Secretion polarity of interferon-β in epithelial cell lines

Secretion polarity of interferon-β in epithelial cell lines

ABB Archives of Biochemistry and Biophysics 402 (2002) 201–207 www.academicpress.com Secretion polarity of interferon-b in epithelial cell lines Kiyo...

369KB Sizes 0 Downloads 69 Views

ABB Archives of Biochemistry and Biophysics 402 (2002) 201–207 www.academicpress.com

Secretion polarity of interferon-b in epithelial cell lines Kiyo Nakanishi,a Yoshihiko Watanabe,b Masato Maruyama,c Fumiyoshi Yamashita,a Yoshinobu Takakura,c and Mitsuru Hashidaa,* a b

Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan Department of Molecular Microbiology, Graduate School of Pharmaceutical Sciences, Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan c Department of Biopharmaceutics and Drug Metabolism, Graduate School of Pharmaceutical Sciences, Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan Received 2 January 2002, and in revised form 2 April 2002

Abstract Epithelial cells are an attractive target for local gene delivery in gene therapy for which cytokine genes such as interferon (IFN) genes are promising. However, how the secretion of the gene products is regulated in epithelial cells has been insufficiently investigated. Here, we have studied the secretion polarity of IFN-b expressed via gene transfection in mouse epithelial Pam-T cells on a bicameral culture system. In transient expression, IFN-b was predominantly secreted from the cell membrane side on which the transfection was carried out. Meanwhile, the secretion of constitutive IFN-b from stable transformants was apparently unpolarized. Interestingly, the transformants displayed a polarized secretion of transiently expressed IFN-b in a transfection-side-dependent manner, their stable IFN-b secretion remaining unpolarized. These results suggest that epithelial cells have at least dual protein sorting–secretion pathways, transient and stable, for the same secretory proteins, such as IFNs. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: Epithelial cells; Interferon-b; Secretion polarity; Secretory protein; Transfection

Epithelial cells cover various tissues and organs in animal bodies [1,2] and form barriers via cell adherent systems, including the tight junction, which separate cell surface membranes into two domains, apical and basolateral. The two domains are distinct from each other in terms of function and composition, or are polarized. Epithelial cells thereby play a fundamental role in the vectorial transport and secretion of proteins in tissues and organs [3–6]. A number of studies on protein secretion in cultured epithelial cells have demonstrated various modes of secretion: apical, basolateral, or nonpolar secretion. The mode of secretion is primarily determined by sorting signals on the peptide. For instance, N-glycans are thought to act as a signal for apical sorting of secretory proteins [7,8], although this is still controversial [9–11]. Thus, epithelial cells, when transferred with a gene encoding secretory protein, will secrete the gene product from the apical and/or *

Corresponding author. Fax: +81-75-753-4575. E-mail address: [email protected] (M. Hashida).

basolateral membrane, depending on the nature of the protein. Therefore, epithelial cells are considered to be a potential target for local delivery of genes for secretory proteins such as cytokines, including interferon (IFN),1 which have therapeutic activities. Little is known, however, about the secretion polarity of the gene products. IFNs exhibit a wide range of biological activities, including antiviral effects, inhibition of cell proliferation, regulation of cell differentiation, and modulation of the immune system [12,13], and they have been used clinically as antiviral and antitumor reagents. In addition, IFN gene transfer to tumors is being attempted for therapeutic purposes [14,15]. Hence, IFN genes are a promising candidate for gene transfer into epithelial cells for therapeutic purposes. The secretion polarity of

1 Abbreviations used: IFN, interferon; FBS, fetal bovine serum; PBS, phosphate-buffered saline; FITC, fluorescein isothiocyanate; TER, transepithelial electrical resistance; MDCK, Madin–Darby canine kidney.

0003-9861/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 3 - 9 8 6 1 ( 0 2 ) 0 0 0 9 3 - 0

202

K. Nakanishi et al. / Archives of Biochemistry and Biophysics 402 (2002) 201–207

IFN in the gene-transferred cells is one of the most important factors determining its therapeutic potential. In this study, we investigated the secretion mode of mouse or human IFN-b in mouse epithelial Pam-T cells [16] after being exogenously transferred. Interestingly, transient gene expression showed predominant IFN-b secretion from the cell membrane side on which the transfection was performed, while constitutive expression of stable transformants harboring exogenous IFN-b gene was apparently unpolarized. These observations provide a novel insight into the process of protein sorting–secretion.

