An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells

An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells

JIM-12035; No of Pages 11 Journal of Immunological Methods xxx (2015) xxx–xxx Contents lists available at ScienceDirect Journal of Immunological Met...

3MB Sizes 0 Downloads 23 Views

JIM-12035; No of Pages 11 Journal of Immunological Methods xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim

An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells Joana Cerveira b,1, Julfa Begum b,1, Rafael Di Marco Barros c, Annemarthe G. van der Veen d, Andrew Filby a,b,⁎ a

Flow Cytometry Core Facility, Newcastle Biomedicine, Newcastle University, Newcastle-upon-Tyne NE1 7RU, UK FACS Laboratory, London Research Institute, Cancer Research UK, 44 Lincoln's Inn Fields, Holborn WC2A 3LY, UK Immuno Surveillance Laboratory London Research Institute, Cancer Research UK, 44 Lincoln's Inn Fields, Holborn, London WC2A 3LY, UK d Immunobiology Laboratory London Research Institute, Cancer Research UK, 44 Lincoln's Inn Fields, Holborn, London WC2A 3LY, UK b c

a r t i c l e

i n f o

Article history: Received 15 November 2014 Received in revised form 22 March 2015 Accepted 30 April 2015 Available online xxxx Keywords: Imaging flow cytometry Calcium flux T cells NFAT translocation

a b s t r a c t Calcium ions (Ca2+) are a ubiquitous transducer of cellular signals controlling key processes such as proliferation, differentiation, secretion and metabolism. In the context of T cells, stimulation through the T cell receptor has been shown to induce the release of Ca2+ from intracellular stores. This sudden elevation within the cytoplasm triggers the opening of ion channels in the plasma membrane allowing an influx of extracellular Ca2+ that in turn activates key molecules such as calcineurin. This cascade ultimately results in gene transcription and changes in the cellular state. Traditional methods for measuring Ca2+ include spectrophotometry, conventional flow cytometry (CFC) and live cell imaging techniques. While each method has strengths and weaknesses, none can offer a detailed picture of Ca2+ mobilisation in response to various agonists. Here we report an Imaging Flow Cytometry (IFC)-based method that combines the throughput and statistical rigour of CFC with the spatial information of a microscope. By co-staining cells with Ca2+ indicators and organelle-specific dyes we can address the spatiotemporal patterns of Ca2+ flux in Jurkat cells after stimulation with anti-CD3. The multispectral, high-throughput nature of IFC means that the organelle co-staining functions to direct the measurement of Ca2+ indicator fluorescence to either the endoplasmic reticulum (ER) or the mitochondrial compartments without the need to treat cells with detergents such as digitonin to eliminate cytoplasmic background. We have used this system to look at the cellular localisation of Ca2+ after stimulating cells with CD3, thapsigargin or ionomycin in the presence or absence of extracellular Ca2+. Our data suggest that there is a dynamic interplay between the ER and mitochondrial compartments and that mitochondria act as a sink for both intracellular and extracellular derived Ca2+. Moreover, by generating an NFAT–GFP expressing Jurkat line, we were able to combine mitochondrial Ca2+ measurements with nuclear translocation. In conclusion, this method enables the high throughput study of spatiotemporal patterns of Ca2+ signals in T cells responding to different stimuli. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Calcium (Ca2+) serves as an intracellular secondary messenger in a panoply of different cell types and regulates a wide variety of key cellular processes (Berridge et al., 2000). In the immune system, this ubiquitous mode of signalling is required for virtually all cellular functions downstream of antigen and Fc receptor stimulation (Hogan et al., 2010). Thus, pharmacological inhibition of the Ca2+ signalling pathway achieves potent immunosuppression (Sieber and Baumgrass, 2009) and individuals with mutations in Ca2+ signalling machinery exhibit severe combined immunodeficiency (SCID) (Feske et al., 2006; Oh-Hora et al.,

⁎ Corresponding author at: Flow Cytometry Core Facility, Newcastle Biomedicine, Newcastle University, Newcastle-upon-Tyne NE1 7RU, UK. Tel.: + 44 191 241 8831/ 8826; fax: +44 191 241 8666. E-mail address: andrew.fi[email protected] (A. Filby). 1 Denotes equal contribution by authors.

2008; Picard et al., 2009). Downstream of the T cell antigen receptor (TCR), Ca2+ signalling is implicated in regulating cell survival, proliferation, differentiation (Bueno et al., 2002; Neilson et al., 2004; Melichar et al., 2013; Oh-Hora et al., 2013), motility (Bhakta et al., 2005), effector functions and functional non-responsiveness (Sloan-Lancaster et al., 1993, 1996; Feske et al., 2001; Macian et al., 2002). However, comparatively less is understood about how this ubiquitous mode of signalling can regulate such a diverse range of cellular processes. In this regard, there is still much to be gained from the study of T cells (Lewis, 2001) as they are an extremely amenable cellular system. In T cells, storeoperated Ca2+ entry (SOCE) represents the canonical mechanism that initiates extracellular Ca2 + influx (Hogan et al., 2010; Irvine et al., 2002; Huse et al., 2007). At resting state, the intracellular Ca2+ ‘store’ is predominantly sequestered within the lumen of the endoplasmic reticulum (ER) and upon TCR stimulation, is released into the cytoplasm (Smith-Garvin et al., 2009). This in turn causes conformational changes in stromal interaction molecules (STIM1/2), which cluster in regions of

http://dx.doi.org/10.1016/j.jim.2015.04.030 0022-1759/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

