Measurement of the Extracellular pH of Adherently Growing

high spatial resolution, and none of them is widely used in laboratories around the world. Cell biologists very often limit themselves to measuring th...
0 downloads 0 Views 850KB Size
Subscriber access provided by Kaohsiung Medical University

Article

Measurement of the extracellular pH of adherently growing mammalian cells with high spatial resolution using a voltammetric pH microsensor Raluca-Elena Munteanu, Luciana Stanica, Mihaela Gheorghiu, and Szilveszter Gaspar Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b01124 • Publication Date (Web): 07 May 2018 Downloaded from http://pubs.acs.org on May 7, 2018

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 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 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.

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 8 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

Measurement of the extracellular pH of adherently growing mammalian cells with high spatial resolution using a voltammetric pH microsensor Raluca-Elena Munteanu, Luciana Stǎnicǎ, Mihaela Gheorghiu, Szilveszter Gáspár * International Centre of Biodynamics, 1B Intrarea Portocalelor, 060101 – Bucharest, Romania *corresponding author: [email protected]; Tel.: + 40 21 3104354

ABSTRACT: There are only a few tools suitable for measuring the extracellular pH of adherently growing mammalian cells with high spatial resolution, and none of them is widely used in laboratories around the world. Cell biologists very often limit themselves to measuring the intracellular pH with commercially available fluorescent probes. Therefore, we have built a voltammetric pH microsensor and investigated its suitability for monitoring the extracellular pH of adherently growing mammalian cells. The voltammetric pH microsensor consisted of a 37 µm diameter carbon fiber microelectrode modified with reduced graphene oxide and syringaldazine. While graphene oxide was used to increase the electrochemically active surface area of our sensor, syringaldazine facilitated pH sensing through its pH-dependent electrochemical oxidation and reduction. The good sensitivity (60 ± 2.5 mV / pH unit), reproducibility (coefficient of variation ≤ 3% for the same pH measured with 5 different microsensors), and stability (pH drift around 0.05 in 3 h) of the built voltammetric pH sensors were successfully used to investigate the acidification of the extracellular space of both cancer cells and normal cells. The results indicate that the developed pH microsensor, and the perfected experimental protocol based on Scanning Electrochemical Microscopy, can reveal details of the pH regulation of cells not attainable with pH sensors lacking spatial resolution or which cannot be reproducibly positioned in the extracellular space.

Why is measuring the extracellular pH (pHe) of mammalian cells important? Why is measuring pHe with high spatial resolution also desirable? To answer the first question, it is enough to mention that cancer cells regulate their pH in a different manner than normal cells 1. Through few exclusive mechanisms, they succeed to maintain a neutral intracellular pH even while their pHe reaches values as low as 6.5. This ability is intensively studied while developing anticancer drugs, which can terminate it 2, or drug delivery and cancer imaging solutions, which beneficially exploit it 3. To answer the second question, it is enough to mention that tumors often survive chemotherapy because of the heterogeneity of cancer cells even within the same tumor 4. Although a given therapy kills most cancer cells within a tumor, there are a few cells which are different enough to survive and cause further problems. This heterogeneity of cancer cells is completely neglected and lost when cells are investigated with analytical tools which read signals averaged from millions of cells. To be taken into account, this heterogeneity requires analytical tools characterized by good (ideally single cell) spatial resolution. In addition to cancer, there are several other pathological states in which the pH regulation of cells is perturbed (e.g. quite many neurological diseases 5). The study of these pathological states at cellular level could also benefit from a tool able to measure pHe with high spatial resolution. However, there are only a few tools suitable for measuring the pHe of adherently growing mammalian cells with high spatial resolution and none of them is perfect or widely used in laboratories around the world. A very promising fluorescent pH probe was recently described 6. The probe, obtained by attaching a pHsensitive fluorescent dye to a pH low insertion peptide

