The following chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA): D-(+)–glucose (G-5767), sodium bicarbonate (S5761), L-arginine (Arg, A-6969), L-glutamine (G-5792), L-leucine (L-8000), L-proline (P-0380), human insulin (I3536), N${}^{\omega }$-nitro-L-arginine methyl ester hydrochloride (L-NAME, N5751), forskolin (F3917), Sp-Adenosine 3${}^{\prime }$, 5${}^{\prime }$-cyclic monophosphorothioate triethylammonium salt hydrate [sp-cAMP (an analogue of cAMP), A166] and H-89 dihydrochloride hydrate [H-89 (an inhibitor of PKA), B1427]. The DMEM/F12 (Gibco, Merelbeke, Belgium), custom-made Arg-free Dulbecco’s modified Eagle’s Ham medium (DMEM) (Formula, #08-5009EF), penicillin/streptomycin solution (P/S), fetal bovine serum (FBS), 0.25% trypsin-ethylene diamine tetraacetic acid (EDTA), and phosphate buffered saline (PBS) were obtained from Gibco. The Arg-free DMEM contained concentrations of other amino acids that are normally present in the plasma of gestating swine [3]. Isotope water stock (${}^3$H${}_2$O, stock concentration at 1 mCi/mL) was purchased from American Radiolabeled Chemicals.

Porcine trophectoderm (pTr2) cells were derived from elongated porcine blastocysts obtained from gilts on day 12 of pregnancy [27]. Cells were cultured in 75-cm${}^2$ flasks (Corning, NY, USA) in DMEM/F12 medium supplemented with 5% fetal bovine serum (FBS), 1% penicillin/streptomycin (PS) and 0.10 U/mL insulin at 37 ${}^{\circ }$C under an atmosphere of 5% CO${}_2$ in air. The medium was changed every 2 days.

At confluence, the cells were collected using 0.25% trypsin-EDTA, and seeded at a density of 1 $\times$ 10${}^5$ cells/mL in 6-well, or 2 $\times$ 10${}^4$ cells/mL in four-well culture plates (Lab-Tek® Chamber Slide™ System, Nunc International). Cells were cultured overnight (16 h) in complete DMEM/F-12 containing 5% FBS, 1% PS and 0.10 U/mL insulin, and then incubated with the Arg-free DMEM containing 5% FBS, 1% PS and 0.10 U/mL insulin for 6 h, as previously described [27]. To determine an appropriate concentration of Arg supplementation, cells were treated with 0.00 mM (control), 0.25 mM, or 0.50 mM Arg for 24 or 48 h, respectively. The Arg concentrations used in the present study were selected based on the previous study that Arg concentrations in the plasma of non-supplemented gilts are 240 $\mu$M (approximately 0.25 mM) at 1 h after feeding and reach a peak value of 500 $\mu$M (0.50 mM) at 1 h after feeding in arginine-supplemented gilts [28].

To investigate the potential role of NO, L-NAME, an inhibitor of NO synthase (NOS), was added to inhibit the Arg-NO pathway. Briefly, pTr2 cells were cultured in Arg-free DMEM containing 5% FBS, 1% PS and 0.10 U/mL insulin for 6 h, and then in fresh medium supplemented with either 0.50 mM Arg or 0.50 mM Arg plus 3.00 mM L-NAME (0.50 mM Arg + L-NAME) for 24 h.

Additionally, Forskolin (10 $\mu$M; a cell-permeable activator of adenylyl cyclase), a cAMP analogue (sp-cAMP, 50 $\mu$M), or H-89 [10 $\mu$M; which inhibits cAMP-dependent protein kinase (PKA)] was used to determine the role of the cAMP signaling pathway in mediating the effect of Arg on AQP expression, as previously described [29]. Briefly, the pTr2 cells were cultured in Arg-free DMEM containing 5% FBS, 1% PS and 0.10 U/mL insulin for 6 h, and then in fresh medium supplemented with 10 $\mu$M Forskolin, 50 $\mu$M sp-cAMP, or 10 $\mu$M H-89 (H-89) for 2 h. Thereafter, the cells were cultured in fresh medium containing 0.50 mM Arg for another 48 h. Thus, the four treatments were Control (0.50 mM Arg), Forskolin (10 $\mu$M Forskolin + 0.50 mM Arg), sp-cAMP (50 $\mu$M sp-cAMP + 0.50 mM Arg), and H-89 (10 $\mu$M H-89 + 0.50 mM Arg). The treated cells were harvested for the determination of mRNA expression, protein expression, NO production, and cAMP concentrations (see below). All experiments were performed in triplicate using cells from passages 10–20.