2

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

juxtaposition between the plasma membrane and the ER and directly interact with ORAI1 proteins to assemble the Ca2+-release-activated Ca2+ (CRAC) channels to initiate Ca2+ influx (Liou et al., 2005; Penna et al., 2008). As CRAC channels are highly sensitive to cytosolic Ca2+, they are rapidly inactivated by the accumulation of Ca2+ ions in the adjacent cytosolic space. However, in order to sustain Ca2+ influx, it has been shown that mitochondria relocate to focal regions of TCR stimulation, where they buffer the excess accumulation of cytosolic Ca2+ and prevent premature inhibition of the Ca2+ signal. In support of this, inhibition of mitochondrial re-localization and Ca2+ uptake have both been shown to reduce the duration of Ca2 + signalling, in addition to impairing T cell activation (Hoth et al., 2000; Quintana et al., 2007, 2011). In T cells, there are clear correlations between the strength of TCR stimulation and the amplitude, duration and oscillatory frequency of the downstream Ca2+ signal (Donnadieu et al., 1992). These correlations were verified in numerous experimental systems by altering the concentration of antigenic peptides (Rotnes and Bogen, 1994), modulating their affinity to the TCR when presented on major histocompatibility (MHC) molecules (Sloan-Lancaster et al., 1996; Wulfing et al., 1997) and by analysing the signal-amplifying effects of co-stimulation through bona fide co-stimulatory receptors (Michel et al., 2001; Nedellec et al., 2010). Later studies investigated the role of these features of Ca2+ signalling in the regulation of downstream gene expression. This was assessed by directly imposing Ca2+ currents across the membrane of Jurkat T cell lines that expressed Ca2 +-sensitive, NFκB and NFAT-dependent reporter genes (Dolmetsch et al., 1998). Using this system, it was found that at low levels of stimulation, oscillating Ca2 + currents were significantly more efficient than stable Ca2 + currents at inducing NFAT-dependent gene transcription and while low frequency Ca2+ oscillations were permissive for the activation of NFκB, high frequency oscillations were critical for the sustained activation of NFAT. Together, these two studies independently identified a molecular mechanism by which the efficiency and specificity of nuclear signalling can be regulated by distinct profiles of Ca2+ signalling (Lewis, 2003). Additional insight into T cell Ca2+ signals has more recently been gained via the analysis of thymic T cell development. T cell precursors (thymocytes) are positively or negatively selected on the basis of TCR signal strength, which is determined by the TCRs' affinity to selfpeptide:MHC complexes displayed on thymic epithelial cells (Lo and Allen, 2013). Notably, both negative and positive selections have been shown to require Ca2 + signalling (Bueno et al., 2002; Neilson et al., 2004; Macian, 2005). Two analyses of T cell selection in vitro have found that low affinity/positively selecting ligands elicit a modest but sustained increase in intracellular Ca2+, whereas high affinity/negative selecting ligands elicit large amplitude Ca2+ pulses, which are often more transiently sustained (Melichar et al., 2013; Lo et al., 2012). In summary, these studies demonstrate how the context of TCR stimulation can determine the nature of the downstream Ca2+ signal and that this in turn has consequences to the cellular processes that are initiated (e.g. survival vs. apoptosis). Given this contribution of the Ca2+ signalling profile to the outcome of TCR engagement, there is substantial interest in the development of methods that can fully characterize the spatiotemporal features of these signals. The introduction of Ca2+-sensitive fluorescent dyes such as fluo4, indo-1 and fura-2 has revolutionized our understanding of Ca2+ signalling as it has enabled the analysis of Ca2+ signalling by microscopy (Bootman et al., 2013) and by conventional flow cytometry (CFC) (Burchiel et al., 2000). Each of these methods have their respective advantages and disadvantages, but it is worth noting that while CFC offers a high throughput method of analysing Ca2+ signals in heterogeneous cell preparations, it fails to provide the spatial information afforded by microscopy, albeit at a comparatively lower throughput. In this respect, imaging flow cytometry (IFC) offers a unique opportunity to gather spatial information of the Ca2+ signal, while preserving the high throughput nature of CFC (Filby and Davies, 2012; Filby et al., 2011; Hawkins et al., 2013; Thaunat et al., 2012). The ability of IFC to combine these advantageous features makes it an attractive

addition to the immunologists' ‘signalling toolkit’. The ability to simultaneously analyse the spatial and kinetic features of Ca2+ signalling is of particular interest as it likely differs in duration, amplitude, oscillatory frequency and spatiotemporal kinetics based on the context of stimulation (Berridge et al., 2000). To this end, we have developed a methodology whereby we labelled T cells with vital dyes that mark the ER and mitochondria in conjunction with intensiometric reporters of intracellular Ca2+ levels. This allows us to then activate these cells under various conditions and acquire samples on an IFC system to address the spatiotemporal nature of the Ca2+ flux. 2. Material and methods 2.1. Labelling of Jurkat cells with vital organelle and Ca2+ sensor dyes E6.1 Jurkat cells (FHCRC-derived clone, Cell Services, CRUK) were cultured in Ca2+, sufficient RPMI media (Cat no 31870-082, Life Technologies Inc., USA) containing 10% FBS, penicillin/streptomycin, glutamine and 2-mercaptoethanol, counted and checked for viability using a Vi-Cell counter (Beckman Coulter Inc., USA). Cells were washed once in serum free RPMI prior to labelling with combinations of ER Tracker Red (ERTR, Ex 587 nm, Em 615 nm, E34250), MitoTracker Deep Red (MitoDR, Ex 644 nm Em 665 nm, M22426), Fluo4-AM (Ex 488 nm, Em 530 nm, F-14201), Mag-Fluo4-AM (MagF4, Ex 488 nm, Em 530 nm, M-14206), and Rhod2-AM (Ex 552 nm, Em 581 nm, R-1245MP) at a cell density of 2 × 106/ml for 20 min at 37 °C. All dyes were purchased from Life Technologies, Paisley, UK. In the initial experiments, we found that Rhod2 did not readily accumulate in mitochondria (Fig. S1A) therefore we used Dihydro-Rhod2 (DiH-Rhod2). DiH-Rhod2 was synthesised immediately prior to use by adding a small excess of sodium borohydride (NaBH4) dissolved in methanol until the Rhod2 solution became colourless (Bowser et al., 1998). The final optimised dye concentrations used were: 2 μM Fluo4, 5 μM MagF4, 6 μM DiH-Rhod2, 0.5 μM ERTR and 5 nm MitoDR. Cells were washed post-labelling a minimum of three times into Ca2+ free DMEM media (21068-028) and adjusted to a final cell density of 30 × 106/ml. For stimulation conditions that required Ca2+, cells were washed back into standard RPMI media. 2.2. Generation of an NFAT-1 GFP expressing Jurkat line HA-tagged murine NFAT1 fused to eGFP was obtained from the lab of Dr. Rao via Addgene (plasmid 11107) as described (Aramburu et al., 1999) and subcloned into a retroviral plasmid (pMSCV-PGK-puromycin, Clonetech). Retrovirus was produced by transfecting a 100 mm dish of low passage 293 T cells (ATCC) with 3 μg of VSV-G, 3 μg of Gag-Pol, and 4 μg of HA–NFAT1–eGFP using Lipofectamine-2000 (11668027, Life Technologies) according to the manufacturer's instructions. Retrovirus was harvested at 48 and 72 h post-transfection, passed through a 0.45 μm filter and pooled. About 1 × 106 Jurkat cells were infected in a 6-well plate with 1 ml of viral supernatant plus 1 ml of medium supplemented with 4 μg/ml polybrene (Sigma), spun at 1180 g for 90 min at room temperature and placed for 3–4 h. in the incubator before medium replacement after which antibiotic selection (1 μg/ml puromycin) was initiated. Cells were sorted for GFP positivity using an INFLUX cell sorter (Becton Dickinson, USA) and maintained without selection in full RPMI media. The percentage of GFP positive cells was monitored by CFC and stabilised at ~ 60% (Fig. S2) in the absence of puromycin. We consciously maintained a GFP negative population to serve as an internal control for Ca2+ flux measurements. 2.3. Measurement of Ca2+ flux by CFC Ca2 + flux measurement by CFC was conducted using a BD LSRFortessa system (Becton Dickinson, Carlsbad, USA) configured with a 355 nm trigon, 405 nm octagon, 488 nm trigon, 561 nm octagon and

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

633 nm trigon laser excitation lines and associated detector arrays. Briefly samples of Fluo4-labelled Jurkat cells were acquired using a PMT voltage in the 530/30 Blue detector of 480 V for a minimum of 60 s to establish a baseline for resting cytoplasmic Ca2 + levels. DAPI was added prior to this step to identify dead cells (Violet 450/40). The sample tube was removed from the sample inlet port (SIP) to enable the addition of the experimentally-determined concentration of antiCD3 (OKT3, LEAF purified, 317304, Biolegend, UK) and the cytometer was left in acquisition mode to allow for the rapid return of the stimulated sample to the SIP. Samples were then recorded for a further 360 s to measure any increase in Fluo4 intensity. As a positive control to check for complete Fluo4 loading and to determine the maximal possible Ca2+ response, 1 μM ionomycin (Cat no, SIGMA) was added to the samples before being returned once again to the SIP for a further 180 s. FCS files were analysed using the kinetic platform within FlowJo 9.75 (Treestar Inc., USA) after first gating on live (DAPI negative), intact (monotonic for FSC-A/SSC-A) single (SSC-W) cells.

