Molecular and Cellular Endocrinology, 82 ( 199 1) R 1- R5 a 1YYl Elsevier Scientific Publishers Ireland. Ltd. 0303-7207/91/$03.50
MOLCEL
Rl
02662
Rapid Paper
Regulation of thyroid follicular volume by bidirectional ion transport A.S. Yap, J.W. Armstrong,
E.J. Cragoe, Jr. *, J.R. Bourke, and S.W. Manley
transepithelial G.J. Huxham
Department of Physiology and Pharmacology, The Unil~er.sityof Queensland, Brisbane 4072. Australia
Key worrist Thyroid
follicle:
Chloride
(Received
2X August
transport;
Sodium
1991; accepted
transport;
2 September
Bumetanide:
1991)
Phenamil:
Goitre
Summary Previous studies have shown that thyroid cells in monolayer culture exhibit bidirectional ion transport and oppositely directed bumetanide-sensicomprising apical-to-basal amiloride-sensitive Na+ transport tive Cll transport. We have now investigated the role of ion transport in the regulation of thyroid follicular size using follicular primary porcine thyroid cell cultures. Bumetanide (10 FM) added at the beginning of culture inhibited the formation of follicular lumina and caused a fall in follicle height when added to 3-day-old cultures. In contrast, phenamil (1 PM; an amiloride analog1 increased follicle size both in freshly isolated and 3-day-old cultures. The effect of bumetanide was prevented by the prior addition of phenamil. Micropuncture studies showed that follicles had a lumen-negative, basal-positive transepithelial potential difference which was progressively reduced in magnitude by the serial addition of bumetanide (10 PM) and phenamil (1 PM). We conclude that thyroid follicles possess a bidirectional ion transport system which transports Na+ in an apical-to-basal direction and Cl- in the opposite direction. The balance between these two processes determines net solute flux and hence follicular size. A physiological role of ion transport in the thyroid may be to regulate follicular volume suggesting that abnormalities of ion transport may be responsible for disorders of follicular size.
Introduction Thyroid epithelial cells are arranged as follicles whit lumina containing thyroglobulin-rich colloid, various ions and water (Halmi, 1986). In monolayer culture it has been shown that the thyroid epithelium transports fluid through the
Address for correspondence: Associate Professor S.W. Manley. Department of Physiology and Pharmacology, The University of Queensland. Brisbane 4072 Australia. * P.O. Box 631548, Nacogdoches. TX 75963-1548, U.S.A.
osmotic forces generated by transepithelial ion transport (Bourke et al., 1987; Chambard et al., 1987; Pearson et al., 1988). To date two such ion transport mechanisms have been identified. In the first, and best characterized, Nat absorption occurred in an apical-to-basal direction (Bourke et al., 1987; Pearson et al., 1988; Penel et al., 19891, involving amiloride-inhibitable Nat channels in the apical membrane (Matainaho et al., 1989) and a- basal ouabain-sensitive Na+/K+: ATPase (Pearson et al., 1988). More recently we have identified active transport of Cl- in the opposite (i.e. basal-to-apical) direction (Arm-
R2
strong et al., 1991, submitted). Cl- uptake was mediated by a basal NaKCl, symporter and inhibited by the loop diuretic bumetanide. Furthermore, both forms of ion transport were stimulated by cyclic AMP agonists and thus potentially under hormonal control (Bourke et al., 1987; Armstrong et al., 1991, submitted). In the present study we have investigated the physiological significance of ion transport using follicular thyroid cell cultures. We have identified bidirectional ion transport in follicles and present evidence that this fluid transport system regulates thyroid follicular volume. Material
and methods
Cell cultures. Porcine thyroid cells were isolated by digestion of tissue with neutral protease (Dispase; Boehringer-Mannheim, Sydney, NSW, Australia) and collagenase (Worthington type 1; Flow Laboratories, Sydney, NSW, Australia) as previously described (Bourke et al., 1981; Yap et al., 1987). Cells were cultured in medium containing 10% (v/v> heat-inactivated calf serum (Commonwealth Serum Laboratories, Melbourne, Australia) at a density of 3 x 10h cells/35 X 10 mm tissue culture dish (Corning Plastics, New York, NY, U.S.A.). Bovine TSH (256 pU/ml; Thytropar; Armour Pharmaceuticals, Sydney, Australia) was added at the beginning of incubation. Prior to use in experiments cultures were washed into