- Email: [email protected]
Prolactin, alone or in combination with glucocorticoids, enhances tight junction formation and expression of the tight junction protein occludin in mammary cells
Molecular and Cellular Endocrinology 156 (1999) 55 – 61 www.elsevier.com/locate/mce
Prolactin, alone or in combination with glucocorticoids, enhances tight junction formation and expression of the tight junction protein occludin in mammary cells Kerst Stelwagen *, Holli A. McFadden, Jerome Demmer Dairy Science, Ruakura Research Centre, AgResearch, Pri6ate Bag 3123, Hamilton, New Zealand Received 19 April 1999; accepted 28 June 1999
Abstract Tight junctions (TJ) between adjacent epithelial cells play an important role in maintaining mammary function in the differentiated mammary gland. Mouse mammary cell lines (HC11 and Comma-1D) were used to investigate the effect of the lactogenic hormones prolactin (PRL) and glucocorticoids on the formation of mammary TJ. TJ formation was assessed by an increase in transepithelial electrical resistance and a decrease in paracellular flux of radiolabeled inulin. Both PRL and the synthetic glucocorticoid dexamethasone (DEX) stimulated TJ formation. The biggest effect on TJ formation was observed when both hormones were used in combination, but only when cells were pretreated with DEX. The effects of PRL and DEX are mediated, at least in part, via expression of the transmembrane TJ protein occludin. In summary, these data are the first to show an effect of PRL on mammary TJ formation and the expression of TJ proteins, and confirm the TJ-stimulating effects of glucocorticoids that have been reported previously. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Prolactin; Tight junction; Occludin; Mammary gland; HC11 cells
1. Introduction The tight junction (TJ) is the most apical member of the junctional complex between epithelial or endothelial cells. The TJ acts as a barrier between adjacent cells, and as such regulates the paracellular transport of ions and small molecules (‘barrier function’), but it also separates the plasma membrane of the cell into basolateral and apical domains of distinct composition and function (‘fence function’) (Schneeberger and Lynch, 1992). TJ were long considered static structures, and of unknown composition, but recently much insight has been gained into the molecular make-up of the TJ. The first true transmembrane TJ protein identified is occludin (Furuse et al., 1993), and recently two other Abbre6iations: DEX, dexamethasone; PRL, prolactin; TER, transepithelial electrical resistance; TJ, tight junction(s). * Corresponding author. Present address: Research Station for Cattle, Sheep and Horse Husbandry (PR), PO Box 2176, 8203 AD Lelystad, The Netherlands. Tel.: + 31-320-293211; fax: + 31-320241584. E-mail address: [email protected] (K. Stelwagen)
proteins have been isolated from junction fractions, claudin-1 and -2 (Furuse et al., 1998). Little is known about the possible role of claudins, however occludin is intimately connected to the cytoskeleton of the cell via TJ-associated proteins such as ZO-1 (Fanning et al., 1998). The linkage of TJ to the cytoskeleton suggests that TJ may play an important role in cell functioning. Indeed, there is a link between TJ and the Rho GTPase-mediated signaling pathway in the cell (Nusrat et al., 1995; Gopalakrishnan et al., 1998). Consistent with a role of TJ in cell functioning, epithelial TJ in the mammary gland play an important role in maintaining milk synthesis, as demonstrated by the fact that induced loss of barrier function in the lactating mammary gland results in a significant decrease in milk synthesis (Stelwagen et al., 1995, 1997). However, the exact mechanism by which the loss of TJ barrier function affects milk synthesis is not fully understood. In the mammary gland TJ form during lactogenesis when the gland switches from a growth state to a differentiated state, i.e. when milk synthesis and secretion are initiated (Linzell and Peaker, 1974). This
0303-7207/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 3 - 7 2 0 7 ( 9 9 ) 0 0 1 4 5 - 8
56
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
switch coincides with major changes in systemic endocrine profiles at that time, such as rapidly decreasing levels of progesterone during the immediate pre-partum period combined with a sharp increase in the levels of prolactin (PRL) and glucocorticoids (Convey, 1974). Indeed, recent studies have shown that mammary TJ are under endocrine regulation. Progesterone, involved in maintenance of gestation, and thus preventing mammary cells undergoing differentiation, inhibits TJ formation (Nguyen and Neville, 1998), whereas the rapid decrease in progesterone during lactogenesis allows for TJ formation. Moreover, glucocorticoids have been shown to have a direct effect on the formation and maintenance of mammary TJ in vitro (Zettl et al., 1992; Singer et al., 1994) and in vivo (Thompson, 1996; Stelwagen et al., 1998). The effects of PRL on mammary TJ are much less defined. PRL in combination with insulin or with insulin and the synthetic glucocorticoid dexamethasone (DEX) appear to have no effect on TJ in 31EG4 mouse-derived mammary cells (Zettl et al., 1992). In contrast, in in vivo studies, where animals received exogenous PRL (Linzell et al., 1975) or where endogenous PRL release was blocked by bromocriptine (Flint and Gardner, 1994), corresponding changes in the levels of sodium, potassium and lactose in milk were observed that suggest that PRL may act upon mammary TJ. However, it can not be ruled out that these in vivo effects of PRL may be on cell maintenance or survival and thereby preventing alveolar cell loss, which would result in similar changes in milk composition. Therefore, the objective of the present study was to examine whether PRL, alone or in combination with glucocorticoids, could enhance the formation of mammary TJ, and if so, was this effect mediated via the TJ protein occludin and/or TJ-associated proteins
2. Materials and methods
2.1. Materials HC11 (Ball et al., 1988) and COMMA-1D (Danielson et al., 1984) mouse mammary cells were gifts from Dr B. Groner (Chemotherapeutisches Forschungsinstitut, Frankfurt am Main, Germany) and Dr D. Medina (Baylor College of Medicine, Houston, TX, USA), respectively. Media for HC11 (RPMI 1640) and COMMA-1D (DMEM-F12) were from Gibco BRL (Life Technologies, Auckland, New Zealand) and Sigma (St. Louis, MO, USA), respectively. Fetal calf serum came from Life Technologies. Ovine PRL, DEX, murine submaxillary gland epidermal growth factor, bovine pancreatic insulin, penicillin – streptomycin, bovine serum albumin were all purchased from Sigma. 3 H-inulin was obtained from Amersham New Zealand
(Auckland, New Zealand). Cell culture inserts were from Biolab Scientific, Auckland New Zealand (HC11: Millipore PCF, 12 mm, 0.4 mm pore size) and Life Technologies (COMMA-1D: Nunc Anopore, 10 mm, 0.02 mm pore size). The ZO-1 monoclonal antibody, developed by D.A. Goodenough, was purchased from the Developmental Studies Hybridoma Bank (University of Iowa, Iowa City, USA) and the occludin antibody came from Zymed Laboratories Inc. (BioScientific, Gymes 2227, N.S.W., Australia). Secondary antibodies were from Amersham Life Science (Auckland, New Zealand). The BCA protein assay kit was from Pierce, Rockford, IL, USA. Protease inhibitor tablets were purchased from Boehringer Mannheim, Auckland, New Zealand. All other reagents were of the highest available purity.