2.4. Measurement of Ca2+ flux by IFC Ca2 + flux measurement by IFC was conducted using either a FlowSight (FS) or dual camera ImageStream X MKII (ISx) system with multi-mag capabilities (Amnis, Seattle, USA). Both systems were equipped with a 488 nm, 405 nm, 561 nm and 642 nm excitation lasers. Bright-field (BF) illumination was collected in channels 1 (camera 1, 430 nm–48 nm) and 9 (camera 2, 560 nm–595 nm). All systems were fully ASISST calibrated with additional quality control performed using Cyto Cal 6 peak beads (FC3MV, Thermo Fisher, USA). FS magnification was fixed at 20×, whereas ISx data was collected at 60×. The 20× limited FlowSight was used in initial experiments to determine the ability of the common IFC fluidics system to cope with the kinetics of Ca2 + flux measurement. Subsequently the ISx platform was used at 60 × magnification in order to provide the resolution capable of determining the spatial context. In all cases the high sensitivity fluidics mode was selected within the INSPIRE acquisition software. Fluo4/MagF4 emission was measured from the 488 nm laser in CH2 (505 nm–560 nm), DiHRhod2 emission was measured from the 488 nm/561 nm lasers in CH3 (560 nm–595 nm), ERTR emission was measured from the 561 nm laser in CH4 (595 nm–642 nm), MitoDR emission was measured from the 642 nm laser in CH11 (642 nm–740 nm) and Live Dead Near Infra-red (LD-NIR, L10119, Life Technologies) emission was measured from the 654 nm laser in CH12 (740 nm–800 nm). The excitation laser powers used were typically 20 mW (488 nm), 30 mW (561 nm) and 150 mW (642 nm) and were set so as to avoid saturation of the 12-bit CCD camera (raw max pixel values below 4096). An acquisition storage gate was set on a plot of CH1 BF versus CH1 BF aspect ratio to exclude small debris from the file. Single stained controls were collected with BF illumination and scatter laser turned off in order to generate a compensation matrix post-acquisition as described previously (Filby and Davies, 2012; Filby et al., 2011). Fully stained samples (3–4 million cells in 100 μl) were first acquired for 60 s in order to determine the baseline. Acquisition was then stopped and the file saved before the sample was unloaded. The sample loading script was then initiated and once completed (~45 s) various stimulating factors were added to the tube as indicated. These included anti-CD3 (Biolegend), 1 μM thapsigargin (T9033-5MG, SIGMA, Gillingham, UK) or Phorbol 12myristate 130 acetate (PMA, P1585, SIGMA). All stimulatory solutions were made up in Ca2+ free DMEM. In certain experiments, cells were pre-incubated with 1 μM thapsigargin/ionomycin or 1 μM Carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP, C2920, SIGMA). Data was collected for a minimum of 6 min after which samples were unloaded again to allow for the addition of 1 μM ionomycin (SIGMA) with or without 1 mM NaCl (SIGMA). The sample line was cleaned between samples by first loading a 1% bleach solution then DI H2O to ensure no carryover of stimulatory compounds.

3

2.5. Measurement of NFAT translocation and Ca2+ flux by IFC Jurkat cells expressing a GFP–NFAT 1 were labelled as described with DiH-Rhod2 and MitoDR. Dye Cycle Violet (DCV, V35121, Life Technologies) was added to the cells at a concentration of 1 μM for 5 min prior to acquisition to identify the nucleus. Baseline acquisition for 60 s followed by the addition of various stimuli was conducted exactly as described above with the additional use of the 405 nm laser to excite DCV (emission in CH7, 430 nm–505 nm). Mitochondrial Ca2+ flux was measured for a 10–15 minute period then the sample was unloaded and moved to 37 °C/5% CO2 for a total of 24 h. after which it was acquired on an IFC system to determine the degree of NFAT–GFP nuclear occupancy at assay end point. 2.6. Data analysis and feature extraction using IDEAS Raw image files (.rif) were compensated using the wizard embedded within the IDEAS V6.1 (Amnis) as described previously (Filby and Davies, 2012; Filby et al., 2011). An example compensation matrix generated from the single stained .rif samples is shown in Fig. S3. Once compensated, the image-based data was reduced using a variety of intensiometric and morphometric parameters (features) as shown in Fig. S4 using the default channel masks and pre-calculated features with the IDEAS software. Briefly dead cells were eliminated using a plot of BF Area (CH1) versus LD-NIR intensity. Focused cells from within the live gate were selected using the gradient RMS feature and setting a gate N 45 au. Single cells were then identified based on the area and aspect ratio of the BF image in CH1. Finally total integrated channel fluorescence was used to identify cells that were positive for all vital stains. Once live, in focus, single, fluorescently stained cells were identified through reductionist gating, we adapted the default channel masks for the ERT-Red (CH4, M04) and the MitoDR (CH11, M011) so as to better delineate pixels that corresponded to the ER and mitochondria respectively with the aim of restricting the measurement of various Ca2 + indicator dyes to these organelles only. The adapted masks are shown in Fig. S5 along with the IDEAS naming strings for reproduction purposes. We then asked the IDEAS software to calculate the intensity feature for each Ca2+ indicator dye from within these new masks. For the measurement of GFP–NFAT nuclear translocation, the nuclear translocation wizard was used from within the IDEAS software (George et al., 2006). GFP positivity was first determined by gating on sample of GFPnon-expressing Jurkat cells stained with all other fluorochromes ensuring that no more than 0.1% of cells were within this gate (99.9% confidence interval). These values, along with the time parameter, were exported to EXCEL for further analysis. For the measurement of signal correlation, we used the bright-detail similarity (BDS) feature within the IDEAS software calculated from the MC default mask. We chose this feature because it is best suited to situations where one is looking for correlation between two potentially discrete signals that may be aligned on the x and y axes of a 2D image but could be in different focal planes (see Fig. S1B). The BDS feature only considers pixel information that is within the plane of focus (so at 60 × magnification, a focal depth of roughly 1 μm). All other light is considered to be out of focus and as such has no weight in the correlation scoring. The equation used is such that values are never negative unlike the similarity score feature; however BDS scores of 0–1.5 are considered not or weakly correlated and whereas values of 1.5 and above are considered to be correlated. 2.7. Data analysis using EXCEL and PRISM Each time a sample was acquired (baseline or after various stimulations), a single .rif file was generated with a time stamp starting from 0. This meant that even if we merged the .rif files for a given sample using IDEAS there would not be a contiguous time series and the data would suffer from a temporal overlap. To this end we took the exported values

