Surface Plasmon Resonance for Measuring Exocytosis from

Subscriber access provided by UNIVERSITY OF SOUTH CAROLINA LIBRARIES

Article

Surface Plasmon Resonance for Measuring Exocytosis from Populations of PC12 Cells: the Mechanisms of Signal Formation and Assessment of Analytical Capabilities Beatriz Moreira, Jani Tuoriniemi, Naghmeh Kouchak Pour, Lýdia Mihal#íková, and Gulnara Safina Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b04811 • Publication Date (Web): 13 Feb 2017 Downloaded from http://pubs.acs.org on February 15, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10

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 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Surface Plasmon Resonance for Measuring Exocytosis from Populations of PC12 Cells: the Mechanisms of Signal Formation and Assessment of Analytical Capabilities Beatriz Moreira†#, Jani Tuoriniemi†#, Naghmeh Kouchak Pour†, Lýdia Mihalčíková†, Gulnara Safina†,‡,* †

Department of Chemistry and Molecular Biology, University of Gothenburg, Kemigården 4, 412 96 Gothenburg, Sweden Division of Biological Physics, Department of Physics, Chalmers University of Technology, Kemigården 1, 412 96 Gothenburg, Sweden

‡

#

These authors contributed equally * To whom correspondence should be addressed: [email protected], Tel +46 766 22 90 61 KEYWORDS: Surface plasmon resonance; Exocytosis; PC12 cells; Dopamine ABSTRACT: Due to cell to cell variation it is difficult to obtain statistically significant data on the frequency of exocytosis events (Rexocytosis, t-1m-2) with traditional single cell electrophysiological or fluorescence microscopy based methods. Here we take the first steps towards a rapid cost effective surface plasmon resonance (SPR) based method for measuring the Rexocytosis for populations of PC12 cells. The conditions for culturing confluent monolayers on the sensor slides were optimized, and neurotransmitter exocytosis was evoked by injecting solutions with elevated [K+]. Exocytosis caused a shift of the resonance angle (∆θr) that was linearly proportional to Rexocytosis. The ∆θr was mainly due to elevated concentration of secretory vesicles close to the cell membrane. The increased vesicle concentration thus acted as a proxy for the Rexocytosis that could not be measured directly. The ∆θr was calibrated for Rexocytosis using single cell amperometry on parallel cell cultures. The cell populations were large enough for variation in responses between sensor slides to only reflect actual differences in biological condition. The applicability for drug screening is demonstrated by studying the effects of EGTA, reserpine, and prolonged stimulation by K+.

There is a renewed interest in the pharmaceutical industry for using in vivo screening where drug candidates are tested for response on cell cultures, or even organisms instead of the predominating conventional schemes that only assess their effect on a single target molecule1-2. Designing readout methods capable of quantifying complex biological processes while still being replicable on a mass scale requires ingenuity. Common bases for such methods are automated confocal fluorescence microscopy or measurement of the fluorescence of suspended cells in a flow cytometry set up3-4. One of the most interesting cellular processes for pharmacological manipulation is exocytosis where the cell excretes e.g. toxins, peptides, hormones or neurotransmitters loaded inside vesicles that merge with the cell membrane (Figure 1)5. Impaired rate of secretion of insulin and dopamine (DA) are for instance characteristic for diabetes and Parkinson’s disease. Studying the dynamics of exocytosis demands methods6 capable of following the course of secretion in real time. Electrophysiological methods7 can resolve individual exocytosis events on a single cell basis. The capacitance increments upon merging of vesicles with the cell membrane can be detected in a patch clamp setup. However, no spatial or chemical information can be obtained. In single cell amperometry or fast scan cyclic voltammetry using carbon fiber microelectrodes (CFME), an electrode with a diameter of a few μm is brought in contact with the cell membrane and the secretion events are recorded as current spikes. Spatial information from

