Fluoxetine suppresses calcium signaling in human T lymphocytes through depletion of intracellular calcium stores

Accepted Manuscript Title: Fluoxetine suppresses calcium signaling in human T lymphocytes through depletion of intracellular calcium stores Author: V. Gobin M. De Bock B.J.G. Broeckx M. Kiselinova W. De Spiegelaere L. Vandekerckhove K. Van Steendam L. Leybaert D. Deforce PII: DOI: Reference: S0143-4160(15)00098-6 http://dx.doi.org/doi:10.1016/j.ceca.2015.06.003 YCECA 1682 To appear in: Cell Calcium Received date: Revised date: Accepted date: 6-1-2015 1-6-2015 6-6-2015 Please cite this article as: V. Gobin, M. De Bock, B.J.G. Broeckx, M. Kiselinova, W. De Spiegelaere, L. Vandekerckhove, K. Van Steendam, L. Leybaert, D. Deforce, Fluoxetine suppresses calcium signaling in human T lymphocytes through depletion of intracellular calcium stores, Cell Calcium (2015), http://dx.doi.org/10.1016/j.ceca.2015.06.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. an us c ri *Graphical Abstract M Antigen TCR ed PLCγ nucleus ce Ca2+ Ac Ca2+ PIP2 pt IL2 IP3 + DAG Ca2+ IP3R Ca2+ Ca2+ caffeine RyR ER fluoxetine SERCA Ca2+ TG Page 1 of 34 *Highlights (for review) ep t ed M an us cr ip t Fluoxetine suppresses the rise in [Ca2+]i after T cell receptor activation The effect is due to depletion of intracellular Ca2+ stores No effect of fluoxetine was observed on capacitative calcium entry Fluoxetine inhibits Ca2+ release in response to inositol trisphosphate or caffeine Inhibition of Ca2+ signaling is likely at the basis of fluoxetine-induced immunosuppression Ac c      Page 2 of 34 *Manuscript 1 Fluoxetine suppresses calcium signaling in human T lymphocytes through depletion of intracellular 2 calcium stores 3 V. Gobin1, M. De Bock2, B. J.G. Broeckx1, M. Kiselinova3, W. De Spiegelaere3, L. Vandekerckhove3 , K. 4 Van Steendam1, L. Leybaert2,*, D. Deforce1,* us cr ip t 5 6 7 1 8 2 9 3 Department of Internal Medicine, Ghent University, Ghent, Belgium *these authors contributed equally. ed 11 15 16 Ac c 14 ep t 12 13 an Physiology Group, Department of Basic Medical Sciences, Ghent University, Ghent, Belgium M 10 Laboratory of Pharmaceutical Biotechnology, Ghent University, Ghent, Belgium 17 Corresponding author: 18 Dieter Deforce, Laboratory of Pharmaceutical Biotechnology, Faculty of Pharmaceutical Sciences, 19 Ghent University, Ottergemsesteenweg 460, 9000 Ghent, Belgium, e-mail: Dieter.Deforce@Ugent.be, 20 Ph: +32 (0)9 264 80 67, Fax: +32 (0)9 220 66 88. 1 Page 3 of 34 Abstract 2 Selective serotonin reuptake inhibitors, such as fluoxetine, have recently been shown to exert anti- 3 inflammatory and immunosuppressive effects. Although the effects on cytokine secretion, 4 proliferation and viability of T lymphocytes have been extensively characterized, little is known about 5 the mechanism behind these effects. It is well known that Ca2+ signaling is an important step in the 6 signaling transduction pathway following T cell receptor activation. Therefore, we investigated if 7 fluoxetine interferes with Ca2+ signaling in Jurkat T lymphocytes. Fluoxetine was found to suppress 8 Ca2+ signaling in response to T cell receptor activation. Moreover, fluoxetine was found to deplete 9 intracellular Ca2+ stores, thereby leaving less Ca2+ available for release upon IP3- and ryanodine- 10 receptor activation. The Ca2+-modifying effects of fluoxetine are not related to its capability to block 11 the serotonin transporter, as even a large excess of 5HT did not abolish the effects. In conclusion, 12 these data show that fluoxetine decreases IP3- and ryanodine-receptor mediated Ca2+ release in 13 Jurkat T lymphocytes, an effect likely to be at the basis of the observed immunosuppression. ed M an us cr ip t 1 14 Ac c 16 ep t 15 2 Page 4 of 34 1 1. Introduction Selective serotonin reuptake inhibitors (SSRIs) have been shown to exert anti-inflammatory and 3 direct immunosuppressive effects such as suppression of T cell activation, cytokine secretion and 4 proliferation and induction of apoptosis in vitro and in vivo [1-3]. Although it has been shown that 5 these compounds have a high affinity for the serotonin transporter (SERT) in the central nervous 6 system, it is not clear whether the immunological effects of SSRIs are mediated by inhibition of SERT- 7 mediated serotonin (5HT) uptake in lymphocytes. On the contrary, several arguments oppose to the 8 involvement of 5HT and SERT in the immunosuppressive effects of SSRIs, especially the discrepancy 9 between the concentration needed for blockage of 5HT uptake on the one hand (nM range) and for 10 in vitro immunosuppression on the other hand (µM range) [4, 5]. The actual mechanism underlying 11 the immunosuppressive effects of SSRIs has not been elucidated yet. 12 Elevation of the cytoplasmic free Ca2+ concentration ([Ca2+]i) is one of the key triggering signals for T- 13 cell activation. The [Ca2+]i is regulated through an intimate interplay between Ca2+ in the extracellular 14 space and intracellular storage sites such as the endoplasmic reticulum (ER). Ca2+ signaling 15 mechanisms mostly rely on Ca2+ release from the ER through inositol 1,4,5 trisphosphate receptors 16 (IP3R) and ryanodine receptors (RyR) following the activation of G-protein coupled receptors on the 17 plasma membrane. Subsequent depletion of the ER triggers store-operated, capacitative Ca2+ entry 18 to replenish the ER [6]. SSRIs have been shown to affect Ca2+ signaling in several cell types. Fluoxetine 19 inhibited ATP-induced Ca2+ increases in PC12 cells through inhibition of both influx of extracellular 20 Ca2+ and release of Ca2+ from intracellular stores [7]. Whereas fluoxetine has also been shown to 21 suppress Ca2+ spikes in cultured rat hippocampal neurons, two other SSRIs, namely paroxetine and 22 citalopram, did not [8]. Furthermore, chronic exposure of astrocytes to fluoxetine diminished RyR- 23 and IP3R-mediated Ca2+ release as well as the subsequent capacitative Ca2+ entry [9]. In microglia, 24 pretreatment with paroxetine or sertraline reduced the amplitude of the Ca2+ increase induced by 25 interferon-gamma (IFNγ) [10]. Oppositely, sertraline induced a Ca2+ rise in MG63 osteosarcoma cells Ac c ep t ed M an us cr ip t 2 3 Page 5 of 34 [11]. Fluoxetine, paroxetine and citalopram induced a rise in [Ca2+]i in Burkitt lymphoma cells [12]. In 2 platelets, SSRIs (sertraline, paroxetine, fluoxetine) potentiated thrombin-mediated increases in 3 intracellular Ca2+ [13]. Clearly, SSRIs are capable of interfering with Ca2+ signaling in a wide variety of 4 cell types. Furthermore, it has been suggested that fluoxetine interferes with mitogen-induced Ca2+ 5 influx in murine and human T lymphocytes as fluoxetine exerted similar effects as the Ca2+ ionophore 6 A23187 on T cell proliferation, protein kinase C (PKC) degradation and cAMP levels [14, 15]. 