Total RNA was extracted from pTr2 cells using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The concentrations and purity of the RNA samples were determined using a NanoDrop 2000 UV-Vis spectrophotometer (Thermo, Wilmington, MA, USA). The integrity of the RNA was assessed with 1.5% agarose gel in Tris acetate-EDTA buffer. The ratio of absorbance at 260 nm and 280 nm was approximately 2.0 for the total RNA isolated from pTr2 cells. Total RNA was reverse-transcripted into cDNA using the PrimerScript® RT reagent Kit with gDNA Eraser (Takara Biotechnology Co., Dalian, China). Real-time PCR was performed in the CFX Connect TM Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) using the SYBR® Premix Ex Taq TM II (Takara Bio, Otsu, Japan). The specific sequences of forward and reverse primers and targeted amplicons for AQP-1 (221 bp), AQP-3 (212 bp), AQP-9 (62 bp), and AQP-11 (229 bp), as well as the housekeeping gene $\beta$-actin (208 bp) are provided as previously described [30]: AQP1:5${}^{\prime }$-TGACCTTCGCTATGTGCTTCC${}^{\prime }$ and 5${}^{\prime }$-GTCCAAGTGTCCAGAGGGGTAG-3${}^{\prime }$, AQP3: 5${}^{\prime }$-TGACCTTCGCTATGTGCTTCC-3${}^{\prime }$ and 5${}^{\prime }$- GTCCAAGTGTCCAGAGGGGTAG-3${}^{\prime }$, AQP9: 5${}^{\prime }$-TGTCATTGGCCTCCTGATTG-3${}^{\prime }$ and 5${}^{\prime }$-TGGCACAGCCACTGTTCATC-3${}^{\prime }$, AQP11: 5${}^{\prime }$-CGTCTTGGAGTTTCTGGCTACC-3${}^{\prime }$ and 5${}^{\prime }$-CCTGTCCCTGACGTGATACTTG-3${}^{\prime }$, and $\beta$-actin: 5${}^{\prime }$-TCCACCGCAAATGCTTCTAG-3${}^{\prime }$ and 5${}^{\prime }$-TGCTGTCACCTTCACCGTTC-3${}^{\prime }$, respectively. The primers were synthesized by Sangon Biotech Co. (Shanghai, China). The relative expression of mRNAs for targeted genes after treatments was calculated using the 2${}^{-\mathrm{\Delta }\mathrm{\Delta }Ct}$ method, while the specific expression of AQPs in pTr2 cells cultured without Arg was calculated using the 2${}^{-\mathrm{\Delta }Ct}$ method. The experiments were run in triplicates and the relative expression of targeted genes were normalized to the expression of the housekeeping gene ($\beta$-actin).

For the measurements of water transport, the pTr2 cells were cultured on a snap-well insert (polycarbonate with a pore size of 4 microns; Cat. no. P3901, Physiologic Instruments) to 100% confluency. The snap-well insert with a confluent monolayer of cells was mounted to a special slider (Cat. no. P2305, Physiologic Instruments), and then transferred to four sets of Ussing chambers. Both sides of the chambers in each set contained the same volume of Krebs biocarbonate buffer (pH 7.4, 37 ${}^{\circ }$C) and were constantly gassed with 95% O${}_2$ and 5% CO${}_2$. Krebs biocarbonate buffer contained 119 mM NaCl, 4.8 mM KCl, 2.5 mM CaCl${}_2$.2H${}_2$O, 1.2 mM MgSO${}_4$, 1.2 mM KH${}_2$PO${}_4$, 25 mM NaHCO${}_3$, 20 mM Hepes, and 5 mM glucose. Before assays, the quality of electrodes used in the chambers was assessed with a blank snap-well insert mounted into the slider to ensure the electric current of 60–68 $\mu$A. Thereafter, the blank insert was replaced with a snap-well insert containing confluent cells. Water transport was measured by adding 20 $\mu$L of a ${}^3$H${}_2$O solution [prepared by mixing 500 $\mu$L Krebs biocarbonate buffer with 50 $\mu$L ${}^3$H${}_2$O stock (American Radiolabeled Chemicals Inc)] to 5 mL of oxygenated (95% O${}_2$/5% CO${}_2$) Krebs bicarbonate medium (containing 5 mM D-glucose and 0.00, 0.25, or 0.50 mM Arg) in the “mucosal side” of the Ussing chamber. A sample (50 $\mu$L) was collected from the “serosal side” of the Ussing chamber every 5 min over a 20-min period, and measured for radioactivity by a liquid scintillation spectrometer (Perkinelmer, USA). The specific radioactivity of ${}^3$H${}_2$O in the solution of the “mucosal side” was used to calculate the rate of water transport per well (the confluent monolayer of cells).