areas of a few hundred of μm in size can be obtained by electrode arrays. CMFE techniques have been important for furthering the understanding of secretion of catecholamine (CA) based neurotransmitters that are easily oxidized on bare carbon surfaces. However, detection of e.g. insulin from pancreatic βcells requires functionalizing the carbon surface with materials, on which the oxidation would occur at sufficiently low voltages8. Furthermore, electrophysiological experiments demand highly skilled personnel, the mastering of patch clamp techniques often requiring long and extensive training. Individual exocytosis events can be visualised with total internal reflection fluorescence microscopy (TIRFM)9 and interference reflection microscopy (IRM)10. In TIRFM the vesicles are stained with fluorescent labels, and those near the cell membrane that is in contact with the microscope slide are visualized by excitation with an evanescent field. As with other techniques requiring fluorescent labelling, phototoxicity, photobleaching, and possible interferences from the dyes is of concern. In IRM the sample is placed on top of a glass coverslip and illuminated at an oblique angle from below. The image is formed from the reflected light below the sample. Light reflected from objects within 100 nm from the glass surface interfere destructively with the light reflected by the coverslip creating enough contrast to visualize vesicles approaching the membrane10. Cells are viable only for a limited time outside incubators restricting the time frame for experiments to < 1 h. It is diffi-

ACS Paragon Plus Environment

Analytical Chemistry

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 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

cult to collect information from a statistically significant number of cells within this time frame using the techniques mentioned above. For drug screening applications, it would also be more relevant to simultaneously measure the average rate of exocytosis for a large population of cells rather than gathering detailed information on individual exocytosis events. The simple option to quantify the released molecules in outflow from a chamber containing a cell culture was described by Kasai et.al.11 They measured the ATP and CA secreted by bovine chromaffin cells using a colorimetric and amperometric method respectively. A better time resolution and sensitivity can be achieved by culturing the cells on electrode surfaces of sensors designed for this purpose12. However, electrodes are prone to fouling by cell extrudates13, and the difficulties of quantifying molecules of interest in a complex cell culture fluid has so far been deterring from the use of potentially more sensitive mass spectrometry methods. Quartz crystal microbalance has been shown to detect the changes in viscoelastic properties upon exocytosis of a cell layer grown on a sensor slide, although no method for extracting quantitative information was devised14. Here a surface plasmon resonance (SPR) based method is developed for measurement of the rate of K+ evoked CA exocytosis events per membrane area (Rexocytosis, s-1m-2) of PC12 cells. SPR15 measures the resonance angle (θr) that is approximately linearly dependent on the refractive index (n) of a cultured cell layer within an evanescent field reaching a few hundred nanometres from the sensor slide surface (Figure 1). Additional illustrations giving more detailed explanation of the experimental set up and SPR technique are found in Supporting Information (Figures 1S and 2S). The averaged n measured for a heterogeneous sample is weighted by the intensity of the exponentially decaying evanescent field16. Among earlier SPR cytometry studies are measurement of the response of mast cells to antigens17, the MDCKII cells volume as a function of osmolality18, and their uptake of propranolol and D-mannitol19, and nanoparticles20 using SPR with visible light. An infrared SPR instrument that probes deeper into the cells has been used for e.g. detecting cholesterol depletion and enrichment in HeLa cell membranes upon exposure to pure cyclodextrins, or their complex with cholesterol21. Shinohara et al. showed that exocytosis from PC12 cells is detectable by SPR microscopy22. Here we go further by establishing a quantitative relationship between the resonance angle shift (∆θr) and Rexocytosis by calibration with CFME. The accuracy and reproducibility are assessed, and the biological origin of the measured signal is investigated. Finally, pharmacological manipulation is carried out with EGTA and reserpine to demonstrate the technique’s usefulness.