(pHLIP), can decorate the outer surface of cells with pH sensitive fluorescent dye. The probe reported pH values as low as 6.0 at the surface of cancer cells (i.e. pH values 1.4 pH units lower than the pH of the bulk). This approach carries however the problems of fluorescence-based methods (e.g. quenching and photo bleaching). Electrochemical microsensors were also developed to measure pH with high spatial resolution. It is important to mention that, unlike the fluorescence-based methods which require a dye to be added onto cells, electrochemical pH microsensors advantageously investigate cells in their natural state. Most of these pH microsensors fall into three categories: i.) glass-based microsensors, ii.) ionophorebased microsensors, and iii.) metal oxide-based microsensors. However, only a few of them were used to measure the pHe of cells 7–12, and even fewer to measure the pHe of mammalian cells 7,9,10. Potentiometric microsensors, made with protonselective glass or liquid ionophore, were used to measure pH under adherently growing macrophages and osteoclasts 7. pH values as low as 3 were measured in the attachment zone between osteoclasts and the base of the culture dish. More recently, a potentiometric pH microsensor was built by loading a commercial proton ionophore into a glass micropipette and used to measure the pH in close proximity of bacteria biofilms 12 . pH was evaluated not only above but also inside hundreds of microns thick bacteria biofilms. Glass pH microsensors from Unisense A/S 11 or iridium oxide-based potentiometric pH microsensors 8 were used for this purpose. Planar, on-chip, metal oxide pH microsensors were also developed and used to observe the pHe of cells adherently growing next to (or all over) them 9,10. The glass nanopipette used in Scanning Ion Conductance Microscopy was enhanced with pH sensing abili-

1 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

ties 13,14. However, the resulting pH sensors were not yet used to investigate pHe. Potentiometric, glass-based or ionophore-based pH microsensors are very fragile and have relatively long response times. To be used in the extracellular space, they require not only electrochemical instrumentation but also an optical microscope (as their positioning relative to the cells is done under visual observation). This makes the experimental set-up for measuring pHe relatively complex. The lifetime of pH microsensors with liquid ionophore membranes is also somewhat limited because of the ionophore leaking out of the microsensor. Planar, on-chip, metal oxide pH microsensors eliminate several of these disadvantages. The position of such pH microsensors relative to the cells is however poorly controlled and impossible to change during the experiment. The present work shows that a reduced graphene oxide- and syringaldazine-based pH microsensor can be used to reliably measure pHe with good spatial resolution, to capture cellular heterogeneity, and to reveal differences between cancer and normal cells. In comparison to probably the most similar work 12 , the present study puts forward not a potentiometric but a voltammetric pH microsensor and investigates pHe of mammalian cells instead of bacteria (while mammalian cells build up smaller pH gradients than bacterial biofilms). The use of syringaldazine within a voltametric pH nanosensor was already reported 15. However, enhancing the syringaldazinebased pH sensor with reduced graphene oxide and the use of the novel reduced graphene oxide- and syringaldazine-based pH sensor at cellular level are described here for the first time.

buffer pH 7.4, containing 130 mM NaCl, 5 mM KCl, 0.9 mM CaCl2, 1.2 mM MgCl2, 15 mM NaH2PO4 2H2O and 5 mM glucose. pHe values were recorded in the extracellular space for times not longer than 85 min. Cells grown to 100% confluence were used in all experiments, in order to avoid making measurements in positions without cells (as experiments were not observed through an optical microscope). Fabrication and calibration of the pH microsensor. Carbon fiber microelectrodes were first fabricated and then modified with reduced graphene oxide and syringaldazine. The microelectrodes were fabricated by encapsulating carbon fibers into the tapered end of glass micropipettes obtained by pulling glass capillaries with a P-97 puller from Sutter Instruments Inc. (USA). Microelectrodes fabricated in this way were first polished mechanically on a EG-401 microgrinder (Narishige Group, Japan) and then cleaned electrochemically (30 s at +2 V, 10 s at -1 V, and 25 cycles from 0 V to +1 V at 0.1 V s-1 in 0.5 M H2SO4). The carbon fiber microelectrodes used throughout this work featured an electrochemically active disk of 37 µm in diameter and also Rg factors in between 2 and 3.3 (the Rg factor being the ratio in between the radius of the electrode and the radius of the electrochemically active disk). For modification with reduced graphene oxide and with syringaldazine, such carbon fiber microelectrodes were dipped 4-5 times into an aqueous solution of 4 mg mL-1 graphene oxide (and let dry in between dips), and then 3 times into an ethanolic solution of 0.85 mg mL-1 syringladazine (and let dry in between dips). The graphene oxide adsorbed onto the microelectrodes was then electrochemically reduced by 10 potential cycles from 0 V to -0.9 V at 0.01 V s-1. The microelectrodes modified with reduced graphene oxide and syringaldazine gave cyclic voltammograms with a pair of current peaks corresponding to the electrooxidation and electroreduction of syringaldazine. The peak potentials of these two current peaks were found to depend on the pH of the solution bathing the modified microelectrodes. Therefore, the two currents peaks facilitated the use of the syringaldazine-modified microelectrodes as voltammetric pH microsensors. Before using a given pH microsensor at cellular level, it was always calibrated using 6 solutions with biologically relevant pH values. All electrochemistry experiments were carried out using a PGSTAT128N potentiostat from Metrohm Autolab BV (The Netherlands) and a three electrode configuration with a Ag/AgCl, 3 M KCl reference electrode and a glassy carbon counter electrode from Metrohm AG (Switzerland). All potentials mentioned throughout this work are versus this Ag/AgCl, 3 M KCl reference electrode. Use of the pH microsensor in the extracellular space. A Scanning Electrochemical Microscope (SECM) from Sensolytics GmbH (Germany) was used to reproducibly position the pH microsensors in the extracellular space. For positioning on the vertical direction, the microsensor was polarized to -0.550 V (a potential where oxygen reduction occurs) and then slowly lowered until the cell-covered surface started blocking the access of oxygen to the electrochemically active area of the microsensor. Lowering the microsensor towards the cells was always stopped when the current given by the oxygen reduction reaction decreased to 80% of its value in the bulk solution. Then, a cyclic voltammogram was recorded at this distance from the cells and used to calculate the pHe close to the cells (pHclose). Next, the pH microsensor was lifted with 500 µm and another cyclic voltammogram was recorded. This voltammogram was used to calculate the pHe far from the cells

