not fixed once follicles have formed. Instead, thyroid cells spread and migrate, converting follicles to monolayer, if deprived of ongoing TSH stimulation (16, 20, 21). This observation implies that follicular organization must be actively maintained, both by processes that promote follicle formation (such as cadherin-mediated cell adhesion) (6, 7) and by the tonic inhibition of cellular mechanisms that favor monolayer (such as cell locomotion) (20). Presumably, these antagonistic cellular processes are also kept in balance by strict physiological regulation. Indeed, follicle to monolayer conversion after withdrawal of TSH stimulation is inhibited by cAMP analogs, implying that it reflects the loss of tonic cAMP signaling activity. However, the observation that agents such as epidermal growth factor (22) and phorbol esters (18) can induce monolayer formation from follicles despite the presence of TSH suggests the existence of additional regulatory processes capable of influencing thyroid epithelial morphology in culture. In the present study we sought to identify additional fundamental cellular mechanisms mediating follicle to monolayer conversion. We report that removal of TSH stimulation from established follicles is associated with increased cellular phosphotyrosine, particularly at sites of cell-substrate adhesion (focal adhesions) and cell-cell adhesion. Inhibition of tyrosine kinases by genistein largely blocked the accumulation of phosphotyrosine and prevented both follicle to monolayer conversion and focal adhesion assembly. We conclude that signaling pathways that use tyrosine phosphorylation are likely to play an important role in determining thyroid epithelial organization in culture, at least in part by regulating focal adhesion assembly.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell cultures
Primary porcine thyroid cell cultures were prepared as
previously described (6, 16, 19). In brief, cells were extracted from
freshly collected tissue by discontinuous incubation with 1 g/liter
neutral protease (Dispase; grade II; 0.5 U/mg) and 0.1 g/liter
collagenase (type I, Worthington Biochemical Corp., Freehold, NJ; 200
U/mg) in Ca- and Mg-free Spinner’s salt solution. Cells were isolated
by filtration and centrifugation, and washed in incubation medium
consisting of MEM supplemented with HEPES (20 mM),
L-glutamine (1 mM), NaHCO3 (10
mM), nystatin (50,000 U/liter), gentamicin (50 mg/liter),
porcine insulin (68 µM), cortisol sodium succinate (0.5
µM), and heat-inactivated newborn calf serum (10%,
vol/vol). For immunofluorescence microscopy, cells were then plated
onto glass coverslips coated with rat tail collagen prepared as
previously described (4). Cells were established at a density of
3.2 x 105 cells/cm2 in incubation medium
supplemented with TSH (256 µU/ml). To induce monolayer formation,
established follicular cultures were washed into fresh incubation
medium lacking TSH; controls were washed into fresh medium containing
TSH (256 µU/ml). For studies using genistein, the drug was added to
medium from a stock solution of 100 mM in
dimethylsulfoxide; controls received vehicle alone.
Immunochemicals
The primary immunochemicals were as follows. 1) Mouse monoclonal
antibody VIN-11–5 (Sigma) directed against chicken gizzard vinculin
was used at a dilution of 1:100 from ascites fluid. 2) Rabbit
polyclonal antibody against phosphotyrosine (ICN Biomedical, Costa
Mesa, CA) was used at a dilution of 1:100. 3) Rat monoclonal antibody
R40.76 directed against ZO-1 (23) was used at a dilution of 1:100 from
ascites fluid. 4) Mouse monoclonal antibody 3G8 directed against
E-cadherin (24) (a gift from Dr. Warren Gallin, University of Alberta,
Edmonton, Canada) was used at a dilution of 1:5 from hybridoma
supernatant. 5) Phalloidin conjugated with fluorescein isothiocyanate
(FITC-phalloidin, Sigma) was used at a concentration of 0.4
µM. Biotinylated antirat, antirabbit, and antimouse IgG
were obtained from Vector Laboratories (Burlingame, CA), and FITC- and
Texas red-conjugated streptavidin were purchased from Amersham
Australia (Castle Hill, Australia). All antibodies were diluted in
blocking buffer (10% calf serum in PBS, pH 7.4).