Page 2 of 10

Figure 1. Measuring exocytosis by SPR cytometry. Cells grown on top of the SPR slide are probed by surface plasmons (red curved line). The inset illustrates the stages of exocytosis: 1) Vesicle trafficking. The vesicle is approaching the cell membrane; 2) Vesicle docking. High vesicle concentration close to the cellgold interface may result in a detectable increase of n; 3) Vesicle fusion. The vesicles fuse with the cell membrane, and the vesicle content is released into the space outside the cell.

EXPERIMENTAL SECTION Reagents and solutions. The chemicals were of analytical grade and used as received. Milli-Q water (18.2 MΩ·cm) was used as dilution media. Isotonic buffer with pH 7.4 was prepared using 150 mM NaCl, 5 mM KCl, 1.2 mM MgCl2, 2 mM CaCl2, 5 mM glucose and 10 mM HEPES. Stimulation solutions with 10, 20, 40, 50, and 100 mM K+ were of the same composition as the isotonic buffer, except for that Na+ concentration was adjusted to maintain constant osmolality. Stock solution of reserpine was prepared by dissolving the solid compound (Sigma-Aldrich, Germany) in isotonic buffer. Hypotonic solutions were prepared by diluting the isotonic buffer (317 mOsm/kg) to 50% (160 mOsm/kg), 10% (28.3 mOsm/kg), 1% (3.09 mOsm/kg), and 0.1% (0.312 mOsm/kg) of its original concentration. The osmolality was measured using the Fiske® model 210 micro-sample osmometer (Advanced Instruments Inc., MA, USA). Culturing of PC12 cells. PC12 cells were purchased from the American Type Culture Collection (Manassas, VA). Cells were grown on 25 cm2 mouse collagen IV coated flasks BD BioCoatTM (Fisher Scientific, MA, USA), and maintained in phenol red-free RPMI-1640 culture medium (Lonza, Belgium) supplemented with 10% donor horse serum and 5% fetal bovine serum. The cells were kept in an incubator at 37 °C in 7% CO2 and 100% humidity. The medium was replaced every 1-2 days throughout the lifetime of the cell culture. Cells were split every 7 days and propagated using a standard trypsinization procedure with 0.05% trypsin/EDTA (Thermo Fisher, UK). Culturing of PC12 cells on SPR slides. Gold-coated SPR glass slides (12×20 mm) were purchased from Bionavis Ltd. (Finland). Prior to use, the slides were cleaned for 10 min in a boiling mixture of 25% NH3, 30% H2O2 and Milli-Q water (v/v 1:1:1). This was followed by rinsing with ethanol, Milli-Q water, and drying with N2. Prior to cell culture, the slides were sterilized by UV radiation for 20 min. The slides were immersed into 50 µg mL-1 mouse collagen IV (Sigma-Aldrich, Germany) for 6 hours, followed by rinsing with sterile Dulbecco's PBS (Thermo Fisher, UK), and dried at 25 °C in a sterile hood. Prior to transfer to the SPR slide, PC12 cells were removed from the culture flasks by trypsinization, suspended in culture medium and diluted to 4×105 cells mL-1. Two mL of this cell suspension was placed on the collagen coated slides. The slides were kept in the incubator for 4-5 days until their surfaces became covered with a confluent layer of cells. Culturing of PC12 cells for amperometry. For amperometry, cells were cultured on collagen IV coated Petri dishes (VWR International AB, Sweden) during 4–5 days with daily replacement of the cell medium. Microscope examination of the cell covered SPR slide. All slides were examined for cell viability and confluency using phase contrast microscopy prior to SPR measurements. The

2

ACS Paragon Plus Environment

Page 3 of 10

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 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

cell coverage in micrographs was measured by Fiji/ImageJ software (www.fiji.sc). To quantify the fraction of uncovered areas, their boundaries were defined manually using the freehand selection tool. The program then determined the number of pixels contained within the highlighted areas which could be compared with the total number of pixels in the image. The surface density of cells was calculated by manually counting them within a predefined area. Cell viability test. Cell viability was for selected slides examined using the trypan blue exclusion test. Cells were trypsinized and resuspended in 0.4 w/v% trypan blue (SigmaAldrich, Germany). Cell viability was calculated under 40× magnification using a hematocytometer and expressed as
 