Materials and methods Cell lines. Mouse epithelial squamous cell carcinoma Pam-T cells [16,17] were cultured in RPMI 1640 medium (Nissui Pharmaceutical, Tokyo, Japan) supplemented with 100 U/ml penicillin, 100 lg/ml streptomycin (GIBCO–Invitrogen, Carlsbad, CA), and 10% fetal bovine serum (FBS). Mouse L and human FL cells were maintained in Eagle’s minimum essential medium (Nissui Pharmaceutical) supplemented with 100 U/ml penicillin, 100 lg/ml streptomycin, and 6% FBS and used for the bioassay of mouse and human IFNs, respectively. Plasmids and IFNs. Human IFN-b cDNA was cloned by reverse transcriptase-polymerase chain reaction (not shown) and ligated into the BamHI/XhoI site of pcDNA3(+) (GIBCO–Invitrogen). The resultant plasmid was named pCMV-Hub. Mouse IFN-b cDNA [18] was inserted into the XhoI site of pcDNA3(+) and the constructed plasmid was designated pCMV-Mub [19]. Highly purified human IFN-b was kindly donated by Toray Basic Research Laboratories (Kamakura, Japan) and purified mouse IFN-b was prepared as described [20]. Establishment of transformants expressing mouse or human IFN-b. Stable cell lines expressing mouse or human IFN-b were established from Pam-T cells transfected with pCMV-Mub or pCMV-Hub (1 lg/ml)– Lipofectin (GIBCO–Invitrogen) (8 lg/ml) complexes followed by selection with medium containing 1 mg/ml G418 (Geneticin; Sigma, St Louis, MO). G418-resistant single colonies were picked up and examined for their IFN-b secretion. Sublines that produced substantial amounts of IFN-b were used for further experiments. Such sublines showed cell growth similar to that of the parental cells, albeit exhibiting somewhat slow cell growth, and formed an impermeable monolayer. Secretion polarity experiments. For transient gene expression, Pam-T cells were seeded on Transwell filters (1-cm2 culture area) (Costar, Cambridge, MA) at a cell density of 2  105 cells per Transwell and cultured for 5–10 days, replacing the medium every day with fresh medium, to establish impermeable cell sheets. The

monolayers were thoroughly washed twice with serumfree medium followed by treatment with IFN expression plasmid (1 lg/ml) complexed with Lipofectin (8 lg/ml) [19] from the upper or lower compartment for 4 h. The medium in both compartments was then replaced with growth medium. For stable gene expression, the IFN-b gene transformants were inoculated onto the Transwells at a cell dose of 2  105 per well and cultured for 5–7 days, with the medium being replaced each day with fresh medium. In either case, 1 h after the last medium replacement, a small amount of human or mouse IFN-b (final concentration about 200 U/ml) was added to the upper compartments, as a permeability marker, respectively, for mouse or human IFN-b gene expression [21]. After a further 24-h incubation, the culture fluids in both compartments were separately recovered and their IFN activities were assessed by a bioassay (see blow). IFN bioassay. The IFN activity of the culture fluids was measured as described previously [20,22] by monitoring the reduction in the cytopathic effect of vesicular stomatitis virus on mouse L cells or of encephalomyocarditis virus on human FL cells, for mouse or human IFN, respectively. Microplates for IFN assay were stained with 0.3% (w/v) crystal violet in 0.2 M citric acid–10% formaldehyde–10% ethanol. The titers were expressed in reference units (U) as calibrated against NIH reference mouse IFN-a=b (G002-904-511) or human IFN-a (G023-901-527). These bioassays allow distinct measurement of individual IFN activities of a mixture of mouse and human IFN-b’s, provided that the content of the heterospecies IFN is not so high that it overcomes the cross-species sensitivity of the indicator cells. Transepithelial electrical resistance. Pam-T cells were inoculated on Transwell filters (1-cm2 culture area) at a cell density of 2  105 cells/culture on day 0. Every 24 h the transepithelial electrical resistance (TER) of each cell sheet was measured with a Millicell-ERS meter (Millipore, Bedford, MA) and expressed in ohm  cm2 by subtracting values of blank filters inoculated with medium without cells. In parallel, 25 ll of 0.5% trypan blue solution was added to the upper compartments to observe the permeation of dye into the lower compartments through the monolayer. Laser scanning confocal microscopy. Pam-T cells (2  105 cells/culture) were inoculated on Transwell filters (1-cm2 per filter) and grown for 5 days. Cells were then rinsed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde–0.01% glutaraldehyde in PBS at room temperature for 20 min. Cells were washed with PBS and permeabilized with 0.2% Triton X-100–PBS for 20 min, followed by treatment with 15 lg/ml RNase A (Roche Diagnostics, Indianapolis, IN) at 37 °C for 30 min. After being washed with PBS, the cell nuclei were stained by incubation with 0:5 lg/ml