7 Given the importance of Ca2+ signaling in T cell activation [6], we investigated whether interference 8 with Ca2+ signaling might be at the basis of the immunosuppressive effects of fluoxetine in Jurkat T 9 lymphocytes. In addition, we investigated whether the observed effects on Ca2+ signaling are related us cr ip to the inhibition of 5HT uptake. 2. Experimental procedures 12 2.1. Cell culture M 11 an 10 t 1 Jurkat T cells, clone E6-1, were cultured at 37°C and 5% CO2 in RPMI supplemented with 10% heat- 14 inactivated fetal bovine serum, 1% glutamine and 1% penicillin/streptomycin (100 U/ml penicillin G; 15 100 µg/ml streptomycin). All cell culture reagents were purchased from Life technologies (Carlsbad, 16 CA, USA). ep t 2.2. Buffers and chemicals Ac c 17 ed 13 18 Krebs HEPES buffer contained 133.5 mM NaCl, 5.9 mM KCl, 1.2 mM MgCl2, 11.6 mM HEPES, 11.5 mM 19 glucose and 1.5 mM CaCl2, pH 7,4. In Ca2+-free Krebs buffer, CaCl2 was replaced by 4.5 mM EGTA (pH 20 7.4) preserving equimolarity with the Krebs HEPES buffer. No external 5HT was added to the buffer. 21 1,2-Bis(2-aminophenoxy) ethane-N,N,N′,N′-tetraacetic acid tetrakis-acetoxymethyl ester (BAPTA- 22 AM), D-myo-inositol 1,4,5-trisphosphate, P4(5)-1-(2- itrophe l eth l ester NPE-caged IP3 , dextran 23 Texas Red 10000 MW (DTR), fluo3-AM and thapsigargin were purchased from Molecular Probes, Life 4 Page 6 of 34 1 technologies. Caffeine was from Sigma-Aldrich (St. Louis, MO, USA), ryanodine from Abcam 2 (Cambridge, UK) and fluoxetine from ABC chemicals (Woutersbrakel, Belgium). 3 2.3. Visualization of intracellular Ca2+ Dynamic changes in [Ca2+]i were monitored using fluo3-AM. Cells in suspension were loaded with 5 5 µM fluo3-AM at 2x106/ml in Krebs buffer for 1h at room temperature and subsequently washed 3x in 6 Krebs buffer. Thereafter, cells (0.5-1x106) were allowed to adhere on poly-L-lysine (0.1%) coated 7 18mm diameter glass coverslips and left for 30 min at room temperature for de-esterification and 8 settling on the dish. Cells were washed once in Krebs buffer to remove any unbound cells before 9 imaging. an us cr ip t 4 Intracellular Ca2+ imaging was performed in Krebs buffer at room temperature and was carried out 11 using a Nikon Eclipse TE300 inverted epifluorescence microscope (Nikon Belux, Brussels, Belgium), 12 equipped with a 40x oil-immersion objective (Plan Fluor, NA 1.30; Nikon) and an EM-CCD camera Qua tEM™ 5 “C CCD a era, Photo etri s, Tu so , A) . We used a Lambda DG-4 filterswitch ed 13 M 10 (Sutter Instrument Company, Novato, CA) to deliver excitation at 482 nm and captured emitted light 15 via a 505-nm long-pass dichroic mirror and a 535 nm bandpass filter (35 nm bandwidth). Images (1/s) 16 were generated with custom-generated QuantEMframes software written in Microsoft Visual C++ 6.0. 17 Fluo3 fluorescence-intensity changes were analyzed with custom-developed FluoFrames software 18 (generated by L.L., Ghent University, Belgium). Background fluorescence was subtracted from all 19 images. Traces of individual cells were obtained by point analysis in Fluoframes software. Ac c 20 ep t 14 2.4. Electroporation loading with NPE-caged IP3 and photoliberation 21 In order to study IP3-mediated Ca2+ release from the ER, T cells were loaded with NPE-caged IP3 22 through electroporation, as described elsewhere [16]. Briefly, cells seeded on coverslips were 23 washed 3x with a low conductivity electroporation buffer (4.02 mM KH2PO4, 10.8 mM K2HPO4, 1.0 24 mM MgCl2, 300 mM sorbitol, 2.0 mM HEPES, pH 7.4). The coverslips were then placed on the 5 Page 7 of 34 microscopic stage, 400 µm underneath a parallel wire Pt-Ir electrode and electroporated in the 2 presence of 10 µl electroporation buffer containing 100 µM NPE-caged IP3 and 100 µM DTR to 3 visualize the electroporation zone. Electroporation was done with 50 kHz bipolar pulses applied as 4 trains of 10 pulses of 2 ms duration each and repeated 15 times. The field strength was 100V peak- 5 to-peak applied over a 500 µm electrode separation distance. After electroporation, cells were 6 thoroughly washed with Krebs buffer. Electroporation was performed after fluo3 loading and did not 7 result in loss of fluo3 from the cells [17]. 8 After loading with NPE-caged IP3, coverslips were transferred to the microscope stage for Ca2+ 9 imaging. Photoliberation of IP3 as do e spot μ us cr ip t 1 diameter) illumination with 1-kHz pulsed UV light (349 nm UV laser Explorer, Spectra-Physics, Newport, Utrecht, The Netherlands) applied during 11 50 ms (50 pulses of 9 μJ e erg easured at the e tra e of the i ros ope epifluores e e tu e . M 12 an 10 2.5. Activation with anti-CD3/CD28 beads T cell receptor activation was achieved by adding magnetic particles coated with antibodies against 14 CD3 and CD28 (Dynabeads® Human T-Activator CD3/CD28, Life technologies) at a concentration of 15 25 µl per 106 cells (1:1 bead:cell ratio). Cells were visually inspected under the microscope at the end 16 of each experiment to determine which cells were making contact with at least one magnetic bead. ep t 2.6. Data analysis Ac c 17 ed 13 18 The statistical analysis was conducted in R [18]. Homoscedasticity and normality of residuals were 19 visually checked using residuals vs fitted plots and QQ plots. If necessary, power transformations 20 were applied, using a Box-Cox plot for guidance [19]. The datasets in section 3.1, 3.4. and 3.7 were 21 analyzed using a one-way ANOVA or two-way ANOVA (to correct for the possible influence of time if 22 the experiment was conducted for > 1 day). The other datasets (section 3.2, 3.3, 3.5 and 3.6) were 23 analyzed using the non-parametric Kruskal-Wallis test. “tude t s T tests para etri 24 rank sum tests (non-parametric) were used for post-hoc testing with a Holm-Bonferroni correction or Wil o o 6 Page 8 of 34 for multiple testing being applied. Data are presented as mean ± SD or median and range, for 2 parametric and non-parametric data, respectively. Significance was set at p < 0.05, all tests were two- 3 tailed. 4 Data are visually presented as boxplots showing median, first and third quartile. Whiskers represent 5 lowest and highest data within 1,5 interquartile range (IQR). Data exceeding the 1.5 IQR were 6 omitted from the graphs for clarity. 