Aquaporin-1 plays important role in proliferation by affecting cell cycle progression

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ Esta es la versión aceptada del artículo publicado en: Journal of Cellular Physiology This is a accepted manuscript of a paper published in: Journal of Cellular Physiology Journal of Cellular Physiology (2015): 16 June 2015 DOI: https://doi.org/10.1002/jcp.25078 Copyright: © 2015 Wiley Periodicals, Inc. El acceso a la versión publicada del artículo puede requerir la suscripción de la revista. Access to the published version may require subscription. “This is the peer reviewed version of the following article: Galán-Cobo, A., Ramírez-Lorca, R., Toledo-Aral, J.J. and Echevarría, M. (2016), Aquaporin-1 plays important role in proliferation by affecting cell cycle progression. J. Cell. Physiol., 231: 243-256. . which has been published in final form at https://doi.org/10.1002/jcp.25078 This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. This article may not be enhanced, enriched or otherwise transformed into a derivative work, without express permission from Wiley or by statutory rights under applicable legislation. Copyright notices must not be removed, obscured or modified. The article must be linked to Wiley’s version of record on Wiley Online Library and any embedding, framing or otherwise making available the article or pages thereof by third parties from platforms, services and websites other than Wiley Online Library must be prohibited." 1 Aquaporin-1 plays important role in proliferation by affecting cell cycle progression Ana Galán-Cobo1, Reposo Ramírez-Lorca1, Juan José Toledo-Aral1,2 and Miriam Echevarría*,1,3. 1 Institute of Biomedicine of Seville (IBiS), Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla (Departamento de Fisiología Médica y Biofísica), Seville, Spain. 2 Biomedical Research Centre Network for Neurodegenerative Diseases (CIBERNED). 3 Biomedical Research Centre Network for Respiratory Diseases (CIBERES). *Miriam Echevarría Instituto de Biomedicina de Sevilla (IBIS), Hospital Universitario Virgen del Rocío/ CSIC/ Universidad de Sevilla, Av. Manuel Siurot s/n, Seville 41013, Spain Tel: +34 955 923036 Fax: +34 955 923101 E-mail: [email protected] 2 Overexpression of AQP1 increases cell proliferation Key words: • Aquaporin-1 • Proliferation • Hypoxia • Carotid body • BrdU • Cell cycle • Apoptosis • Nocodazole • Affymetrix Total number of text figures: 7 Total number of tables: 4 Total number of supplementary figure: 1 Grant information: This work was funded by a grant from the “Instituto de Salud Carlos III” (Exp. PS09/00605), and by grants from “La Junta de Andalucía, Consejería de Innovación Ciencia y Empresa” (P08-CTS-03574) and Consejería de Salud (PI0298-2010) to ME. 3 Abstract Aquaporin-1 (AQP1) has been associated with tumor development. Here, we investigated how AQP1 may affect cell proliferation. The proliferative rate of adult carotid body (CB) cells, known to proliferate under chronic hypoxia, was analyzed in wild-type (AQP1 +/+) and knock out (AQP1 -/-) mice, maintained in normoxia or exposed to hypoxia while BrdU was administered. Fewer numbers of total BrdU+ and TH-BrdU+ cells were observed in AQP1 -/- mice, indicating a role for AQP1 in CB proliferation. Then, by flow cytometry, cell cycle state and proliferation of cells overexpressing AQP1 were compared to those of wild-type cells. In the AQP1-overexpressing cells, we observed higher cell proliferation and percentages of cells in phases S and G2/M and fewer apoptotic cells after nocodazole treatment were detected by annexin V staining. Also in these cells, proteomic assays showed higher expression of cyclin D1 and E1 and microarray analysis revealed changes in many cell proliferation-related molecules, including, Zeb 2, Jun, NF-kβ, Cxcl9, Cxcl10, TNF, and the TNF receptor. Overall, our results indicate that the presence of AQP1 modifies the expression of key cell cycle proteins apparently related to increases in cell proliferation. This contributes to explaining the presence of AQP1 in many different tumors. 4 Introduction It has been extensively demonstrated that AQPs are widely distributed in nature and are present in practically all cell types in the human body (Carbrey and Agre, 2009;Takata et al, 2004). These proteins are involved in transcellular and transepithelial water movement, but more recent studies implicated them also in functions as migration and proliferation (Verkman, 2011). In fact abnormal overexpression of AQPs have been reported in several different types of tumors (Longatti et al, 2006;Oshio et al, 2005;Saadoun et al, 2002) and specific labeling of AQP1 has been shown in both tumor cells and proliferating tumor microvessels in many human cancers (Vacca et al, 2001;Lopez-Campos et al, 2011). A clear role for this protein in angiogenesis was confirmed in experiments in which melanoma-B16F10 cells were subcutaneously implanted in AQP1-null mice and there was observed to be considerably less tumor growth than in wild-type (Wt) animals, due to impaired microvessel formation and thus faster tumor necrosis (Saadoun et al, 2005). Similarly, lower migratory capacity was detected in melanoma cells that naturally do not express AQP1 compared to AQP1-expressing cells (Papadopoulos et al, 2008). Thus, there is already a consensus on the important role of AQP1 in angiogenesis and cell migration, both necessary processes for tumorigenesis, but less is known about the role of this protein in the cell proliferation process itself. In the cell cycle, cells normally increase their volume and complexity before dividing. A simplistic view of this process would be to suppose that overexpression of AQPs is only necessary to favor this increase in cell volume; however, the precise role of AQP1 during the cell cycle