EXPERIMENTAL SECTION Materials. C3005 carbon fiber from World Precision Instruments (USA), EPO-TEK 302-3M epoxy from Epoxy Technology Inc. (USA), and borosilicate glass capillary (1.5 mm O.D. x 0.86 mm I.D.) from Harvard Apparatus Ltd (UK) were used to build carbon fiber microelectrodes. Graphene oxide from Graphenea (Spain) and syringaldazine from Sigma Aldrich Inc. (USA) were used to turn such microelectrodes into pH microsensors. Ringer solution, made with NaCl, KCl, CaCl2, and glucose from Sigma Aldrich Inc. (USA), MgCl2 from Chimopar SA (Romania) and NaH2PO4 2H2O from Carl Roth GmbH (Germany), was used both for sensor calibration experiments and experiments at cellular level. Ultrapure water from a Direct-Q 3 UV water purification system (from Millipore) was used in all aqueous solutions. Cells. The pHe of human colorectal adenocarcinoma cells (HT-29 cells, ATCC no. HTB38), human embryonic kidney cells (HEK-293 cells, ATCC no. CRL1573), and transfected human embryonic kidney cells (Flp-In-293, from Thermo Fisher Scientific Inc.) was investigated. The cells were grown in Dulbecco-modified Eagle medium supplemented with 10% fetal bovine serum and penicillin-streptomycin (100 IU mL-1 0.1 mg mL-1). Cells were seeded at a concentration of 2.25 x 105 cells mL-1 on 3.5 cm diameter, plastic Petri dishes (Nunc, Thermo Fisher Scientific) and grown until confluence at 37°C in a 5% CO2 humidified incubator (MCO-20AIC Sanyo, Japan). All culture media and supplements were purchased from Thermo Fisher Scientific Inc. and Sigma Aldrich Inc. The pHe of the confluent cells was investigated at about 20 min after replacing the cell culture medium with prewarmed Ringer

2 ACS Paragon Plus Environment

Page 2 of 8

Page 3 of 8 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 (pHfar). The two pHe values were used to calculate the pH gradient (pHfar - pHclose) characterizing the confluent cell monolayer in the respective position. The pH gradient was measured in a similar manner in 7 - 8 positions of each cell monolayer (Petri dish) taken into work. A number of at least 7 cell monolayers (Petri dishes) were investigated for each cell type. The number of cells interrogated in each position of a cell monolayer was approximated to be in between 4 and 6 (by comparing the dimensions of the developed pH microsensor with the dimensions of the investigated cells).