7 2.7. Digital droplet (dd)PCR us cr ip t 1 mRNA was isolated using the RNeasy® mini kit (QIAGEN, Hilden, Germany) and on column digestion 9 of genomic DNA was performed using the RNase-free DNase set (QIAGEN). RNA concentrations were 10 determined with the Quant-it™ Ribogreen® RNA assay kit (Bio-Rad, Hercules, CA, USA). RNA quality 11 assessment using microfluidic capillary electrophoresis (Experion RNA HighSens Chip, Bio-Rad) 12 showed good quality RNA samples as determined by 18S/28S rRNA ratios (RNA quality index [RQI] 13 8.5 for positive control, 9.3-9.4 for representative T lymphocyte samples; a RQI >7 indicates good 14 quality). mRNA was transcribed to cDNA using the iScript™ advanced cDNA synthesis kit (Bio-Rad). 15 cDNA concentrations were subsequently estimated using the Quant-it™ oligreen® ssDNA kit (Life 16 technologies). All kits were used a ordi g to the 17 available Taqman® assay (Hs00984349_m1; Life technologies) was used for amplification of serotonin 18 transporter cDNA. 20 µl reactions were prepared containing 5 µl of cDNA (500 ng input material), 10 19 µl 2x ddPCR™ super mix for probes (Bio-Rad), 1 µl Taqman® assay and 4 µl water. ddPCR assays were 20 performed as described previously [20]. Briefly, droplets were generated in 8-channel cartridges 21 a ufa turer s i stru tio s. A commercially Ac c ep t ed M an 8 o tai i g the µl sa ples plus 5 µl droplet ge erati g oil usi g the QX ™ droplet ge erator 22 (Bio-Rad). Subsequently, droplet-in-oil suspensions were transferred to 96 well plates and placed into 23 aT 24 cycles of 95°C for 15 sec and 60°C for 60 sec. Subsequently, the droplets were automatically read by 25 the QX ™ Ther al C ler Bio-Rad). Cycling conditions were as follows: 95°C for 5 min, followed by 40 ™ droplet reader Bio-‘ad a d the data ere a al zed ith the Qua ta“oft™ a al sis 7 Page 9 of 34 1 software 1.2.10.0 (Bio-Rad). All samples were tested in duplicate. No-template controls (NTCs) were 2 included in every ddPCR run. HEK 293 cells stably transfected with hSERT (kind gift from Randy 3 Blakely, Vanderbilt University) were used as a positive control. 4 2.8. Western blotting 107 cells were lysed in 1 ml radioimmunoprecipitation (RIPA) buffer containing 150 mM sodium 6 chloride, 1.0% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8.0 and supplemented 7 with 5 mg/ml Complete Mini protease inhibitor cocktail (Roche, Basel, Switzerland), 10 µl/ml 8 phosphatase inhibitor cocktails 2 and 3 (Sigma-Aldrich) and 1 µl/ml benzonase® nuclease (Sigma- 9 Aldrich). 50 µg of total protein was subjected to 10% sodium dodecyl sulphate polyacrylamide 10 gelelectrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membrane using 11 Tris-glycine buffer (25 mM Tris base, 190 mM glycine, 0.05% SDS), as described elsewhere [21]. 12 Serotonin transporter protein was detected with a 1:5000 dilution of ST51-1 (aa51-66) mouse 13 monoclonal anti-human serotonin transporter antibody (Santa Cruz Biotechnology, CA, USA) in PBS + 14 0.3% Tween-20 + 10% nonfat dry milk overnight at room temperature. Incubation with secondary 15 goat anti-mouse poly-HRP antibody (Thermo Fisher Scientific, Waltham, Massachusetts, USA) (1:1000 16 dilution) was performed during 1h at room temperature in PBS + 0.3% Tween-20 + 5% nonfat dry 17 milk. Protein bands were detected with enhanced chemiluminescence. HEK 293 cells, stably 18 transfected with hSERT were used as a positive control. Specificity of the primary antibody was 19 confirmed with a blocking peptide (PSPGAGDDTRHSIPAT; Thermo Fisher Scientific, Waltham, 20 Massachusetts, USA). The ST51 antibody was incubated for 2h at room temperature with the 21 blocking peptide at 5- fold excess in PBS, as described elsewhere [21]. Then, the antibody-peptide 22 mixture was further diluted in PBS containing 0.3% Tween-20 and 5% nonfat dry milk and added to 23 the membrane. 24 Ac c ep t ed M an us cr ip t 5 2.9. Detection of T cell activation 8 Page 10 of 34 In order to detect T cell activation, T cells were stimulated with anti-CD3/CD28 beads in a 1:1 2 bead:cell ratio for 5h at 37°C and 5% CO2. Fluoxetine or BAPTA-AM were added 30 minutes before 3 addition of the T cell stimulus and were maintained in the culture medium throughout the 4 experiment. After 5h incubation, cells were stained with anti-human CD69 PECy7 and anti-human 5 CD3 PECy5 (eBioscience, San Diego, CA, USA) for 30 minutes in PBS + 1% bovine serum albumin and 6 0.1% NaN3, washed once and analyzed on a FC500 (Beckman coulter, Fullerton, CA, USA). us cr ip t 1 7 3. Results 8 3.1. Fluoxetine suppresses Ca2+ signaling in response to T cell receptor activation In order to analyze the effect of fluoxetine on Ca2+ signaling when T cells are activated through the T 10 cell receptor (TCR), we activated Jurkat T cells with magnetic particles coated with antibodies against 11 CD3 and CD28 and analyzed the resulting changes in [Ca2+]i through labeling with the fluorescent Ca2+ 12 dye fluo3-AM. The fluorescent images obtained and the response of T cells to the anti-CD3/CD28 13 beads are illustrated in figure 1A. After each experiment, T cells making contact with at least one 14 magnetic bead were visualized and selected for analysis. The changes in fluorescent signal over time 15 (arbitrary units, A.U.), which relate to changes in [Ca2+]i, were plotted and result in a Ca2+ trace. The 16 majority of cells in the control samples (83%) responded to contact with a bead with a short period 17 of Ca2+ oscillations followed by a sustained increase in [Ca2+]i. A small percentage of cells showed an 18 oscillatory pattern (11%) or transient response (6%) after contact with a bead (figure 1B). The same 19 types of responses were found in T cells pre-incubated with 10 µM fluoxetine. A slight shift from 20 sustained responses (75%) towards oscillatory (13%) and transient (12%) responses was observed, 21 but no significant changes were detected as compared to control (Chi-square test, p=0.28). However, 22 fluoxetine did affect the magnitude of the response to TCR activation. T cells that were pre-incubated 23 for 30 min with 10 µM fluoxetine (F10) responded with oscillations with smaller amplitude and a 24 weaker sustained increase in [Ca2+]i. At 100 µM fluoxetine (F100), the response to TCR activation was 25 almost completely absent (figure 1C). Viability of the cells was assessed by trypan blue staining at the Ac c ep t ed M an 9 9 Page 11 of 34 end of the experiment and no increased cell death was observed in samples pre-incubated with 2 fluoxetine compared to control samples (data not shown). The absence of a response in samples 3 incubated with 100 µM fluoxetine was thus not due to loss of viability. In order to quantify the 4 different responses, we calculated the difference between the maximum of the Ca2+ peak (maximum 5 value of 0-720s) and the baseline (mean of 0-120s). Whereas the mean ± SD peak height of control T 6 cells was 59.18 ± 30.38 (A.U.), the mean peak height