and hence, in the cell proliferation process, has not yet been explored in detail. A few authors have 5 described the involvement of some AQPs in cell cycle progression and cell death by apoptosis. For instance, expression of AQP2 in collecting duct cells has been associated with the rapid regulatory volume decrease that allows these cells to enter into apoptosis (Flamenco et al, 2009), and was also involved in the acceleration of cell proliferation per se, by increasing the rate of cell cycle progression in these renal cells (Di Giusto et al, 2012;Rivarola et al, 2010). A more recent study confirmed that AQP5 expression in esophageal squamous cell carcinoma (ESCC) may affect cell proliferation by modulating the expression of genes involved in cell cycle progression. In that study, authors demonstrated that the knockdown of AQP5 with siRNA inhibits cell cycle transition from the G1 to S phase (Shimizu et al, 2014), similar to pioneering research demonstrating that the level of AQP1 expression in A5 epithelial cells (Delporte et al, 1996) and AQP4 in astrocytes (Yoneda et al, 2001) was highest at the G0/G1-phase during cell cycle progression. Additionally, in previous work, we demonstrated that stable overexpression of AQP1, 3 and 5 in PC12 increases the cell proliferation rate in response to hypoxia (Galán-Cobo et al, 2013), and that specific inhibition of AQP3 with the gold-based compound Auphen (Martins et al, 2012;Serna et al, 2014) significantly reduced cell proliferation, by arresting cells in the S-G2/M phases (Serna et al, 2014). In this context, the aim of the present study was to explore the mechanisms by which AQP1 may affect the cell proliferation process and their molecular basis. For this, we used two experimental approaches: one, an animal model with deletion of the AQP1 gene, and the other, a cell line with stable overexpression of this protein. Our results show that proliferation of carotid body (CB) TH+ cells in response to hypoxia is lower in knock-out (KO) than in Wt mice. In addition, 6 cells with overexpression of AQP1 have a higher proliferative rate, higher resistance to synchronization by nocodazole and dramatically lower level of apoptosis than Wt cells. Western blot and Affymetrix assays confirm changes in gene expression of proteins relevant for cell cycle progression and cell volume was higher in AQP1-overexpressing cells. Consistently in both scenarios, loss and gain of AQP1 expression, our results demonstrate that AQP1 has an important role in cell proliferation and not only associated with cell volume regulatory activity, but also by affecting gene expression of proteins needed for cell cycle progression. Material and methods Animal assays Animals: C57Bl/6, knock-out (AQP1 -/-) mice kindly provided by Dr. W. Boron (Case Western Reserve University, Cleveland, USA) {Cooper, 2002 5 /id} and Wt littermate mice were housed at a controlled temperature (22±1°C) in a 12-h light/dark cycle with ad libitum access to food and water. Genotyping of KO animals was done as indicated previously {Ma, 1998 2 /id}. At all times, the mice were treated in accordance with the animal care guidelines of the European Directive on the protection of animals used for scientific purposes (86/609/EEC). All procedures used in the study were approved by the Animal Research Committee of the Virgen del Rocío University Hospital (University of Seville). In vivo hypoxic treatments and BrdU administration: When 2 to 3 months old, the mice were maintained either in normal conditions or chronically Eliminado: To maintain the colony mice were genotyped as Eliminado: in Eliminado: publication Eliminado: 7 exposed to a 10% O2 environment using a specially-designed hermeticallysealed chamber (Coy Laboratory Products, Inc., Grass Lake, MI) with gas concentrations, temperature and humidity conditions under continuous monitoring and control. To evaluate the CB cell proliferation induced by chronic hypoxia, mice were injected intraperitoneally with 50 mg/kg of BrdU (Sigma Aldrich, USA) every 7 days during the experiments and also supplemented with 1 mg/ml BrdU in their drinking water. Histological analyses of CB bifurcations: Following exposure to normoxia/hypoxia, animals were anesthetized with 100 mg/kg ketamine (Pfizer, New York, NY, USA) and 10 mg/kg xylazine (Bayer, Leverkusen, Germany), intracardially perfused with 4% paraformaldehyde (PFA; Sigma), and the CB bifurcations were dissected and post-fixed in PFA for 2 h. After overnight cryoprotection in 30% sucrose solution (Sigma), they were embedded in OCT compound (Sakura Finetek, Torrance, CA) and snap frozen by immersion in liquid nitrogen. Then, the bifurcations were cut into 10-μm thick sections using a Leica cryostat. To detect BrdU in these tissue sections, DNA was denatured in 2M HCl for 45 min at room temperature and neutralized with 0.1 M sodium borate. Sections were then pre-blocked for 1 h at room temperature in blocking solution (PBS containing 10% goat serum, 0.1% BSA, and 0.3% Triton X-100; Sigma). Primary rat anti-BrdU (1:100; Accurate Chemical, Westbury, NY) diluted in blocking solution was incubated overnight at 4°C, and then with fluorescein isothiocyanate (FITC)-conjugated anti-rat secondary antibodies (1:200; Jackson Immunoresearch, West Grove, PA) for 4 h at room temperature. Other antibodies used, following the same procedure as for BrdU except for the DNA denaturing/neutralizing steps, were primary rabbit anti- 8 mouse TH (1:1000; Pel-Freez, Roger, AR), followed by Alexa 568-conjugated goat anti-rabbit secondary antibodies (1:500; Molecular Probes, Eugene, OR). Finally, slides were counter-stained in 2.5 mg/ml DAPI (Sigma) for 10 min at room temperature, and then mounted in DAKO (Invitrogen, Carlsbad, CA) and preserved at 4°C. Cell culture Wild type mammalian PC12 cells (PC12-Wt) were