Immunohistochemistry and microscopy
Cells were washed in PBS, fixed with 4% paraformaldehyde (in
PBS, pH 7.4; 20 min) and permeabilized with Triton X-100 (0.25%; 10
min). After incubation with blocking buffer for 2–4 h, specimens were
incubated overnight with primary antibodies. For single label studies
the primary antibodies were visualized by sequential incubation with
biotinylated secondary antibodies (1:100; 90 min) and then
streptavidin-conjugated fluorophores (1:100; 90 min), with extensive
washing between incubations. In double labeling studies for both
vinculin and phosphotyrosine, vinculin localization was detected with
antimouse IgG conjugated to Texas red (Jackson ImmunoResearch
Laboratories, Westgrove, PA), and phosphotyrosine staining was
visualized with biotinylated antirabbit IgG followed by
FITC-streptavidin. The antimouse antibody was added in the final
incubation step. For double labeling vinculin and F-actin, vinculin was
localized with biotinylated antimouse IgG and streptavidin conjugated
to Texas red. FITC-phalloidin was added in the final incubation step.
All incubations were conducted at room temperature. Coverslips were
mounted on cavity slides in PBS-buffered glycerol. Specimens were
viewed with a Bio-Rad MRC600 confocal laser scanning microscope
(Bio-Rad, Richmond, CA) mounted on a Zeiss Axioskop (Zeiss, New York,
NY) equipped with Zeiss Plan-APOCHROMAT x40 and x63 oil immersion
objectives. Specific filter blocks were used for FITC and Texas red
fluorescence; no significant bleed-through occurred when the filters
were exchanged in single label studies.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) or cAMP analogs
(not shown) (20). In contrast, when established follicular cultures
were washed into TSH-free medium supplemented with FCS, cells spread
from aggregates and migrated upon substrate, resulting in the
replacement of follicles by patches of monolayer (Fig. 1B), which later
coalesced to form a confluent sheet (not shown) (20). Follicles also
converted to monolayer when cultures were washed into TSH-free medium
lacking calf serum, although this generally occurred more slowly than
in the presence of serum (not shown). Follicle to monolayer conversion
was inhibited by genistein, a potent inhibitor of protein tyrosine
kinases (25). As shown in Fig. 1C, 24 h after being washed into
TSH-free medium containing genistein (100 µM),
phase-bright cell aggregates were preserved, and monolayer patches were
not detectable. These observations strongly suggested that protein
tyrosine phosphorylation could influence the supracellular organization
of cultured thyroid cells.
|
) nor at sites of
cell-cell contact (not shown). However, 6 h after washing cells
into TSH-free medium, numerous vinculin-rich focal adhesions were
detected in cells attached to substrate at the bases of follicles (Fig. 2C). These increased in prominence by 24 h (Fig. 2E), when many
cells in follicles had spread to monolayer. In contrast, although focal
vinculin staining was occasionally detected at sites of cell-cell
contact by 6 h after withdrawal of TSH, consistent vinculin
staining was not detected in cell-cell contacts until 48 h, when
follicles had been replaced by monolayer patches. At this time, linear
vinculin staining encircled cells (Fig. 2G), a pattern consistent with
localization in adherens junctions (28).
|
, cells adherent to substrate at
the bases of established follicles displayed F-actin staining in
cell-cell junctions and short filopodia at their free peripheries.
After washing into TSH-free medium, the filopodia were replaced by
broad lamellipodia that are characteristic of motile spreading cells
(26) and that stained intensely for F-actin (Fig. 2, D and F). In
addition, microfilaments reorganized into prominent stress fibers (Fig. 2, D and F), which terminated in focal adhesions. By 48 h, cells
within monolayer patches also displayed a perijunctional actin ring
(Fig. 2H) that colocalized with vinculin at cell-cell contacts (Fig. 2G). These findings indicated that withdrawal of TSH support induced
both the assembly of focal adhesions and the coordinate reorganization
of the actin cytoskeleton.
To assess further the relationship between tyrosine phosphorylation and
junctional assembly, we examined cells for both vinculin and
phosphotyrosine by double label immunofluorescence microscopy. No
significant localized phosphotyrosine staining was detectable in
established, TSH-stimulated follicles (not shown). However, within
6 h of washing into TSH-free medium, phosphotyrosine staining was
detected in spreading cells at the bases of aggregates (Fig. 3B), where it colocalized with vinculin deposits in
nascent focal adhesions (Fig. 3A). Phosphotyrosine staining became
increasingly prominent in the succeeding 42 h as follicles
converted to monolayer patches. At all times, phosphotyrosine
colocalized with vinculin in focal adhesions (Fig. 3, E and F).