of T cells pre-incubated with F10 was 50.97 ± 7 28.16 (p=0.041) (figure 1D). Cells pre-incubated with F100 showed a dramatically reduced response 8 to TCR activation, with a mean peak height of only 7.45 ± 7.05 (p<0.0001). Similar results were 9 obtained when the area under the [Ca2+]i trace was analyzed instead of the peak [Ca2+]i change (data 10 not shown). These results show that fluoxetine inhibits the Ca2+ signaling pathway following TCR 11 activation in a concentration-dependent manner in T lymphocytes. us cr ip an 3.2. Interference of fluoxetine with intracellular stores or capacitative Ca2+ entry M 12 t 1 We next questioned whether the observed suppression of [Ca2+]i responses by fluoxetine was due to 14 either inhibition of capacitative Ca2+ entry or interference with the release of Ca2+ from intracellular 15 stores. To this end we added thapsigargin (TG), a selective inhibitor of sarco-endoplasmic reticulum 16 Ca2+-ATPase (SERCA) that prevents reuptake of Ca2+ into the ER, to the cells in Ca2+-free buffer 17 containing EGTA. Ca2+ exits the ER through a yet unidentified basal leak system, and blockage of 18 SERCA consequently results in depletion of the ER. After a five-minute incubation period with TG that 19 allowed the ER to be completely emptied (confirmed by subsequent addition of 50 µM cyclopiazonic 20 acid or 100 µM thapsigargin, data not shown), Ca2+-containing buffer (1.5 mM Ca2+) was added, thus 21 allowing the cells to refill their ER with Ca2+ through capacitative Ca2+ entry (figure 2A). The impact of 22 fluoxetine on both steps was analyzed. Interestingly, 30 min pre-incubation of fluoxetine reduced 23 the magnitude of the TG-induced Ca2+ release in a concentration-dependent manner (control median 24 7.90, range [-75.54 – 192.90]; F10 3.35, [-58.45 – 180.27], p<0.0001; F100 1.12, [-22.14 – 99.43], 25 p<0.0001; figure 2B). No significant differences could be detected with respect to the magnitude of Ac c ep t ed 13 10 Page 12 of 34 the peak after addition of Ca2+-containing buffer (control median 38.00, range [7.0-217.3]; F10 44.55, 2 [5.2-196.1]; F100 41.00, [3.3-208.7]; p=0.423; figure 2C). Similar results were obtained when using 3 anti-CD3/CD28 beads instead of TG to stimulate the cells. Whereas the initial response to the beads 4 in Ca2+-free buffer was suppressed by fluoxetine (control median 2.69, range [-6.2-85.2]; F10 1.84, 5 range [-2.2-50.1], p=0.47; F100 0.00, range [-6.7-46.1], p<0.0001), the capacitative Ca2+ entry was not 6 significantly affected by fluoxetine (control median 42.26, range [2.4-179.3]; F10 47.06, range [12.0- 7 135.5], p=0.24; F100 31.50, range [2.3-120.8], p=0.058)(figure 2D-F). These data suggest that 8 fluoxetine might interfere with Ca2+ release from intracellular stores. In contrast, fluoxetine does not 9 appear to affect capacitative Ca2+ entry in T lymphocytes. us cr ip 3.3. Effect of fluoxetine on intracellular Ca2+ stores an 10 t 1 The observed inhibition of Ca2+ release from intracellular stores could be due to either depletion of 12 the stores or direct inhibition of Ca2+ channels responsible for Ca2+ release from the ER. To 13 differentiate between both options, a series of experiments was conducted studying the direct effect 14 of fluoxetine on [Ca2+]i. Addition of fluoxetine to resting T cells induces a rise in [Ca2+]i both in Ca2+- 15 containing and Ca2+-free buffer (figure 3A and 3B). These data indicate that fluoxetine partially 16 depletes the intracellular stores, leaving less Ca2+ available for release upon stimulation. If fluoxetine 17 depletes ER stores, it could be expected that capacitative Ca2+ entry, which is activated upon 18 emptying of the ER, might also be affected by fluoxetine. However, the peak height of the 19 capacitative Ca2+ response after exposure of T cells to fluoxetine was not significantly different from 20 the response elicited in control samples (control median 43.2, range [9.4-83.7]; F100 44.9, [23.8- 21 161.1], p=0.19, figure 3B), confirming previous observations that fluoxetine pre-incubation does not 22 alter capacitative Ca2+ entry. Finally, it was confirmed that acute depletion of the ER stores with 23 fluoxetine is followed by a lower Ca2+ peak in response to TG (figure 3C). From these data and those 24 above, it can be concluded that fluoxetine depletes intracellular Ca2+ stores. 25 Ac c ep t ed M 11 3.4. Interference of fluoxetine with IP3-induced Ca2+ release 11 Page 13 of 34 In order to study in more detail the functional effect of fluoxetine on the ER, T cells were loaded with 2 NPE-caged IP3 through electroporation, and IP3 was released during imaging through flash photolysis. 3 The height of the resulting [Ca2+]i peak was measured. In accordance with the results of fluoxetine 4 partially depleting intracellular Ca2+ stores, fluoxetine reduced the height of the Ca2+ peak after 5 release of IP3, although statistical significance was only reached at 100 µM fluoxetine (mean control 6 11.43 ± 11.81; F10 9.98 ± 12.94, p=0.11, F100 2.13 ± 2.65, p<0.0001 (figure 4). Thus, fluoxetine 7 suppresses IP3-mediated Ca2+ release from the ER. us cr ip 8 t 1 3.5. Interference of fluoxetine with ryanodine receptor-mediated Ca2+ release In addition to IP3R, ryanodine receptors (RyR) are equally known to regulate Ca2+ release from 10 intracellular stores. Ca2+ released by IP3R may in turn activate RyR resulting in Ca2+-induced Ca2+ 11 release [22]. The functional effect of fluoxetine on RyR-mediated Ca2+ release was analyzed by 12 addition of caffeine, which is known to activate RyR [23]. Preliminary experiments to select the most 13 suitable concentration of caffeine showed a concentration-dependent increase in [Ca2+]i upon 14 exposure to caffeine in a range from 10 to 50 mM (data not shown). As 50 mM caffeine induced the 15 strongest effect, we selected this concentration to study the impact of fluoxetine hereon (figure 5A). 