2.2. Cell culture HC11 cells were cultured as previously described (Stelwagen and Ormrod, 1998). Briefly, cells were seeded on permeable Millipore PCF inserts at a density of 4× 105 cells per insert. Cells were grown in growth medium (RPMI 1640), supplemented with 10% fetal calf serum, penicillin (5 U/ml) and streptomycin (5 mg/ml), insulin (5 mg/ml), and epidermal growth factor (10 ng/ml). Once cells became confluent they were maintained for 24 h in priming medium (same as growth medium, but with no epidermal growth factor). Cells were then kept in differentiation medium, i.e. priming medium supplemented with PRL (1 or 5 mg/ml) and/or DEX (1 mM). COMMA-1D cells were seeded on Nunc Anopore membranes at a density of 1.8×105 cells per insert. Cells were grown to confluence in growth medium (DMEM-F12), supplemented with 2% fetal calf serum, serum albumin (300 ng/ml), penicillinstreptomycin (as for HC11), insulin (5 mg/ml), and epidermal growth factor (5 ng/ml). Subsequently, cells were kept in priming (24 h) and differentiation media. Both media were based on the growth medium, and hormone omission and/or supplementation were as described for HC11 cells. Media was changed every 24 h, and cells were culture at 37°C in an atmosphere of 5% CO2 and 95% air.
2.3. TJ formation Formation of TJ was assessed by transepithelial electrical resistance (TER) and paracellular movement of the 3H-inulin (5200 MW) inert marker across cell monolayers. An increase in TER represents increased TJ formation. TER was measured across cell monolayers every 24 h using a Millicell-ERS voltohmmeter (Millipore, Bedford, MA, USA) after transferring an insert to an Endohm-12 chamber (World Precision Instruments, Sarasota, FL, USA). TER values of insert
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
57
membrane and medium (i.e. no cells, medium only), and membrane surface area (i.e., 0.6 cm2) of the insert were used to correct TER measurements. TER is expressed as V cm2. After 48 h of hormone treatment (details in figure legend) 3H-inulin was added to the apical medium (2.5 mCi/ml; specific activity 0.50 Ci/mmol). Then, 24 h later, a 100-ml sample was taken from the basolateral medium, added to 1 ml scintillation fluid (OptiPhase ‘HiSafe’, Science & Technology New Zealand, Auckland, New Zealand) and counted for 2 min (Wallac 1409 Liquidscintillation Counter, Science & Technology New Zealand).
2.4. Western blot analyses for ZO-1 and occludin HC11 cells were grown in 260-ml culture flasks (Nunc, Life Technologies) under the same conditions as above (Section 2.2). Following treatment, cells were harvested in lysis buffer (0.01 M Tris, 1% SDS, 1 minitablet of EDTA-free protease inhibitors), and the protein content of cell lysates was determined using the BCA assay. Sample lysates, normalized for protein content were subjected to 7.5% (ZO-1) or 10% (occludin) SDS-polyacrylamide gel electrophoresis according to the method of Laemmli (1970). Following electrophoresis, separated proteins were transferred (1 h at 0.8 mA/cm2) to nitrocellulose membranes (OptitranBA-S 85, Schleichen and Schuell, Dassel, Germany). Membranes were then probed with either a monoclonal rat anti-mouse ZO-1 (1:50) or a polyclonal rabbit antihuman occludin (1:2000) antibody. The secondary antibodies used with ZO-1 and occludin were sheep anti-rat (1:1000) and donkey anti-rabbit (1:1000), respectively, each conjugated to horseradish peroxidase; signals were detected by DAB (ZO-1) or chemiluminescence (occludin).
Fig. 1. Effect of dose of PRL on TER in HC11 cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h all cells were exposed to DEX (1 mM) and either 0 (open bars), 1 (solid bars), or 5 (narrow-hatched bars) mg/ml of PRL at indicated times. Data are also shown for cells that were treated with 1 mg/ml of PRL only after 48 h (wide-hatched bars). Means represent the average of n = 5 inserts per treatment. (a, b) Means with different letters per time point differ at P B0.05.
5 mg/ml of PRL on TER in HC11 (Fig. 1). Treatment of HC11 cells with PRL resulted in a significant increase in TER of approximately 20% compared to cells receiving only DEX. There was no difference in response between cells exposed to either 1 or 5 mg/ml of PRL, therefore, in all subsequent experiments PRL was used at 1 mg/ml. PRL was also able to enhance TJ formation during advanced TJ formation (i.e. \48 h), because when PRL was added to DEX-treated cells after 48 h of culture, PRL caused TER to increase to
2.5. Statistical analyses Data were analyzed by ANOVA (SAS Systems for Windows, Release 6.11, 1996; SAS, Cary, NC, USA). Differences between means were considered significant at P B 0.05, in which case a Fisher protected least significant differences test was performed to compare means. When appropriate, TER values from the last day that cells were in growth medium, were used as covariate in the analyses. All values are expressed as means 9 SEM.
3. Results
3.1. Effect of PRL and DEX on TER in HC11 cells The first experiment examined the effect of 0, 1 and
Fig. 2. Effect of PRL and DEX on TER in HC11 cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h cells were then exposed to priming medium alone, or medium supplemented with PRL (1 mg/ml), DEX (1 mM), or a combination of PRL and DEX at indicated times. Means represent the average of n =5 inserts per treatment. (a, b) Means with different letters per time point differ at P B0.05.
58
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
the same level as in cells that had received PRL from the beginning. The separate and combined effects of PRL and DEX on TER in HC11 cells is shown in Fig. 2. Addition of either PRL or DEX alone to the medium caused a small, but significant, increase in TER between 12 and 14%, thus indicating a positive effect on mammary TJ formation. The effect was similar for each hormone separately, whereas the combination of PRL and DEX appeared to give an additional increase in TER. The total increase in TER for the combined treatments was approximately 20% and of a magnitude similar to that observed in the first experiment (Fig. 1).
3.2. Effect of PRL and DEX on 3H-inulin mo6ement between HC11 cells In order to confirm that the increase in TER was related to TJ formation, the paracellular movement of the inert radiolabeled size-marker inulin from the apical to the basolateral side of the cells was measured in HC11 cells. Both lactogenic hormones, as well as their combination, reduced the paracellular leakage of the tracer compared to cells exposed to priming medium only (priming medium vs. PRL vs. DEX vs. PRL+ DEX: 24 462a vs. 19 342b vs. 20 891b vs. 21 093b 91032 cpm; abPB0.05, n= 5 inserts per treatment). Data for membranes without cells were 104 76196268 cpm (n= 3). These results are in agreement with the TER data, which indicated that both PRL and DEX stimulate mammary TJ formation.
3.3. Effect of PRL on TER in HC11 cells with or without pretreatment with DEX. To investigate if the effects of PRL on TJ formation were affected by glucocorticoids, we studied the effect of PRL on TER with or without DEX (Fig. 3). This experiment confirmed the results from the earlier experiments that PRL and DEX both increase TER, and that the combined treatment gives the largest increase. Further, when after 24 h the DEX-treated cells received PRL there was a further increase in TER, as was also observed in experiment 1 (Fig. 1). However, in contrast, when PRL-treated cells received DEX after 24 h, TER did not exhibit a further increase.
Fig. 3. Effect of DEX-pretreatment on the effect of PRL on TER in HC11 cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h cells were exposed to medium supplemented with PRL (1 mg/ml), DEX (1 mM), or a combination of PRL and DEX at indicated times. Means represent the average of n = 5 inserts per treatment. (a, b) Means with different letters per time point differ at P B0.05.
similar in the two cell lines. In COMMA-1D cells PRL and DEX each elicited significant increases in TER (Fig. 4), however, in these cells the effect of DEX appeared to be double that of PRL. Furthermore, the effects of PRL and DEX appeared to be additive, with the largest TER values being attained with a combined PRL and DEX treatment.