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

4

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

for each sample (baseline, stimulated, ionomycin treated etc.) and used a custom EXCEL sheet (Microsoft, Seattle, USA) to firstly bin the events into whole second integers (Fig. S6) using the “=ROUND” argument. This now meant that we had multiple intensity measurements for each whole second of acquisition (typically in the order of 20 cells per whole second integer). To obtain an average intensity value for each whole second, we used the “AVERAGEIF” argument within EXCEL. We then reconstructed a contiguous time line in seconds and exported the values to GraphPad Prism (V6.0d). Smoothing was applied using the 6th order (4 nearest neighbours) option within the software. In certain situations, MagF4 and DiH-Rhod2 intensities were expressed as a % of the initial values. In all cases, values were expressed as a function of time in seconds. 3. Results 3.1. IFC-systems are able to record the initiation of Ca2+ flux as long as the stimulus is titrated appropriately It is well established that Ca2+ flux occurs rapidly after ligation of the TCR by various agonists such as anti-CD3 antibodies (Lovatt et al., 2006) and pMHC complexes (Filby et al., 2007). CFC systems are generally able to cope well with such rapid signalling as stimulation can be achieved with minimal impact on the sample acquisition or in certain systems stimulus can be added without interrupting sample flow (Vines et al., 2010). IFC systems require the full sample to be loaded meaning that any stimulus addition will require an unloading step followed by another loading cycle. We routinely find that the whole process from unloading a sample, running the load sample script, adding the stimulus and re-loading the sample can take up to 100 s before data acquisition can begin again. As such we began by assessing the ability of an IFC system to be able to record the upturn in the Ca2+ flux response while factoring in the lag time of unloading and re-loading the sample. Classically the ratiometric dye Indo-1 is used as it fluoresces in both Ca2+ bound and unbound states helping to determine if cells have been evenly

loaded. Due to the fixed filter configuration of IFC systems, we were unable to use Indo-1 as there is currently no way to differentiate the bound or unbound emission. Instead we began by using the intensiometric dye Fluo-4 that increases its fluorescence when bound to Ca2+. This makes determining if all cells have been loaded evenly with dye difficult, however by using a strong stimulus such as the Ca2+ ionophore ionomycin loading uniformity can be established through measuring the maximal response. To this end we loaded Jurkat T cells with 2 μM of Fluo-4 and measured CD3-induced Ca2 + flux using both CFC and IFC systems over a range of CD3 concentrations. We initially chose to use the FlowSight platform (20× restricted) as we simply wanted to determine if an increase in the total Fluo4 integrated fluorescence could be recorded under conditions of sample loading and un-loading. Fig. 1A shows that the CFC system was able to capture the initiation of the flux over the full range of CD3 concentrations tested and that ionomycin produced a maximal response. In contrast, the delay in sample un-loading and re-loading (around 1 min to resuming data collection) meant that FlowSight IFC only captured the contraction of the Ca2+ flux in response to 10 μg/ml CD3. However, at both 1 and 0.1 μg/ml, we were able to capture the entire flux. Interestingly, the Fluo4 response to ionomycin recorded by IFC was very different to that of CFC and may reflect the gross difference in sample handling between the two platforms. CFC draws the sample continuously from a pressurised tube whereas the IFC system loads the entire sample into an internal “holding” area ready to be pushed into the flow via syringe pressure. Based on these data we concluded that the effect of sample re-loading times could be mitigated by properly titrating the stimulus and used 1 ug/ml CD3 for all future experiments. 3.2. IFC combined with vital organelle dyes reveals the Juxtaposition of the ER and mitochondria in Jurkat T cells We specifically wanted to use IFC to measure the spatial as well as temporal aspect to CD3-induced Ca2 + mobilisation. The major

Fig. 1. Fluo4 serves as an effective reporter of CD3 induced Ca2+ mobilisation by IFC as long as the stimulus is properly titrated. Fluo4-labelled Jurkat cells were activated with the indicated CD3 concentrations and acquired on (A) an LSRFortessa CFC system or (B) a FlowSight IFC system (20× magnification). Plotted values are expressed as the average median Fluo4 intensity on a log scale at whole second integers. Data is representative of 3 independent experiments.

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

intracellular compartments involved in Ca2+ signalling are the ER and mitochondria (Smith-Garvin et al., 2009; Hoth et al., 2000) as such we used ERTR and MitoDR to fluorescently label each organelle respectively. If titrated correctly these dyes are highly specific but if used at relatively high concentrations they have been shown to lose specificity (personal observation). At this stage, we switched to using the ISx system with the 60× objective selected because we wanted to improve our spatial resolution in the x, y and z axes. First we performed something that we describe as a “spatial titration” whereby cells were dual labelled with varying concentrations of ERTR and MitoDR. We then measured the degree of spatial overlap using the bright-detail similarity (BDS) feature within the IDEAS analysis software in concert with the overall balance between the intensities of the two stains being compared. We have noted previously that the BDS score is highly influenced by the relative intensities of the two dyes being compared. Fig. 2A shows that the lowest level of signal overlap in concert with the most balanced staining intensity between the two dyes was achieved using 0.5 μM of ERTR and 5 nM of MitoDR (sample tube 7). Sample condition 1 gave a similar BDS score as tube 7 however we felt that the fact that we achieved a more balanced staining between the two dyes made this the favourable concentrations to take forward. However, a BDS value of ~1.3 shows that there was some degree of relatively weak signal overlap that was likely to the limitations of magnification even at 60× (Fig. 2B). This highlights the previously reported close proximity between the ER and mitochondria within Jurkat cells. 3.3. Fluo4 is not suited to measuring either ER or mitochondrial Ca2+ levels Having established the conditions for ERTR and MitoDR labelling we sought to determine the best way of measuring Ca2+ levels within these organelles. It has been suggested that many Ca2+ indicators are unable

5

to enter certain sub-cellular organelles or if they do, are unable to function due to high local Ca2+ concentrations (Lambert, 2006). As such we wanted to determine whether Fluo4 could report ER and mitochondrial Ca2+ dynamics. We began by labelling resting Jurkat cells with Fluo4 and ERTR, the idea being that in un-stimulated cells bright Fluo4 signal should be restricted to the Ca2+ rich ER. We again used the BDS score due to the fact that we were looking at the degree of signal overlap between at least one of which should always be discrete in nature. Fig. 3A shows that the Fluo4 and ERTR signals showed no correlation (mean BDS score of 0.4) meaning that it was either unable to enter and accumulate in the ER, or it could do so but was unable to fluoresce under such Ca2+ rich conditions possibly due to the fact that it has such high affinity for Ca2 +. Furthermore when we labelled Jurkat cells with Fluo4 and MitoDR and this time activated them with 1 μg/ml CD3 to induce a flux (Fig. 3B) we again noted that they were not correlated (mean BDS score of 0.3) indicating that Fluo4 is also unable to report mitochondrial Ca2+ levels. 3.4. MagFluo4 accumulates in the ER of Jurkat cells and reports stimulation-induced Ca2 + release Having established that Fluo4 was not suited to measuring organelle Ca2 + dynamics, we switched our focus to two other Ca2 + indicator dyes. The low affinity Ca2+ indicator MagF4 has been shown to preferentially accumulate in the ER lumen of several cell types (Lambert, 2006) making it a particularly strong candidate for our purposes. Fig. 4A shows that MagF4 and ERTR signals correlated with one another both in qualitative (image panel) and quantitative terms (BDS median of. 2.5). Fig. 4B shows that within the ERTR-defined cell area, resting cells had high levels of MagF4 fluorescence but after stimulation with either 1 μg/ml CD3 or the SERCA pump inhibitor thapsigargin, MagF4

Fig. 2. Optimisation of ERTR and MitoDR labelling conditions based on IFC-derived intensity and BDS feature values. (A) A graph and table showing the range of dye concentrations tested. The delta in the median fluorescence for each dye channel is plotted (left y-axis) with the associated BDS score overlaid (right y-axis). The values shown on the x-axis (1–9) correspond to the sample labelling concentrations shown in the table below the graph. (B) Multispectral images (upper panels) taken at 60× using an ISx system showing the spatial relationship between ERTR (0.5 μM) and MitoDR (5 nm) signals from sample tube 7. The BDS value for each image set is shown on the right side of the panels. A bi-variate plot of ERTR versus MitoDR intensities (lower left panel) and a histogram with population median of the associated BDS score between the two signals. (lower right panel) is also shown.