16 Responses of similar magnitude but shorter duration were obtained when the experiment was 17 repeated in Ca2+-free buffer, indicating that caffeine indeed released intracellular Ca2+ followed by 18 capacitative Ca2+ entry (data not shown). In order to confirm that the rise in [Ca2+]i induced by 19 caffeine was due to RyR stimulation, cells were pre-incubated with an antagonistic concentration of 20 ryanodine (200 µM). As shown in figure 5, ryanodine completely suppressed the rise in [Ca2+]i induced 21 by caffeine. Fluoxetine suppressed the rise in [Ca2+]i in a concentration-dependent manner (figure 22 5A). In order to quantify the effect of fluoxetine on the RyR-mediated Ca2+ release, the difference 23 between the maximal and minimal [Ca2+]i was calculated. Control cells showed a median peak height 24 of 17.5, range [2.8 - 74.9], F10 12.3, [0.4 - 50.8], p<0.0001 and F100 4.1, [0.0 - 57.8], p<0.0001 (figure 25 5B). Thus, fluoxetine inhibits RyR-induced Ca2+-release in a concentration-dependent manner. Ac c ep t ed M an 9 12 Page 14 of 34 1 3.6. Involvement of 5HT and SERT in the fluoxetine-induced effects on Ca2+ signaling As fluoxetine is known to inhibit serotonin uptake through the SERT, we next questioned whether 3 the observed effects on Ca2+ signaling could be initiated by SERT inhibition. Therefore, we first 4 analyzed whether Jurkat T lymphocytes express SERT. As shown in figure 6, T lymphocytes express 5 SERT both at the mRNA level and the protein level. Whereas the positive control (hSERT transfected 6 HEK293 cells) showed a protein band at ~80 and 60 kDa, only the 60 kDa band was found in T cells. 7 As described previously, the ~80 kDa band presumably represents a highly glycosylated form of SERT, 8 whereas the 60 kDa band is most likely the unmodified SERT protein [21]. Further, it should be noted 9 that the large difference in expression levels between the positive control and the T cells is due to 10 overexpression of SERT in the hSERT transfected cell line [21]. Specificity of the primary antibody was 11 confirmed by incubation with a blocking peptide. The results of these experiments indicate that T 12 lymphocytes do express SERT. 13 To investigate if the effects of fluoxetine on Ca2+ signaling are mediated by SERT inhibition, we 14 analyzed the influence of a large excess (1 mM) 5HT on the fluoxetine-induced suppression of RyR- 15 mediated Ca2+ release (100 µM fluoxetine). If inhibition of SERT by fluoxetine causes the observed 16 decrease in Ca2+ signaling, it can be expected that 5HT reverses this effect by competing with 17 fluoxetine for binding to SERT. As shown in figure 6C, 5HT did not inhibit the suppression of RyR- 18 mediated Ca2+ release by 10 or 100 µM fluoxetine (control median 21.70, [0.0-70.7] vs 5HT 20.70, 19 [3.1-71.1], p=0.61; F10 18.60, [0.9-60.3] vs F10 + 5HT 20.20, [-0.1-100], p=0.35; F100 7.30, [-0.9-62.2] 20 vs F100 + 5HT 5.25, [-0.6-93.6], p=0.18). Thus, it is not likely that fluoxetine depletes Ca2+ from the ER 21 through blockage of 5HT uptake by SERT. Notably, addition of 5HT (1 mM) did not induce Ca2+ 22 changes on its own (data not shown). Ac c ep t ed M an us cr ip t 2 23 3.7. The effect of fluoxetine on T cell activation is mimicked by buffering of intracellular Ca2+ 24 In order to investigate whether the observed effect of fluoxetine on intracellular Ca2+ stores is at the 25 basis of its immunosuppressive effect, we analyzed the effect of fluoxetine on CD69 expression, an 13 Page 15 of 34 1 early activation marker. Incubation of T cells with anti-CD3/CD28 beads during 5h in the absence of 2 fluoxetine induced a strong upregulation of CD69 expression (figure 7 . No sti ulated N“ 3 showed a mean fluorescence intensity (MFI) of 10.29 ± 0.97, whereas stimulated cells “ had a MFI 4 of 107.0 ± 3.0. Fluoxetine (100 µM) decreased the MFI to 9.77 ± .97 p< . 5 same suppressive effect was found when cells were incubated with BAPTA-AM (50 µM), an 6 intracellular Ca2+ chelator added to silence cytoplasmic Ca2+ changes, demonstrating that 7 interference with intracellular Ca2+ signals after TCR stimulation indeed impairs T cell activation (MFI 8 18.27 ± 0.21, p<0.0001 o pared to “ ). Thus, the inhibitory effect of fluoxetine on CD69 expression 9 can be mimicked by buffering the intracellular Ca2+ of the cells with BAPTA-AM. These data show that 10 interference with Ca2+ signaling in T lymphocytes results in impaired T cell activation, as estimated 11 from CD69 expression, and that the effect of fluoxetine is comparable in magnitude to the effect of 12 buffering [Ca2+]i with BAPTA-AM. M an us cr ip t o pared to “ . The 4. Discussion ed 13 ells In this report, we investigated the impact of fluoxetine on Ca2+ signaling in Jurkat T lymphocytes. 15 Previous research has demonstrated that fluoxetine and other SSRIs exert anti-inflammatory and 16 immunosuppressive effects on T lymphocytes [3, 24]. Similar suppressive effects have been 17 described in Jurkat T lymphocytes [25]. Although several hypotheses on the mechanism behind the 18 observed effects were investigated (reviewed in [2]), the exact mechanism by which fluoxetine 19 suppresses T cell activation and proliferation was not clarified. SSRIs have been shown to affect Ca2+ 20 signaling in several cell types including neurons [8], astrocytes [9], microglia [10], osteosarcoma cells 21 [11], platelets [13] and adrenal medulla PC12 cells [7, 26]. Since elevation of intracellular Ca2+ plays a 22 major role in the pathway leading to T cell activation in response to antigens [6], we investigated if 23 SSRIs, in particular fluoxetine, interfere with this signaling pathway in T cells. 24 In the case of T lymphocytes, Ca2+ is stored in the ER and release from the ER is mediated 25 predominantly by binding of IP3 to IP3R, and is further regulated by RyR [27]. The majority of research Ac c ep t 14 14 Page 16 of 34 conducted on the effect of antidepressants, including SSRIs, on Ca2+ signaling in other cell types 2 suggests interference with intracellular Ca2+ stores [7, 9, 11, 13]. In accordance with these data, we 3 demonstrated that fluoxetine interferes with the ER Ca2+ stores in T lymphocytes. As opposed to 4 tricyclic antidepressants, we found that fluoxetine inhibits IP3-induced Ca2+ release [28]. More 5 specifically, we demonstrated that fluoxetine suppresses the rise in [Ca2+]i in response to TCR 6 activation. Additionally, we showed that the decreased Ca2+ signaling is due to the inhibition of Ca2+ 7 release from ER stores, rather than the blockage of capacitative Ca2+ entry. There are two possible 8 explanations for the inhibition of the Ca2+ release from intracellular stores: either fluoxetine causes a 9 depletion of stored Ca2+ thus leaving less Ca2+ available for release after IP3R or RyR activation, or 10 fluoxetine directly interferes with the Ca2+ channels blocking the Ca2+ release in response to IP3R or 11 RyR activation. In accordance to Serafeim et al., who found that fluoxetine and other SSRIs induced a 12 rise in [Ca2+]i in malignant B cells [12], the addition of fluoxetine to resting T cells resulted in an 13 increase of the cytoplasmic Ca2+ concentration. Subsequent addition of TG resulted in a significantly 14 lower amount of Ca2+ being released from the ER. Therefore, these data suggest that fluoxetine 15 depletes the ER stores, thereby leaving less Ca2+ available for release after IP3R or RyR activation 16 (figure 8). 