fresh serum-free medium containing 1 mu/ml TSH. Measurement of fluid transport. Fluid transport was quantitated by change in follicle height in a computer-controlled microscope apparatus as described for the measurement of monolayer dome height (Bourke et al., 1987). Readings were made at 15 min intervals and drugs added 30 min after beginning the experiment. Where both phenamil and bumetanide were used, phenamil was added at the beginning of the experiment. At the end of each experiment individual follicle heights were plotted against time. To reduce the variance due to differences in initial follicle height, the data for each follicle were normalized by subtraction of the height at 45 min from other readings for that follicle. In addition, the initial rate of change in height (velocity; pm/h) was
calculated by regression analysis for each follicle from the four sets of readings after normalization. Data were pooled from 4-5 dishes. Statistical comparisons were performed on the initial velocities by Student’s t-test. Measurement of follicular transepithelial potential difference. Micropuncture studies were performed by a modification of the method of Bourke et al. (1990). The follicular transepithelial potential (TEP) was measured by penetration of the cell layer with a 40-60 MR single-barrelled glass microelectrode (containing 0.5 M KCI) and a Narishige micromanipulator. Criteria of acceptance were that the potential transition was immediate, the TEP was constant for 10 s and on withdrawal the electrode voltage offset was less than 2 mV. The results presented were pooled from three separate cultures. Scanning electron microscopy. Follicles were torn open by micromanipulation with a glass microelectrode. Cultures were then fixed in 2.5% glutaraldehyde, dehydrated in ethanol, dried in liquid CO, in a critical-point dryer, coated with gold in vacua and examined by secondary electron emission in a Jeol JSM 35C scanning electron microscope. Drugs. Phenamil was synthesized as previously described (Cragoe et al., 1967). Phenamil and bumetanide (a gift from Astra Pharmaceuticals, Sydney, Australia) were dissolved in dimethyl sulphoxide (DMSO, Sigma) as 10 mM and 100 mM stock solutions respectively. Control experiments included DMSO in equal final concentrations. Results Freshly isolated thyroid cells formed phasebright aggregates in which follicular lumina were first evident after 24 h incubation with TSH (256 pU/ml). By 72 h numerous follicles were seen (Fig. 1 A). Scanning electron microscopy showed that in mechanically disrupted follicles microvilli were present on the luminal, but not the basal, surface of follicles (Fig. 2), indicating cell polarization similar to that in the normal gland (Halmi, 1986). In dishes incubated with 10 PM bumetanide in addition to TSH, cell aggregates were present but
R3
Fig. 2. Scanning electron micrograph of a disrupted follicle. Numerous microvilli are seen on the tuminal surface (centre of section) but not on the basal surface of the follicle (to left and right). Bar is 10 Wm.
lumina were conspicuously larger than those in control dishes with TSH alone (Fig. 1C). The effect of drugs on established follicles was studied in 72-h-old cultures. Follicle height in control cultures exposed to vehicle alone (DMSO; l/10’ v/v) showed little change during the 150 min experimental period (velocity - 1.29 f 1.67 pm/h, mean _t SE, n = 75, Fig. 31. Bumetanide (10 PM) produced a fall in follicle height fveloc-
F
Fig. 1. Effects of bumetanide and phenamil upon follicular reorganization in culture. Phase contrast photomicrographs of primary porcine thyroid cell cultures incubated for 3 days with TSH (256 ,~~U/rnf) alone (A), in the presence of 10 FM bumetanide (B) or with 1 PM phenamil (CL Bar represents 100 urn.
follicuIar contrast, amiloride man and
lumina were not observed (Fig. 1Bl. In in the presence of 1 PM phenamil (an analog specific for Na ’ channels (KleyCragoe, 19881) and TSH the follicular
4
Time
(min)
Fig. 3. Change in thyroid foilicular height in response to bumetanide and phenamil. Follicle height was measured in 3-day-old cultures. Bumetanide (B; 10 FM) or phenamil (P; 1 PM) were added at 30 min. Where both drugs were present (PB) phenamil (1 PM) was added at 0 min and bumetanide (10 FM) at 30 min. Control cultures (C) received DMSO (l/lo4 v/v)_ Data are shown as means (*SEM) of 50-75 follicles pooled from 4-S dishes.