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

6

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

Fig. 3. Fluo4 does not report ER or mitochondrial Ca2+ levels. (A) Multispectral images taken at 60× using an ISx system of resting Jurkat cells labelled with Fluo4 and ERTR. A bi-variate plot of Fluo4 versus ERTR intensities (lower left panel) and a histogram of the associated BDS score with population median between the two signals (lower right panel) are also shown. (B) As described for A, except showing the relationship between Fluo4 and MitoDR signals after activation with 1 μg/ml CD3.

levels sharply decreased within the ER over the first 40 seconds poststimulation. When ionomycin was added we noted a further decrease in the ER signal but also an increase in the cytoplasmic fluorescence levels (Fig. 3C). This highlighted the importance of having the spatial information to dissect the ER and non-ER MagF4 levels as any decrease in the ER may be masked by an increase in the cytoplasmic signals (data not shown) or at the very least, confirm the specificity of the staining. 3.5. Di-hydro Rhod2 is able to report CD3-induced mitochondrial Ca2 + dynamics In order to measure mitochondrial Ca2+ levels we used DiH-Rhod 2 as we found that Rhod2 failed to specifically load into the mitochondrial compartment of Jurkat T cells (Fig. S1, mean BDS score of 1.29). In contrast DiH-Rhod2 showed an excellent correlation with MitoDR signal (Fig. 5A) both in qualitative (image panel) and quantitative terms (mean BDS score of 3.1). Furthermore, CD3 stimulation (1 μg/ml) of DiH-Rhod2 and MitoDR labelled Jurkat cells showed a rapid increase (within 40 s) in DiH-Rhod2 fluorescence within the MitoDR-defined area over time compared to the un-stimulated baseline levels. Furthermore, the addition of ionomycin led to a second wave of Ca2+ entry into the mitochondria (Fig. 5B). In qualitative terms, the multi-spectral imagery (Fig. 5C) showed that the majority of DiH-Rhod2 signal resided within the MitoDR defined cellular area. 3.6. Simultaneous measurements of ER and mitochondrial Ca2+ levels a dynamic interplay between the two organelles Having established that MagF4 and DiH-Rhod2 could report ER and mitochondrial Ca2+ levels respectively we combined all stains into a single panel that was spectrally compatible with both IFC systems. One very interesting observation from the data in Fig. 5B was that mitochondria

could seemingly load Ca2+ even in the absence of extracellular sources suggesting that they were able to respond to ER-derived release. This is contradictory to the widely held belief that in T cells mitochondria only act to buffer the extracellular derived Ca2+ through close proximity to the CRAC (Hoth et al., 1997, 2000; Rizzuto et al., 2000). To directly test this hypothesis using our novel method we labelled Jurkat cells with MagF4, DiH-Rhod2, ERTR and MitoDR and washed several times (3+) into Ca2 + free DMEM. To certain tubes, we added the mitochondrial transport inhibitor FCCP in order to block mitochondrial Ca2+ uptake prior to CD3 stimulation (Hoth et al., 1997). This would prove that DiH-Rhod2 was specifically reporting mitochondrial Ca2+ levels. Furthermore, to prove the direct relationship between ER Ca2+ release and mitochondrial uptake, we pre-depleted the ER stores using thapsigargin and ionomycin for 15 min prior to CD3 stimulation. As part of the stimulation timeline, we also added back extracellular Ca2+ after a period of time to check organelle function. Fig. 6A shows that under Ca2+ free conditions, again anti-CD3 induced a modest decrease in the ER-Ca2 + levels followed by a more profound decrease when ionomycin was added. As expected, the addition of extracellular Ca2+ leads to the re-charging of the ER. Within the same sample, we found that DiH-Rhod2 fluorescence increased upon CD3 stimulation in line with the drop in ER levels (Fig. 6B) and began to decrease again around 250 seconds post-stimulus. The addition of ionomycin induced a second flux in the mitochondria that again correlated with the further decrease in ER signal. As expected, the addition of extracellular Ca2 + caused a strong increase in mitochondrial DiH-Rhod2 levels supporting a role in buffering CRAC channel Ca2 + levels (Demaurex et al., 2009). Pretreating cells with FCCP had no effect on CD3 or ionomycin-induced Ca2+ release from the ER but did completely abrogate the increase in mitochondrial DiH-Rhod2 fluorescence. The addition of extracellular Ca2+ did lead to a modest increase in DiH-Rhod2 fluorescence suggesting that mitochondria could still respond however it was much less than

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

7

Fig. 4. MagFluo4 accumulates in the ER of resting Jurkat cells and can report stimulus induced ER Ca2+ release. (A) Multispectral images from an ISx system (60×, left panel) showing resting Jurkat cells labelled with MagF4. A bi-variate plot showing the intensities of ERTR and MagF4 (middle panel) and a histogram of the associated BDS score between the two signals with population median included (right panel). (B) A kinetic plot of MagF4 intensity (% of initial value) from within the ERTR defined cell area at baseline levels (BL), after the stimulation with 1 μg/ml CD3 or thapsigargin as well as after the addition of ionomycin. (C) Example multispectral images taken at 60× from within each aligned period of measurement. Data is representative of at least three independent experiments.

in controls. The most interesting observation with the FCCP treatment was that inhibition of mitochondrial Ca2+ uptake affected the ability of extracellular Ca2+ to recharge the ER stores, suggesting that Ca2+ movement is a two way process between these organelles. Pre-treatment with thapsigargin and ionomycin emptied the ER stores as shown by the baseline MagF4 levels (Fig. 6A). In the absence of ER stored Ca2+, CD3 stimulation or ionomycin treatment failed to induce an increase in DiH-Rhod2 fluorescence in the mitochondria, however the addition of extracellular Ca2+ did elicit a mitochondrial response. As expected, the inhibitory effect of thapsigargin on the ER SERCA pump blocked the recharging of the ER by extracellular Ca2+. Collectively, these data highlight the power of our methodology in measuring the spatiotemporal characteristics of Ca2+ mobilisation and reveal a previously unappreciated interplay between ER and mitochondrial Ca2+ stores in activated T cells.