17 Jurkat and primary T lymphocytes have been shown to express several types of 5HT receptors 18 (5HT1A, 5HT1B, 5HT2A, 5HT3 and 5HT7), as well as tryptophan hydroxylase indicating that these cells 19 are capable of synthesizing and responding to 5HT [29, 30]. Furthermore, T cells are capable of 20 releasing 5HT into the extracellular space in response to stimulation [30]. Although the precise role of 21 5HT in T lymphocyte function has not been elucidated, 5HT has been identified as an important 22 factor in T cell activation and proliferation [31]. Fluoxetine was designed to selectively inhibit the 23 serotonin transporter, which is responsible for uptake of 5HT into the cell. Although no external 5HT 24 was added to the incubation buffer in our experiments, T cells have been demonstrated to secrete 25 5HT themselves and therefore it is possible that 5HT was present in the microenvironment during the 26 experiments. Given the presumed importance of 5HT in T cell activation and proliferation, it could be Ac c ep t ed M an us cr ip t 1 15 Page 17 of 34 expected that the anti-proliferative effects of fluoxetine might be related to its capability to inhibit 2 5HT uptake in T cells. Here we show that fluoxetine depletes Ca2+ from intracellular stores, thereby 3 disturbing the main signaling transduction pathway leading to T cell activation. Nevertheless, we 4 demonstrated that the depletion of ER stores is not mediated through blockage of 5HT transport by 5 SERT since addition of even a large excess of 5HT did not abrogate the effect of fluoxetine on Ca2+ 6 signaling. Instead, it has been proposed that fluoxetine, being a highly lipophilic molecule, interacts 7 with the membrane lipid bilayer and thereby influences the ion channel structure and function [7]. 8 Future research will be needed to elucidate how fluoxetine interacts with Ca2+ channels at the 9 molecular level. us cr ip t 1 Finally, we demonstrated that the immunosuppressive effects of fluoxetine - under the form of 11 decreased CD69 expression in response to TCR activation – can be mimicked by buffering of 12 intracellular Ca2+ with BAPTA-AM. Others have shown that inhibition of IP3- or RyR- mediated Ca2+ 13 release downregulates Jurkat T cell proliferation and IL2 production [27]. In primary human T cells, 14 inhibition of RyR equally inhibited T cell proliferation [32]. These data suggest that inhibition of IP3- 15 and RyR-mediated Ca2+ release from ER stores plays an important role in the immunosuppressive 16 effects of fluoxetine, although it cannot be excluded that other mechanisms contribute to the 17 immunosuppressive outcome. 18 It should be noted that the concentrations of fluoxetine used in this report are considerably higher 19 than the plasma concentrations found in depressive patients. Whereas plasma concentrations of 20 fluoxetine are usually below 1 µM, we applied concentrations of 10-100 µM to study the effects of 21 fluoxetine on Ca2+ signaling. The applied concentrations are based on previous reports on in vitro T 22 cell immunosuppression by SSRIs [3]. However, since SSRIs are lipophilic compounds that accumulate 23 in tissues, significantly higher concentrations in organs than in plasma can occur. In that respect, it 24 has been demonstrated that SSRIs can reach 10-fold higher concentrations in spleen than in plasma 25 [33]. As the meeting of a naïve T cell and its antigen occurs in lymphoid tissue such as the spleen or Ac c ep t ed M an 10 16 Page 18 of 34 lymph nodes, it can be expected that T lymphocytes going through the activation process in lymphoid 2 tissue are actually exposed to fluoxetine concentrations up to 10 µM, a concentration which we have 3 demonstrated to exert acute inhibitory effects on Ca2+ signaling in vitro. Furthermore, it has been 4 shown that the effects of fluoxetine on IP3R and RyR are time- and concentration dependent [9]. The 5 EC50 for the chronic effects of fluoxetine in astrocytes was almost 10 times lower than for the acute 6 effects, suggesting that the potency of fluoxetine to interfere with Ca2+ signaling increases with 7 longer exposure time. Therefore, chronic exposure of T lymphocytes to fluoxetine might result in 8 immunosuppression at lower concentrations (0.5-1 µM) which are within the same range as plasma 9 concentrations found in depressive patients. us cr ip t 1 Finally, we selected fluoxetine to study the effects on Ca2+ signaling in T lymphocytes. As other SSRIs 11 also induce immunosuppressive effects in T lymphocytes [3], it would be interesting to investigate 12 whether these compounds also affect Ca2+ signaling in T lymphocytes. 13 In conclusion, these data show that fluoxetine suppresses intracellular Ca2+ signaling in Jurkat T 14 lymphocytes through depletion of Ca2+ from intracellular stores, an effect likely to be at the basis of 15 the observed immunosuppression. 16 Acknowledgements 17 The authors thank Evi Standaert (Ghent University) for assistance with the detection of serotonin 18 transporter expression. Randy Blakely (Vanderbilt University) is acknowledged for the kind gift of 19 hSERT transfected HEK cells. Ac c ep t ed M an 10 20 17 Page 19 of 34 1. 2. 3. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. an 7. M 6. ed 5. us cr ip t 4. Janssen, D.G., et al., A psychoneuroimmunological review on cytokines involved in antidepressant treatment response. Hum Psychopharmacol, 2010. 25(3): p. 201-15. Gobin, V., et al., Selective serotonin reuptake inhibitors as a novel class of immunosuppressants. Int Immunopharmacol, 2014. 20: p. 148-156. Gobin, V., et al., Fluoxetine reduces murine graft-versus-host disease by induction of T cell immunosuppression. J Neuroimmune Pharmacol, 2013. 8(4): p. 934-43. Schuster, C., et al., Selective serotonin reuptake inhibitors--a new modality for the treatment of lymphoma/leukaemia? Biochem Pharmacol, 2007. 74(9): p. 1424-35. Cloonan, S.M., et al., The antidepressants maprotiline and fluoxetine have potent selective antiproliferative effects against Burkitt lymphoma independently of the norepinephrine and serotonin transporters. Leuk Lymphoma, 2010. 51(3): p. 523-39. Lewis, R.S., Calcium signaling mechanisms in T lymphocytes. Annu Rev Immunol, 2001. 19: p. 497-521. Kim, H.J., et al., Fluoxetine inhibits ATP-induced [Ca(2+)](i) increase in PC12 cells by inhibiting both extracellular Ca(2+) influx and Ca(2+) release from intracellular stores. Neuropharmacology, 2005. 49(2): p. 265-74. Kim, H.J., et al., Fluoxetine suppresses synaptically induced [Ca(2)(+)]i spikes and excitotoxicity in cultured rat hippocampal neurons. Brain Res, 2013. 1490: p. 23-34. Li, B., et al., Effects of chronic treatment with fluoxetine on receptor-stimulated increase of [Ca2+]i in astrocytes mimic those of acute inhibition of TRPC1 channel activity. Cell Calcium, 2010. 