R4
7ime Fig. 1. Effect
of bumetanide
transepithelial
potential
sured by micropuncture (filled
circles),
experiment. triangles)
(TEP).
\minj
and phenamil Follicular
in J-day-old
the TEP
was stable
upon TEPs
then
phenamil
mea-
cultures. In control dishes for the duration
In test dishes first bumetanide and
follicular
were
(1 PM)
(IO FM)
(open
circles)
of the (open were
added and the TEP measured.
ity - 19.0 k 1.80 pm/h, n = 75; P < 0.001 vs. control). In contrast, phenamil (1 PM) caused follicle height to increase (velocity + 5.34 f 2.87 pm/h, n = 60; P < 0.05 vs. control). In the presence of both phenamil and bumetanide no significant change in height occurred (velocity - 3.24 & 1.75 pm/h, n = 73, P > 0.2 vs. control). Follicles in 3-day-old cultures exhibited a TEP of - 19.0 + 0.5 mV (lumen negative, n = 51, Fig. 4). In control dishes TEP remained stable for the duration of the experiment, but addition of bumetanide (10 PM) and phenamil (1 PM) reduced the magnitude of the TEP. In the presence of bumetanide the TEP was - 10.1 + 0.3 mV tn = 51, P < 0.001 vs. control), and with bumetanide and phenamil was -2.6 & 0.1 mV tn = 94. P < 0.001 vs. bumetanide alone). Discussion Like other transporting epithelia, the thyroid epithelium exhibits an electrical potential difference between the apical and basal surfaces (TEP) the sign and magnitude of which reflect the ionic transport mechanisms and paracellular shunt pathways in the epithelium (Ma&night et al., 1980; Pearson et al., 1988). From first principles
the lumen-negative, basal-positive TEP that we observed in follicles could be due to the outward (i.e. apical-to-basal) transport of cations or to the inward (i.e. basal-to-apical) transport of anions. The serial addition of bumetanide (to inhibit Cltransport (Schlatter et al., 1983)) and phenamil (to inhibit Naf transport (Matainaho et al., 1989)) produced a stepwise reduction in the TEP consistent with active transport of Cll towards the follicular lumen and of Na+ away from the lumen. We conclude that thyroid follicles exhibit bidirectional ion transport analogous to that of the tracheal epithelium which absorbs Na+ and secretes Cl- (Liedtke, 1989). The fall in follicle height induced by bumetanide indicated that inward Cl- transport was associated with fluid flux. As this fall was prevented by prior addition of phenamil it is unlikely that the decrease in follicle height was attributable simply to the elastic recoil of the epithelium. Instead, this observation suggests that when Cll transport was inhibited, the follicles were being emptied by unopposed outwardly-directed Na+ transport. Conversely, the increase in follicle height that occurred when Naf transport alone was blocked is consistent with ongoing inward Cll transport. The direction of bumetanide-sensitive and phenamil-sensitive fluid transport corresponds to the orientation of Cland Nat active transport indicated by their contribution to the basal-positive TEP. These data support a model for ion transport in the thyroid where follicle volume is determined by the balance between inward Cl- transport and outward Na+ transport (Fig. 5). Changes in either transport process sufficient to perturb net fluid flux across the epithelium are likely to cause a change in follicular volume. However, the stable height of control cultures after 72 h incubation suggests that in established follicles NaC and Cll transport normally exist in balance. Vectorial transport in epithelia depends upon the maintenance of normal epithelial cell polarity (Simons and Fuller, 1985). Although under some conditions, cultured thyroid follicles may show cell polarity that is the reverse of normal (Nitsch and Wollman, 19801, in this study the presence of microvilli on the luminal, but not the basal, surface of follicles indicated that normal epithelial
R5
zation, Cannon Hill, Brisbane, for access to equipment and help with scanning electron micrographs and Ms. V. Cooper and Mr. K. Abel for excellent technical assistance. This research was supported by the National Health and Medical Research Council of Australia. A.Y. was the recipient of an NH&MRC Postgraduate Medical Scholarship and J.A. of an Australian Postgraduate Research Award.