RESULTS AND DISCUSSION Principle of the pH microsensor and the role of graphene in the pH microsensor. Syringaldazine is insoluble in water. Therefore, once syringaldazine is adsorbed onto the surface of the carbon fiber microelectrodes from a solution made in ethanol, it will not desorb from there as long as the microelectrodes are used in aqueous solutions. Syringaldazine is also relatively easy to electrochemically oxidize and reduce. As a consequence, the cyclic voltammograms of syringaldazinemodified electrodes show an anodic current peak at 254 ± 8 mV, due to the oxidation of syringaldazine, and a cathodic current peak at 198 ± 9 mV, due to the reduction of syringaldazine (at pH = 7.36 ± 0.04). Moreover, the electrochemical processes of syringaldazine involve two protons in addition to two electrons 16. These two protons make the electrochemistry of syringaldazine pH-sensitive. In cyclic voltammetry this pH sensitivity translates into a shift of the current peaks to more positive potentials when the pH is decreasing. Therefore, a syringaldazine-modified electrode and the formal potential of the adsorbed syringaldazine (calculated as the average of anodic and cathodic peak potentials) can be used to determine the pH of the solution bathing the electrode. pH microsensors with syringaldazine but without graphene were initially fabricated. The cyclic voltammograms of such pH microsensors were characterized by two small, often poorly-defined, current peaks due to oxidation and reduction of adsorbed syringaldazine. Next, as there is a significant amount of literature on the beneficial effects of graphene on electrochemical sensors 17,18, pH microsensors with syringaldazine and reduced graphene oxide were fabricated. The cyclic voltammograms of such microsensors revealed that the reduced graphene oxide increases the amplitude of the current peaks given by the oxidation and reduction of syringladazine (see Figure 1) most probably by increasing the electrochemically active surface area of the microelectrodes. The larger, bettershaped peaks make reading of peak potentials easier and more precise. Reduced graphene oxide has also increased the peak potential difference from 18 ± 1 to 36 ± 8 mV. However, it had only minor impact on the total width at half-height (which increased from 81 ± 1 to 86 ± 4 mV) and on the ratio of current peaks (which increased from 0.96 ± 0.03 to 0.99 ± 0.04). The voltammograms remained characteristic to a surfaceadsorbed, quasi-reversible redox couple in spite of these small changes induced by the reduced graphene oxide. The analytical performances of the pH microsensor. Six solutions with known, biologically relevant pH were used to calibrate every single pH microsensor before it was used at cellular level. Typical cyclic voltammograms recorded during a calibration experiment are shown in Figure 2 together with

Figure 1. Cyclic voltammograms of syringaldazine-modified electrodes built with or without reduced graphene oxide. Experimental conditions: Ringer buffer at pH = 7.28 was used as supporting electrolyte; the potential scan rate was 20 mV s-1; the resulting calibration curve that links the formal potential of the syringaldazine adsorbed onto the microelectrode to the pH of the solution. According to the linear fit shown in Figure 2B, the pH sensitivity of the developed pH microsensors is very close to that predicted by the Nernst equation for a redox couple exchanging two electrons and two protons (57 mV versus 59 mV). Next, the reproducibility of the pH microsensors was investigated. Figure 3 shows average calibration curves obtained for five pH microsensors built in parallel in the same day and for five pH microsensors built one-by-one in different days. The pH microsensors are characterized by excellent reproducibility both when built in parallel in the same day and when built one-by-one in different days. The largest coefficient of variation was observed for sensors built one-by-one in different days used in the solution with the highest investigated pH (i.e. pH = 7.49). However, the coefficient of variation characterizing the pH microsensors was still only 3%. This good reproducibility is in a great extent due to the insensitivity of the formal potential to experimental conditions other than pH (e.g. the formal potential recorded at a given pH value, just as expected, did not change with the amount of syringaldazine adsorbed onto the pH microsensors). Important to mention, each pH microsensor was calibrated before being used at cellular level and, in spite of the good reproducibility, no average calibration curves were used when investigating pHe. This approach was preferred due to the small pH changes expected to be observed in the extracellular space of mammalian cell monolayers. After determining the sensitivity and reproducibility of the pH microsensors, their operational stability was also investigated. A good stability is needed because the pH microsensors are being used to read pH in several distant positions of the extracellular space of adherently growing cells and moving the pH microsensors from one position to the other is done slowly (in order to not significantly perturb the pH gradients built up by the cells). In these conditions, the experiments at cellular level (about 14 - 16 pH measurements made at spots 500 µm from each other) are expected to take about 85 minutes. Figure 4 shows how one of the pH microsensors is reporting in time the pH of two different solutions with biologically relevant pH values which were previously determined with a commercial pH electrode. The microsensor initially reports a pH of 7.34

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

A.