50(1): p. 42-53. Horikawa, H., et al., Inhibitory effects of SSRIs on IFN-gamma induced microglial activation through the regulation of intracellular calcium. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 2010. 34(7): p. 1306-1316. Lin, K.L., et al., Effect of sertraline on [Ca2+](i) and viability of human MG63 osteosarcoma cells. Drug Chem Toxicol, 2013. 36(2): p. 231-40. Serafeim, A., et al., Selective serotonin reuptake inhibitors directly signal for apoptosis in biopsy-like Burkitt lymphoma cells. Blood, 2003. 101(8): p. 3212-9. Helmeste, D.M., et al., Serotonin uptake inhibitors modulate intracellular Ca2+ mobilization in platelets. Eur J Pharmacol, 1995. 288(3): p. 373-7. Edgar, V.A., et al., Fluoxetine action on murine T-lymphocyte proliferation: participation of PKC activation and calcium mobilisation. Cell Signal, 1998. 10(10): p. 721-6. Edgar, V.A., et al., Role of protein kinase C and cAMP in fluoxetine effects on human T-cell proliferation. Eur J Pharmacol, 1999. 372(1): p. 65-73. Monaco, G., et al., Alpha-Helical Destabilization of the Bcl-2-BH4-Domain Peptide Abolishes Its Ability to Inhibit the IP3 Receptor. PLoS One, 2013. 8(8): p. e73386. De Bock, M., et al., Connexin 43 hemichannels contribute to cytoplasmic Ca2+ oscillations by providing a bimodal Ca2+-dependent Ca2+ entry pathway. J Biol Chem, 2012. 287(15): p. 12250-66. R core team, R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing. 2014; Available from: http://www.R-project.org. Venables, W.N., Ripley, B.D., Modern Applied Statistics with S. Fourth ed. 2002, New York: Springer. Kiselinova, M., et al., Comparison of droplet digital PCR and seminested real-time PCR for quantification of cell-associated HIV-1 RNA. PLoS One, 2014. 9(1): p. e85999. Chamba, A., et al., Characterisation of the endogenous human peripheral serotonin transporter SLC6A4 reveals surface expression without N-glycosylation. J Neuroimmunol, 2008. 204(1-2): p. 75-84. ep t 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 References Ac c 1 18 Page 20 of 34 27. 28. 29. 30. 31. 32. 33. t us cr ip 26. an 25. M 24. ed 23. Ricard, I., et al., A caffeine/ryanodine-sensitive Ca2+ pool is involved in triggering spontaneous variations of Ca2+ in Jurkat T lymphocytes by Ca2+-induced Ca2+ release (CICR) mechanism. Cellular Signalling, 1997. 9(2): p. 197-206. Zucchi, R. and S. Ronca-Testoni, The sarcoplasmic reticulum Ca2+ channel/ryanodine receptor: modulation by endogenous effectors, drugs and disease states. Pharmacol Rev, 1997. 49(1): p. 1-51. Taler, M., et al., Immunomodulatory effect of selective serotonin reuptake inhibitors (SSRIs) on human T lymphocyte function and gene expression. Eur Neuropsychopharmacol, 2007. 17(12): p. 774-80. Amit, B.H., et al., Proapoptotic and chemosensitizing effects of selective serotonin reuptake inhibitors on T cell lymphoma/leukemia (Jurkat) in vitro. Eur Neuropsychopharmacol, 2009. 19(10): p. 726-34. Hahn, S.J., et al., Inhibition by fluoxetine of voltage-activated ion channels in rat PC12 cells. Eur J Pharmacol, 1999. 367(1): p. 113-8. Dadsetan, S., et al., Store-operated Ca2+ influx causes Ca2+ release from the intracellular Ca2+ channels that is required for T cell activation. J Biol Chem, 2008. 283(18): p. 12512-9. Shimizu, M., et al., Ca2+ release from inositol 1,4,5-trisphosphate-sensitive Ca2+ store by antidepressant drugs in cultured neurons of rat frontal cortex. J Neurochem, 1993. 60(2): p. 595-601. Levite, M., Neurotransmitters activate T-cells and elicit crucial functions via neurotransmitter receptors. Curr Opin Pharmacol, 2008. 8(4): p. 460-71. Chen, Y., et al., T lymphocytes possess the machinery for 5-HT synthesis, storage, degradation and release. Acta Physiol (Oxf), 2015. 213(4): p. 860-7. Ahern, G.P., 5-HT and the immune system. Curr Opin Pharmacol, 2011. 11(1): p. 29-33. Thakur, P., S. Dadsetan, and A.F. Fomina, Bidirectional coupling between ryanodine receptors and Ca2+ release-activated Ca2+ (CRAC) channel machinery sustains store-operated Ca2+ entry in human T lymphocytes. J Biol Chem, 2012. 287(44): p. 37233-44. Uhr, M., M.T. Grauer, and F. Holsboer, Differential enhancement of antidepressant penetration into the brain in mice with abcb1ab (mdr1ab) P-glycoprotein gene disruption. Biol Psychiatry, 2003. 54(8): p. 840-6. ep t 33 22. Ac c 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 19 Page 21 of 34 Figure captions 2 Figure 1: Effect of fluoxetine on Ca2+ signaling in response to TCR activation. T cells were stimulated 3 with magnetic beads coated with anti-CD3 and anti-CD28 antibodies. Figure 1A shows a series of 4 images taken after addition of anti-CD3/CD28 beads (beads were added at 120s). The upper right 5 image is a bright field image taken at the end of the experiment, in which the location of the beads 6 can be seen. The lower right image is a magnification of the marked area above. Scale bars are 50 7 µm. Color scale shows the pseudocolors in the RGB color mode assigned to the 256 shades of gray in 8 an 8-bit image. B) Representative Ca2+ responses induced by anti-CD3/CD28 beads in individual cells. 9 Contact with a bead triggered sustained (top), transient (middle) or oscillatory (bottom) responses. 10 The arrow indicates the addition of the beads. C) representative traces of T cells activated with anti- 11 CD3/CD28 beads in Krebs buffer (contr), 10 µM fluoxetine (F10), 100 µM fluoxetine (F100) or without 12 beads (negative control). For the negative control, the arrow indicates addition of an equal amount 13 of buffer without beads. D) Peak height of the Ca2+ response of T cells when stimulated with anti- 14 CD3/CD28 beads in Krebs buffer (contr) or fluoxetine (10 µM, F10 and 100 µM, F100). Peak height 15 was calculated as the difference between the maximum and the baseline. Each condition was 16 repeated at least three times and data were pooled for analysis. In total, 89-146 bead-bound cells 17 per condition were analyzed. * = p<0.05. *** = p<0.0001. 