basal
Fig. 5. Hypothetical orientation of transepithelial ion transport systems in the thyroid follicle. Apical-to-basal Naf transport is orientated in an outward direction from the lumen to the medium (or extracelluiar fluid space) whereas basal-to-apical Cl- transport directs fluid flux inwards to the lumen of the follicle. A combination of outward cation transport and inward anion transport generates an overall basalpositive transepithelial potential difference (TEP). This model predicts that unopposed Na+ transport would lead to a decrease in follicle lumen (and hence overall follicle) size while unopposed Cl transport would increase follicle size.
cell polarity has been preserved (Halmi, 1986). Consequently we infer that the directions of Cland Naf transport demonstrated in these cultured follicles correspond to those in the intact tissue in vivo. We conclude that bidirectional ion transport is present in cultured thyroid follicles and influences follicular size by determining fluid accumulation in the follicuiar lumen. Furthermore, the change in follicle size seen when cells are incubated with drugs from the beginning of culture suggests that ion transport may be instrumental in follicular reorganization. Since abnormalities of follicle size are seen in a range of thyroid diseases, including simple colloid and multinodular goitres (Studer et al., 1989) a possible role for disturbed fluid transport in their pathogenesis warrants investigation. Acknowledgments
We thank Dr. D. Morton of the Commonwealth Scientific and Industrial Research Organi-
References Bourke. J.R., Carseldine. K.L., Ferris, S.H., Huxham, G.J. and Manley, S.W. (19811 J. Endocrinol. 88. 187-196. Bourke. J.R., Matainaho. T., Huxham, G.J. and Manley, S.W. (1987) J. Endocrinol. 115. 19-X. Bourke. J.R., Cragoe, Jr.. E.J., Huxham, G.J.. Pearson. J.V. and Manley, SW. 11990) J. Endocrinol. 127, 107-202. Chamhard. M.. Mauchamp. J. and Chabaud. 0. (1987) J. Cell. Physiol. 133, 37-4s. Cragoe. Jr., E.J.. Woltersdorf. Jr., O.W., Bicking, J.B., Kwong, S.F. and Jones, J.F. 11967) J. Med. Chem. IO. 66-75. Eagle. H. (1959) Science 130, 432-437. Halmi, N.S. (1986) in Werner‘s The Thyroid, A Fundamental and Clinical Text (Ingbar, S.H. and Braverman. L.E., eds.). pp. 24-36. Lippincott, Philadelphia, PA. Kleyman, T.R. and Cragoe. Jr.. E.J. (1988) J. Membr. Biol. 105, 1-21. Leidtke. C.M. (1989) Annu. Rev. Physiol. 51, 143-160. Lissitzky, S.. Fayet, G.. Giraud, A., Verrier, B. and Torresani, J. (1971) Eur. J. Biochem. 24. 88-99. MacKnight. A.D.C.. DiBona, D.R. and Leaf, A. (1980) Physiol. Rev. 60. 615-715. Matainaho. T., Cragoe, Jr., E.J.. Manley, S.W.. Huxham, G.J., Pearson. J.V. and Bourke, J.R. (1989) J. Endocrinol. 123, 93-97. Nitsch. L. and Wollman, S.H. (1980) J. Cell Biol. 86, X75-880. Pearson. J.. Bourke. J.R., Manley. S.W.. Huxham, G.J., Matainah~~. T.. Gerard, C.. Verrier. B. and Mauchamp, J. (1988) J. Endocrinol. 119.309-314. Penel. C., Gerard. C.. Mauchamp, J. and Verrier. B. (1989) Pfliig. Arch. 313, 509-515. Schlatter. E.. Greger. R. and Weidtke. C. (1983) Pfliig. Arch. 396.210-217. Simons. K. and Fuller, SD. (1085) Annu. Rev. Cell Biol. 1, 243-288. Studer, H.. Peter. H.J. and Gerber. H. (1489) Endocr. Rev. IQ, 325-135. Yap, AS.. Bourke. J.R. and Manley. SW. (1987) J. Endocrinol. 113, 2233229.