Page 4 of 8

B.

Figure 2. Cyclic voltammograms of a pH microsensor (built with reduced graphene oxide and syringaldazine) immersed into solutions with known, biologically relevant pH (A) and resulting calibration curve linking the formal potential of the adsorbed syringaldazine to the pH of the solution (B). Experimental conditions: the pH of the Ringer buffer used as supporting electrolyte was adjusted to the indicated values using 1 M HCl; the potential scan rate was 20 mV s-1;

A.

B.

Figure 3. Average calibration curves obtained with five pH microsensors built in parallel in the same day (A) and with five pH microsensors built one-by-one in different days (B). Error bars are standard deviations;

for the solution with an actual pH of 7.34 and a pH of 6.97 for the solution with an actual pH of 6.98.The microsensor then starts reporting slightly lower values of pH for the two solutions. However, during the almost 3 h of the experiment, the pH values reported by the microsensor for the two solutions were only 0.05 pH units away from the initially measured pH values. It is thus clear that the developed pH microsensors are suitable to work in Ringer buffer for up to 3 h and to document pH changes larger than 0.05 pH units. The developed microsensor measures pH based on a cyclic voltammogram recorded in a potential window suitable to oxidize and reduce syringaldazine. Such a voltammogram can be recorded in 45 s when using a potential scan rate of 20 mV s-1. Response times shorter than 45 s can be achieved using faster potential scan rates. The analytical performances of the developed microsensor are compared to those of previously reported pH microsensors in Supporting Information.

Investigation of pHe using the developed pH microsensors. Once satisfactory analytical performances were observed for the developed pH microsensors, they were used to measure the pH gradient developed by groups of 4 - 6 cells randomly selected out of monolayers of either cancer or normal cells. As cancer cells, human colorectal adenocarcinoma cells (HT-29 cells) were investigated in two states: grown in normoxic conditions and after they have been subjected to 12 h of hypoxia. Hypoxia was shown to elevate carbonic anhydrase IX levels in these cells 19 and this was also confirmed by us in our experimental conditions (see Supporting Information). Important, carbonic anhydrase IX is a major contributor to the extracellular acidosis of cancer cells and is currently investigated as a potential therapeutic target 20. As normal cells, human embryonic kidney cells (HEK-293) were investigated also in two states: without and with a transfection preparing them for protein expression. However, transfection with the pFRT/lacZeo2

4 ACS Paragon Plus Environment

Page 5 of 8 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 by the oxygen reduction reaction on the pH microsensor, decreases during positioning not only because of the blocking effect of the cells but also because of the consumption of the oxygen by the cells (via respiration) 22. In some cases the consumption of oxygen by the cells was obvious as it imposed a simple (drift) correction on the current signal recorded during positioning. In other cases there were no clear signs of it. As already mentioned in the Experimental Section, the second value of pHe (i.e. pHfar) was acquired after lifting the pH microsensors with additional 500 µm. pHe was measured in more than 230 positions of confluent monolayers of the above-mentioned cells. When pHclose values were pulled together and averaged, the HEK-293 cells were characterized by a mean pHe of 7.18, the Flp-In-293 cells by a mean pHe of 7.21, the normoxic HT-29 cells by a mean pHe of 7.13, and finally the hypoxic HT-29 cells by a mean pHe of 7.03. When these mean values were analyzed using ANOVA, the mean pHe of hypoxic HT-29 cells (7.03) was identified as significantly different from all the other mean pHe values (at a significance level of 0.01). The mean pHe of normoxic HT-29 cells (7.13) was found to be significantly different from the mean pHe of Flp-In-293 cells (at a significance level of 0.05). Mean pHe values of the normal cells (i.e. HEK-293 and FlpIn-293) were found to be not significantly different. These mean pHe values tend to confirm expectations: cancer cells have a more acidic extracellular space than normal cells, and cancer cells with an elevated level of carbonic anhydrase IX (i.e. cancer cell subjected to hypoxia) have an even more acidic extracellular space than normoxic cancer cells. It is important to note that, these mean pHe values were recorded at short times (≤ 85 min), while the pH of the bulk solution was still practically unaffected by cellular activity. Therefore, these pHe values cannot be observed with classic, large pH electrodes dipped into the solution bathing the cells. Figure 5 shows a more detailed picture of the pHe of the investigated cells. This picture also cannot be obtained with classic, large pH electrodes. According to Figure 5, the majority of the HEK-293, Flp-In-293 and normoxic HT-29 cells develop a pH gradient smaller than 0.1 pH units. However, the majority of the investigated hypoxic HT-29 cells developed a pH gradient between 0.1 and 0.2 pH units confirming the literature regarding the acidosis of hypoxic tumors 23 and our expectations detailed above. Normal cells never develop a pH gradient larger than 0.4 pH units. As a distinctive feature, cancer cells (both the normoxic and the hypoxic ones) are clearly more heterogeneous regarding the pH gradient they develop. About 22% of investigated normoxic HT-29 cells developed a pH gradient larger than 0.4 pH units. Interestingly, only 10% of the hypoxic HT-29 cells developed a pH gradient larger than 0.4 pH units, their lower mean pHe being thus the result of many cells having a pH gradient in between 0.2 and 0.3 pH units. Thus, the mean pHe values are clearly hiding important differences between the investigated cells.