cultured in DMEM medium (Invitrogen) with 10% horse serum (HS), 5% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37°C and 10% CO2. The medium was changed every 2-3 days with 1:5 dilution passages every 5 days (Galán-Cobo et al, 2013). Aquaporin-1 overexpressing PC12 cells (PC12-AQP1) were stably transfected with full-length cDNA of AQP1 for stable overexpression of this protein. Protocols to obtain this cell line have been described previously (Echevarría et al, 2007). These cells were cultured in the same medium as that used for the PC12-Wt cell line, namely, DMEM plus 2 mg/ml Geneticin, and the culture was maintained at 37°C with a 10% CO2 atmosphere, with medium replacements and dilution passages as described above. Proliferation assays Proliferation measurement by counting live cells in a hemocytometer chamber: PC12-Wt and PC12-AQP1 cell lines were plated in 60-mm diameter dishes (Nunc, Thermo Fisher, Waltham, MA) at initial concentrations of 0.9 x 106 cells and cultured for either: 0, 24, 48, 72 or 96 h. Cells were then 15 of mice to a hypoxic atmosphere induced a marked CB enlargement (~2- to 3- fold) caused by dilation and multiplication of blood vessels as well as expansion of the parenchyma (Figure 1-A and B), with an increased number of TH+ glomus cell clusters (Figure 1-C). After 12 and 21 days in hypoxia, we observed numerous new BrdU+-TH+ cells in all cases (Figure 1-D), but quantification of these cells revealed that there were fewer in the KO than the Wt animals, indicating less cell proliferation in the CB when mice lack AQP1. Likewise, quantifying the total number of BrdU+ cells (Figure 1-E), there was less overall CB cell proliferation in the KO mice at all time points, though differences were only statistically significant at 12 days of hypoxia. Looking closely the differences in number of cells in the two groups of mice after the hypoxia treatment, we found around 500 fewer new TH+ cells (Figure 1-D), and a total of ~1000 fewer BrdU+ cells (Figure 1-E) in the KO than the Wt mice. These results again indicate less proliferation in the absence of AQP1, and not only for the glomus cells of the KO mice, but also in other type of proliferative cells in the CB exposed to chronic hypoxia. Measurements of cell proliferation in a cell line with overexpression of AQP1 Cell proliferation rate was compared between the PC12-AQP1 and PC12-Wt cell lines using two methods: BrdU incorporation and cell counting. BrdU incorporation, as measure of new DNA synthesis, was assessed after a short exposure (2 h) to BrdU, and BrdU+ cells were quantified by flow cytometry. The results (Figure 2-A and 2-B) revealed a higher percentage of BrdU+ cells in 16 PC12-AQP1 (~10% higher) than PC12-Wt cells. Cell counting experiments after 24, 48 and 72 h of culture also found that there were more cells in the PC12- AQP1 than the Wt control line (Figure 2-C), although differences were only significant (p ≤ 0.01) after a seedtime of 72 h. Both results indicate that overexpression of AQP1 in the PC12 cell line tends to increase cell proliferation capacity. Cell cycle analysis To improve our understanding of the pro-proliferative effect of AQP1 overexpression, we investigated whether overexpression of this protein had an effect on the distribution of cells in the different cell cycle phases. The results shown in Figure 3 demonstrate that AQP1 overexpression produces marked changes in the cell cycle pattern, significantly increasing the number of cells in the more proliferative phases S and G2/M (p ≤ 0.001 and p ≤ 0.01, respectively; Figure 3-A), compared to the distribution in the PC12-Wt control cell line. Consistently, cells that overexpress AQP1 showed significantly fewer cells in the G0/G1 phase of the cell cycle (p ≤ 0.001; Figure 3-A). Additionally, in asynchronous cells, Western blot analysis of levels of cyclin D1 and E1, a key protein for cell cycle progression and a representative protein of the S phase respectively, revealed markedly high levels of both proteins (Figure 3-B) in the PC12-AQP1 cell line, contrasting with the low levels detected in PC12-Wt control cells. By other hand, not differences in the levels of cyclin B1, specific cyclin for the M phase, were detected among cells, indicating normal regulatory capacity for division and discarding cell arrest in the asynchronous cells. 17 Cell cycle analysis was also performed after incubation with the synchronization drug, sodium butyrate (Figure 3-C). Quantification revealed that PC12-AQP1 cells significantly resist being synchronized in the G1 phase after treatment with butyrate compared with the response in PC12-Wt controls, the number of cells in the G1 phase increasing by only ~5% in the case of PC12-AQP1 cells and by more than 10% in controls. Comparative analysis of the speed of the G1-S transition in the two cell lines was complicated to interpret due to the different start points in the synchronized cell cycle patterns after butyrate treatment (Figure 3-C). Cell cycle analysis with nocodazole We also studied the cell cycle after nocodazole treatment to synchronize cells in the mitotic prophase. The analysis showed clearly less synchronization in the G2/M phase for PC12-AQP1 than for PC12-Wt control cells (Figure 4-A and 4- B). Here, the most notable result was the emergence of an extra peak in the cell cycle pattern corresponding to a new population of cells called SubG1, which corresponds to dead cells. This new population also was smaller in the PC12- AQP1 than the PC12-Wt cell line (Figure 4-A). In more detail, the quantification of the percentage of cells at different phases in the cell cycle revealed that in PC12-Wt cells nocodazole induced (Figure 4-B) a high percentage of cells to remain in G2/M and S phases, with a consequent reduction in the number of cells in the G0/G1 phase, compared