× 100%. The trypan blue stained cells were con
  

sidered non-viable. The morphology of non-viable PC12 cells was in most cases also altered by shrinkage, dehydration, and blebbing. This allowed for a rapid rough assessment of viability prior to the SPR measurements that were always conducted using slides where >97% of the cells seemed viable. SPR measurements. The measurements were conducted using the SPR Navi 220A analyzer (Bionavis Ltd., Finland). Isotonic buffer served as a running buffer in which the cells were immersed unless exposed to a stimulation solution. All solutions were thermostated, and the SPR measurements were conducted at 37±0.25 °C. The cell coated slides were transferred immediately from the incubator to the SPR instrument after cleaning the back (glass) side with 70% ethanol and Milli-Q water. The solutions were injected into the 100 µL flow cell (Bionavis Ltd., Finland) during 30 s at a flow rate of 200 µL min-1. The wavelength of the used laser was 785 nm. The angular scan range during the experiments was 39-78°. Data acquisition and analysis was performed using MP-SPR Navi™ software integrated with the analyzer. HPLC analysis. DA concentrations were measured using a Perkin Elmer 200 series HPLC system with a Waters Symmetry C18 column (3.9×150 mm, particle size 5 µm) and an UV detector set at 210 nm. The mobile phase consisted of Milli-Q water acidified to pH 3 by HCl, because DA is more stable to oxidation in acidic solutions. Stock solution of DA standard (10 mM) was prepared by dissolving solid DA (Sigma Aldrich, MO, USA) in the mobile phase, and then was consecutively diluted with mobile phase to desired concentrations immediately before measurements. The sample injection volume was 100 µL. The separation was conducted using a flow rate of 1.0 mL min-1, at 25±0.25 °C. The acquired data were processed in LabView based bespoke software. Standard addition calibration was used to quantify the DA released from the PC12 cells. Cyclic voltammetry. Cyclic voltammetry (CV) was performed using a CHI 1030B potentiostat (CH Instruments, Austin, TX, USA). A standard three electrode configuration was used with a Pt working electrode (3 mm diameter), Ag|AgCl (3 M KCl) reference electrode (BASi, West Lafayette, IN, USA) and platinum wire as counter electrode. CVs of dopamine were recorded between -0.2 and 0.6 V vs. Ag|AgCl at scan rate 0.05 V s-1. Standard solutions of DA and effluent collected from cells were carefully purged with nitrogen prior to experiments. The concentration of DA was calculated from a standard addition calibration curve constructed using the current values measured at 0.15 V.