Occasional foci of phosphotyrosine and vinculin staining were seen at
sites of cell-cell contact shortly after withdrawal of TSH stimulation
(Fig. 3, C and D), but extensive intercellular staining was not
observed until 48 h, when linear deposits of vinculin (Fig. 3G)
and phosphotyrosine (Fig. 3H) colocalized throughout cell-cell contact
zones in monolayer patches.
|
). Twenty-four hours after cells
were washed into fresh medium supplemented with genistein (100
µM), scant staining for vinculin and phosphotyrosine was
detected at sites of cell-substrate adhesion (Fig. 4, A and B) or
cell-cell contact (Fig. 4, C and D). Genistein-treated follicles showed
only short, rudimentary focal adhesions at the peripheries of cells
(Fig. 4, A and B). Although detectable, staining for both vinculin and
phosphotyrosine was much weaker in genistein-treated cells than in
control cells depleted of TSH (compare staining in Fig. 4, A and B
with that in Fig. 3, E and F, which are controls analyzed and
reproduced under identical conditions). Taken together, these findings
indicate that tyrosine phosphorylation accumulated in and was necessary
for the assembly or maturation of focal adhesions and cell-cell
adherens junctions.
|
) and ZO-1 (Fig. 5B) staining encircling the cells,
consistent with mature adhesive and tight junctions, respectively. In
contrast, 24 h after cells were washed into TSH-free medium,
E-cadherin staining was found in a punctate distribution both at sites
of cell-cell contact and in cytoplasmic vesicles (Fig. 5C), whereas
ZO-1 staining was lost from many regions of cell-cell contact (Fig. 5D). Only occasional regions of linear E-cadherin (Fig. 5C) and ZO-1
(Fig. 5D) staining could be identified. However, by 48 h, when
follicles had converted to monolayer patches, linear E-cadherin (Fig. 5E) and ZO-1 (Fig. 5F) staining was again found in all cell-cell
contact zones. These changes suggested that as follicles converted to
monolayer they initially disassembled, and then reassembled, both
adhesive cell-cell junctions and tight junctions.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our present studies two lines of evidence now identify protein tyrosine phosphorylation as an important regulatory signal that is activated upon withdrawal of TSH/cAMP signaling. Firstly, we observed a significant increase in cellular phosphotyrosine, generally at or near the cell surface, using immunofluorescent staining techniques. This is also supported by preliminary immunoblotting studies, which demonstrated that as thyroid follicles convert to monolayer, multiple proteins become tyrosine phosphorylated (Yap, A. S., M. J. Waters, and S. W. Manley, unpublished observations). Secondly, genistein, a potent tyrosine kinase inhibitor (25), blocked both the accumulation of phosphotyrosine and the conversion of follicles to monolayer despite withdrawal of TSH from cultures. Taken together, these observations therefore indicate that, in addition to loss of cAMP signaling, activation of tyrosine phosphorylation-dependent regulatory pathways was necessary for follicles to convert to monolayer upon withdrawal of TSH.
Ultimately, signaling pathways in the thyroid must exert their morphogenetic influence by regulating the cellular effector mechanisms that determine epithelial organization. In an earlier study we found that follicle to monolayer conversion was mediated by two distinct processes: an initial phase of cell spreading, which converted follicular aggregates to monolayer patches, and subsequent cell locomotion, which drove the process to confluence (20). The observation that genistein inhibited thyroid cell spreading from follicles suggested that increased tyrosine phosphorylation might activate a mechanism involved in cell spreading. It was consequently striking to find that the earliest sites of phosphotyrosine accumulation coincided precisely with nascent focal adhesions, cell-substrate junctions that exert an influence on cell morphology.