with that seen in untreated asynchronous cells (AS). Additionally, a considerable SubG1 population appeared, even at the lowest concentration of nocodazole used, and the percentage of cells in this 18 population rose with nocodazole concentration while the percentage of G2/M cells fell. Conversely, in PC12-AQP1 cells treated with nocodazole (Figure 4-B), even though there was also a higher percentage of cells in G2/M and SubG1 phases than in untreated cells (AS), the blockade in the G2/M phase is less strong than that seen in PC12-Wt control cells, and the dose-dependent effects in the SubG1 phase disappeared, the cell cycle pattern remaining the same even at the highest nocodazole dose (10 μM) (Figure 4-B). Hence, AQP1 overexpression seems to confer a resistance to cell death induction by nocodazole treatment. Apoptosis analysis To confirm the putative resistance to cell death with nocodazole treatment observed in the previous cell cycle analysis, we assessed staining with annexin V, a specific cell surface marker of apoptosis, in PC12-Wt and PC12-AQP1 cell lines. Cells were treated with 7.5 μM nocodazole for 24 h, stained with annexin V and PI, and then analyzed and quantified by flow cytometry (Figure 5-A). Quantification of different cell populations in the absence of treatment showed a high number of live cells for both cell lines (over 90%; annexin V- and PI-, Figure 5-B). In contrast, after nocodazole treatment, both cell lines showed an increase in apoptotic and late-apoptotic populations (annexin V+ and annexin V+ plus PI+ respectively), though this increase was significantly (p ≤ 0.001) smaller in the PC12-AQP1 cell line. Specifically, in the PC12-Wt control cell line, the live cell population fell to ~20% and numbers of both apoptotic and late apoptotic cells rose, to ~65% and ~15% respectively; while in PC12-AQP1, the numbers of live 19 cells only fell by around 15% and apoptotic and late apoptotic populations remained very small (~15% and ~5% respectively). Additionally, reduced induction of apoptosis was confirmed by Western blot analysis of PARP protein levels. PARP fractions were used as a readout for the cell apoptosis process. Results (Figure 5-C) revealed a significantly lower level of cleaved PARP after nocodazole treatment in PC12-AQP1 than in PC12-Wt. This confirms the idea that overexpression of AQP1 endows cells with considerable resistance to apoptosis induced by nocodazole treatment. Cell size and volume evaluation Since cell volume is an important factor related to cell cycle changes and progression, we analyzed whether volume and related cell characteristics, such as size and complexity, might be associated with the differences observed between PC12-Wt and PC12-AQP1 cell lines. Two different methods were used for this. First, we analyzed the relative cell size of cells by flow cytometry using the FSC parameter (see Material and methods). These experiments revealed a 40% larger cell size in the PC12-AQP1 cell line (Figure 6-A). Secondly, cell volume calculations were performed based on microscopy images of cell suspensions (Figure 6-B and 6-C). In agreement with flow cytometry results, quantification again showed a 40% larger volume in PC12-AQP1 than PC12-Wt control cells, the values being 4500 and 3000 μm3, respectively. The cellular complexity was evaluated from the SSC flow cytometry parameter (Supplementary Figure S1), 20 and we confirmed a higher intracellular complexity inside PC12-AQP1 cells. Taken together, these results strengthen the idea that AQP1 overexpression changes cell morphology facilitating cell progression through the cell cycle. Gene expression profiling in PC12-AQP1 cells To look for genes possibly involved in the molecular mechanisms by which AQP1 participates in cell proliferation, we generated microarray gene expression profiles for each cell type and analyzed the results using bioinformatic approaches. The microarray analysis showed 824 genes in PC12- AQP1 cells with fold changes ≥ 2. Of these genes, 364 were up-regulated, and 460 were down-regulated (Figure 7-A). Table 2 shows genes associated with cell proliferation-related process that had the most strongly up- or downregulated expression levels. Among the 824 genes with altered expression in the AQP1-overexpressing cells, many had cell proliferation-related functions (cellular growth and proliferation, 183 genes; cell morphology, 151 genes; cellular movement, 127 genes; cellular assembly and organization, 118 genes; and cellular function and maintenance, 108 genes) (Table 3). Ingenuity Pathway Analysis showed that the top-ranked associated network functions related to AQP1 overexpression were Cell-To-Cell Signaling and Interaction, Tissue Development, Cellular Movement with a score of 42 (Table 4). These results indicate that AQP1 overexpression influences cell growth and proliferation genes. Levels of expression of the most relevant genes modified in the Affymetrix analysis and related to proliferation process were validated by qPCR 21 and their fold changes are shown in Figure 7-B. The results obtained confirm changes previously seen in Affymetrix analysis of this set of genes. Discussion While several studies have confirmed the participation of AQPs in angiogenesis and cell migration, and tentative mechanistic models have been proposed for its role in this process (Saadoun et al, 2005;Stroka et al, 2014), less is known about the function these proteins may have in cell proliferation and cell cycle progression. Here we present for the first time clear evidence, from in vivo and in vitro experiments, that expression of this protein in both situations, loss and gain of AQP1 expression, directly affects cell proliferation by altering key proteins essential for cell cycle progression. The adrenal medulla, from which pheochromocytoma PC12 cells