3.5. Effect of PRL and DEX on occludin protein expression The previous experiments showed that both PRL and DEX increased mammary TJ formation. However, it was not known through which components of the TJ
3.4. Effect of PRL and DEX on TER in COMMA-1D cells The effects of PRL and DEX on TER were not exclusive to HC11 cells, because these hormones were able to increase TER in COMMA-1D cells, the cell line from which HC11 cells were derived. COMMA-1D cells developed much higher TER values than HC11 cells, but the response pattern for these hormones was
Fig. 4. Effect of PRL and DEX on TER in COMMA-1D cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h cells were then exposed to priming medium alone, or medium supplemented with PRL (1 mg/ml), DEX (1 mM), or a combination of PRL and DEX at indicated times. Means represent the average of n =5 inserts per treatment. (a, b) Means with different letters per time point differ at PB 0.05.
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
59
4. Discussion
Fig. 5. A representative Western analyses of the effects of PRL and DEX on the expression of the tight junction protein occludin in HC11 cells. Cells had been treated for 24 h with PRL (1 mg/ml; lane 1), DEX (1 mM; lane 2), a combination of PRL and DEX (lane 3), or no PRL and DEX (priming medium; lane 4).
these hormones acted. We therefore examined the effect of either PRL and/or DEX on the expression of occludin, the major characterized TJ integral membrane protein. A representative Western analysis is shown in Fig. 5. PRL alone induced the expression of occludin (lane 1), albeit that the effect was smaller than that of DEX alone (lane 2). Again the effects of PRL and DEX appeared to be additive, as the highest expression was found when both PRL and DEX were present in the medium (lane 3). In contrast, the absence of both PRL and DEX resulted in very low levels of occludin being expressed (lane 4).
3.6. Effect of PRL on ZO-1 protein expression On the cytoplasmic side occludin is linked to the cytoskeleton via a complex arrangement of TJ-associated proteins, including ZO-1 (Fanning et al., 1998). In order to understand further how PRL affects mammary TJ we examined whether PRL had an effect on the expression of ZO-1. As a result of alternative RNA splicing ZO-1 appears as two isoforms (Balda and Anderson, 1993), an a+-isoform of 223 kDa and an a−-isoform of 214 kDa. Our antibody recognized both isoforms (Fig. 6). Consistent with TER results there appeared to be no advantage in increasing the dose of PRL from 1 to 5 mg/ml. However, compared to cells treated with DEX alone (lanes 5 – 6), the addition of PRL appeared to induce a small increase in the expression of the a−-isoform relative to the a+- isoform (lanes 1–4).
Fig. 6. Western analyses of the effects of PRL and DEX on the expression of tight junction-associated protein ZO-1 in HC11 cells. Cells had been treated for 24 h with DEX (1 mM) and 1 (lanes 1, 2), 5 (lanes 3, 4), or 0 mg/ml of PRL (lanes 5, 6).
Conditions that cause a sudden decrease in the rate of milk removal from the mammary gland or complete cessation of milk removal have been shown to induce an increase in permeability of mammary TJ (Fleet and Peaker, 1978; Stelwagen et al., 1997). Interestingly, these conditions are also associated with decreased PRL concentrations, and when PRL secretion is blocked with bromocriptine, mammary TJ become ‘leaky’ (Flint and Gardner, 1994). The main objective of the present study was therefore to establish whether PRL plays a role in the regulation of mammary TJ formation, and if such involvement is mediated via the true transmembrane TJ protein occludin, and/or via the TJ-associated protein ZO-1. Electrical resistance (TER) measurements across mammary epithelial cells have been successfully used both in vitro (Zettl et al., 1992; Stelwagen and Ormrod, 1998) and in vivo (Peaker, 1977; Stelwagen et al., 1994) to measure mammary TJ, with an increase in resistance indicating TJ formation. Typically, in in vitro experiments there is an increase in TER over time, until it reaches a plateau after several days. Zettl et al. (1992) demonstrated that the synthetic glucocorticoid DEX causes an additional increase in TER in 31EG4 mouse mammary epithelial cells. In the present study we confirmed a similar effect of DEX on TER in mouse mammary epithelial HC11 and COMMA-1D cells. Our data also demonstrated that PRL caused a small, but significant, increase in TER in addition to the effect of DEX. Moreover, 1 and 5 mg/ml doses of PRL gave similar results, suggesting that 1 mg/ml of PRL was sufficient to elicit a TJ-response. Although we did not examine lower doses of PRL, this dose level is identical to that at which PRL has its maximal effect on mammary b-casein expression (Ball et al., 1988). Because both events, TJ formation and subsequent casein synthesis, are characteristics of differentiated mammary cells, combining these data suggest that 1 mg/ml is the optimum dose for PRL to elicit a biological response in differentiated cells. In HC11 cells the effect of PRL alone on TER was similar to that of DEX by itself, but the combination of the two hormones gave the largest increase. Interestingly, although PRL caused a further increase in TER in cells that had been pretreated with DEX, no additional TER response was observed in cells that had been exposed solely to PRL and subsequently also to DEX. Thus, while PRL increases TJ formation, the maximum response can only be obtained when cells are pretreated with DEX or are maintained in the presence of both hormones from the start. These data imply that the hormonal interactions underlying the activity of PRL in the regulation of TJ formation are quite similar to its role in milk protein synthesis, because both
60
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
Doppler et al. (1990), Lee et al. (1998) report that PRL induces b-casein gene expression in mammary cells, but only following pretreatment with DEX. HC11 cells are derived from COMMA-1D mouse mammary epithelial cells (Ball et al., 1988). The difference between the two cell lines is that COMMA-1D cells require the presence of an extracellular matrix in order to become differentiated (e.g. express b-casein), whereas HC11 cells do not require such a matrix to become differentiated. Our results showed that the effects of PRL and DEX on mammary TJ were very similar in the two cell lines. Thus, while both TJ and the extracellular matrix are linked to the cytoskeleton (Fanning et al., 1996, 1998; Ingber, 1997), the effects of PRL and DEX on TJ appear to be mediated independently from the extracellular matrix. Although changes in TER are commonly accepted to indicate changes in TJ status, TER is an instantaneous measurement and Balda et al. (1996) report in one study with kidney epithelial cells that a high TER (indicating tight TJ) corresponded with a decrease in paracellular movement of marker molecules (indicating loose TJ), which is a measure of permeability over time. Therefore, we also measured the paracellular movement of radiolabeled inulin from the apical side to the basolateral side in the presence of PRL and DEX. These results were in full agreement with those of the TER experiments. Thus, taken together, our results clearly indicated that PRL was involved in the regulation of mammary TJ formation. Such a role is in keeping with systemic changes in prolactin concentrations and concurrent changes in TJ permeability that are observed in vivo during lactogenesis, i.e. when TJ form in the mammary gland. However, they did not reveal which components of the TJ were affected by PRL and DEX. Therefore, we examined the effects of PRL and DEX on the synthesis of the structural components of the mammary TJ. Both hormones independently induced the expression of the true transmembrane TJ protein occludin, albeit that the effect of DEX was greater than that of PRL, and the combined hormone treatment gave the highest expression. These results are the first to show that occludin is expressed in mammary cells, and that both PRL and DEX act to increase the synthesis of occludin. In an earlier study, DEX was shown to increase the level of TJ-associated protein ZO-1 (Singer et al., 1994), but had no effect on cytoplasmic distribution of ZO-1 as examined by immunofluorescence microscopy (Zettl et al., 1992). Unlike occludin, ZO-1 is not an integral TJ protein, but is intimately associated with TJ by linking occludin to the cytoskeleton (Fanning et al., 1998). Compared to cells treated with DEX only, in those receiving DEX and PRL, PRL appeared to induce a small increase in the expression of the a−-isoform relative to the a+- isoform. This is an interesting
observation because expression of the a−- isoform has previously been associated with cells that have structurally dynamic TJ, such as in podocytes and Sertoli cells (Balda and Anderson, 1993). In such cells TJ can rapidly and actively open and reseal. We have recently shown in vivo that mammary TJ can open and seal quickly (i.e. 53 h) in response to changing physiological stimuli (Stelwagen et al., 1997). Combining these data therefore seem to suggest that mammary TJ can also be classified as structurally dynamic junctions. Thus, PRL appears to act on TJ formation by increasing the expression of true TJ components as well as TJ-associated proteins at the cytoplasmic interface. In conclusion, PRL stimulates mammary TJ formation by enhancing expression of the transmembrane protein occludin and the a−-isoform of the TJ-associated protein ZO-1. In addition we confirmed the TJstimulating effect of glucocorticoids and showed that these effects also appear to be due to an increase in occludin expression.