Focal adhesions are macromolecular complexes that link actin filaments to clusters of cell-substrate adhesion receptors (integrins and proteoglycans) (26, 32). As detected by antibodies to the sensitive marker, vinculin (26, 27), focal adhesions are found in many cultured cells, including thyroid cells grown as monolayers de novo (19). By contrast, in this and an earlier study (19), we could not detect focal adhesions in established thyroid follicles. Instead, focal adhesions first appeared shortly after removal of TSH, in cells spreading from follicles, and increased in number and prominence as follicles converted to monolayer. This indicated that focal adhesion assembly is regulated in thyroid cells and correlates closely with epithelial organization. It was, therefore, significant that not only did nascent focal adhesions colocalize with sites of phosphotyrosine accumulation, but inhibition of tyrosine phosphorylation by genistein also blocked focal adhesion assembly. This indicated that tyrosine phosphorylation was necessary for the formation of focal adhesions. Similarly, earlier studies demonstrated that focal adhesions were the major sites of tyrosine phosphorylation within cells (26, 33) and that tyrosine phosphorylation was essential for their assembly (34, 35). Indeed, a number of kinases have been identified in focal adhesions (32, 36), and tyrosine kinase activation is likely to be a major mechanism for assembling focal adhesions after integrin adhesion molecules bind to matrix ligands (37). Taken together, these findings strongly suggest that the assembly of focal adhesions in thyroid cells is regulated by the increased tyrosine phosphorylation that occurs upon withdrawal of TSH/cAMP stimulation.
Importantly, a large body of evidence implicates focal adhesions as critical mechanisms for cell spreading upon substrate. Thus, the appearance of focal adhesions typically correlates with cell spreading (26), and spreading itself is blocked by a variety of maneuvers that inhibit focal adhesion assembly (38), including disruption of vinculin expression (39, 40) or function (41). Focal adhesions probably facilitate spreading by both increasing cell-substrate adhesion (42) and coupling actin microfilaments to sites of cell adhesion, thereby allowing the motor apparatus of the cytoskeleton to exert traction and alter cell shape (26). The appearance of stress fibers provides internal evidence for this in our studies. Stress fibers are prominent bundles of actin filaments that form through the isometric contraction of microfilament meshworks against tightly anchored regions of the cell surface (26, 43, 44, 45). A characteristic feature of many cultured cells, stress fibers were seldom found in established thyroid follicles, appearing only in spreading cells deprived of TSH support, in association with, and terminating in, nascent focal adhesions. This finding is consistent with the productive coupling of cytoskeletal and adhesive function mediated by focal adhesions that is necessary for cell spreading to occur.
Based on these and earlier findings (4, 19, 20, 35), we propose that the regulated assembly of focal adhesions is an important determinant of thyroid epithelial organization. In this scheme, focal adhesion assembly triggered by increased tyrosine phosphorylation facilitates cell spreading from follicles by increasing cell-substrate adhesion coupled to the contractile apparatus of the actin cytoskeleton. Spreading, in turn, initiates the locomotor activity that completes the conversion of follicles to monolayer (20). In an earlier study we found that protein tyrosine phosphorylation was necessary for focal adhesion assembly and spreading in isolated thyroid cells (35). Our present findings extend this to strongly suggest that regulation of focal adhesion assembly can act across coherent aggregates of thyroid cells to influence supracellular organization and not just the morphology of individual cells. Focal adhesion assembly may, therefore, play an important role in linking tyrosine phosphorylation signals to the cellular processes that determine thyroid epithelial morphogenesis. This model also implies that TSH/cAMP signaling, which inhibits spreading (20) and supports follicular organization, should inhibit focal adhesion assembly in established follicles. Indeed, cAMP induced the disassembly of focal adhesions in cultured fibroblasts (26) and decreased the number and prominence of focal adhesions in established thyroid monolayers (19). Our data do not yet allow us to identify the mechanism by which cAMP antagonizes tyrosine phosphorylation and focal adhesion assembly. It is possible that normally a balance exists between TSH and growth factors found in serum; withdrawal of TSH would then leave unopposed growth factor-driven signaling pathways that stimulate tyrosine phosphorylation (46). However, follicles convert to monolayer, and focal adhesions assemble even when cultures are washed into serum-free medium suggesting that cAMP may inhibit constitutive tyrosine kinases, which then become active upon withdrawal of TSH.