K. Nakanishi et al. / Archives of Biochemistry and Biophysics 402 (2002) 201–207

propidium iodide in 0.1 M Tris–HCl (pH 7.4) containing 0.1 M NaCl for 20 min. If necessary, cell sheets were further stained with a monoclonal anti-mouse zonula occuludin-1 (ZO-1) rat IgG (Chemicon International, CA) at 1:100 dilution or an anti-IFN-c receptor rabbit serum (C-20) (Santa Cruz Biotechnology, CA) at 1:100 dilution and then with secondary antibodies, FITCconjugated anti-rat IgG goat serum or FITC-conjugated anti-rabbit IgG goat serum (ICN Pharmaceuticals, OH), respectively, at 1:200 dilution. Emissions for optical x–y sections were collected using a confocal microscope (MRC-1024; Bio-Rad, CA) equipped with a Nikon Optiphot 2 microscope (Nikon, Tokyo, Japan) using 40 oil immersion objective. Collected x–y image segments were processed to obtain x–z images.

203

A

B

Results Formation of impermeable Pam-T cell monolayers Pam-T cells seeded on Transwell filters grow to form tight monolayers which do not allow the solute in the medium to cross from one side compartment to the other through the sheet of cells. The process of the monolayer formation was monitored by measuring the TER and observing penetration through the monolayer of trypan blue added into the upper medium as a permeability marker, as shown in Fig. 1A. The TERs increased gradually and reached a plateau after 7–8 days of culture in a cell-dose-dependent fashion. The cell sheet became impermeable several days prior to the attainment of a steady state. The cells at days 5–7 were used for transient gene expression. As some epithelial cell lines showed

Fig. 1. TER during the monolayer formation of Pam-T cells and the effects of IFN-c treatment. (A) Pam-T cells (2  105 cells per culture) were seeded on Transwells and the TER was measured every day as described under Materials and methods. The marks near the points, ‘‘+’’ and ‘‘)’’, indicate ‘‘permeable’’ and ‘‘impermeable’’ for the cell sheets, respectively. (B) Experiments were carried out as in A except for IFN-c treatment: the gray rectangle indicates the duration of treatment (from days 9 to 12) with mouse IFN-c (500 U/ml). Closed circles, closed squares, and open circles, respectively, represent no treatment, IFN-c treatment from the apical surface, and IFN-c treatment from the basolateral surface. The values are the means of three determinations with the corresponding SD.

Fig. 2. Confocal microscopy of a Pam-T cell monolayer. Pam-T cells (2  105 cells per culture) were seeded on Transwell filters and cultured for 5 days. Cell nuclei were stained with propidium iodide. Cells were stained further with anti-ZO-1 antibody (C) or anti-IFN-c receptor antiserum (D) and then with FITC-conjugated secondary antibodies. (A and B) The x–y and x–z fields, respectively, for the same cell sheet. (C and D) x–z fields. Arrowheads indicate the site of the filter.

204

K. Nakanishi et al. / Archives of Biochemistry and Biophysics 402 (2002) 201–207

different IFN-c susceptibility for the apical and basolateral cell surfaces by TER [23], the Pam-T cell sheet was examined on the IFN-c response of TER: the TER decreased only when the sheet was treated from the basolateral surface, not from the apical surface (Fig. 1B). Vertical section images of the cell sheet on day 5 obtained by confocal microscopy displayed a linear arrangement of cells in direct contact with each other, localization of ZO-1, a component of tight junctions, at the apical cell borders, and basolateral localization of the IFN-c receptor (Fig. 2). These indicate polarization of Pam-T cell sheet.

Polarized secretion of transiently expressed IFN-b Fig. 3 shows the secretion mode of human IFN-b expressed after gene transfer into Pam-T cells cultured in the Transwell system. When the plasmids were applied to the apical membrane, IFN activity was predominantly detected in the upper compartment. Inversely, when the plasmid was applied to the basal side, IFN activity was selectively detected in the lower compartment. A similar secretion mode was seen in mouse IFN-b gene transfer (Fig. 4). Therefore, in Pam-T, the transiently expressed IFN-b was preferentially secreted

Fig. 3. IFN-b activities expressed following transfection of human IFN-b into Pam-T cells. Human IFN-b gene was transfected onto apical and basal sides. In (A), typical assay wells are shown. Human IFN-b activities expressed in upper and lower compartments are shown in (B), together with recovery of marker IFN added to upper compartments. The values represent the means of three determinations with SD.