Acknowledgements Financial support for the research presented was provided by the Foundation for Research, Science and Technology of New Zealand, Wellington, New Zealand.
References Balda, M.S., Anderson, J.M., 1993. Two classes of tight junctions are revealed by ZO-1 isoforms. Am. J. Physiol. 264, C918 –C924. Balda, M.S., Whitney, J.A., Flores, C., Gonza´lez, S., Cereijido, M., Matter, K., 1996. Functional dissociation of paracellular permeability and transepithelial electrical resistance and disruption of the apical-basolateral itramembrane diffusion barrier by expression of a mutant tight junction membrane protein. J. Cell Biol. 134, 1031 – 1049. Ball, R.K., Friis, R.R., Schoenenberger, C.A., Doppler, W., Groner, B., 1988. Prolactin regulation of b-casein gene expression and of a cytosolic 120-kd protein in a cloned mouse mammary epithelial cell line. EMBO J. 7, 2089 – 2095. Convey, E.M., 1974. Serum hormone concentrations in ruminants during mammary growth, lactogenesis, and lactation: a review. J. Dairy Sci. 57, 905 – 917. Danielson, K.G., Oborn, C.J., Durban, E.M., Butel, J.S., Medina, D., 1984. Epithelial mouse mammary cell line exhibiting normal morphogenesis in vivo and functional differentiation in vitro. Proc. Natl. Acad. Sci. USA 81, 3756 – 3760. Doppler, W., Ho¨ck, W., Hofer, P., Groner, B., Ball, R.K., 1990. Prolactin and glucocorticoid hormones control transcription of the ß-casein gene by kinetically distinct mechanisms. Mol. Endocrinol. 4, 912 – 919. Fanning, A.S., Jameson, B.J., Jesaitis, L.A., Anderson, J.M., 1998. The tight junction protein ZO-1 establishes a link between the transmembrane protein occludin and the actin cytoskeleton. J. Biol. Chem. 273, 29745 – 29753.
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61 Fanning, A.S., Lapierre, L.A., Brecher, A.R., Van Itallie, C.M., Anderson, J.M., 1996. Protein interactions in the tight junction: the role of MAGUK proteins in regulating tight junction organization and function. Curr. Topics Membr. 43, 211–235. Fleet, I.R., Peaker, M., 1978. Mammary function and its control at the cessation of lactation in the goat. J. Physiol. 279, 491 – 507. Flint, D.J., Gardner, M., 1994. Evidence that growth hormone stimulates milk synthesis by direct action on the mammary gland and that prolactin exerts effects on milk secretion by maintenance of mammary deoxyribonucleic acid content and tight junction status. Endocrinology 135, 1119–1124. Furuse, M., Fujita, K., Hiiragi, T., Fujimoto, K., Tsukita, S., 1998. Claudin-1 and -2: novel integral membrane proteins localizing at tight junctions with no sequence similarity to occludin. J. Cell Biol. 141, 1539 – 1550. Furuse, M., Hirase, T., Itoh, M., Nagafuchi, A., Yonemura, S., Tsukita, S., Tsukita, S., 1993. Occludin: a novel integral membrane protein localizing at tight junctions. J. Cell Biol. 123, 1777 – 1788. Gopalakrishnan, S., Raman, N., Atkinson, S.J., Marrs, J.A., 1998. Rho GTPase signaling regulates tight junction assembly and protects tight junctions during ATP depletion. Am. J. Physiol. 275, C798 – C809. Ingber, D.E., 1997. Tensegrity: the architectural basis of cellular mechanotransduction. Annu. Rev. Physiol. 59, 575–599. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature 227, 680– 685. Lee, C.S., Kim, K., Yo, D.Y., Lee, K.K., 1998. Pretreatment with glucocorticoid is essential for lactogenic induction of the bovine ß-casein/CAT expression in HC11 cells. Endocr. Res. 24, 65 – 77. Linzell, J.L., Peaker, M., 1974. Changes in colostrum composition and in the permeability of the mammary epithelium at about the time of parturition in the goat. J. Physiol. 243, 129–151. Linzell, J.L., Peaker, M., Tayler, J.C., 1975. The effects of prolactin and oxytocin on milk secretion and on the permeability of the mammary epithelium in the rabbit. J. Physiol. 253, 547–563. Nguyen, D.A.D., Neville, M.C., 1998. Tight junction regulation in the mammary gland. J. Mamm. Gl. Biol. Neopl. 3, 233–246. Nusrat, A., Giry, M., Turner, J.R., Colgan, S.P., Parkos, C.A., Carnes, D., Lemichez, E., Boquet, P., Madara, J.L., 1995. Rho
.
61
protein regulates tight junctions and perijunctional actin organization in polarized epithelia. Proc. Natl. Acad. Sci. USA 92, 10629– 10633. Peaker, M., 1977. Mechanism of milk secretion: milk composition in relation to potential difference across the mammary epithelium. J. Physiol. 270, 489 – 505. Schneeberger, E.E., Lynch, R.D., 1992. Structure, function, and regulation of cellular tight junctions. Am. J. Physiol. 262, L647– L661. Singer, K.L., Stevenson, B.R., Woo, P.L., Firestone, G.L., 1994. Relationship of serine/threonine phosphorylation/dephosphorylation signaling to glucocorticoid regulation of tight junction permeability and ZO-1 distribution in nontransformed mammary epithelial cells. J. Biol. Chem. 269, 16108 – 16115. Stelwagen, K., Davis, S.R., Farr, V.C., Prosser, C.G., Sherlock, R.A., 1994. Epithelial cell tight junction integrity and mammary blood flow during an extended milking interval in goatsl. J. Dairy Sci. 77, 426 – 432. Stelwagen, K., Farr, V.C., Davis, S.R., Prosser, C.G., 1995. EGTAinduced disruption of epithelial cell tight junctions in the lactating caprine mammary gland. Am. J. Physiol. 269, R848 – R855. Stelwagen, K., Farr, V.C., McFadden, H.A., Prosser, C.G., Davis, S.R., 1997. Time course of milk accumulation-induced opening of mammary tight junctions and blood clearance of milk components. Am. J. Physiol. 273, R379 – R386. Stelwagen, K., Ormrod, D.J., 1998. An anti-inflammatory component derived from milk of hyper-immunised cows reduces tight junction permeability in vitro. Inflam. Res. 47, 384 – 388. Stelwagen, K., Van Espen, D.C., Verkerk, G.A., McFadden, H., Farr, V.C., 1998. Elevated plasma cortisol reduces permeability of mammary tight junctions in the lactating bovine mammary epithelium. J. Endocrinol. 159, 173 – 178. Thompson, G.E., 1996. Cortisol and regulation of tight junctions in the mammary gland of the late-pregnant goat. J. Dairy Res. 63, 305 – 308. Zettl, K.S., Sjaastad, M.D., Riskin, P.M., Parry, G., Machen, T.E., Firestone, G.L., 1992. Glucocorticoid-induced formation of tight junctions in mouse mammary epithelial cells in vitro. Proc. Natl. Acad. Sci. USA 89, 9069 – 9073.