3.7. IFC uniquely allows for the simultaneous kinetic measurement of Ca2+ mobilisation and NFAT translocation after CD3 stimulation Having established we could measure the spatiotemporal Ca2+ dynamics in CD3 activated Jurkat cells by IFC, we wanted to further exploit the imaging capabilities of the system. Sustained increase in intracellular Ca2+ results in the de-phosphorylation of NFAT allowing it to enter the nucleus and drive gene expression (Macian, 2005) and IFC can be used to quantify nuclear translocation (George et al., 2006; Maguire et al., 2011). To this end we generated a Jurkat line stably expressing GFP in around 60% of all cells (Fig. S2). To test if the NFAT–GFP was able to translocate to the nucleus we treated cells with PMA and ionomycin and measured the degree of similarity (translocation) between the NFAT–GFP and nuclear DCV images. Reassuringly, we were

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

8

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

Fig. 5. DiH-Rhod2 is able to report mitochondrial Ca2+ dynamics by IFC. (A) Multispectral images taken by an ISx system (60×, left panel) showing Jurkat cells labelled with DiH-Rhod2 and MitoDR and stimulated with 1 μg/ml CD3. A bi-variate plot showing the intensities of MitoDR and DiH-Rhod2 (middle panel) and a histogram of the associated BDS score between the two signals with population median included (right panel). (B) A kinetic plot of DiH-Rhod2 intensity (% of initial value) from within the MitoDR defined cell area at baseline levels (BL), after the indicated stimuli (1 μg/ml CD3) and after the addition of ionomycin. (C) Example multispectral images taken at 60× (ISx) from within each aligned period of measurement. Data is representative of at least three independent experiments.

able to detect a rapid increase in translocation (Fig. S7) conforming that our GFP tag did not in any way inhibit the translocation function of NFAT in response to strong signals that bypass the TCR. We then labelled the GFP expressing cells with DiH-Rhod2, MitoDR and DCV to stain the nucleus in live cells with minimal effect on viability (Begum et al., 2013). We then activated these cells with 1 μg/ml CD3 in the presence of Ca2 + and measured DiH-Rhod2 fluorescence within the MitoDRdefined cell area and correlated this with the degree of NFAT–GFP nuclear translocation for 20 min (Fig. 7A). As expected, we were able to measure a mitochondrial flux that peaked within the first 60 s prior to stimulation and did not return to baseline levels for the duration of measurement (~800 seconds). Unsurprisingly, based on the reported kinetic

of NFAT translocations (Maguire et al., 2013) in the majority of cells NFAT remained excluded from the nucleus within this short time frame. However, maybe more surprisingly we monitored NFAT translocation for up to 6 h and were unable to detect significant nuclear occupancy (data not shown, manuscript in preparation). In contrast when we analysed the same sample again by IFC 24 h later we were able to detect an increase in the population NFAT–GFP translocation score corresponding to cellular images with clear NFAT–GFP signal within the nucleus (Fig. 7B). Collectively these data show that by using our IFCbased method, we can directly correlate the spatiotemporal nature of Ca2+ mobilisation with TCR-distal signalling events after delivering a defined stimulus to our T cells.

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

9

Fig. 6. The multi-spectral properties of IFC allow simultaneous measurement of ER and mitochondrial Ca2+ levels, revealing a dynamic interplay between the two organelles. (A) A kinetic plot of MagF4 intensity within the ERTR-defined cell area for the indicated pre-treatment conditions and in response to the indicated periods of stimulation with 1 μg/ml CD3. (B) As for A, but plotting the DiH-Rhod2 intensity within the MitoDR-defined cell area as a % of initial values within the same cells as in A. (C) Example multispectral images taken at 60× using an ISx system from within each aligned period of measurement. Data is representative of at least three independent experiments.

4. Discussion The aim of this study was to develop a method able to combine the high throughput cellular acquisition rates afforded by CFC with the localisation of signal provided by a microscope to measure not only the temporal, but spatial aspect of the CD3-induced Ca2+ response. As such IFC was a perfect technology platform for this purpose due to its ability to rapidly acquire 12 spectrally decomposed images of each cell at up to 60× magnification. We show that controlled activation conditions allow for kinetic intensiometric measurements to be made with similar efficiency compared to CFC. One major technical issue that we faced however was the delay in measurement time caused by having to unload and re-load samples for stimulus addition. We found that titrating the stimulus appropriately largely mitigated this problem. We have tested different ways in which an ant-TCR stimulus could be introduced with slower activation kinetics such as CD3-coated microspheres (data not shown) or through conjugate formation with antigen presenting cells (Filby et al., 2007). However for cell types such as platelets where activation-induced Ca2 + flux occurs within seconds (Davies et al., 1988) our IFC method may not be suitable. Our IFC-based method excelled was when we began to analyse the multispectral imagery output from the ISx system. The idea of a “spatial titration” whereby we evaluated not only the total fluorescence of the ERTR and MitoDR signal but also how they related in location to one another within a large

population of single cells is as far as we know unique. The imaging and throughput capabilities of IFC allowed us to carry out such an optimisation step that is impossible using CFC and impractical by traditional microscopy. As such we were able to rapidly find the right balance of the two dyes to ensure minimal non-specific staining and to the direct the measurement of Ca2 + to defined cellular structures. Importantly, our method allowed us to determine which intensiometric Ca2+ indicators were best suited for measuring cytoplasmic, ER or mitochondrial Ca2+ levels. Our data highlighted that Fluor 4 does not work well as a reporter of either ER or mitochondrial Ca2+ levels. Two possible explanations for this observation are that Fluo4 cannot access these organelles or, if able to enter cannot cope with the relatively high levels of Ca2+ that may be present in the ER for example (Lambert, 2006). Fortunately we found that MagF4 was highly specific for the ER and performed well as a reporter of ER-localised Ca2 + levels. In fact we reason that it could be used in conjunction with CFC to report changes in ER-Ca2+ levels and would be preferential to using a cytoplasmic resident indicator under Ca2+ free conditions as this makes the assumption that all Ca2+ detected in the cytoplasm comes from the ER and not from other potential intracellular sources. Furthermore it also overcomes the criticism often levelled at intensiometric dyes that it is impossible to determine if equal loading has been achieved. In the case of MagF4, resting cells should all be strongly fluorescent the only issue would be if the increase in cytoplasmic Ca2+ levels was sufficient to cause an increase in MagF4

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

10

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

Fig. 7. IFC allows for the simultaneous measurement of mitochondrial Ca2+ dynamics and the nuclear translocation of NFAT. (A) A kinetic plot of DiH-Rhod2 intensity within the MitoDRdefined area (left y-axis) and the NFAT–GFP nuclear translocation score for the same cells (right y-axis) of Jurkat cells activated with 1 μg/ml CD3 and acquired on an ISx system. The break in the x-axis corresponds to the period where the cells were unloaded and then incubated for 24 h before being analysed again by IFC for translocation. (B) Example multispectral images from within each aligned period of measurement. Data is representative of at least three independent experiments.