18 Figure 2. Effect of fluoxetine on intracellular Ca2+ release and capacitative Ca2+ entry. Krebs buffer 19 was replaced with Ca2+-free buffer immediately before imaging and 10 µM TG (A-C) or 25 µl anti- 20 CD3/CD28 beads (D-F) were added at 120s. At 420s, Ca2+-free buffer was replaced with Ca2+- 21 containing buffer (1.5 mM Ca2+). Cells were pre-incubated with fluoxetine (10 µM, F10 and 100 µM, 22 F100) in Krebs buffer for 30 minutes, and fluoxetine was maintained in all added solutions. All 23 conditions were repeated at least three times. Data were pooled for analysis. A) Mean traces of cells 24 in Ca2+-free Krebs buffer (contr), 10 µM fluoxetine (F10) and 100 µM fluoxetine (F100). Arrows 25 indicate addition of thapsigargin (TG, peak 1) and Ca2+-containing buffer (Ca2+, peak 2). B) Peak height Ac c ep t ed M an us cr ip t 1 20 Page 22 of 34 of the Ca2+ change induced by TG (peak 1). Peak height was calculated as the difference between the 2 maximum and the baseline. C) Peak height of Ca2+ change induced by re-introduction of Ca2+- 3 containing buffer (peak 2). D) Representative traces of cells in Ca2+-free Krebs buffer (contr), 10 µM 4 fluoxetine (F10) and 100 µM fluoxetine (F100). Arrows indicate addition of anti-CD3/CD28 beads 5 (beads, peak 1) and Ca2+-containing buffer (Ca2+, peak 2). E) Peak height of the Ca2+ change induced 6 by anti-CD3/CD28 beads (peak 1). Peak height was calculated as the difference between the 7 maximum and the baseline. F) Peak height of Ca2+ change induced by re-introduction of Ca2+- 8 containing buffer (peak 2). *** = p<0.0001. 9 Figure 3. Fluoxetine depletes intracellular calcium stores. A) T cells were incubated in Ca2+- 10 containing Krebs buffer and 100 µM fluoxetine (F100) was added at 120s. B) Krebs buffer was 11 replaced with Ca2+-free buffer immediately before imaging and 100 µM fluoxetine (F100) or buffer 12 (contr) was added at 120s. At 420s, Ca2+-free buffer was replaced for Ca2+-containing Krebs buffer 13 (1.5 mM Ca2+). C) Krebs buffer was replaced with Ca2+-free buffer immediately before imaging and 14 100 µM fluoxetine (F100) or Ca2+-free buffer (contr) was added at 120s. Thapsigargin was added at 15 420s. The traces depict the mean response of at least 50 cells. 16 Figure 4. Effect of fluoxetine on IP3-mediated Ca2+ release. T cells were electroporated with NPE- 17 caged IP3 and IP3 was released by flash photolysis after 120s of imaging. Imaging was continued for 5 18 minutes. Fluoxetine (10 µM, F10 and 100 µM, F100) was added 30 minutes before the start of the 19 experiment. The peak height of the Ca2+ change after photolytic release of IP3 was calculated as the 20 difference between the maximum and the baseline. A total of 61 – 111 cells per group were 21 analyzed. A) mean traces of electroporated cells within the flash zone in Krebs buffer (contr), 10 µM 22 fluoxetine (F10) and 100 µM fluoxetine (F100). B) Calculated peak heights of the recorded Ca2+ 23 changes after photolytic release of IP3. *** = p<0.0001. 24 Figure 5. Effect of fluoxetine on RyR-mediated Ca2+ release. T cells were stimulated with 50 mM 25 caffeine in the presence of different concentrations of fluoxetine. The experiments were performed Ac c ep t ed M an us cr ip t 1 21 Page 23 of 34 in calcium-containing Krebs solution. A) Individual traces of cells in Krebs buffer (contr), 10 µM (F10), 2 100 µM (F100) fluoxetine or 200 µM ryanodine and stimulated with 50 mM caffeine at 120s (arrow 3 on the graph). Fluoxetine was added 30 minutes before the start of the experiment and maintained 4 in all added solutions. B) Calculated peak heights of the recorded Ca2+ changes. Per sample, 100 5 arbitrary cells were analyzed. Each condition was performed in duplicate and results were pooled for 6 analysis. *** = p<0.0001. 7 Figure 6. Analysis of SERT expression in Jurkat T lymphocytes and involvement of 5HT in the 8 fluoxetine-induced effects. A) SERT mRNA expression was detected with ddPCR. Samples were 9 analyzed in duplo and compared to a positive control (hSERT transfected HEK293 cells); B) SERT 10 protein expression in hSERT HEK cells (positive control) and T cells. Left: detection with anti-SERT; 11 Right: detection with anti-SERT and a blocking peptide; C) Results of competition experiments with 12 5HT (1 mM) on fluoxetine (10 and 100 µM) inhibition of caffeine-induced Ca2+ release. Cells were 13 incubated with fluoxetine and/or 5HT 30 minutes before the start of the experiment. All conditions 14 were analyzed in triplicate and data were pooled for analysis. Peak height expresses the caffeine- 15 induced change in Ca2+. ns = not significant. 16 Figure 7. Effect of fluoxetine and BAPTA-AM on CD69 expression in activated T cells. T cells were 17 activated with anti-CD3/CD28 beads. BAPTA-AM (50 µM) or fluoxetine (100 µM, F100) were added 18 30 minutes before addition of the beads and the cells were incubated for 5h at 37°C and 5% CO2. 19 Cells were stained with anti-human CD3 PECy5 and CD69 PECy7 and analyzed by flow cytometry. NS 20 = non-stimulated cells. S = cells stimulated with anti-CD3/CD28 beads. Mean ± SD of mean 21 fluorescent intensities (MFI) are shown. Each condition was analyzed in triplicate. *** = p<0.0001. ns 22 = not significant. 23 Figure 8. Schematic representation of fluoxetine-induced effects on Ca2+ signaling in T lymphocytes. 24 In T cell activation, binding of a a tige to the TC‘ results i a ti atio of PLCγ, hi h o erts PIP2 25 to IP3 and DAG. IP3 induces Ca2+ release from the ER through activation of IP3R. Ca2+ may in turn Ac c ep t ed M an us cr ip t 1 22 Page 24 of 34 activate RyR thereby further stimulating Ca2+ release. RyR agonists such as caffeine also induce Ca2+ 2 release by RyR. Secondary to the release of Ca2+ from the ER, influx of Ca2+ through the plasma 3 membrane is induced. The rise in cytoplasmic Ca2+ eventually leads to transcription of a pleotropic 4 set of genes, including IL2, resulting in T cell activation and proliferation. Fluoxetine depletes 5 intracellular Ca2+ stores, thereby leaving less Ca2+ available for release upon IP3R or RyR activation. As 6 Ca2+ release from intracellular stores is an indispensable step in the pathway leading to T cell 7 activation, inhibition of Ca2+ signaling by fluoxetine results in impaired T cell activation and 8 proliferation. Intermediate steps in the signaling transduction pathway were omitted for clarity. TCR 9 = T ell re eptor, PLCγ = phospholipase Cγ; PIP2 = phosphatidylinositol 4,5-bisphosphate; IP3 = 10 phosphatidylinositol 3,4,5-trisphosphate; DAG = diacylglycerol; ER = endoplasmic reticulum; TG = 11 thapsigargin. an us cr ip t 1 Ac c ep t ed M 12 23 Page 25 of 34 Ac ce pt ed M an us cr ip t Figure 1 Page 26 of 34 Ac ce pt ed M an us cr ip t Figure 2 Page 27 of 34 Ac ce pt ed M an us cr i Figure 3 Page 28 of 34 Ac ce pt ed M an us cr i Figure 4 Page 29 of 34 Ac ce pt ed M an us cr i Figure 5 Page 30 of 34 Ac ce pt ed M an us cr i Figure 6 Page 31 of 34 Ac ce pt ed M an us cr i Figure 7 Page 32 of 34 an us c ri Figure 8 M Antigen TCR ed PLCγ nucleus ce Ca2+ Ac Ca2+ PIP2 pt IL2 IP3 + DAG Ca2+ IP3R Ca2+ Ca2+ caffeine RyR ER fluoxetine SERCA Ca2+ TG Page 33 of 34 *Conflict of Interest Statement Conflict of interest statement Ac c ep t ed M an us cr ip t The authors declare no conflicts of interest. Page 34 of 34 View publication stats