Prolactin, alone or in combination with glucocorticoids, enhances tight junction formation and expression of the tight junction protein occludin in mammary cells Kerst Stelwagen *, Holli A. McFadden, Jerome Demmer Dairy Science, Ruakura Research Centre, AgResearch, Pri6ate Bag 3123, Hamilton, New Zealand Received 19 April 1999; accepted 28 June 1999
Abstract Tight junctions (TJ) between adjacent epithelial cells play an important role in maintaining mammary function in the differentiated mammary gland. Mouse mammary cell lines (HC11 and Comma-1D) were used to investigate the effect of the lactogenic hormones prolactin (PRL) and glucocorticoids on the formation of mammary TJ. TJ formation was assessed by an increase in transepithelial electrical resistance and a decrease in paracellular flux of radiolabeled inulin. Both PRL and the synthetic glucocorticoid dexamethasone (DEX) stimulated TJ formation. The biggest effect on TJ formation was observed when both hormones were used in combination, but only when cells were pretreated with DEX. The effects of PRL and DEX are mediated, at least in part, via expression of the transmembrane TJ protein occludin. In summary, these data are the first to show an effect of PRL on mammary TJ formation and the expression of TJ proteins, and confirm the TJ-stimulating effects of glucocorticoids that have been reported previously. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Prolactin; Tight junction; Occludin; Mammary gland; HC11 cells
1. Introduction The tight junction (TJ) is the most apical member of the junctional complex between epithelial or endothelial cells. The TJ acts as a barrier between adjacent cells, and as such regulates the paracellular transport of ions and small molecules (‘barrier function’), but it also separates the plasma membrane of the cell into basolateral and apical domains of distinct composition and function (‘fence function’) (Schneeberger and Lynch, 1992). TJ were long considered static structures, and of unknown composition, but recently much insight has been gained into the molecular make-up of the TJ. The first true transmembrane TJ protein identified is occludin (Furuse et al., 1993), and recently two other Abbre6iations: DEX, dexamethasone; PRL, prolactin; TER, transepithelial electrical resistance; TJ, tight junction(s). * Corresponding author. Present address: Research Station for Cattle, Sheep and Horse Husbandry (PR), PO Box 2176, 8203 AD Lelystad, The Netherlands. Tel.: + 31-320-293211; fax: + 31-320241584. E-mail address: [email protected] (K. Stelwagen)
proteins have been isolated from junction fractions, claudin-1 and -2 (Furuse et al., 1998). Little is known about the possible role of claudins, however occludin is intimately connected to the cytoskeleton of the cell via TJ-associated proteins such as ZO-1 (Fanning et al., 1998). The linkage of TJ to the cytoskeleton suggests that TJ may play an important role in cell functioning. Indeed, there is a link between TJ and the Rho GTPase-mediated signaling pathway in the cell (Nusrat et al., 1995; Gopalakrishnan et al., 1998). Consistent with a role of TJ in cell functioning, epithelial TJ in the mammary gland play an important role in maintaining milk synthesis, as demonstrated by the fact that induced loss of barrier function in the lactating mammary gland results in a significant decrease in milk synthesis (Stelwagen et al., 1995, 1997). However, the exact mechanism by which the loss of TJ barrier function affects milk synthesis is not fully understood. In the mammary gland TJ form during lactogenesis when the gland switches from a growth state to a differentiated state, i.e. when milk synthesis and secretion are initiated (Linzell and Peaker, 1974). This
0303-7207/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 3 - 7 2 0 7 ( 9 9 ) 0 0 1 4 5 - 8
56
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
switch coincides with major changes in systemic endocrine profiles at that time, such as rapidly decreasing levels of progesterone during the immediate pre-partum period combined with a sharp increase in the levels of prolactin (PRL) and glucocorticoids (Convey, 1974). Indeed, recent studies have shown that mammary TJ are under endocrine regulation. Progesterone, involved in maintenance of gestation, and thus preventing mammary cells undergoing differentiation, inhibits TJ formation (Nguyen and Neville, 1998), whereas the rapid decrease in progesterone during lactogenesis allows for TJ formation. Moreover, glucocorticoids have been shown to have a direct effect on the formation and maintenance of mammary TJ in vitro (Zettl et al., 1992; Singer et al., 1994) and in vivo (Thompson, 1996; Stelwagen et al., 1998). The effects of PRL on mammary TJ are much less defined. PRL in combination with insulin or with insulin and the synthetic glucocorticoid dexamethasone (DEX) appear to have no effect on TJ in 31EG4 mouse-derived mammary cells (Zettl et al., 1992). In contrast, in in vivo studies, where animals received exogenous PRL (Linzell et al., 1975) or where endogenous PRL release was blocked by bromocriptine (Flint and Gardner, 1994), corresponding changes in the levels of sodium, potassium and lactose in milk were observed that suggest that PRL may act upon mammary TJ. However, it can not be ruled out that these in vivo effects of PRL may be on cell maintenance or survival and thereby preventing alveolar cell loss, which would result in similar changes in milk composition. Therefore, the objective of the present study was to examine whether PRL, alone or in combination with glucocorticoids, could enhance the formation of mammary TJ, and if so, was this effect mediated via the TJ protein occludin and/or TJ-associated proteins
2. Materials and methods
2.1. Materials HC11 (Ball et al., 1988) and COMMA-1D (Danielson et al., 1984) mouse mammary cells were gifts from Dr B. Groner (Chemotherapeutisches Forschungsinstitut, Frankfurt am Main, Germany) and Dr D. Medina (Baylor College of Medicine, Houston, TX, USA), respectively. Media for HC11 (RPMI 1640) and COMMA-1D (DMEM-F12) were from Gibco BRL (Life Technologies, Auckland, New Zealand) and Sigma (St. Louis, MO, USA), respectively. Fetal calf serum came from Life Technologies. Ovine PRL, DEX, murine submaxillary gland epidermal growth factor, bovine pancreatic insulin, penicillin – streptomycin, bovine serum albumin were all purchased from Sigma. 3 H-inulin was obtained from Amersham New Zealand
(Auckland, New Zealand). Cell culture inserts were from Biolab Scientific, Auckland New Zealand (HC11: Millipore PCF, 12 mm, 0.4 mm pore size) and Life Technologies (COMMA-1D: Nunc Anopore, 10 mm, 0.02 mm pore size). The ZO-1 monoclonal antibody, developed by D.A. Goodenough, was purchased from the Developmental Studies Hybridoma Bank (University of Iowa, Iowa City, USA) and the occludin antibody came from Zymed Laboratories Inc. (BioScientific, Gymes 2227, N.S.W., Australia). Secondary antibodies were from Amersham Life Science (Auckland, New Zealand). The BCA protein assay kit was from Pierce, Rockford, IL, USA. Protease inhibitor tablets were purchased from Boehringer Mannheim, Auckland, New Zealand. All other reagents were of the highest available purity.
2.2. Cell culture HC11 cells were cultured as previously described (Stelwagen and Ormrod, 1998). Briefly, cells were seeded on permeable Millipore PCF inserts at a density of 4× 105 cells per insert. Cells were grown in growth medium (RPMI 1640), supplemented with 10% fetal calf serum, penicillin (5 U/ml) and streptomycin (5 mg/ml), insulin (5 mg/ml), and epidermal growth factor (10 ng/ml). Once cells became confluent they were maintained for 24 h in priming medium (same as growth medium, but with no epidermal growth factor). Cells were then kept in differentiation medium, i.e. priming medium supplemented with PRL (1 or 5 mg/ml) and/or DEX (1 mM). COMMA-1D cells were seeded on Nunc Anopore membranes at a density of 1.8×105 cells per insert. Cells were grown to confluence in growth medium (DMEM-F12), supplemented with 2% fetal calf serum, serum albumin (300 ng/ml), penicillinstreptomycin (as for HC11), insulin (5 mg/ml), and epidermal growth factor (5 ng/ml). Subsequently, cells were kept in priming (24 h) and differentiation media. Both media were based on the growth medium, and hormone omission and/or supplementation were as described for HC11 cells. Media was changed every 24 h, and cells were culture at 37°C in an atmosphere of 5% CO2 and 95% air.