fluorescence that masked the decrease in the ER. In terms of reporting mitochondria Ca2+ levels, we would caution against using Rhod2 on non-imaging systems to infer mitochondrial Ca2+ levels as it has been shown to also load into the cytoplasm and nucleoli (Lambert, 2006) and in our hands gave a BDS correlation score indicating no correlation with the MitoDR signal. Fortunately conversion of Rhod2 to DiH-Rhod2 largely circumvented this issue however there were still some experiments where specificity for the mitochondria was not absolute (data not shown) meaning that imagery was absolutely required in order to make accurate measurements of mitochondria Ca2+ levels. Interestingly, when we compared the correlation of each respective indicator and organelle we found that DiH-Rhod2/MitoDR gave higher BDS values than MagF4/ERTR. This is most likely due to the fact that not all of the ER is able to store calcium (Koch, 1990). One of the most interesting observation from this study came when we analysed the response of mitochondria to CD3 stimulation in the absence of extracellular Ca2+. It is a widely accepted theory that mitochondria act as a sink only for Ca2+ that enters T cells through the CRAC channels (Hoth et al., 1997, 2000). The fact that we could still detect a significant Ca2+ response in mitochondria in the absence of extracellular Ca2+ led us to postulate that they may in fact be able to respond to ER-derived Ca2+ release as well. By staining cells with both mitochondrial and ER dyes and the associated Ca2+ reporters, we were indeed able to elegantly show that there was in fact dynamic interplay between the two organelles. Firstly, depletion of ER Ca2+ levels abrogated the observed increase in mitochondrial levels seen in the untreated controls. Reassuringly, the addition of extracellular Ca2+ showed that the mitochondria were still fit to respond. Secondly, we noted that pre-treating cells with the mitochondrial inhibitor FCCP not only prevented mitochondrial Ca2+ uptake as expected, but seemed to affect the ability of the ER to recharge itself

in response to the addition of extracellular Ca2+. It should be noted that our current data does not preclude the possibility that FCCP exerts an effect on the ability of the ER to load Ca2+ that is independent of its effect on mitochondria. However in many ways, such dynamic interplay is not unexpected due to the close proximity between the two organelles and the plasma membrane in Jurkat cells. This leads us to the question what effect disrupting the mitochondrial and ER proximity would exert on Ca2+ mobilisation. It is known in B cells that disruption of the actin cytoskeleton by cytochalasin D initiates a Ca2+ flux (Mattila et al., 2013). We also noted that treatment of cell with FCCP seemed to induce a more diffuse mitochondrial-staining pattern (data not shown) and this may explain why the ER failed to recharge properly if proximity to the mitochondria is required. It is possible that certain cell types in certain circumstances may position their mitochondria differently with respect to the ER and plasma membrane and that this could affect the nature and outcome of Ca2+ flux. Moreover, cells may even have different relative levels of ER-stored Ca2+ and in combination with differential expression of various ion channels could explain why certain T cell subsets have very different Ca2+ flux characteristics (Orban et al., 2014). Finally the ability to combine spatiotemporal measurements of Ca2+ dynamics with nuclear translocation of NFAT will be a powerful tool in elucidating how differential Ca2+ flux can drive gene transcription and changes in cellular function and fate (Hoth et al., 2000; Dolmetsch et al., 1998; Lewis, 2003). To this end we are constructing lymphocyte lines expressing NFAT isoforms tagged to spectrally distinct fluorescent protein tags to see if we get differential nuclear translocation and to be able to include MagF4 into the panel without any conflict with GFP fluorescence. We are also using our method to look at primary lymphocytes and how the spatiotemporal nature of Ca2 + fluxes may differ depending on the context of stimulation and cell lineage and

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030

J. Cerveira et al. / Journal of Immunological Methods xxx (2015) xxx–xxx

combining this with vector-based introductions of fluorescently tagged NFAT into primary T cells so we can combine all measurements into a single system. 5. Conclusions Ca2+ is a key process in cell signalling, notably in lymphocytes. There still remains much debate as to how such a ubiquitous response can potentially elicit very different functional outcomes. One compelling argument is that the spatiotemporal nature of the Ca2+ flux may dictate a particular fate. We feel that our IFC-derived method offers the potential to look directly at the involvement of the ER, mitochondrial and cytoplasmic Ca2 + dynamics and also potentially correlate this with functional outcomes such as nuclear translocation of factors such as NFAT or NFκB. It could also be extended to look at other markers of T cell fate such as cytokines due to the multi-parameter nature of IFC. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jim.2015.04.030. Acknowledgements JC, JB, RDMB, AVDV and AF are funded by Cancer Research UK. AF also acknowledges the support from the ISAC SRL emerging leaders programme. We would also like to thank James Patterson from the LRI Cell Cycle laboratory for assistance with the figures. References Aramburu, J., et al., 1999. Affinity-driven peptide selection of an NFAT inhibitor more selective than cyclosporin A. Science 285 (5436), 2129. Begum, J., et al., 2013. A method for evaluating the use of fluorescent dyes to track proliferation in cell lines by dye dilution. Cytometry A 83 (12), 1085. Berridge, M.J., Lipp, P., Bootman, M.D., 2000. The versatility and universality of calcium signalling. Nat. Rev. Mol. Cell Biol. 1 (1), 11. Bhakta, N.R., Oh, D.Y., Lewis, R.S., 2005. Calcium oscillations regulate thymocyte motility during positive selection in the three-dimensional thymic environment. Nat. Immunol. 6 (2), 143. Bootman, M.D., et al., 2013. Ca2+-sensitive fluorescent dyes and intracellular Ca2+ imaging. Cold Spring Harb. Protoc. 2013 (2), 83. Bowser, D.N., et al., 1998. Role of mitochondria in calcium regulation of spontaneously contracting cardiac muscle cells. Biophys. J. 75 (4), 2004. Bueno, O.F., et al., 2002. Defective T cell development and function in calcineurin A betadeficient mice. Proc. Natl. Acad. Sci. U. S. A. 99 (14), 9398. Burchiel, S.W., et al., 2000. Analysis of free intracellular calcium by flow cytometry: multiparameter and pharmacologic applications. Methods 21 (3), 221. Davies, T.A., et al., 1988. Flow cytometric measurements of cytoplasmic calcium changes in human platelets. Cytometry 9 (2), 138. Demaurex, N., Poburko, D., Frieden, M., 2009. Regulation of plasma membrane calcium fluxes by mitochondria. Biochim. Biophys. Acta 1787 (11), 1383. Dolmetsch, R.E., Xu, K., Lewis, R.S., 1998. Calcium oscillations increase the efficiency and specificity of gene expression. Nature 392 (6679), 933. Donnadieu, E., et al., 1992. Imaging early steps of human T cell activation by antigenpresenting cells. J. Immunol. 148 (9), 2643. Feske, S., et al., 2001. Gene regulation mediated by calcium signals in T lymphocytes. Nat. Immunol. 2 (4), 316. Feske, S., et al., 2006. A mutation in Orai1 causes immune deficiency by abrogating CRAC channel function. Nature 441 (7090), 179. Filby, A., Davies, D., 2012. Reporting imaging flow cytometry data for publication: why mask the detail? Cytometry A 81 (8), 637. Filby, A., et al., 2007. Fyn regulates the duration of TCR engagement needed for commitment to effector function. J. Immunol. 179 (7), 4635. Filby, A., et al., 2011. An imaging flow cytometric method for measuring cell division history and molecular symmetry during mitosis. Cytometry A 79 (7), 496. George, T.C., et al., 2006. Quantitative measurement of nuclear translocation events using similarity analysis of multispectral cellular images obtained in flow. J. Immunol. Methods 311 (1–2), 117. Hawkins, E.D., et al., 2013. Regulation of asymmetric cell division and polarity by Scribble is not required for humoral immunity. Nat. Commun. 4, 1801. Hogan, P.G., Lewis, R.S., Rao, A., 2010. Molecular basis of calcium signaling in lymphocytes: STIM and ORAI. Annu. Rev. Immunol. 28, 491. Hoth, M., Fanger, C.M., Lewis, R.S., 1997. Mitochondrial regulation of store-operated calcium signaling in T lymphocytes. J. Cell Biol. 137 (3), 633.