MOLCEL
Rl
02662
Rapid Paper
Regulation of thyroid follicular volume by bidirectional ion transport A.S. Yap, J.W. Armstrong,
E.J. Cragoe, Jr. *, J.R. Bourke, and S.W. Manley
transepithelial G.J. Huxham
Department of Physiology and Pharmacology, The Unil~er.sityof Queensland, Brisbane 4072. Australia
Key worrist Thyroid
follicle:
Chloride
(Received
2X August
transport;
Sodium
1991; accepted
transport;
2 September
Bumetanide:
1991)
Phenamil:
Goitre
Summary Previous studies have shown that thyroid cells in monolayer culture exhibit bidirectional ion transport and oppositely directed bumetanide-sensicomprising apical-to-basal amiloride-sensitive Na+ transport tive Cll transport. We have now investigated the role of ion transport in the regulation of thyroid follicular size using follicular primary porcine thyroid cell cultures. Bumetanide (10 FM) added at the beginning of culture inhibited the formation of follicular lumina and caused a fall in follicle height when added to 3-day-old cultures. In contrast, phenamil (1 PM; an amiloride analog1 increased follicle size both in freshly isolated and 3-day-old cultures. The effect of bumetanide was prevented by the prior addition of phenamil. Micropuncture studies showed that follicles had a lumen-negative, basal-positive transepithelial potential difference which was progressively reduced in magnitude by the serial addition of bumetanide (10 PM) and phenamil (1 PM). We conclude that thyroid follicles possess a bidirectional ion transport system which transports Na+ in an apical-to-basal direction and Cl- in the opposite direction. The balance between these two processes determines net solute flux and hence follicular size. A physiological role of ion transport in the thyroid may be to regulate follicular volume suggesting that abnormalities of ion transport may be responsible for disorders of follicular size.
Introduction Thyroid epithelial cells are arranged as follicles whit lumina containing thyroglobulin-rich colloid, various ions and water (Halmi, 1986). In monolayer culture it has been shown that the thyroid epithelium transports fluid through the
Address for correspondence: Associate Professor S.W. Manley. Department of Physiology and Pharmacology, The University of Queensland. Brisbane 4072 Australia. * P.O. Box 631548, Nacogdoches. TX 75963-1548, U.S.A.
osmotic forces generated by transepithelial ion transport (Bourke et al., 1987; Chambard et al., 1987; Pearson et al., 1988). To date two such ion transport mechanisms have been identified. In the first, and best characterized, Nat absorption occurred in an apical-to-basal direction (Bourke et al., 1987; Pearson et al., 1988; Penel et al., 19891, involving amiloride-inhibitable Nat channels in the apical membrane (Matainaho et al., 1989) and a- basal ouabain-sensitive Na+/K+: ATPase (Pearson et al., 1988). More recently we have identified active transport of Cl- in the opposite (i.e. basal-to-apical) direction (Arm-
R2
strong et al., 1991, submitted). Cl- uptake was mediated by a basal NaKCl, symporter and inhibited by the loop diuretic bumetanide. Furthermore, both forms of ion transport were stimulated by cyclic AMP agonists and thus potentially under hormonal control (Bourke et al., 1987; Armstrong et al., 1991, submitted). In the present study we have investigated the physiological significance of ion transport using follicular thyroid cell cultures. We have identified bidirectional ion transport in follicles and present evidence that this fluid transport system regulates thyroid follicular volume. Material
and methods
Cell cultures. Porcine thyroid cells were isolated by digestion of tissue with neutral protease (Dispase; Boehringer-Mannheim, Sydney, NSW, Australia) and collagenase (Worthington type 1; Flow Laboratories, Sydney, NSW, Australia) as previously described (Bourke et al., 1981; Yap et al., 1987). Cells were cultured in medium containing 10% (v/v> heat-inactivated calf serum (Commonwealth Serum Laboratories, Melbourne, Australia) at a density of 3 x 10h cells/35 X 10 mm tissue culture dish (Corning Plastics, New York, NY, U.S.A.). Bovine TSH (256 pU/ml; Thytropar; Armour Pharmaceuticals, Sydney, Australia) was added at the beginning of incubation. Prior to use in experiments cultures were washed into fresh serum-free medium containing 1 mu/ml TSH. Measurement of fluid