2.3. TJ formation Formation of TJ was assessed by transepithelial electrical resistance (TER) and paracellular movement of the 3H-inulin (5200 MW) inert marker across cell monolayers. An increase in TER represents increased TJ formation. TER was measured across cell monolayers every 24 h using a Millicell-ERS voltohmmeter (Millipore, Bedford, MA, USA) after transferring an insert to an Endohm-12 chamber (World Precision Instruments, Sarasota, FL, USA). TER values of insert
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
57
membrane and medium (i.e. no cells, medium only), and membrane surface area (i.e., 0.6 cm2) of the insert were used to correct TER measurements. TER is expressed as V cm2. After 48 h of hormone treatment (details in figure legend) 3H-inulin was added to the apical medium (2.5 mCi/ml; specific activity 0.50 Ci/mmol). Then, 24 h later, a 100-ml sample was taken from the basolateral medium, added to 1 ml scintillation fluid (OptiPhase ‘HiSafe’, Science & Technology New Zealand, Auckland, New Zealand) and counted for 2 min (Wallac 1409 Liquidscintillation Counter, Science & Technology New Zealand).
2.4. Western blot analyses for ZO-1 and occludin HC11 cells were grown in 260-ml culture flasks (Nunc, Life Technologies) under the same conditions as above (Section 2.2). Following treatment, cells were harvested in lysis buffer (0.01 M Tris, 1% SDS, 1 minitablet of EDTA-free protease inhibitors), and the protein content of cell lysates was determined using the BCA assay. Sample lysates, normalized for protein content were subjected to 7.5% (ZO-1) or 10% (occludin) SDS-polyacrylamide gel electrophoresis according to the method of Laemmli (1970). Following electrophoresis, separated proteins were transferred (1 h at 0.8 mA/cm2) to nitrocellulose membranes (OptitranBA-S 85, Schleichen and Schuell, Dassel, Germany). Membranes were then probed with either a monoclonal rat anti-mouse ZO-1 (1:50) or a polyclonal rabbit antihuman occludin (1:2000) antibody. The secondary antibodies used with ZO-1 and occludin were sheep anti-rat (1:1000) and donkey anti-rabbit (1:1000), respectively, each conjugated to horseradish peroxidase; signals were detected by DAB (ZO-1) or chemiluminescence (occludin).
Fig. 1. Effect of dose of PRL on TER in HC11 cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h all cells were exposed to DEX (1 mM) and either 0 (open bars), 1 (solid bars), or 5 (narrow-hatched bars) mg/ml of PRL at indicated times. Data are also shown for cells that were treated with 1 mg/ml of PRL only after 48 h (wide-hatched bars). Means represent the average of n = 5 inserts per treatment. (a, b) Means with different letters per time point differ at P B0.05.
5 mg/ml of PRL on TER in HC11 (Fig. 1). Treatment of HC11 cells with PRL resulted in a significant increase in TER of approximately 20% compared to cells receiving only DEX. There was no difference in response between cells exposed to either 1 or 5 mg/ml of PRL, therefore, in all subsequent experiments PRL was used at 1 mg/ml. PRL was also able to enhance TJ formation during advanced TJ formation (i.e. \48 h), because when PRL was added to DEX-treated cells after 48 h of culture, PRL caused TER to increase to
2.5. Statistical analyses Data were analyzed by ANOVA (SAS Systems for Windows, Release 6.11, 1996; SAS, Cary, NC, USA). Differences between means were considered significant at P B 0.05, in which case a Fisher protected least significant differences test was performed to compare means. When appropriate, TER values from the last day that cells were in growth medium, were used as covariate in the analyses. All values are expressed as means 9 SEM.
3. Results
3.1. Effect of PRL and DEX on TER in HC11 cells The first experiment examined the effect of 0, 1 and
Fig. 2. Effect of PRL and DEX on TER in HC11 cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h cells were then exposed to priming medium alone, or medium supplemented with PRL (1 mg/ml), DEX (1 mM), or a combination of PRL and DEX at indicated times. Means represent the average of n =5 inserts per treatment. (a, b) Means with different letters per time point differ at P B0.05.
58
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
the same level as in cells that had received PRL from the beginning. The separate and combined effects of PRL and DEX on TER in HC11 cells is shown in Fig. 2. Addition of either PRL or DEX alone to the medium caused a small, but significant, increase in TER between 12 and 14%, thus indicating a positive effect on mammary TJ formation. The effect was similar for each hormone separately, whereas the combination of PRL and DEX appeared to give an additional increase in TER. The total increase in TER for the combined treatments was approximately 20% and of a magnitude similar to that observed in the first experiment (Fig. 1).
3.2. Effect of PRL and DEX on 3H-inulin mo6ement between HC11 cells In order to confirm that the increase in TER was related to TJ formation, the paracellular movement of the inert radiolabeled size-marker inulin from the apical to the basolateral side of the cells was measured in HC11 cells. Both lactogenic hormones, as well as their combination, reduced the paracellular leakage of the tracer compared to cells exposed to priming medium only (priming medium vs. PRL vs. DEX vs. PRL+ DEX: 24 462a vs. 19 342b vs. 20 891b vs. 21 093b 91032 cpm; abPB0.05, n= 5 inserts per treatment). Data for membranes without cells were 104 76196268 cpm (n= 3). These results are in agreement with the TER data, which indicated that both PRL and DEX stimulate mammary TJ formation.
3.3. Effect of PRL on TER in HC11 cells with or without pretreatment with DEX. To investigate if the effects of PRL on TJ formation were affected by glucocorticoids, we studied the effect of PRL on TER with or without DEX (Fig. 3). This experiment confirmed the results from the earlier experiments that PRL and DEX both increase TER, and that the combined treatment gives the largest increase. Further, when after 24 h the DEX-treated cells received PRL there was a further increase in TER, as was also observed in experiment 1 (Fig. 1). However, in contrast, when PRL-treated cells received DEX after 24 h, TER did not exhibit a further increase.
Fig. 3. Effect of DEX-pretreatment on the effect of PRL on TER in HC11 cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h cells were exposed to medium supplemented with PRL (1 mg/ml), DEX (1 mM), or a combination of PRL and DEX at indicated times. Means represent the average of n = 5 inserts per treatment. (a, b) Means with different letters per time point differ at P B0.05.
similar in the two cell lines. In COMMA-1D cells PRL and DEX each elicited significant increases in TER (Fig. 4), however, in these cells the effect of DEX appeared to be double that of PRL. Furthermore, the effects of PRL and DEX appeared to be additive, with the largest TER values being attained with a combined PRL and DEX treatment.
3.5. Effect of PRL and DEX on occludin protein expression The previous experiments showed that both PRL and DEX increased mammary TJ formation. However, it was not known through which components of the TJ
3.4. Effect of PRL and DEX on TER in COMMA-1D cells The effects of PRL and DEX on TER were not exclusive to HC11 cells, because these hormones were able to increase TER in COMMA-1D cells, the cell line from which HC11 cells were derived. COMMA-1D cells developed much higher TER values than HC11 cells, but the response pattern for these hormones was
Fig. 4. Effect of PRL and DEX on TER in COMMA-1D cells. Following a 24-h exposure to priming medium (no PRL or DEX) at 0 h cells were then exposed to priming medium alone, or medium supplemented with PRL (1 mg/ml), DEX (1 mM), or a combination of PRL and DEX at indicated times. Means represent the average of n =5 inserts per treatment. (a, b) Means with different letters per time point differ at PB 0.05.