Figure 4. pH values reported in time by one of the developed pH microsensors for two solutions with biologically relevant pH. Experimental conditions: the pH of the Ringer buffers was adjusted to the indicated values using 1 M HCl; the pH microsensor was first used to determine the pH of the solution with higher pH and then of the solution with lower pH at 10 min intervals;

vector is not expected to impact the pH regulation of HEK-293 cells. Therefore, it can be expected to observe differences between the pHe of cancer and normal cells, and between the pHe of cancer cells kept in normoxic conditions and hypoxic conditions. It is expected not to see any significant differences between the pHe of normal cells and those transfected with the pFRT/lacZeo2 vector (and that is why these cells were taken into study). Cells are usually grown and investigated in solutions which contain a variety of nutrients and are characterized by a certain buffering capacity. Due to this buffering capacity, the metabolic activity of the cells propagates into the extracellular space as a pH change only over relatively short distances (i.e. few hundreds of micrometers 11). Therefore, an important problem that needs to be addressed when measuring pHe is the reproducible positioning of the pH microsensors at small and known distances from the cells. This problem was solved in this study by exploiting one of the advantages of voltammetric pH microsensors: unlike many other electrochemical pH microsensors (e.g. potentiometric sensors based on proton selective glasses or liquid ionophores), voltammetric pH microsensors can be reproducibly positioned using SECM and the approach curves which link the current signal to the distance from the investigated surface. Therefore, SECM was used in our experimental protocol to lower the pH microsensors towards the investigated cells until their current signal (due to the oxygen reduction reaction) decreased to 80% of the current observed in the bulk solution. The exact distance, at which the pHe was measured above the cells, was determined by fitting the approach curves recorded for each experiment to analytical expressions from literature 21 (see details in the Supporting Information). According to these fits, pHe close to the cells (i.e. pHclose) was measured at an average distance of 28 ± 3 µm from normoxic HT-29 cells, 29 ± 4 µm from HEK-293 cells, 28 ± 4 µm from hypoxic HT-29 cells, and 24 ± 2 µm from Flp-In-293 cells. However, it is important to note that these distances are still somewhat approximate. The current, given

CONCLUSIONS A syringaldazine-based voltammetric pH microsensor was improved by using reduced graphene oxide to increase its electrochemically active surface area and thus its current signals. The resulting pH microsensor was characterized by

5 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

Figure 5. pH gradients discovered in the extracellular space of the investigated cells. The percentage of cells with pH gradients above a set threshold of 0.4 pH units is shown in red. The pH gradient developed by cells was evaluated in 59 different positions of HEK-293 cell monolayers, 53 different positions of Flp-In-293 cell monolayers, 73 different positions of normoxic HT-29 cell monolayers, and 48 different positions of hypoxic HT-29 cell monolayers. The positions above the same cell monolayer were 500 µm from each other.