transport. Fluid transport was quantitated by change in follicle height in a computer-controlled microscope apparatus as described for the measurement of monolayer dome height (Bourke et al., 1987). Readings were made at 15 min intervals and drugs added 30 min after beginning the experiment. Where both phenamil and bumetanide were used, phenamil was added at the beginning of the experiment. At the end of each experiment individual follicle heights were plotted against time. To reduce the variance due to differences in initial follicle height, the data for each follicle were normalized by subtraction of the height at 45 min from other readings for that follicle. In addition, the initial rate of change in height (velocity; pm/h) was
calculated by regression analysis for each follicle from the four sets of readings after normalization. Data were pooled from 4-5 dishes. Statistical comparisons were performed on the initial velocities by Student’s t-test. Measurement of follicular transepithelial potential difference. Micropuncture studies were performed by a modification of the method of Bourke et al. (1990). The follicular transepithelial potential (TEP) was measured by penetration of the cell layer with a 40-60 MR single-barrelled glass microelectrode (containing 0.5 M KCI) and a Narishige micromanipulator. Criteria of acceptance were that the potential transition was immediate, the TEP was constant for 10 s and on withdrawal the electrode voltage offset was less than 2 mV. The results presented were pooled from three separate cultures. Scanning electron microscopy. Follicles were torn open by micromanipulation with a glass microelectrode. Cultures were then fixed in 2.5% glutaraldehyde, dehydrated in ethanol, dried in liquid CO, in a critical-point dryer, coated with gold in vacua and examined by secondary electron emission in a Jeol JSM 35C scanning electron microscope. Drugs. Phenamil was synthesized as previously described (Cragoe et al., 1967). Phenamil and bumetanide (a gift from Astra Pharmaceuticals, Sydney, Australia) were dissolved in dimethyl sulphoxide (DMSO, Sigma) as 10 mM and 100 mM stock solutions respectively. Control experiments included DMSO in equal final concentrations. Results Freshly isolated thyroid cells formed phasebright aggregates in which follicular lumina were first evident after 24 h incubation with TSH (256 pU/ml). By 72 h numerous follicles were seen (Fig. 1 A). Scanning electron microscopy showed that in mechanically disrupted follicles microvilli were present on the luminal, but not the basal, surface of follicles (Fig. 2), indicating cell polarization similar to that in the normal gland (Halmi, 1986). In dishes incubated with 10 PM bumetanide in addition to TSH, cell aggregates were present but
R3
Fig. 2. Scanning electron micrograph of a disrupted follicle. Numerous microvilli are seen on the tuminal surface (centre of section) but not on the basal surface of the follicle (to left and right). Bar is 10 Wm.
lumina were conspicuously larger than those in control dishes with TSH alone (Fig. 1C). The effect of drugs on established follicles was studied in 72-h-old cultures. Follicle height in control cultures exposed to vehicle alone (DMSO; l/10’ v/v) showed little change during the 150 min experimental period (velocity - 1.29 f 1.67 pm/h, mean _t SE, n = 75, Fig. 31. Bumetanide (10 PM) produced a fall in follicle height fveloc-
F
Fig. 1. Effects of bumetanide and phenamil upon follicular reorganization in culture. Phase contrast photomicrographs of primary porcine thyroid cell cultures incubated for 3 days with TSH (256 ,~~U/rnf) alone (A), in the presence of 10 FM bumetanide (B) or with 1 PM phenamil (CL Bar represents 100 urn.
follicuIar contrast, amiloride man and
lumina were not observed (Fig. 1Bl. In in the presence of 1 PM phenamil (an analog specific for Na ’ channels (KleyCragoe, 19881) and TSH the follicular
4
Time
(min)
Fig. 3. Change in thyroid foilicular height in response to bumetanide and phenamil. Follicle height was measured in 3-day-old cultures. Bumetanide (B; 10 FM) or phenamil (P; 1 PM) were added at 30 min. Where both drugs were present (PB) phenamil (1 PM) was added at 0 min and bumetanide (10 FM) at 30 min. Control cultures (C) received DMSO (l/lo4 v/v)_ Data are shown as means (*SEM) of 50-75 follicles pooled from 4-S dishes.