For immunofluorescence assays, pTr2 cells were seeded at 2 $\times$ 10${}^4$cells/mL into four-well dishes (Lab-Tek, Nunc). After treatment, cells were fixed with 2% paraformaldehyde for 15 min, and then permeabilized with 1% Triton X-100 for 10 min, washed twice with PBS, and blocked with 5% BSA for 30 min. The samples were incubated with a primary antibody overnight at 4 ${}^{\circ }$C. The primary antibodies were polyclonal rabbit anti-AQP-1 (1:200, 101AP, FabGennix, USA), anti-AQP-3 (1:200, bs-1253R, Bioss, China), anti-AQP-9 (1:50, OABF00907, Aviva Systems Biology, USA), anti-AQP-11 (1:50, 1101AP, FabGennix, USA), anti-iNOS (1:100, ab15323, Abcam, USA), anti-cAMP protein kinase catalytic subunit (1:1000, ab26322, Abcam, USA), anti-PKA $\alpha$/$\beta$/$\gamma$ (1:200, sc-98951, Santa Cruz Technology, USA) and anti-p-PKA $\alpha$/$\beta$/$\gamma$ (1:200, sc-32968, Santa Cruz Technology, USA). Subsequently, the stained cells were washed three times with PBS and incubated with the corresponding secondary antibodies at 1:200 (Alexa Fluor 594 goat anti-rabbit-IgG, Life Technology, USA; Alex Fluor 488 goat anti- rabbit-IgG, Life Technology, USA) at 25 ${}^{\circ }$C for 2 h in the dark. The samples were subsequently counterstained with DAPI/anti-fade mounting medium (ProLong® Gold anti-fade reagent with DAPI, Life Technologies, Carlsbad, CA, USA). Fluorescence signals were observed using a fluorescence microscope (Axio Observer A1, Carl Zeiss, Jena, Germany) and a Confocal Laser Scanning Microscope (LSM 710, Carl Zeiss, Heidenheim, Germany).

After treatment, cells were washed twice with cold PBS and lysed in a lysis buffer (RIPA, BioTeke, Beijing, China) containing 1 mM phenyl-methylsulfonyl fluoride (PMSF) (Sigma, St. Louis, MO, USA), followed by centrifugation at 14,000 $\times$g at 4 ${}^{\circ }$C for 15 min. The concentrations of protein in the supernatant fluid were determined using a BCA Protein Assay Reagent Kit (Thermo Fisher Scientific, Rockford, USA). After boiling for 10 min, the samples (15 $\mu$g each) were separated by 12% SDS-polyacrylamide gel electrophoresis (PAGE) gels and then transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Billerica, MA, USA). After being blocked with 5% BSA in Tris-Tween-buffered saline (TBST) at 25 ${}^{\circ }$C for 3 h, the membranes were incubated with a primary antibody at 4 ${}^{\circ }$C overnight with gentle rocking. The membranes were then incubated with the secondary antibody for 2 h at room temperature. The information on all antibodies used in this study is provided in Table 1. The membranes were treated with the Clarity™ Western ECL Substrate (Bio-Rad, Irvine, CA, USA), and protein bands were visualized and quantitated using the ChemiDoc XRS imaging system and the Image Lab software (version 3.0, National Institutes of Health, Bethesda, MD, USA), as previously described [30].

The pTr2 cells were incubated in ice-cold lysis buffer (0.1 M HCl) for 5 min, and then centrifuged at 1000 $\times$g for 3 min at 4 ${}^{\circ }$C. The concentrations of protein in the cell lysates were determined using the BCA Protein Assay Reagent Kit. The concentrations of cAMP in cell lysates were determined using a commercial enzyme-linked immunosorbent assay (ELISA) kit (Cat. no. CA200, Sigma-Aldrich, Munich, Germany), according to the manufacturer’s guidelines. The intracellular concentrations of cAMP are expressed as pmoL/$\mu$g protein.