Cell effluent collection for HPLC and CV analysis. A cell coated SPR slide was immersed during 10 min in 2 mL of 100 mM [K+] stimulation buffer thermostated at 37 °C. After removing the slide the pH was adjusted to 3, and the sample was filtered through a 0.2 µm sterile syringe filter (MillexTM, Millipore). The HPLC and CV experiments were carried out immediately. In order to calculate the DA concentrations that would have been reached inside the flow cell of the SPR instrument, it was necessary to take into account the differences in cell covered areas (12 mm2 and 240 mm2) and fluid volumes (0.1 mL and 2 mL) between the two experimental setups. Fabrication of microelectrodes. CFME were fabricated as described elsewhere23. Briefly, a 5-µm diameter carbon fiber was inserted into a glass capillary (BF120-69-10, Sutter Instrument Co., Novato, CA). The capillary was pulled with a micropipette puller (PE-21, Narishige, Japan) to form two fiber-containing pipettes, and sealed by dipping into an epoxy resin bath (EpoTek 301, Billerica, MA). The electrodes were cured at 100 °C overnight. The tips were polished to an angle of 45° using a micropipette beveller (Sutter Instrument Co., Novato, CA, USA) and back filled with 3 M KCl. Electrical contact was established by inserting a silver wire into the capillary. Prior to experiments, the electrodes were tested using a 100 µM DA solution. Amperometric measurements. The DA release events from individual PC12 cells were measured using CFME connected to an Axon 200B potentiostat (Molecular Devices, Sunnyvale, CA). A Petri dish with PC12 cells was placed on a thermostated stage (37 °C) of an inverted microscope (IX71, Olympus). The CFME was placed in contact with the cell membrane and a glass micropipette filled with stimulant (K+) was placed at ∼20 µm from the cell. The micropipette was connected to a microinjection system (Femtojet, Eppendorf, Germany). A potential of +0.7 V was applied to the microelectrode with respect to an Ag|AgCl reference electrode (World Precision Instruments, Inc., Sarasota, FL). Individual cells were chosen and stimulated at least three times, during 5 s with an interval of 30 s between the stimulations. The frequency of exocytosis events was calculated by dividing their number within 30 s from the onset of stimulation with the time. The signal was filtered with a lowpass filter at 2 kHz and digitized at 5 kHz. Data analysis. Amperometric traces were analyzed with IgorPro software (Version 6.2.2.0; WaveMetrics, USA). The peaks were identified by finding zero crossings in the differentiated trace. The intensity threshold for peak detection was three times the standard deviation of the noise. RESULTS AND DISCUSSION Culturing monolayers of PC12 cells on the SPR slide. Developing a reliable protocol for growing confluent cell monolayers on the SPR slide is pivotal. The ∆θr evoked by chemical stimuli are likely to depend both on the cells health, and the surface coverage (fcovererage, %) and density of cells within the covered areas (fdensity, m-2). Because it turned out that the PC12 cells could not be efficiently cultured on bare gold surfaces, the slides were coated with collagen that acts as an adhesion promoter. It could be calculated16 from the θr that the collagen layers were thin enough (2-3 nm) to not obscure the ∆θr due to cells. Two different seeding concentrations of 7×106 and

3

ACS Paragon Plus Environment

Analytical Chemistry

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 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

4×105 cells mL-1 were tried. The higher concentration resulted in growth of multilayered cell aggregates. In contrast, the lower concentration grew into almost confluent monolayers in 4-5 days (Figure 2). Longer culturing periods aggravated the cell adherence to the slide surface, and the monolayers were no longer able to withstand the shear during measurements.

Figure 2. Phase contrast micrographs of PC12 cells cultured on the SPR slide during 2 (A) and 5 (B) days. Broken yellow circles indicate cell free (A) and cell covered (B) areas. The insets illustrate the reflectance curves obtained from collagen (A) and cell (B) covered areas. The dotted rectangles in the inset show the onset of TIR region. (The scale bar indicates 100 µm).

The area illuminated by the laser beam is ~0.2 mm2 while the area covered by a single PC12 cell is ~250 µm2. Hence, the ∆θr could reflect the averaged response from ~800 cells. However, the contribution to ∆θr is proportional to the light intensity16. Therefore, not every cell probed by the beam with a Gaussian intensity profile contributes equally to the SPR response. The monolayer quality was always examined under microscope before mounting the slides into the SPR instrument. The fcoverage was on average 88.45 %. The standard deviation (  of fcoverage between slides was 3.76 %. The patches of bare Au surface were considerably larger than the area probed by the beam. Individual cells could only be distinguished for counting in a small fraction of the monolayer coated areas. The average fdensity was 4325 cells mm-2 and the  !" was ~7 % among the 0.01 mm2 areas that could be examined. The  !" would most likely be smaller if it was counted among examined areas with a size comparable to the light spot. As the distance between the cell membrane and the gold surface increases, the exocytosis will be probed by a weaker evanescent field. Cell-surface interactions have been studied with SPR microscopy24. Within an individual cell there can be regions where the membrane is located within i) 100-150 nm (i.e. the absence of close contact), ii) 30-50 nm (i.e. close contact), and iii) 10-15 nm (i.e. focal contact)24. The n measured by SPR for confluent cell monolayers varied between