K. Nakanishi et al. / Archives of Biochemistry and Biophysics 402 (2002) 201–207

205

to pass into the other compartment across the monolayer during incubation in the experiments. Moreover, it is unlikely that the unpolarized secretion was due to any change in the cellular properties as a result of gene transfer, because the transformants still exhibited typical polarized secretion of transiently expressed IFN-b, their stable expression remaining unpolarized (Fig. 6).

Discussion

Fig. 4. IFN-b activities expressed following transfection of mouse IFN-b expression plasmid into Pam-T cells. Mouse IFN-b expression plasmid was applied from upper or lower compartments. Mouse IFNb activities in upper and lower compartments are shown as the means of three determinations with SD.

from the cell membrane side on which the transfection was carried out. Unpolarized secretion of stably expressed IFN-b Next, we examined the mode of secretion of IFN-b constitutively expressed in stable Pam-T transformants transduced with the mouse or human IFN-b gene. Interestingly, the stable IFN-b secretion in the transformants was generally unpolarized; i.e., comparable amounts of IFN activity were detected in the upper and lower media (Fig. 5). It is unlikely that the apparent unpolarized secretion was due a defect in the cell sheet, because the monolayers did not permit IFN-b, added into the apical medium as a permeation marker,

A

B

Fig. 5. Constitutive secretion of IFN-b in stable transformants. In (A), Pam-T(Mub) cells were transduced with the mouse IFN-b gene; in (B), Pam-T(Hub) cells were transduced with the human IFN-b gene. The values represent the means of three determinations with the SD.

In this study, we examined the secretion polarity of IFN-b exogenously expressed in mouse epithelial Pam-T cells using the bicameral culture system. According to a widely accepted model for protein sorting in epithelial cells, glycoproteins are trafficked in the cytoplasm via the endoplasmic reticulum–Golgi vehicle systems, being bridled into the systems by the leader sequence, regulated by other signal sequences for sorting direction, and thereby secreted toward the apical or the basolateral membrane side or otherwise toward both sides. As IFN-b is a paradigm of the secretory glycoprotein with a typical leader sequence (signal sequence for secretion) and N-glycosylated residues [24], but without any other overt sequences for intracellular localization or sorting, it should be secreted in a defined direction through default bulk flow of membrane traffic in polarized epithelial cells. Unexpectedly, however, we found that the mode of secretion of transiently expressed IFN-b was not restricted to one direction but changed depending on the side at which gene transfer was performed: apical secretion for apical transfection and basal secretion for basal transfection. Meanwhile, the secretion of constitutive IFN-b from the stable transformants was unpolarized. It is worth noting that the transfection-side selective secretion of transiently expressed IFN-b was seen in the transformants, while the unpolarized stable secretion remained unchanged. It is unlikely that the variety of the mode of secretion is due to the immediate-early promoter of human cytomegalovirus used for driving the exogenous IFN-b gene, because the promoter would contribute only to the transcription level. It is also unlikely that the gene transfection method is related to the secretion variety because a cationic liposome (Lipofectin) complexed with plasmid DNA can enhance cellular uptake of the plasmid but not affect the protein secretion: the secretion polarity of the transiently expressed IFN-b following transfection onto one side of the cells was not disturbed by simultaneous treatment with liposomes alone as well as plasmid pcDNA3(+)–liposome complexes on the other membrane side (data not shown). This phenomenon is not unique in Pam-T cells and has been also observed in another type of epithelial cells, human intestinal

206

K. Nakanishi et al. / Archives of Biochemistry and Biophysics 402 (2002) 201–207

A

B

Fig. 6. Secretion of IFN-b from Pam-T(Hub) cells following transfection of the mouse IFN-b gene. In (A), the mouse IFN-b gene was applied to the apical side; in (B), the mouse IFN-b gene was applied to the basal side. Open bars indicate mouse IFN-b and shaded bars human IFN-b. The values represent the means of three determinations with SD.

Caco-2 and Madin–Darby canine kidney (MDCK) cells (not shown). Therefore, it seems to be the case that N-glycans of IFN-b do not act as crucial apical sorting signals in those epithelial cells, although more experiments are needed to confirm this. Overall, these findings suggest that epithelial cells have multiple sorting–secretion pathways for the same secretory proteins, at least for IFN-b, which might be distinctly separated without any mutual crossover. A similar stimulation-side dependency of IFN-b production was observed in endogenous IFN gene expression after treatment of Pam-T, Caco-2, and MDCK cells with polyriboinosinic acid:polyribocytidylic acid (poly I:poly C), a potent IFN inducer [25]. In this case, poly I:poly C treatment onto apical and basal sides caused predominant IFN secretion into apical and basal medium, respectively, [21]. Thus, the phenomenon is not specific for exogenous gene expression. Furthermore, essentially the same phenomenon was observed for mouse IFN-c gene expression as well as human IFN-c gene expression (not shown). The mode of secretion of some secretory proteins, at least for IFN-b and IFN-c, seems to be determined not simply by their structural properties, although further experiments are required to verify the underlying mechanism(s). At any rate, our findings may open up new perspectives on cell polarity, which previously had been thought to be a fixed cell property, depending on the nature of the proteins, and also provide useful basic information on gene delivery to epithelial cells in somatic gene therapy.