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
59
4. Discussion
Fig. 5. A representative Western analyses of the effects of PRL and DEX on the expression of the tight junction protein occludin in HC11 cells. Cells had been treated for 24 h with PRL (1 mg/ml; lane 1), DEX (1 mM; lane 2), a combination of PRL and DEX (lane 3), or no PRL and DEX (priming medium; lane 4).
these hormones acted. We therefore examined the effect of either PRL and/or DEX on the expression of occludin, the major characterized TJ integral membrane protein. A representative Western analysis is shown in Fig. 5. PRL alone induced the expression of occludin (lane 1), albeit that the effect was smaller than that of DEX alone (lane 2). Again the effects of PRL and DEX appeared to be additive, as the highest expression was found when both PRL and DEX were present in the medium (lane 3). In contrast, the absence of both PRL and DEX resulted in very low levels of occludin being expressed (lane 4).
3.6. Effect of PRL on ZO-1 protein expression On the cytoplasmic side occludin is linked to the cytoskeleton via a complex arrangement of TJ-associated proteins, including ZO-1 (Fanning et al., 1998). In order to understand further how PRL affects mammary TJ we examined whether PRL had an effect on the expression of ZO-1. As a result of alternative RNA splicing ZO-1 appears as two isoforms (Balda and Anderson, 1993), an a+-isoform of 223 kDa and an a−-isoform of 214 kDa. Our antibody recognized both isoforms (Fig. 6). Consistent with TER results there appeared to be no advantage in increasing the dose of PRL from 1 to 5 mg/ml. However, compared to cells treated with DEX alone (lanes 5 – 6), the addition of PRL appeared to induce a small increase in the expression of the a−-isoform relative to the a+- isoform (lanes 1–4).
Fig. 6. Western analyses of the effects of PRL and DEX on the expression of tight junction-associated protein ZO-1 in HC11 cells. Cells had been treated for 24 h with DEX (1 mM) and 1 (lanes 1, 2), 5 (lanes 3, 4), or 0 mg/ml of PRL (lanes 5, 6).
Conditions that cause a sudden decrease in the rate of milk removal from the mammary gland or complete cessation of milk removal have been shown to induce an increase in permeability of mammary TJ (Fleet and Peaker, 1978; Stelwagen et al., 1997). Interestingly, these conditions are also associated with decreased PRL concentrations, and when PRL secretion is blocked with bromocriptine, mammary TJ become ‘leaky’ (Flint and Gardner, 1994). The main objective of the present study was therefore to establish whether PRL plays a role in the regulation of mammary TJ formation, and if such involvement is mediated via the true transmembrane TJ protein occludin, and/or via the TJ-associated protein ZO-1. Electrical resistance (TER) measurements across mammary epithelial cells have been successfully used both in vitro (Zettl et al., 1992; Stelwagen and Ormrod, 1998) and in vivo (Peaker, 1977; Stelwagen et al., 1994) to measure mammary TJ, with an increase in resistance indicating TJ formation. Typically, in in vitro experiments there is an increase in TER over time, until it reaches a plateau after several days. Zettl et al. (1992) demonstrated that the synthetic glucocorticoid DEX causes an additional increase in TER in 31EG4 mouse mammary epithelial cells. In the present study we confirmed a similar effect of DEX on TER in mouse mammary epithelial HC11 and COMMA-1D cells. Our data also demonstrated that PRL caused a small, but significant, increase in TER in addition to the effect of DEX. Moreover, 1 and 5 mg/ml doses of PRL gave similar results, suggesting that 1 mg/ml of PRL was sufficient to elicit a TJ-response. Although we did not examine lower doses of PRL, this dose level is identical to that at which PRL has its maximal effect on mammary b-casein expression (Ball et al., 1988). Because both events, TJ formation and subsequent casein synthesis, are characteristics of differentiated mammary cells, combining these data suggest that 1 mg/ml is the optimum dose for PRL to elicit a biological response in differentiated cells. In HC11 cells the effect of PRL alone on TER was similar to that of DEX by itself, but the combination of the two hormones gave the largest increase. Interestingly, although PRL caused a further increase in TER in cells that had been pretreated with DEX, no additional TER response was observed in cells that had been exposed solely to PRL and subsequently also to DEX. Thus, while PRL increases TJ formation, the maximum response can only be obtained when cells are pretreated with DEX or are maintained in the presence of both hormones from the start. These data imply that the hormonal interactions underlying the activity of PRL in the regulation of TJ formation are quite similar to its role in milk protein synthesis, because both
60
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61
Doppler et al. (1990), Lee et al. (1998) report that PRL induces b-casein gene expression in mammary cells, but only following pretreatment with DEX. HC11 cells are derived from COMMA-1D mouse mammary epithelial cells (Ball et al., 1988). The difference between the two cell lines is that COMMA-1D cells require the presence of an extracellular matrix in order to become differentiated (e.g. express b-casein), whereas HC11 cells do not require such a matrix to become differentiated. Our results showed that the effects of PRL and DEX on mammary TJ were very similar in the two cell lines. Thus, while both TJ and the extracellular matrix are linked to the cytoskeleton (Fanning et al., 1996, 1998; Ingber, 1997), the effects of PRL and DEX on TJ appear to be mediated independently from the extracellular matrix. Although changes in TER are commonly accepted to indicate changes in TJ status, TER is an instantaneous measurement and Balda et al. (1996) report in one study with kidney epithelial cells that a high TER (indicating tight TJ) corresponded with a decrease in paracellular movement of marker molecules (indicating loose TJ), which is a measure of permeability over time. Therefore, we also measured the paracellular movement of radiolabeled inulin from the apical side to the basolateral side in the presence of PRL and DEX. These results were in full agreement with those of the TER experiments. Thus, taken together, our results clearly indicated that PRL was involved in the regulation of mammary TJ formation. Such a role is in keeping with systemic changes in prolactin concentrations and concurrent changes in TJ permeability that are observed in vivo during lactogenesis, i.e. when TJ form in the mammary gland. However, they did not reveal which components of the TJ were affected by PRL and DEX. Therefore, we examined the effects of PRL and DEX on the synthesis of the structural components of the mammary TJ. Both hormones independently induced the expression of the true transmembrane TJ protein occludin, albeit that the effect of DEX was greater than that of PRL, and the combined hormone treatment gave the highest expression. These results are the first to show that occludin is expressed in mammary cells, and that both PRL and DEX act to increase the synthesis of occludin. In an earlier study, DEX was shown to increase the level of TJ-associated protein ZO-1 (Singer et al., 1994), but had no effect on cytoplasmic distribution of ZO-1 as examined by immunofluorescence microscopy (Zettl et al., 1992). Unlike occludin, ZO-1 is not an integral TJ protein, but is intimately associated with TJ by linking occludin to the cytoskeleton (Fanning et al., 1998). Compared to cells treated with DEX only, in those receiving DEX and PRL, PRL appeared to induce a small increase in the expression of the a−-isoform relative to the a+- isoform. This is an interesting
observation because expression of the a−- isoform has previously been associated with cells that have structurally dynamic TJ, such as in podocytes and Sertoli cells (Balda and Anderson, 1993). In such cells TJ can rapidly and actively open and reseal. We have recently shown in vivo that mammary TJ can open and seal quickly (i.e. 53 h) in response to changing physiological stimuli (Stelwagen et al., 1997). Combining these data therefore seem to suggest that mammary TJ can also be classified as structurally dynamic junctions. Thus, PRL appears to act on TJ formation by increasing the expression of true TJ components as well as TJ-associated proteins at the cytoplasmic interface. In conclusion, PRL stimulates mammary TJ formation by enhancing expression of the transmembrane protein occludin and the a−-isoform of the TJ-associated protein ZO-1. In addition we confirmed the TJstimulating effect of glucocorticoids and showed that these effects also appear to be due to an increase in occludin expression.