Page 4 of 10

1.344 - 1.347. This is closer to the 1.332 of water than the ~ 1.38 of cells25. This indicates that the cells did not form close or focal contact within their whole area. The cell surface interactions are dynamic24 and could explain the drift in θr that was sometimes observed. The monolayer integrity was during measurements monitored by supervising both the θr and the onset of total internal reflection (TIR) region in the reflectance curve. Drift in θr that was not present from the start or attributable to injected buffers was considered as a sign of deteriorating health of the cells. The appearance of the TIR region depends on the fcoverage and fdensity19. When the laser impinges on an area containing no, or a few cells there is a well-defined crest at the critical angle (Figure 2A, inset). For confluent cell layers the reflectance becomes smooth and continuous (Figure 2B, inset). This occurs because the interior of cells contains light scatterers such as organelles, vesicles, and large protein complexes. The imaginary part of n is therefore non-zero and the slope of the reflectance curve becomes continuous instead of approaching infinity at the critical angle26. All experiments were conducted within 50 minutes after the slide was mounted into the SPR instrument, because the cells often started to detach after 45-50 minutes and the fraction of non-viable cells increased after 50 minutes. SPR signal recorded from PC12 cells stimulated with K+. Exocytosis was evoked by two different kinds of sequences for injecting stimulation solutions in order of increasing [K+]. Sensorgrams are shown in Figure 3. In the injection procedure 1 the cells were immersed in stimulation solution for a few minutes and the slide was rinsed with isotonic buffer between each [K+] until the θr returned to the baseline (Figure 3A). This sequence was applied on cells of both passage # 6, and 13. In the procedure 2 that was applied on cells of passage # 13 there were no rinsing between consecutive injections of increasing [K+] (Figure 3B). Detailed information about the two procedures is found in Supporting Information (Figure 3S). Because the shifts in angle are more reliably measured than the θr, the ∆θr in the sensorgrams is shown relative to an arbitrary reference value.

Figure 3. Sensorgrams recorded from PC12 cells stimulated with K+. A: Injection of K+ (black triangles) followed by rinsing with isotonic buffer (asterisks) according to procedure 1. Broken line

4

ACS Paragon Plus Environment

Page 5 of 10

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 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

indicates responses from passage # 6 upon successive injections of 20 and 50 mM K+ . Solid line indicates responses from passage # 13 upon injection of 10, 20, 40, 50, and 100 mM K+. B: Successive injections of 10, 20, 40, 50 and 100 mM K+ (black triangles) without inter-injection rinsing according to procedure 2. Dotted lines in A and B indicate responses from the collagen coated reference surfaces. The insets show the close-up of the individual responses.

The enlarged views of the signal upon stimulation are shown in the insets of Figure 3. The K+ injections in procedure 1 resulted in an instantaneous small dip in ∆θr before it rose, while injection of isotonic buffer caused, in opposite, a brief hike before the signal returned to the baseline (Figure 3A and B, inset, solid circles). Another recurring feature was a notch (Figure 3A and B, inset, dashed circle) occurring a few minutes after depolarization. The ∆θr as a function of [K+] is shown in Figure 4. The positive ∆θr indicates that the n increased at the sensor surface during exocytosis. In the procedure 1 the response was linearly dependent on [K+] (Figure 4A). The procedure 2 differs by that the cells stay depolarized during the whole ~50 min of the injection sequence with no chance for recovery. The ∆θr followed a sigmoidal curve with a steep raise at 40 mM [K+]. The ∆θr also reached higher values for a given [K+] than with procedure 1 (Figure 4B).