Concentrations of nitrite and nitrate (stable oxidation products of NO) in the culture medium of pTr2 cells were determined using a commercial NO assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Briefly, the pTr2 cells were seeded in a 6-well plate at a density of 1 $\times$ 10${}^5$ cells/mL and the culture medium was used to determine the concentrations of nitrite and nitrate according to the instructions of the manufacturer. The cell lysates were also analyzed for the concentrations of protein as noted previously. The concentrations of nitrite and nitrate in culture medium are expressed as pmoL/$\mu$g protein.

Data from at least three independent experiments were analyzed by the SPSS software (version 17.0, IBM, Chicago IL, USA). The results are expressed as means $\pm$ SEM. The Student’s paired t-test was used to determine differences between two treatment groups. Differences among three or more treatment groups were determined using one-way ANOVA, followed by the Duncan’s multiple-range test. In the case of obtaining results over different time points, the data were analyzed by using ANOVA for repeated measurements. Differences in means among treatments were considered statistically significant at p 0.05.

The expression of mRNAs and proteins for AQP1, AQP3, AQP9, and AQP11 in pTr2 cells were determined by quantitative RT-PCR and immunofluorescence analyses, respectively (Fig. 1). AQP1 was localized mainly to the nucleus, whereas AQP3, AQP9 and AQP11 were localized mainly in the plasma membrane of pTr2 cells (Fig. 1B). Importantly, the expression of AQP3 mRNA in pTr2 cells was greater than that for AQP1, AQP9 and AQP11 (p 0.05), indicating a potential role of the abundantly expressed AQP3 in mediating water transport by pTr2 cells (Fig. 1). Therefore, AQP3 was the focus of our subsequent experiments.

Fig. 1.

The expression of AQPs in pTr2 cells. (A) pTr2 cells were cultured in DMEM/F12 containing 5% FBS, 1% penicillin/streptomycin solution (PS) and 0.10 U/mL insulin for 48 h. (A) The expression of AQP1, AQP3, AQP9, and AQP11 mRNAs in pTr2 cells (n = 6). The relative mRNA expression of AQPs was calculated using the 2${}^{-\mathrm{\Delta }Ct}$ method. Data were analyzed by one-way ANOVA followed by the Duncan’s multiple comparison using the SPSS software (version 17.0). The bars represent means $\pm$ SEM and different letters (a–c) indicate significant differences (p 0.05). (B) Immunofluorescent localization of AQP proteins in pTr2 cells was determined using a confocal laser scanning microscope. Representative immunofluorescence images are shown for three independent experiments in which nuclei were labeled with DAPI (blue) and AQPs detected by specific antibodies (red). Scale bars: 20 $\mu$m. The original magnification was at 400$\times$.

The transport of water across the apical and basolateral membranes of pTr2 cells was linear within a 20 min period (Fig. 2). Compared with the control group (0.00 mM Arg), 0.25 and 0.50 mM Arg increased (p 0.05) the transport of water by pTr2 cells during the 20 min period. The rate of water transport was about 30% greater (p 0.05) in the presence of 0.50 mM Arg, as compared to 0.25 mM Arg. For consistency, we mainly used the 0.50 mM Arg concentration for other assays.

Fig. 2.

Effects of arginine on water transport by pTr2 cells. Cells were cultured on a snap-well insert (Physiologic Instruments) until reaching 100% confluency, and then transferred to four sets of Ussing chambers. Water transport was measured over a 20 min period by adding ${}^3$H${}_2$O to 5 mL of oxygenated (95% O${}_2$/5% CO${}_2$) Krebs bicarbonate medium [containing 5 mM D-glucose and 0.00, 0.25, or 0.50 mM L-arginine (Arg)] in the “mucosal side” of the Ussing Chamber. Data were analyzed by one-way ANOVA for repeated measurements, followed by the Duncan’s multiple comparison using the SPSS software (version 17.0). Values, expressed as $\mu$L water/well (monolayer of cells), are means $\pm$ SEM, n = 6. Different letters (a–c) indicate significant differences (p 0.05). At each time point, increasing the concentration of Arg from 0.00 to 0.25 and to 0.50 mM in the solution of the “mucosal side” increased water transport by pTr2 cells in a dose-dependent manner (p 0.05).