Acknowledgements Financial support for the research presented was provided by the Foundation for Research, Science and Technology of New Zealand, Wellington, New Zealand.
References Balda, M.S., Anderson, J.M., 1993. Two classes of tight junctions are revealed by ZO-1 isoforms. Am. J. Physiol. 264, C918 –C924. Balda, M.S., Whitney, J.A., Flores, C., Gonza´lez, S., Cereijido, M., Matter, K., 1996. Functional dissociation of paracellular permeability and transepithelial electrical resistance and disruption of the apical-basolateral itramembrane diffusion barrier by expression of a mutant tight junction membrane protein. J. Cell Biol. 134, 1031 – 1049. Ball, R.K., Friis, R.R., Schoenenberger, C.A., Doppler, W., Groner, B., 1988. Prolactin regulation of b-casein gene expression and of a cytosolic 120-kd protein in a cloned mouse mammary epithelial cell line. EMBO J. 7, 2089 – 2095. Convey, E.M., 1974. Serum hormone concentrations in ruminants during mammary growth, lactogenesis, and lactation: a review. J. Dairy Sci. 57, 905 – 917. Danielson, K.G., Oborn, C.J., Durban, E.M., Butel, J.S., Medina, D., 1984. Epithelial mouse mammary cell line exhibiting normal morphogenesis in vivo and functional differentiation in vitro. Proc. Natl. Acad. Sci. USA 81, 3756 – 3760. Doppler, W., Ho¨ck, W., Hofer, P., Groner, B., Ball, R.K., 1990. Prolactin and glucocorticoid hormones control transcription of the ß-casein gene by kinetically distinct mechanisms. Mol. Endocrinol. 4, 912 – 919. Fanning, A.S., Jameson, B.J., Jesaitis, L.A., Anderson, J.M., 1998. The tight junction protein ZO-1 establishes a link between the transmembrane protein occludin and the actin cytoskeleton. J. Biol. Chem. 273, 29745 – 29753.
K. Stelwagen et al. / Molecular and Cellular Endocrinology 156 (1999) 55–61 Fanning, A.S., Lapierre, L.A., Brecher, A.R., Van Itallie, C.M., Anderson, J.M., 1996. Protein interactions in the tight junction: the role of MAGUK proteins in regulating tight junction organization and function. Curr. Topics Membr. 43, 211–235. Fleet, I.R., Peaker, M., 1978. Mammary function and its control at the cessation of lactation in the goat. J. Physiol. 279, 491 – 507. Flint, D.J., Gardner, M., 1994. Evidence that growth hormone stimulates milk synthesis by direct action on the mammary gland and that prolactin exerts effects on milk secretion by maintenance of mammary deoxyribonucleic acid content and tight junction status. Endocrinology 135, 1119–1124. Furuse, M., Fujita, K., Hiiragi, T., Fujimoto, K., Tsukita, S., 1998. Claudin-1 and -2: novel integral membrane proteins localizing at tight junctions with no sequence similarity to occludin. J. Cell Biol. 141, 1539 – 1550. Furuse, M., Hirase, T., Itoh, M., Nagafuchi, A., Yonemura, S., Tsukita, S., Tsukita, S., 1993. Occludin: a novel integral membrane protein localizing at tight junctions. J. Cell Biol. 123, 1777 – 1788. Gopalakrishnan, S., Raman, N., Atkinson, S.J., Marrs, J.A., 1998. Rho GTPase signaling regulates tight junction assembly and protects tight junctions during ATP depletion. Am. J. Physiol. 275, C798 – C809. Ingber, D.E., 1997. Tensegrity: the architectural basis of cellular mechanotransduction. Annu. Rev. Physiol. 59, 575–599. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature 227, 680– 685. Lee, C.S., Kim, K., Yo, D.Y., Lee, K.K., 1998. Pretreatment with glucocorticoid is essential for lactogenic induction of the bovine ß-casein/CAT expression in HC11 cells. Endocr. Res. 24, 65 – 77. Linzell, J.L., Peaker, M., 1974. Changes in colostrum composition and in the permeability of the mammary epithelium at about the time of parturition in the goat. J. Physiol. 243, 129–151. Linzell, J.L., Peaker, M., Tayler, J.C., 1975. The effects of prolactin and oxytocin on milk secretion and on the permeability of the mammary epithelium in the rabbit. J. Physiol. 253, 547–563. Nguyen, D.A.D., Neville, M.C., 1998. Tight junction regulation in the mammary gland. J. Mamm. Gl. Biol. Neopl. 3, 233–246. Nusrat, A., Giry, M., Turner, J.R., Colgan, S.P., Parkos, C.A., Carnes, D., Lemichez, E., Boquet, P., Madara, J.L., 1995. Rho
.
61
protein regulates tight junctions and perijunctional actin organization in polarized epithelia. Proc. Natl. Acad. Sci. USA 92, 10629– 10633. Peaker, M., 1977. Mechanism of milk secretion: milk composition in relation to potential difference across the mammary epithelium. J. Physiol. 270, 489 – 505. Schneeberger, E.E., Lynch, R.D., 1992. Structure, function, and regulation of cellular tight junctions. Am. J. Physiol. 262, L647– L661. Singer, K.L., Stevenson, B.R., Woo, P.L., Firestone, G.L., 1994. Relationship of serine/threonine phosphorylation/dephosphorylation signaling to glucocorticoid regulation of tight junction permeability and ZO-1 distribution in nontransformed mammary epithelial cells. J. Biol. Chem. 269, 16108 – 16115. Stelwagen, K., Davis, S.R., Farr, V.C., Prosser, C.G., Sherlock, R.A., 1994. Epithelial cell tight junction integrity and mammary blood flow during an extended milking interval in goatsl. J. Dairy Sci. 77, 426 – 432. Stelwagen, K., Farr, V.C., Davis, S.R., Prosser, C.G., 1995. EGTAinduced disruption of epithelial cell tight junctions in the lactating caprine mammary gland. Am. J. Physiol. 269, R848 – R855. Stelwagen, K., Farr, V.C., McFadden, H.A., Prosser, C.G., Davis, S.R., 1997. Time course of milk accumulation-induced opening of mammary tight junctions and blood clearance of milk components. Am. J. Physiol. 273, R379 – R386. Stelwagen, K., Ormrod, D.J., 1998. An anti-inflammatory component derived from milk of hyper-immunised cows reduces tight junction permeability in vitro. Inflam. Res. 47, 384 – 388. Stelwagen, K., Van Espen, D.C., Verkerk, G.A., McFadden, H., Farr, V.C., 1998. Elevated plasma cortisol reduces permeability of mammary tight junctions in the lactating bovine mammary epithelium. J. Endocrinol. 159, 173 – 178. Thompson, G.E., 1996. Cortisol and regulation of tight junctions in the mammary gland of the late-pregnant goat. J. Dairy Res. 63, 305 – 308. Zettl, K.S., Sjaastad, M.D., Riskin, P.M., Parry, G., Machen, T.E., Firestone, G.L., 1992. Glucocorticoid-induced formation of tight junctions in mouse mammary epithelial cells in vitro. Proc. Natl. Acad. Sci. USA 89, 9069 – 9073.