Focal adhesions are unlikely, however, to be the sole mechanisms responsible for determining thyroid epithelial organization. Cell-cell interactions are also important for tissue patterning (3), and it is noteworthy that as cells spread from follicles, E-cadherin and ZO-1 staining was substantially altered in regions of cell-cell contact. As cadherin-mediated adhesion and tight junctions mediate tissue cohesion and epithelial polarity (5, 9, 47), it seems plausible that their disassembly was necessary to allow the cellular rearrangement and change in polarized orientation that occur as follicles convert to monolayer. Interestingly, however, E-cadherin and ZO-1 staining was restored to junctions as cells established monolayer. This suggests that follicle to monolayer conversion may entail the coordinated regulation of cell-substrate and cell-cell junction assembly. We have not yet determined the regulatory mechanisms responsible for the self-limited disassembly and reassembly of cell-cell junctions. Although inhibition of protein tyrosine phosphatases indiscriminately increased tyrosine phosphorylation and caused the disassembly of cell-cell junctions in MDCK cells (48), in our studies the accumulation of phosphotyrosine staining in cell-cell junctions correlated better with the reassembly of junctions containing vinculin in areas of developing monolayer than with the initial loss of E-cadherin and ZO-1.
We conclude that tyrosine phosphorylation exerts a regulatory influence on thyroid epithelial organization in culture. Our findings place focal adhesion assembly, regulated by the interplay of tyrosine phosphorylation- and cAMP-dependent signaling pathways, at a locus that critically influences thyroid epithelial organization in culture. Inasmuch as cell adhesion influences the organization of many tissues (3), our findings may have implications for organs other than the thyroid. Further biochemical studies will be necessary to identify the precise tyrosine kinase signaling pathways involved and the manner in which they may interact with cAMP/protein kinase A signaling. Nonetheless, it is likely that characterization of the manner in which these signaling pathways impinge upon morphogenetic mechanisms will have a broad significance for the understanding of thyroid physiology and disease. The morphogenetic processes identified in cell culture, cell-matrix adhesion and cell locomotility, are also important determinants of epithelial organization in vivo (3) and, therefore, are likely to play significant roles in the native thyroid gland. Moreover, it is now apparent that in many cases of papillary thyroid cancer, cells express a fusion protein encoding the ret tyrosine kinase domain (PTC/ret) (49, 50). Based on our current work, the tyrosine phosphorylation-dependent checkpoints that regulate cell adhesion, locomotion, and tissue morphogenesis in the thyroid are likely to be important factors that contribute to converting the aberrant expression of a tyrosine kinase into the pathological phenotype of papillary thyroid cancer. Analysis of the mechanisms that convert cultured follicles into monolayer may then provide a useful experimental model to study the cellular processes by which oncogenic gene products disrupt normal thyroid epithelial morphogenesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Recipient of a National Health and Medical Research Council
Postgraduate Medical Research Scholarship.
3 Senior Scholar of the Alberta Heritage Foundation for Medical
Research.
Received November 19, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Basuroy, P. Sheth, D. Kuppuswamy, S. Balasubramanian, R. M. Ray, and R. K. Rao Expression of Kinase-inactive c-Src Delays Oxidative Stress-induced Disassembly and Accelerates Calcium-mediated Reassembly of Tight Junctions in the Caco-2 Cell Monolayer J. Biol. Chem., March 28, 2003; 278(14): 11916 - 11924. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Guo, J. N. Rao, L. Liu, M. Rizvi, D. J. Turner, and J.-Y. Wang Polyamines regulate beta -catenin tyrosine phosphorylation via Ca2+ during intestinal epithelial cell migration Am J Physiol Cell Physiol, September 1, 2002; 283(3): C722 - C734. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Smyth, O. Sheils, S. Finn, C. Martin, J. O'Leary, and E. C. Sweeney Real-time Quantitative Analysis of E-cadherin Expression in ret/PTC-1-Activated Thyroid Neoplasms International Journal of Surgical Pathology, October 1, 2001; 9(4): 265 - 272. [Abstract] [PDF]
|
|
|
|
|
|
O. M. Sheils, J. J. O'Leary, V. Uhlmann, K. Luttich, and E. C. Sweeney ret/PTC-1 Activation in Hashimoto Thyroiditis International Journal of Surgical Pathology, July 1, 2000; 8(3): 185 - 189. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