More importantly, 0.50 mM Arg increased the mRNA expression of AQP3 in pTr2 cells, as compared with 0.00 mM and 0.25 mM Arg (Fig. 3A; p 0.05). Based on western blotting and immunofluorescence analyses, AQP3 protein was also increased in response to the treatment of pTr2 cells with Arg (Fig. 3B and C; p 0.05). These results indicated that Arg up-regulated AQP3 expression to enhance water transport by pTr2 cells. Moreover, because AQP3 and other AQPs may work together to regulate water transport by pTr2 cells, the expression of other AQPs has also been shown in Supplementary Fig. 1. We found that the AQP11 mRNA expression was increased (p 0.05) in pTr2 cells treated with 0.25 and 0.50 mM Arg (Supplementary Fig. 1). However, mRNA levels for both AQP1 and AQP9 mRNAs were not significantly affected by 0.25 or 0.50 mM Arg (p $>$ 0.05).

Fig. 3.

Effects of arginine on the expression of AQP3 in pTr2 cells. Cells were cultured for 48 h in arginine (Arg)-free DMEM medium containing 5% FBS, 1% penicillin/streptomycin solution (P/S) and 0.10 U/mL insulin supplemented with 0.00 mM (control), 0.25 mM, and 0.50 mM arginine (Arg). (A) Expression of the AQP3 mRNA in pTr2 cells at 48 h after treatment (n = 6). Data were analyzed by one-way ANOVA. The bars represent means $\pm$ SEM. Different letters (a-b) indicate significant differences (p 0.05). (B) After treatment for 48 h, the abundance of the AQP3 protein was analyzed by western blotting. Data were analyzed by the Student’s paired t-test. Representative results of three independent experiments are shown as means $\pm$ SEM; * p 0.05. (C) Representative immunofluorescence images of cells at 48 h after treatment, labeled with DAPI (blue, nuclei) and AQP3 antibody (red). Scale bars: 20 $\mu$m. Original magnification, 400$\times$.

It was not known if Arg increased expression of AQP3 via an NO-dependent pathway. Results of the present study revealed that 0.50 mM Arg significantly increased NO production by pTr2 cells after 48 h of treatment (Fig. 4A; p 0.05). NO synthesis by pTr2 cells was significantly decreased by the addition of L-NAME plus Arg (0.50 mM Arg + L-NAME group) when compared with the 0.50 mM Arg group (Fig. 4B; p 0.05). Compared with the control group, 0.50 mM Arg enhanced the fluorescence signaling of iNOS in pTr2 cells (Supplementary Fig. 2). L-NAME treatment markedly decreased (p 0.05) the fluorescence intensity of the labeled iNOS protein in the cells (Supplementary Fig. 3).

Fig. 4.

Measurement of NO production by cultured pTr2 cells. (A) Arginine (Arg) supplementation increased NO synthesis by pTr2 cells. (B) The effect of L-NAME supplementation to medium was to decrease NO synthesis by pTr2 cells. Data were analyzed by the Student’s paired t-test. Representative results of three independent experiments are shown as means $\pm$ SEM. Different letters indicate significant differences (p 0.05). L-NAME, N${}^{\omega }$-nitro-L-arginine methyl ester hydrochloride (an inhibitor of NO synthase).

When compared with the 0.50 mM Arg group, pTr2 cells cultured in the presence of 0.50 mM Arg + L-NAME had lower levels of AQP3 mRNA and protein based on results from the qRT-PCR, western blotting, and immunofluorescence analyses (Fig. 5; p 0.05). Collectively, these results indicate that Arg increased AQP3 expression in pTr2 cells partially via an NO-dependent mechanism.

Fig. 5.

The NO-dependent effect of arginine to increase the expression of AQP3 in pTr2 cells. (A) The expression of AQP3 mRNA in pTr2 cells at 24 h after treatment (n = 6) was decreased (p 0.05) in the presence of L-NAME that inhibited the production of NO from arginine (Arg). (B) After a 24 h treatment period, the abundance of AQP3 protein was decreased (p 0.05) in cells treated with 0.50 mM Arg + L-NAME versus 0.50 mM Arg alone. Data were analyzed by the Student’s paired t-test. (C) Representative immunofluorescence images of cells at 24 h after treatment labeled with DAPI (blue, nuclei) and antibody to AQP3 (red) also suggest a decrease in AQP3 protein in cells cultured in the presence of 0.50 mM Arg + L-NAME as compared to treatment with 0.50 mM Arg alone. Scale bars: 20 $\mu$m. Original magnification, 400$\times$. L-NAME, N${}^{\omega }$-nitro-L-arginine methyl ester hydrochloride (an inhibitor of NO synthase).