stimulation buffers could contribute significantly to the ∆θr. The dependence was as in earlier studies18 linear (R2=0.9998) but the slope (-0.0031 degrees mOsm-1 kg) was not steep enough for accidental hypotonicity to have played any role. It is possible, though, that the cell shrinks upon depolarization because of a small decrease in intracellular osmolality when Ca2+ replaces K+ and Na+. A 12 % osmolality difference could explain the positive y-intercept in Figure 4A. HPLC and CV showed that the concentration of DA in the cell media could reach 3 µM upon prolonged stimulation. The local CA concentration at the slide surface at the moment of release was probably even higher. However, tests showed that the SPR detection limit for a small molecule like DA is ~1 mM. Therefore, the released neurotransmitters themselves do not contribute significantly to the ∆θr. It is likely that proteins are also ejected upon exocytosis, although they have never been quantified. However, excreted proteins are probably not any major contributor either, because injecting a 15 wt% protein solution on a collagen coated slide only produced a ∆θr of ~0.1. Shinohara et. al.22 attributed the reflectance changes observed by SPR microscopy upon exocytosis to increased concentrations protein kinase C at the cell membrane. However, the ∆θr is an approximately linear function of protein concentration, and producing a ∆θr of hundreds of millidegrees as show in Figure 4 would require the formation of a compact mono layer of protein molecules at the cell membrane. It is instead argued that the ∆θr is mostly due to increased concentration of vesicles (cv, m-3) at the cell membrane. This will be investigated in depth below, after using amperometry to calibrate the ∆θr for Rexocytosis. Calibration of ∆θr with single cell amperometry. The Rexocyas a function of [K+] was measured by CFME amperometry (Figure 5A). The representative amperometric traces are found in Supporting Information (Figure 4S). No spikes were observed when cells were stimulated with 20 mM K+. Elevating the [K+] (30-100 mM) resulted in an apparently linear increase of the spike frequency. In amperometry the cells were subject to 5 s pulses of elevated [K+] instead of immersion in a constant concentration for minutes. However, the levels of intracellular [Ca2+] that regulate exocytosis5 are attained rapidly, while they decline slowly as the cells might keep secreting for over one minute after stimulation. The amperometric experiments are therefore comparable, although not identical with SPR. It is for instance not known whether the Rexocytosis is equal on the basal and the apical side of the cell layer that is directly exposed the stimulation buffers. A few additional SPR measurements were made for 20, 50, and 100 mM [K+] by injection procedure 1, and the resulting calibration plot of ∆θr for Rexocytosis is shown in Figure 5B. To calculate the Rexocytosis it was assumed that only the exocytosis events occurring under the projected area of the electrodes active surface were detected. Events might be detected a short (~ 600 nm) distance away from the edges of the carbon fiber which would cause the Rexocytosis to appear higher than its actual value28. tosis

Figure 4. Responses from stimulated PC12 cells plotted as a function of [K+]. A: Procedure 1, with every K+ injection followed by rinsing with isotonic buffer, B: Procedure 2 injection of K+ without inter-injection rinsing.

It has been shown that PC12 cells with different passage number respond differently to drugs27. It was recommended to use cells with passage # 9-13, because the response to pharmacological manipulation is most reproducible in this range. Using PC12 cells in this passage # range is also common practice for single cell amperometry studies. The magnitudes of ∆θr were comparable for both cell passage # in Figure 3A. However, the kinetics was different. The younger cells reacted more sluggishly, and ceased to respond to stimuli after the two first depolarizations despite still being viable. Older cells of passage # 16 were also investigated. Their response was poorly reproducible and not linearly dependent on [K+]. All further experiments were done using passage # of 12-13. The shear of repeated injections of isotonic buffer resulted in no response, nor did injecting K+ over a reference slide coated with collagen only. The ∆θr is therefore due to exocytosis. Spikes similar to the ones in the insets in Figure 3 have been observed when cells contracted or expanded when exposed to hyper- and hypotonic solutions18. They were attributed to rapid accumulation and depletion of actine in the vicinity of the cell membrane upon cell volume changes. The ∆θr, was measured as a function of the osmolality to exclude that the