are derived, and the CB are two paraneural organs with the same sympathoadrenal lineage, that have certain characteristics in common, such as their dopaminergic nature and the cluster arrangement of their chemosensory cells (García-Fernández et al, 2007;López-Barneo et al, 2001;Muñoz-Cabello et al, 2003). The CB is an oxygen-sensing organ that grows under chronic hypoxia (Arias-Stella and Valcarcel, 1976;Edwards et al, 1971;McGregor et al, 1984) and that strongly expresses AQP1 (Muñoz-Cabello et al, 2010). Pardal et al. (Pardal et al, 2007), exploring how hypoxia induces cell proliferation in the adult mammalian CB, developed a useful experimental model that prompted us to explore in AQP1- KO mice how absence of this constitutive protein affects CB cell proliferation in vivo. 22 Quantification of the total number of new BrdU+ cells indicates, in general, less proliferation in the absence of AQP1, and the smaller number of new BrdU+- TH+ cells measured after both 12 and 21 days of hypoxic stimuli evidenced specific impairment in the proliferation of CB glomus cells. This reduced proliferation of different CB cell types could, in part, be due to impaired angiogenesis and, in turn, to reduced cell access to or expression/stability of factors necessary for tissue growth such as, for instance, HIF (Galán-Cobo et al, 2013); as well as to a general impairment of the self-renewal process of CB TH+ cells from glia-like sustentacular cells (Pardal et al, 2007). It was previously demonstrated that AQP1 is present in type II CB cells (Muñoz-Cabello et al, 2010), now known to be CB stem cells that form the neurogenic niche for this organ (Pardal et al, 2007). Previous studies have also indicated involvement of AQP1 in the differentiation of stem-like cells in rat bile duct formation (Jia et al, 2011), development of human corneal keratocytes (Chan et al, 2013) and renewal of limbal basal epithelium from a corneal epithelial stem cell niche (Higa et al, 2013). Taking all this into account, it seems plausible that the presence of AQP1 in the CB favors differentiation of glia-like stem cells elicited by hypoxia, to give new BrdU+-TH+ cells, concomitant to the observed hypertrophy of this organ. There is a clear need for more experiments to specifically investigate the participation of AQP1 in cell differentiation in this adult neurogenic center, and it would also be interesting to explore its role in other known neurogenic niches (Calof et al, 1998;Doetsch et al, 1999). To gain insight into the molecular mechanism by which AQP1 participates in the proliferation process, we moved to a cell model, namely, PC12 cells, comparing cell proliferation rate and cell cycle distribution in Wt and AQP1-overexpressing 23 cells. Data presented here confirm that cells that overexpress AQP1 have a higher rate of DNA synthesis and grow faster than cells with no expression of this protein, reflecting a higher proliferative capacity. Furthermore, in asynchronized cells with AQP1 overexpression, we observed an increase in the percentage of cells in the most proliferative phases, S and G2/M, which, together with the larger total number of cells, suggests a more proliferative phenotype. Further, these cells (PC12-AQP1) also had higher levels of cyclins D1 and E1, two cell cycle regulatory proteins crucial for phase transition during cell cycle progression (Dulic et al, 1992;Harper and Brooks, 2004;Koff et al, 1992). Analysis of the cell cycle in the presence of butyrate, which synchronizes cells in G1 by inhibiting cyclin E activity (Abramova et al, 2006), showed that cells that overexpressed AQP1 are more resistant to being synchronized (see Figure 3). A higher expression of cyclins might well explain this phenotype, and this contributes to understanding the greater proliferative capacity of these cells. Interestingly, adding to the findings of the few studies that have explored the role AQPs might play over the course of the cell cycle (Delporte et al, 1996;Di Giusto et al, 2012;Rivarola et al, 2010;Shimizu et al, 2014;Yoneda et al, 2001), we have learned that AQP expression normally rises while cells are in the G0/G1 phase (Delporte et al, 1996;Yoneda et al, 2001), probably by favoring an increase in cell volume. These proteins become important for cells to move into the S and G2/M phases and allow cell cycle progression. Data from our morphological analysis of these cell lines show that cells with AQP1 overexpression have a constitutive larger size and cell volume that may in some way help to favor cell cycle progression. It is known that there is important functional coupling between cell volume, ion channels and 24 transporters over the course of the cell cycle (Di Giusto et al, 2012;Rivarola et al, 2010;Shimizu et al, 2014). Specifically, disturbances in cell osmotic balance and membrane potential have been associated with alterations in cell cycle and proliferation rate (Pedersen et al, 2013;Yang and Brackenbury, 2013); and this might help explain the participation of AQP1 in many different tumor types in humans (Oshio et al, 2005;Papadopoulos et al, 2008;Saadoun et al, 2002;Shimizu et al, 2014). An increment in the synthesis of DNA and others constitutive elements for cell division and proliferation (Evan and Vousden, 2001) would also need larger cell volume and therefore would also contribute to explain the functional meaning of AQP1 overexpression. Additionally, other physiological processes that require rapid acceleration of the cell cycle such as angiogenesis (Saadoun et al, 2002;Vacca et al, 2001), tissue differentiation during embryogenesis (Kim, 1999;Yamamoto, 1997), and stem cell niche differentiation (Higa et al, 2013;Jia et al, 2011), would also depend on the expression of this protein. Our data also show that overexpression of AQP1 made cells resistant to apoptosis