theoretical sensitivity, and good stability and reproducibility. Next, SECM was used to position the pH microsensor at small, known distances (24 – 29 µm) from confluent layers of cellsAfter measuring a first pHe value (i.e. pHclose) at this relatively small distance from the cells, the pH microsensor was lifted 500 µm and a second pHe value (i.e. pHfar) was measured. The two pH values were then used to calculate the pH gradient that characterizes roughly 4 - 6 cells of the confluent monolayer (taking into account the diameter of only 37 µm of the electrochemically active disk of the pH microsensor). The procedure was then repeated more than 230 times with several Petri dishes with either confluent monolayers of cancer cells (HT-29, before and after hypoxia) or confluent monolayers of normal cells (HEK-293, with and without transfection). The mean pHe values, calculated by averaging all pHclose values for a given type of cells, confirm that cancer cells subjected to hypoxia have indeed a more acidic extracellular space than normal or normoxic cancer cells. The mean pHe of the hypoxic HT-29 cells is significantly smaller than the mean pHe values of the other cells (at a significance level of 0.01). However, when analyzing individual pH gradients (i.e. the ones observed

above only 4 – 6 cells) additional differences between cells emerge. For example, it was observed that monolayers of cancer cells develop a wide range of pH gradients (from 0.1 to 0.8 pH units) while monolayers of normal cells never develop pH gradients larger than 0.4 pH units. It was also observed that the average extracellular pH of hypoxic cancer cells is smaller than that of normal cells due to ~27% of the hypoxic cancer cells featuring pH gradients from 0.2 to 0.3 pH units. Meanwhile, the average extracellular pH of normoxic cancer cells is smaller than that of normal cells due to ~22% of the normoxic cancer cells featuring pH gradients larger than 0.4 pH units. The obtained results prove that the developed pH microsensor is suitable to observe pHe with high spatial resolution and thus to reveal details of pH regulation of cells not attainable before. The described pH microsensor and the approach to measure pHe can easily be improved. Single cell measurements at smaller distances between the cell and the pH microsensor are within reach by making the pH microsensor significantly smaller and by using more advanced SECM set-ups (e.g. with better resolution, constant distance mode, etc.).