R4
7ime Fig. 1. Effect
of bumetanide
transepithelial
potential
sured by micropuncture (filled
circles),
experiment. triangles)
(TEP).
\minj
and phenamil Follicular
in J-day-old
the TEP
was stable
upon TEPs
then
phenamil
mea-
cultures. In control dishes for the duration
In test dishes first bumetanide and
follicular
were
(1 PM)
(IO FM)
(open
circles)
of the (open were
added and the TEP measured.
ity - 19.0 k 1.80 pm/h, n = 75; P < 0.001 vs. control). In contrast, phenamil (1 PM) caused follicle height to increase (velocity + 5.34 f 2.87 pm/h, n = 60; P < 0.05 vs. control). In the presence of both phenamil and bumetanide no significant change in height occurred (velocity - 3.24 & 1.75 pm/h, n = 73, P > 0.2 vs. control). Follicles in 3-day-old cultures exhibited a TEP of - 19.0 + 0.5 mV (lumen negative, n = 51, Fig. 4). In control dishes TEP remained stable for the duration of the experiment, but addition of bumetanide (10 PM) and phenamil (1 PM) reduced the magnitude of the TEP. In the presence of bumetanide the TEP was - 10.1 + 0.3 mV tn = 51, P < 0.001 vs. control), and with bumetanide and phenamil was -2.6 & 0.1 mV tn = 94. P < 0.001 vs. bumetanide alone). Discussion Like other transporting epithelia, the thyroid epithelium exhibits an electrical potential difference between the apical and basal surfaces (TEP) the sign and magnitude of which reflect the ionic transport mechanisms and paracellular shunt pathways in the epithelium (Ma&night et al., 1980; Pearson et al., 1988). From first principles
the lumen-negative, basal-positive TEP that we observed in follicles could be due to the outward (i.e. apical-to-basal) transport of cations or to the inward (i.e. basal-to-apical) transport of anions. The serial addition of bumetanide (to inhibit Cltransport (Schlatter et al., 1983)) and phenamil (to inhibit Naf transport (Matainaho et al., 1989)) produced a stepwise reduction in the TEP consistent with active transport of Cll towards the follicular lumen and of Na+ away from the lumen. We conclude that thyroid follicles exhibit bidirectional ion transport analogous to that of the tracheal epithelium which absorbs Na+ and secretes Cl- (Liedtke, 1989). The fall in follicle height induced by bumetanide indicated that inward Cl- transport was associated with fluid flux. As this fall was prevented by prior addition of phenamil it is unlikely that the decrease in follicle height was attributable simply to the elastic recoil of the epithelium. Instead, this observation suggests that when Cll transport was inhibited, the follicles were being emptied by unopposed outwardly-directed Na+ transport. Conversely, the increase in follicle height that occurred when Naf transport alone was blocked is consistent with ongoing inward Cll transport. The direction of bumetanide-sensitive and phenamil-sensitive fluid transport corresponds to the orientation of Cland Nat active transport indicated by their contribution to the basal-positive TEP. These data support a model for ion transport in the thyroid where follicle volume is determined by the balance between inward Cl- transport and outward Na+ transport (Fig. 5). Changes in either transport process sufficient to perturb net fluid flux across the epithelium are likely to cause a change in follicular volume. However, the stable height of control cultures after 72 h incubation suggests that in established follicles NaC and Cll transport normally exist in balance. Vectorial transport in epithelia depends upon the maintenance of normal epithelial cell polarity (Simons and Fuller, 1985). Although under some conditions, cultured thyroid follicles may show cell polarity that is the reverse of normal (Nitsch and Wollman, 19801, in this study the presence of microvilli on the luminal, but not the basal, surface of follicles indicated that normal epithelial
R5
zation, Cannon Hill, Brisbane, for access to equipment and help with scanning electron micrographs and Ms. V. Cooper and Mr. K. Abel for excellent technical assistance. This research was supported by the National Health and Medical Research Council of Australia. A.Y. was the recipient of an NH&MRC Postgraduate Medical Scholarship and J.A. of an Australian Postgraduate Research Award.