induced by nocodazole (Beswick et al, 2006). In these cells, the subG1 population of apoptotic cells remains very small, contrasting with what is observed in Wt cells. Although the mechanism is unknown, similar results have been described with tumor cell lines that became resistant to drugs used for synchronization experiments (Jordan et al, 1993;Rieder and Maiato, 2004;Weaver and Cleveland, 2005). In this context, a term called slippage has been coined to explain a process in which cells are released from arrest entering in G1, without going through cytokinesis. In our AQP1-overexpressing cells, 24 h after nocodazole release, we did never see aneuploidy problems but 31 Saadoun S, Papadopoulos MC, Hara-Chikuma M, Verkman AS. 2005. Impairment of angiogenesis and cell migration by targeted aquaporin-1 gene disruption. Nature 434:786-792. Serna A, Galán-Cobo A, Rodrigues C, Sánchez-Gomar I, Toledo-Aral JJ, Moura TF, Casini A, Soveral G, Echevarría M. 2014. Functional Inhibition of Aquaporin-3 With a Gold-Based Compound Induces Blockage of Cell Proliferation. J Cell Physiol 229:1787-1801. Shimizu H, Shiozaki A, Ichikawa D, Fujiwara H, Konishi H, Ishii H, Komatsu S, Kubota T, Okamoto K, Kishimoto M, Otsuji E. 2014. The expression and role of Aquaporin 5 in esophageal squamous cell carcinoma. J Gastroenterol 49:655- 666. Stroka KM, Jiang H, Chen SH, Tong Z, Wirtz D, Sun SX, Konstantopoulos K. 2014. Water Permeation Drives Tumor Cell Migration in Confined Microenvironments. Cell 157:611-623. Sun W, Tan X, Shi Y, Xu G, Mao R, Gu X, Fan Y, Yu Y, Burlingame S, Zhang H, Rednam SP, Lu X, Zhang T, Fu S, Cao G, Qin J, Yang J. 2010. USP11 negatively regulates TNFa-induced NF-kb activation by targeting on IkBa. Cellular Signal 22:386-394. Takata K, Matsuzaki T, Tajika Y. 2004. Aquaporins: water channel proteins of the cell membrane. Prog Histochem Cyto 39:1-83. Vacca A, Frigeri A, Ribatti D, Nicchia GP, Nico B, Ria R, Svelto M, Dammacco F. 2001. Microvessel overexpression of aquaporin 1 parallels bone marrow angiogenesis in patients with active multiple myeloma. Brit J Haematol 113:415- 421. Verkman AS. 2011. Aquaporins at a glance. J Cell Sci 124:2107-2112. Weaver BAA, Cleveland DW. 2005. Decoding the links between mitosis, cancer, and chemotherapy: The mitotic checkpoint, adaptation, and cell death. Cancer Cell 8:7-12. Yamaguchi T, Kimura J, Miki Y, Yoshida K. 2007. The Deubiquitinating Enzyme USP11 Controls an Ikb Kinase alpha (IKKa)-p53 Signaling Pathway in Response to Tumor Necrosis Factor alpha (TNFa). Journal of Biological Chemistry 282:33943-33948. Yamamoto TS. 1997. Expression of AQP family in rat kidneys during development and maturation. Am J Physiol-Renal 272:F198-F204. Yang M, Brackenbury WJ. 2013. Membrane potential and cancer progression. Front. Physiol 4. Yoneda K, Yamamoto N, Asai K, Sobue K, Fujita Y, Fujita M, Mase M, Yamada K, Nakanishi M, Tada T, Miura Y, Kato T. 2001. Regulation of aquaporin-4 expression in astrocytes. Mol Brain 89:94-102. 32 33 Figure Legends Figure 1. Cell proliferation induced by chronic hypoxia in the carotid body of AQP1-KO and Wt mice. Immunocytochemical analysis of carotid body bifurcations removed from control (Wt) and KO (A) mice exposed to normoxia (Nx) and chronic hypoxia (10% O2). The typical growth of the TH+ glomus cell mass, highlighted with a white dashed line, 12 (Hy 12d) and 21 (Hy 21d) days of hypoxia is shown. The number of cells was quantified in a cell focal plane, such as the example shown in detail at a larger magnification (B). Quantification of the number of TH+ cells (C), TH+ plus BrdU+ cells (D) and BrdU+ cells (E) in the KO and Wt mice is shown for normoxia after 12 and 21 days, and after 12 and 21 days of hypoxia. Data for 0 days in hypoxia corresponds with the average of values obtained in normoxia. Significant differences are indicated as follows: * p ≤ 0.05 and ** p ≤ 0.01. Bars are mean ± SEM from 6-8 experiments. Figure 2. Analysis of cell proliferation rates in cultured PC12 cells. Twodimensional flow cytometry after BrdU incorporation was performed in PC12- AQP1 and PC12-Wt cells. Cells were incubated for 2 h with 10 µM BrdU and a dot plot of the flow cytometry data is shown for the incorporation of BrdU after treatment (A). The rate of BrdU incorporation after 2 h of treatment was significantly higher in the PC12-AQP1 cells (B). Further, cell proliferation was also assessed by hemocytometer counting at 24, 48 and 72 h after seedtime (C). Numbers of live cells not stained with trypan blue were counted in a hemocytometer chamber. Significantly higher cell proliferation (** p ≤ 0.01) was observed in PC12-AQP1 cell line after 72 h of culture. Bars are mean ± SEM from N experiments where N=4 in (B) and N=6-7 in (C). 34 Figure 3. Analysis by flow cytometry of the cell cycle profiles of PC12-Wt and PC12-AQP1 cell lines. The distribution of cells in different phases was analyzed by flow cytometry and propidium iodide-derived cell cycle profiles show a larger percentage of cells in the S and G2/M phases in PC12-AQP1 cells. A summary of the % of cells in each phase is shown in the upper right corner of each plot (A). Representative Western blots of protein levels (upper panel) and summary of the quantification analysis (lower panel) (B). Cell cycle analysis performed after synchronization in G1 phase with sodium butyrate treatment (C). PC12-Wt and PC12-AQP1 cells at 80% of confluence were cultured with sodium butyrate (6 mM, 24 h), and the cell cycle was analyzed before (0 h) and after removal of butyrate and released into the cell cycle for 10, 18, 24 and 30 h. Significant differences are indicated as follows: * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001. Bars are mean ± SEM from N experiments where N=11 in (A), N=4 in (B) and N=5 in (C). Figure 4. Cell cycle analysis performed after