Acknowledgments This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.

References [1] C. Lau, H.E. Soriano, F.D. Ledley, M.J. Finegold, J.H. Wolfe, E.H. Birkenmeier, S.J. Henning, Hum. Gene Ther. 6 (1995) 1145– 1151. [2] J.N. Lozier, J.R. Yankaskas, W.J. Ramsey, L. Chen, H. Berschneider, R.A. Morgan, Hum. Gene Ther. 8 (1997) 1481–1490. [3] J.E. Casanova, G. Apodaca, K.E. Mostov, Cell 66 (1991) 65–75. [4] E. Rodriguez-Boulan, S.K. Powell, Annu. Rev. Cell Biol. 8 (1992) 395–427. [5] S.W. Pimplikar, K. Simons, Nature 362 (1993) 456–458. [6] K. Matter, I. Mellman, Curr. Opin. Cell Biol. 6 (1994) 545–554. [7] Y. Kitagawa, Y. Sano, M. Ueda, K. Higashop, H. Narita, M. Okano, S.-I. Matsumoto, R. Sasaki, Exp. Cell Res. 213 (1994) 449–457. [8] P. Scheiffele, J. Peranen, K. Simons, Nature 378 (1995) 96–98. [9] M.P. Marzolo, P. Bull, A. Gonzalez, Proc. Natl. Acad. Sci. USA 94 (1997) 1834–1839. [10] T. Su, R. Cariappa, K. Stanley, FEBS Lett. 453 (1999) 391–394. [11] J.E. Larsen, G.V. Avvakumov, G.L. Hammond, L.K. Vogel, FEBS Lett. 451 (1999) 19–22. [12] E. DeMaeyer, J. DeMaeyer-Guignard, Interferons and Other Regulatory Cytokines, Wiley, New York, 1988. [13] I. Gresser, M.G. Tovey, Biochim. Biophys. Acta 516 (1978) 231– 247. [14] X.Q. Qin, N. Tao, A. Dergay, P. Moy, S. Fawell, A. Davis, J.M. Wilson, J. Barsoum, Proc. Natl. Acad. Sci. USA 95 (1998) 14411–14416.

K. Nakanishi et al. / Archives of Biochemistry and Biophysics 402 (2002) 201–207 [15] A. Natsume, M. Mizuno, Y. Ryuke, J. Yoshida, Gene Ther. 6 (1999) 1626–1633. [16] Y. Maruguchi, K.I. Toda, K. Fujii, S. Imamura, Y. Watanabe, Cancer Lett. 60 (1991) 41–49. [17] S.H. Yuspa, P. Hawley-Nelson, B. Koehler, J.R. Stanley, Cancer Res. 40 (1980) 4694–4703. [18] Y. Higashi, Y. Sokawa, Y. Watanabe, Y. Kawade, S. Ohno, C. Takaoka, T. Taniguchi, J. Biol. Chem. 258 (1983) 9522–9529. [19] K. Kawabata, M. Kondo, Y. Watanabe, Y. Takakura, M. Hashida, Pharm. Res. 14 (1997) 483–485. [20] Y. Watanabe, Y. Kawade, in: M.J. Clemens, A.G. Morris, A.J.H. Gearing (Eds.), Lymphokines and Interferons: A Practical Approach, IRL Press, Oxford, 1987, pp. 1–14.

207

[21] S. Okamoto, K. Nakanishi, Y. Watanabe, F. Yamashita, Y. Takakura, M. Hashida, Biochem. Biophys. Res. Commun. 254 (1999) 5–9. [22] Y. Watanabe, K. Kuribayashi, S. Miyatake, K. Nishihara, E. Nakayama, T. Taniyama, T. Sakata, Proc. Natl. Acad. Sci. USA 86 (1989) 9456–9460. [23] R.B. Adams, S.M. Planchon, J.K. Roche, J. Immunol. 150 (1993) 2356–2363. [24] C. Weissmann, H. Weber, Prog. Nucleic Acid Res. Mol. Biol. 33 (1986) 251–300. [25] S. Okamoto, Y. Watanabe, Y. Takakura, M. Hashida, J. Biochem. 124 (1998) 697–701.