To further investigate the role of the cAMP-PKA signaling pathway in the Arg-induced increase in AQP3 expression in pTr2 cells, the cAMP signaling pathway was examined. Arg (0.50 mM) increased the abundance of the PKA protein in pTr2 cells as shown by immunofluorescence analyses (Fig. 6A). Of note, treatment with 0.50 mM Arg induced nuclear translocation, confirming the induction of active intracellular signaling. Further, concentrations of intracellular cAMP were increased in pTr2 cells by 0.50 mM Arg when compared to control values (Fig. 6B; p 0.05). Furthermore, when compared with 0.00 mM Arg, the abundances of the PKA protein’s catalytic subunit (Fig. 6C), phosphorylated PKA, phosphorylated PKA $\alpha$/$\beta$/$\gamma$, and phosphorylated CREB (Fig. 6D) were all increased in response to 0.50 mM Arg (p 0.05).

Fig. 6.

Effects of arginine on the cAMP signaling pathway in pTr2 cells. Cells were cultured in arginine (Arg)-free DMEM [containing 5% FBS, 1% penicillin/streptomycin solution (P/S) and 0.10 U/mL insulin] supplemented with 0.00 mM Arg (control) or 0.50 mM Arg. (A) Representative immunofluorescence images of increases in PKA in pTr2 cells at 48 h after treatment, labeled with DAPI (blue, nuclei) and the PKA catalytic subunit antibody (green). Scale bars: 20 $\mu$m. Original magnification, 400$\times$. (B) cAMP concentrations in pTr2 cells increased in a dose-dependent manner (p 0.05) at 48 h after treatment, measured by enzyme-linked immuno sorbent assay (ELISA). Different letters (a-b) indicate significant differences (p 0.05). (C,D) The expression of cAMP-dependent protein kinase (PKA) catalytic subunit (panel C) and components of the PKA/CREB pathway related proteins (panel D) increased (p 0.05) in pTr2 cells after 48 h treatment as determined by western blotting analysis. Data were analyzed by the Student’s paired t-test. Representative results of three independent experiments are shown as means $\pm$ SEM; * p 0.05.

The effects of Forskolin, sp-cAMP, and H-89 confirmed that changes in the expression of AQP3 protein were coordinate with changes in the PKA catalytic subunit as determined using immunofluorescence microscopy and western blot analyses (Fig. 7). Additional evidence showed that cAMP concentrations in pTr2 cells after the treatment with 0.50 mM Arg were increased (p 0.05) by Forskolin, but was decreased (p 0.05) by H89 (Supplementary Fig. 4). These results indicate that the cAMP-PKA signaling pathway positively enhanced the expression of AQP3 mRNA and protein in pTr2 in response to Arg treatment that stimulated the transport of water by these cells.

Fig. 7.

Effects of the cAMP signaling pathway on AQP3 expression in pTr2 cells. Cells were cultured in arginine (Arg)-free DMEM containing 5% FBS, 1% penicillin/streptomycin solution (P/S) and 0.10 U/mL insulin for 6 h, followed by culture for 2 h in the presence of fresh medium supplemented with 10 $\mu$M Forskolin (a cell-permeable activator of adenylyl cyclase), 50 $\mu$M sp-cAMP (an analogue of cAMP), or 10 $\mu$M H-89 (an inhibitor of PKA). Thereafter, cells were cultured in the presence of 0.50 mM Arg for 48 h. Thus, the four treatments were Control (0.50 mM Arg), Forskolin (10 $\mu$M Forskolin + 0.50 mM Arg), sp-cAMP (50 $\mu$M sp-cAMP + 0.50 mM Arg), and H-89 (10 $\mu$M H-89 + 0.50 mM Arg). (A,B) Representative immunofluorescence images of cells after the treatment, labeled with DAPI (blue) and AQP3 antibody (red) (A) and PKA catalytic subunit antibody (green) (B), are shown. Scale bars: 20 $\mu$m. Original magnification, 400$\times$. (C) The abundances of PKA catalytic subunit protein and AQP3 were determined by western blotting analysis and revealed significant effects of treatments on the expression of AQP3. The bars represent means $\pm$ SEM. Different letters (a–d) indicate significant differences (p 0.05). (D) Western blotting analysis of proteins related to the PKA/CREB signaling pathway in pTr2 cells revealed significant effects of treatment on the expression of AQP3. Data were analyzed by one-way ANOVA followed by the Duncan’s multiple comparison test using the SPSS software (version 17.0). Representative results of three independent experiments are shown as means $\pm$ SEM. Means not sharing the same letter differ (p 0.05).