nocodazole treatment. PC12- Wt and PC12-AQP1 cells at 80% of confluence were cultured without (AS) or with nocodazole (1.25, 2.5, 5, 7.5 and 10 µM, 24 h) and the distribution of cells in different phases was analyzed by flow cytometry. A representative example of the cell cycle pattern after treatment with 7.5 µM of nocodazole is shown for both cell lines (A). Summary of the percentage of cells counted in each phase of the cell cycle with different concentrations of the drug (B). Significant differences are indicated as follows: * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001. Bars are mean ± SEM from 4-6 experiments. Figure 5. Induction of cell apoptosis by nocodazole treatment. Cells at 80% of confluence were cultured without (AS) or with nocodazole (7.5µM, 24 h) and 35 the distribution of cells in different phases was analyzed by flow cytometry. Dot plot for the evaluation of apoptosis induction by nocodazole by flow cytometry analysis of annexin V/PI-stained cells (A). Summary of the percentage of cells annexin V-/PI- (alive), annexin V+/PI- (apoptotic), annexin V+/PI+ (late apoptotic or dead) and annexin V-/PI+ (necrotic) (B). Representative Western blots of levels of full PARP and cleaved PARP protein (upper panel) and summary of the quantification analysis (C). Error bars are mean ± SEM from N experiments where N=4 in (B) and N= 4 in (C). Significant differences are indicated as follows: *p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001. Figure 6. Evaluation of relative cell size and volume. Summary of relative cell size measurements evaluated by flow cytometry using the FSC parameter in PC12-Wt and PC12-AQP1 cell lines (A). Summary of the volume quantification from cell area evaluated from microscopy images (B). Light microscope images of PC12-Wt and PC12-AQP1 cells in suspension to measure cell area (C). Bars in panel A and B are mean ± SEM from N experiments where N=6 in A and N=221-242 in B, and significant differences are indicated with *** p ≤ 0.001. Figure 7. Analysis of gene expression levels. Volcano plot of microarray data analyzed using Affymetrix technology (A). The fold changes between the two groups compared (PC12-AQP1 vs PC12-Wt cells) are represented on the x-axis and the log P-value calculated using the Transcriptome Analysis Console software on the y-axis. In the plot, each gene is represented by a point and up- (red) and down- (green) regulated genes appear symmetrically distributed around a fold change of 1. The expression level of the AQP1 gene in the analysis is highlighted in blue. The most relevant genes modified in the 36 Affymetrix analysis and related to proliferation process were validated by qPCR and their fold changes are shown (B). Name and biological function of these genes appear described in Table 2. All data shown were statistically significant, p ≤ 0.05. Bars are mean ± SEM from 3 experiments. Supplementary Figure 1 (S1). Evaluation of relative cell complexity. Summary of relative cell complexity evaluated by flow cytometry using the SSC parameter in PC12-Wt and PC12-AQP1 cell lines. Bars are mean ± SEM from 6 experiments. Table 1: Primer sequences used for the validation by qPCR amplification of different genes. Gene Primer Forward Primer Reverse CADM1 S: 5’-CCAGCAGTTCACGGCCT A-3’ AS: 5’-CGAGAATCTGAGAGATCCTCACTGT-3’ CDC14B S: 5’-GCGGTGCATTGCAAAGCT-3’ AS: 5’-TGCTCTCGGCTGCTGTCAT-3’ CDH22 S: 5’-ACAGTGACCATCGTGGTTACTGA-3’ AS: 5’-TCCTGTATGCTGAACTGGTACATCTT-3’ CXCL9 S: 5’-AATCCCTCAAAGACCTCAAACAGT-3’ AS: 5’-TCTTCAGTGTAGCGATGATTTCAGT-3’ CXCL10 S: 5’-CTGTCCGCATGTTGAGATCATT-3’ AS: 5’-ATGGCCTCAGATTCCGGATT-3’ IGFBP5 S: 5’-AGAAAGCAGTGCAAGCCTTCTC-3’ AS: 5’-GGCAGCTTCATCCCATACTTGT-3’ IL6R S: 5’-CCCAGAGGAGCCCAAGCT-3’ AS: 5’-GCTTGGATGCCACTCACAAA-3’ JUN S: 5’-TGATCATCCAGTCCAGCAATG-3’ AS: 5’-TGCTCGTCGGTCACGTTCT-3’ JUNB S: 5’-GACACAGGCGCATCTCTGAAG-3’ AS: 5’-GATCACGCCGTTGCTGTTG-3’ MMP19 S: 5’-CGTGGAATCAGTCTTGGATGTTC-3’ AS: 5’-CAAGGTGACATTAGCTGAGGATGA-3’ NFKB2 S: 5’-GACGTTCATAAACAGTATGCCATTG-3’ AS: 5’-ACACCGTTACAGGCCTCTCAA-3’ S100A11 S: 5’- GCCATCCCAAATGGAGCAT-3’ AS: 5’-CTCAGGTCCTCCTTTGTCA AGTAGT-3’ S100A10 S: 5’-GCCATCCCAAATGGAGCAT-3’ AS: 5’-CTCAGGTCCTCCTTTGTCA AGTAGT-3’ SCRT S: 5’-CCAGTGGCACCTTCAAGAGTTC-3’ AS: 5’-TTGCTCAGTGCTAGCCTTGGT-3’ USP11 S: 5’-TGGAGACCCGAAACAAAGATG-3’ AS: 5’-CAGATTGGTAAGGCCACA GATG-3’ ZEB2 S: 5’-GAAAGCTTCCCAGGTCCTATCC-3’ AS: 5’-GGACCGCCTTGATCTCTTCA-3’ AQP1 S: 5’-CCATTGACCAC GGCATAGTACA -3’ AS: 5’-AGTGTCCT ACCTGTGAAGTGAGTA-3’ CICLO S: 5’-GCACTGGTGGCAAGTCCAT-3’ AS: 5’-GCCAGGACCTGTATGCTTCAG-3’ 37 Symbol Name Fold change Biological function CXCL9 Chemokine (C-X-C motif) ligand 9 38.520 Cytokine that affects the growth, movement, or activation state of cells that participate in immune and inflammatory response. Chemotactic for activated T-cells. Bin ds to C XC R 3. CXCL10 Chemokine (C-X-C motif) ligand 10 21.020 Chemotactic for monocytes and T-ly mphocy tes. Binds to CXCR3. TNFSF18 Tumor necrosis factor (ligand) superfamily, member 18 16.750 Cytokine that binds to TNFRSF18/AITR/GITR. Regulates T-cell responses. Can fun c tio n a s c os timula tor an d lo w er the th r es hold for T-cell activation and T-cell proliferation. Important for interactions between activated T-ly mphoc ytes and endothelial cells. Mediates activation of NF-kappa-B. ZEB2 Zinc finger E-box binding homeobox 2 8.440 Trans c riptional inhibitor that binds to DNA sequence 5'-CACCT-3' in different promoters. Represses transcription of E-cadherin. Prl8a2 Pro la c tin family 8 6.500 Role in cell survivor and apoptosis suppression. FGF16 Fibroblast growth factor 16 3.410 FGF family membe rs pos s es s br oa d mito ge nic and cell survival activities, and are involved in a variety of biologic al processes, inc luding embry onic development, c e ll growth, morphogenesis, tissue repair, tumor growth and invasion. CDC42E P2 CDC42 