basal
Fig. 5. Hypothetical orientation of transepithelial ion transport systems in the thyroid follicle. Apical-to-basal Naf transport is orientated in an outward direction from the lumen to the medium (or extracelluiar fluid space) whereas basal-to-apical Cl- transport directs fluid flux inwards to the lumen of the follicle. A combination of outward cation transport and inward anion transport generates an overall basalpositive transepithelial potential difference (TEP). This model predicts that unopposed Na+ transport would lead to a decrease in follicle lumen (and hence overall follicle) size while unopposed Cl transport would increase follicle size.
cell polarity has been preserved (Halmi, 1986). Consequently we infer that the directions of Cland Naf transport demonstrated in these cultured follicles correspond to those in the intact tissue in vivo. We conclude that bidirectional ion transport is present in cultured thyroid follicles and influences follicular size by determining fluid accumulation in the follicuiar lumen. Furthermore, the change in follicle size seen when cells are incubated with drugs from the beginning of culture suggests that ion transport may be instrumental in follicular reorganization. Since abnormalities of follicle size are seen in a range of thyroid diseases, including simple colloid and multinodular goitres (Studer et al., 1989) a possible role for disturbed fluid transport in their pathogenesis warrants investigation. Acknowledgments
We thank Dr. D. Morton of the Commonwealth Scientific and Industrial Research Organi-
References Bourke. J.R., Carseldine. K.L., Ferris, S.H., Huxham, G.J. and Manley, S.W. (19811 J. Endocrinol. 88. 187-196. Bourke. J.R., Matainaho. T., Huxham, G.J. and Manley, S.W. (1987) J. Endocrinol. 115. 19-X. Bourke. J.R., Cragoe, Jr.. E.J., Huxham, G.J.. Pearson. J.V. and Manley, SW. 11990) J. Endocrinol. 127, 107-202. Chamhard. M.. Mauchamp. J. and Chabaud. 0. (1987) J. Cell. Physiol. 133, 37-4s. Cragoe. Jr., E.J.. Woltersdorf. Jr., O.W., Bicking, J.B., Kwong, S.F. and Jones, J.F. 11967) J. Med. Chem. IO. 66-75. Eagle. H. (1959) Science 130, 432-437. Halmi, N.S. (1986) in Werner‘s The Thyroid, A Fundamental and Clinical Text (Ingbar, S.H. and Braverman. L.E., eds.). pp. 24-36. Lippincott, Philadelphia, PA. Kleyman, T.R. and Cragoe. Jr.. E.J. (1988) J. Membr. Biol. 105, 1-21. Leidtke. C.M. (1989) Annu. Rev. Physiol. 51, 143-160. Lissitzky, S.. Fayet, G.. Giraud, A., Verrier, B. and Torresani, J. (1971) Eur. J. Biochem. 24. 88-99. MacKnight. A.D.C.. DiBona, D.R. and Leaf, A. (1980) Physiol. Rev. 60. 615-715. Matainaho. T., Cragoe, Jr., E.J.. Manley, S.W.. Huxham, G.J., Pearson. J.V. and Bourke, J.R. (1989) J. Endocrinol. 123, 93-97. Nitsch. L. and Wollman, S.H. (1980) J. Cell Biol. 86, X75-880. Pearson. J.. Bourke. J.R., Manley. S.W.. Huxham, G.J., Matainah~~. T.. Gerard, C.. Verrier. B. and Mauchamp, J. (1988) J. Endocrinol. 119.309-314. Penel. C., Gerard. C.. Mauchamp, J. and Verrier. B. (1989) Pfliig. Arch. 313, 509-515. Schlatter. E.. Greger. R. and Weidtke. C. (1983) Pfliig. Arch. 396.210-217. Simons. K. and Fuller, SD. (1085) Annu. Rev. Cell Biol. 1, 243-288. Studer, H.. Peter. H.J. and Gerber. H. (1489) Endocr. Rev. IQ, 325-135. Yap, AS.. Bourke. J.R. and Manley. SW. (1987) J. Endocrinol. 113, 2233229.