11

Hoth, M., Button, D.C., Lewis, R.S., 2000. Mitochondrial control of calcium-channel gating: a mechanism for sustained signaling and transcriptional activation in T lymphocytes. Proc. Natl. Acad. Sci. U. S. A. 97 (19), 10607. Huse, M., et al., 2007. Spatial and temporal dynamics of T cell receptor signaling with a photoactivatable agonist. Immunity 27 (1), 76. Irvine, D.J., et al., 2002. Direct observation of ligand recognition by T cells. Nature 419 (6909), 845. Koch, G.L., 1990. The endoplasmic reticulum and calcium storage. Bioessays 12 (11), 527. Lambert, D.G., 2006. Calcium signaling protocols. 2nd ed. Methods in Molecular BiologyHumana Press, Totowa, N.J. (x, 359 pp.). Lewis, R.S., 2001. Calcium signaling mechanisms in T lymphocytes. Annu. Rev. Immunol. 19, 497. Lewis, R.S., 2003. Calcium oscillations in T-cells: mechanisms and consequences for gene expression. Biochem. Soc. Trans. 31 (Pt 5), 925. Liou, J., et al., 2005. STIM is a Ca2+ sensor essential for Ca2+-store-depletion-triggered Ca2+ influx. Curr. Biol. 15 (13), 1235. Lo, W.L., Allen, P.M., 2013. Self-awareness: how self-peptide/MHC complexes are essential in the development of T cells. Mol. Immunol. 55 (2), 186. Lo, W.L., Donermeyer, D.L., Allen, P.M., 2012. A voltage-gated sodium channel is essential for the positive selection of CD4(+) T cells. Nat. Immunol. 13 (9), 880. Lovatt, M., et al., 2006. Lck regulates the threshold of activation in primary T cells, while both Lck and Fyn contribute to the magnitude of the extracellular signal-related kinase response. Mol. Cell. Biol. 26 (22), 8655. Macian, F., 2005. NFAT proteins: key regulators of T-cell development and function. Nat. Rev. Immunol. 5 (6), 472. Macian, F., et al., 2002. Transcriptional mechanisms underlying lymphocyte tolerance. Cell 109 (6), 719. Maguire, O., et al., 2011. Quantifying nuclear p65 as a parameter for NF-kappaB activation: correlation between ImageStream cytometry, microscopy, and Western blot. Cytometry A 79 (6), 461. Maguire, O., et al., 2013. Nuclear translocation of nuclear factor of activated T cells (NFAT) as a quantitative pharmacodynamic parameter for tacrolimus. Cytometry A 83 (12), 1096. Mattila, P.K., et al., 2013. The actin and tetraspanin networks organize receptor nanoclusters to regulate B cell receptor-mediated signaling. Immunity 38 (3), 461. Melichar, H.J., et al., 2013. Distinct temporal patterns of T cell receptor signaling during positive versus negative selection in situ. Sci. Signal. 6 (297), ra92. Michel, F., et al., 2001. CD28 as a molecular amplifier extending TCR ligation and signaling capabilities. Immunity 15 (6), 935. Nedellec, S., et al., 2010. NKG2D costimulates human V gamma 9 V delta 2 T cell antitumor cytotoxicity through protein kinase C theta-dependent modulation of early TCR-induced calcium and transduction signals. J. Immunol. 185 (1), 55. Neilson, J.R., et al., 2004. Calcineurin B1 is essential for positive but not negative selection during thymocyte development. Immunity 20 (3), 255. Oh-Hora, M., et al., 2008. Dual functions for the endoplasmic reticulum calcium sensors STIM1 and STIM2 in T cell activation and tolerance. Nat. Immunol. 9 (4), 432. Oh-Hora, M., et al., 2013. Agonist-selected T cell development requires strong T cell receptor signaling and store-operated calcium entry. Immunity 38 (5), 881. Orban, C., et al., 2014. Different calcium influx characteristics upon Kv1.3 and IKCa1 potassium channel inhibition in T helper subsets. Cytometry A 85 (7), 636. Penna, A., et al., 2008. The CRAC channel consists of a tetramer formed by Stim-induced dimerization of Orai dimers. Nature 456 (7218), 116. Picard, C., et al., 2009. STIM1 mutation associated with a syndrome of immunodeficiency and autoimmunity. N. Engl. J. Med. 360 (19), 1971. Quintana, A., et al., 2007. T cell activation requires mitochondrial translocation to the immunological synapse. Proc. Natl. Acad. Sci. U. S. A. 104 (36), 14418. Quintana, A., et al., 2011. Calcium microdomains at the immunological synapse: how ORAI channels, mitochondria and calcium pumps generate local calcium signals for efficient T-cell activation. EMBO J. 30 (19), 3895. Rizzuto, R., Bernardi, P., Pozzan, T., 2000. Mitochondria as all-round players of the calcium game. J. Physiol. 529 (Pt 1), 37. Rotnes, J.S., Bogen, B., 1994. Ca2+ mobilization in physiologically stimulated single T cells gradually increases with peptide concentration (analog signaling). Eur. J. Immunol. 24 (4), 851. Sieber, M., Baumgrass, R., 2009. Novel inhibitors of the calcineurin/NFATc hub — alternatives to CsA and FK506? Cell Commun. Signal 7, 25. Sloan-Lancaster, J., Evavold, B.D., Allen, P.M., 1993. Induction of T-cell anergy by altered Tcell-receptor ligand on live antigen-presenting cells. Nature 363 (6425), 156. Sloan-Lancaster, J., Steinberg, T.H., Allen, P.M., 1996. Selective activation of the calcium signaling pathway by altered peptide ligands. J. Exp. Med. 184 (4), 1525. Smith-Garvin, J.E., Koretzky, G.A., Jordan, M.S., 2009. T cell activation. Annu. Rev. Immunol. 27, 591. Thaunat, O., et al., 2012. Asymmetric segregation of polarized antigen on B cell division shapes presentation capacity. Science 335 (6067), 475. Vines, A., McBean, G.J., Blanco-Fernandez, A., 2010. A flow-cytometric method for continuous measurement of intracellular Ca(2+) concentration. Cytometry A 77 (11), 1091. Wulfing, C., et al., 1997. Kinetics and extent of T cell activation as measured with the calcium signal. J. Exp. Med. 185 (10), 1815.

Please cite this article as: Cerveira, J., et al., An imaging flow cytometry-based approach to measuring the spatiotemporal calcium mobilisation in activated T cells, J. Immunol. Methods (2015), http://dx.doi.org/10.1016/j.jim.2015.04.030