effector protein (Rho GTPase binding) 2 2.940 Pro b ably involved in the organization of the actin cytoskeleton. May act downstream of CDC42 to induce actin filament assembly leading to cell shape changes. Induces pseudopodia formation in fibroblasts in a CDC42-dependent manner. IL6 R Inte r le uk in 6 Re c e p tor 2.460 The IL6 rec eptor is a protein c omplex c onsisting of this protein and interleukin 6 signal transducer (IL6ST/GP130/IL6-beta), a receptor subunit also shared by man y other cytokines. Dysregulated production of IL6 and this receptor are implicated in the pathogenesis of many diseases, such as multiple myeloma, autoimmune diseases and prostate cancer. MMP19 Matrix metalloproteinas e 19 2.310 Endop e ptidas e that deg ra des va r ious c ompone nts of the ex tr ac ellular ma trix . NFKB2 Subun its of the tran s c r iptio n fac tor complex nuclear factor-kappa-B (p49/p100) 2.250 This gene enc odes one of the s ubunits of the transcription factor complex nuclear fa c to r -kappa-B ( N Fk B). The NFkB transcription factor complex is expressed in numerous cell types and functions as a centr al ac tiv ato r of ge ne s in vo lv ed in inflammation and immune func tion. JUN Jun proto-oncogen 2.130 JUN (jun proto-oncogene) is a protein-coding gene. Diseases associated with JUN include skeletal muscle cancer, and sarcoma, an d among its related superpathways. Development HGF signaling pathway and TNFR1 Pathway. BCL2L1 BC L 2-lik e 1 2.070 Potent in h ibitor of c ell death. Inhibits activation of caspases (By similarity). Appears to r eg ula te c ell d ea th b y b loc k in g th e v olta ge -dependent anion channel (VDAC) b y binding to it and preventing the release of the caspase activator, CYC1, from the mitoc hondrial membrane. Als o acts as a regulator of G2 checkpoint and progression to c y tok in es is dur in g mito s is . JUNB Jun B proto-oncogen 2.030 JUNB (jun B proto-oncogene) is a protein-coding gene. USP13 Ubiquitin specific peptidase 13 (isopeptidase T-3) -2.000 Deubiquitinates USP10, an essential regulator of p53/TP53 stability. CADM1 Cell adhesion molecula 1 -3.550 Mediates homophilic cell-cell adhesion in a Ca(2+) independent manner. Acts a s a tumor s up pr es s or in n on -sma l l-ce ll lu ng c a n cer ( NSC L C ) c e lls . ADAM23 AD AM metallopeptidase domain 23 -3.580 May play a role in c ell-cell and cell-matrix interac tions . This is a non-catalytic metalloprotease-like Protein. USP11 Ubiquitin specific peptidase 11 -3.620 Pro te a s e that c an r emov e c onju g ated ubiq u itin fr om tar get pr oteins and polyub iquitin chains. Inhibits the degradation of target proteins by the proteasome. Plays a role in the regulation of pathways leading to NF-kappa-B ac tiv ation. Pla y s a role in the regulation of DNA repair after double-stranded DNA breaks. S100A11 S100 C alc ium Bin ding Pr otein A11 -5.010 The protein enc oded by this gene is a member of the S100 family of proteins containing 2 EF-hand calcium-binding motifs. S100 proteins are loc aliz ed in the cytoplasm and/or nucleus of a wide range of cells, and involved in the regulation of a number of cellular processes such as cell cycle progression and differentiation. This protein may function in motility, inv as ion, and tubulin polymerization. Chromos omal rearrangements and altered expression of this gene have been implicated in tumo r metastas is . ADAM22 AD AM me tallo peptida s e domain 22 -5.370 Pro b able ligand for integr in in the bra in. Th is is a non-catalytic metalloprotease-like protein. Involved in regulation of cell adhes ion and spreading and in inhibition of ce ll proliferation. OCLN Occludin -6.050 May play a role in the formation and regulation of the tight junction (TJ ) parac ellular permeability barrier. It is able to induce adhesion when expressed in cells lacking tigh t jun c tio ns . S100A10 S100 C alcium binding protein A10 -24.940 The protein enc oded by this gene is a member of the S100 family of proteins containing 2 EF-hand calcium-binding motifs. S100 proteins are localized in the cytoplasm and/or nucleus of a wide range of cells, and involved in the regulation of a number of cellular processes such as cell cycle progression and differentiation. SC TR Sec retin rec eptor -28.300 The protein encoded by this gene is a G protein-coupled receptor and belongs to the glucagon-VIP-secretin receptor family. It binds secretin which is the most potent regulator of pancreatic bicarbonate, electrolyte and volume secretion. Secretin an d its rec eptor are suggested to be inv olved in pancreatic canc er and autism. Table 2: Most relevant genes that displayed the greatest changes in their expressions in the comparative assay performed in PC12-AQP1 vs PC12-wt cells. 38 Molecular and Cellular Functions p-value Molecules Cellular Growth and Proliferation 1.30x 10-6 -8.14x10-3 183 Cell Morphology 1.73x 10-9 -8.17x10-3 151 Cellular Movement 8.50x 10-7 -8.17x10-3 127 Cellular Function and Maintenance 1.73x 10-9 -8.17x10-3 118 Cellular Assembly and Organization 1.73x 10-9 -7.30x10-3 108 Table 3: Top-ranked molecular and cellular functions related to AQP1 overexpression according to Ingenuity Pathway Analysis (IPA) obtained after Affymetrix analysis. Associated Network Functions Score Cell-To-Cell Signaling and Interaction, Tissue Development, Cellular Movement 42 Neurological Disease, Cardiovascular System Development and Function, Embryonic Development 39 Reproductive System Disease, Post-Translational Modification, Tissue Morphology 39 Neurological Disease, Tissue Morphology, Cellular Movement 33 Humoral Immune Response, Protein Synthesis, Cell-To-Cell Signaling and Interaction 31 Table 4: Top-ranked signaling network functions related to AQP1 overexpression according to Ingenuity Pathway Analysis (IPA) obtained after Affymetrix analysis. 39 Figure 1. 40 Figure 2