6 ACS Paragon Plus Environment

Page 6 of 8

Page 7 of 8 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 (8) VanHoudt, P.; Lewandowski, Z.; Little, B. Iridium Oxide pH Microelectrode. Biotechnol. Bioeng. 1992, 40, 601–608. (9) Ges, I. A.; Baudenbacher, F. Microfluidic Device to Confine Single Cardiac Myocytes in Sub-Nanoliter Volumes for Extracellular pH Measurements. J. Exp. Nanosci. 2008, 3, 63–75. (10) Wu, C.-C.; Lin, W.-C.; Fu, S.-Y. The Open Container-Used Microfluidic Chip Using IrOx Ultramicroelectrodes for the in Situ Measurement of Extracellular Acidification. Biosens. Bioelectron. 2011, 26, 4191–4197. (11) Hou, J.; Liu, Z.; Zhou, Y.; Chen, W.; Li, Y.; Sang, L. An Experimental Study of pH Distributions within an Electricity-Producing Biofilm by Using pH Microelectrode. Electrochimica Acta 2017, 251, 187–194. (12) Joshi, V. S.; Sheet, P. S.; Cullin, N.; Kreth, J.; Koley, D. RealTime Metabolic Interactions between Two Bacterial Species Using a Carbon-Based pH Microsensor as a Scanning Electrochemical Microscopy Probe. Anal. Chem. 2017, 89, 11044–11052. (13) Morris, C. A.; Chen, C.-C.; Ito, T.; Baker, L. A. Local pH Measurement with Scanning Ion Conductance Microscopy. J. Electrochem. Soc. 2013, 160, H430–H435. (14) Nadappuram, B. P.; McKelvey, K.; Al Botros, R.; Colburn, A. W.; Unwin, P. R. Fabrication and Characterization of Dual Function Nanoscale PH-Scanning Ion Conductance Microscopy (SICM) Probes for High Resolution pH Mapping. Anal. Chem. 2013, 85, 8070–8074. (15) Michalak, M.; Kurel, M.; Jedraszko, J.; Toczydlowska, D.; Wittstock, G.; Opallo, M.; Nogala, W. Voltammetric pH Nanosensor. Anal. Chem. 2015, 87, 11641–11645. (16) Hapiot, P.; Pinson, J.; Neta, P.; Rolando, C. Electrochemical Behaviour of Syringaldazine, a Colorimetric Redox Reagent. J. Electroanal. Chem. 1993, 353, 225–235. (17) Ratinac, K. R.; Yang, W.; Gooding, J. J.; Thordarson, P.; Braet, F. Graphene and Related Materials in Electrochemical Sensing. Electroanalysis 2011, 23, 803–826. (18) Bahadır, E. B.; Sezgintürk, M. K. Applications of Graphene in Electrochemical Sensing and Biosensing. TrAC Trends Anal. Chem. 2016, 76, 1–14. (19) Carlin, S.; Khan, N.; Ku, T.; Longo, V. A.; Larson, S. M.; Smith-Jones, P. M. Molecular Targeting of Carbonic Anhydrase IX in Mice with Hypoxic HT29 Colorectal Tumor Xenografts. PLOS ONE 2010, 5 (5), e10857. (20) Dubois, L.; Peeters, S. G. J. A.; van Kuijk, S. J. A.; Yaromina, A.; Lieuwes, N. G.; Saraya, R.; Biemans, R.; Rami, M.; Parvathaneni, N. K.; Vullo, D.; Marc, V.; Supuran, C.T.; Jean-Yves, W.; Philippe, L. Targeting Carbonic Anhydrase IX by Nitroimidazole Based Sulfamides Enhances the Therapeutic Effect of Tumor Irradiation: A New Concept of Dual Targeting Drugs. Radiother. Oncol. 2013, 108, 523– 528. (21) Cornut, R.; Lefrou, C. A Unified New Analytical Approximation for Negative Feedback Currents with a Microdisk SECM Tip. J. Electroanal. Chem. 2007, 608, 59–66. (22) Nebel, M.; Grützke, S.; Diab, N.; Schulte, A.; Schuhmann, W. Visualization of Oxygen Consumption of Single Living Cells by Scanning Electrochemical Microscopy: The Influence of the Faradaic Tip Reaction. Angew. Chem. Int. Ed. 2013, 52, 6335–6338. (23) Chiche, J.; Brahimi-Horn, M. C.; Pouysségur, J. Tumour Hypoxia Induces a Metabolic Shift Causing Acidosis: A Common Feature in Cancer. J. Cell. Mol. Med. 2010, 14, 771–794.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Expression of carbonic anhydrase IX by HT-29 cells subjected to hypoxia; Reproducible positioning of the pH microsensors in the extracellular space using SECM; Cellular viability in the experimental conditions of the pHe measurements; Analytical performances of electrochemical pH microsensors used to investigate biological systems (PDF).

AUTHOR INFORMATION ORCID Raluca-Elena Munteanu: 0000-0003-4629-2435 Luciana Stǎnicǎ: 0000-0001-6459-3085 Mihaela Gheorghiu: 0000-0002-4620-6130 Szilveszter Gáspár: 0000-0002-8401-6207 Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT The authors thank the Romanian Executive Unit for Higher Education, Research, Development and Innovation Funding for funding through PN-III-P2-2.1-PED-2016 grants GrapHtool (contract no. 110PED / 2017) and SensCell (contract no. 161PED / 2017) and Flag-ERA grant Graphtivity (contract no. 40 / 2016).

REFERENCES (1) Webb, B. A.; Chimenti, M.; Jacobson, M. P.; Barber, D. L. Dysregulated pH: A Perfect Storm for Cancer Progression. Nat. Rev. Cancer 2011, 11, 671. (2) Neri, D.; Supuran, C. T. Interfering with pH Regulation in Tumours as a Therapeutic Strategy. Nat. Rev. Drug Discov. 2011, 10, 767. (3) Mura, S.; Nicolas, J.; Couvreur, P. Stimuli-Responsive Nanocarriers for Drug Delivery. Nat. Mater. 2013, 12, 991. (4) Dexter, D. L.; Spremulli, E. N.; Fligiel, Z.; Barbosa, J. A.; Vogel, R.; VanVoorhees, A.; Calabresi, P. Heterogeneity of Cancer Cells from a Single Human Colon Carcinoma. Am. J. Med. 1981, 71, 949– 956. (5) Obara, M.; Szeliga, M.; Albrecht, J. Regulation of pH in the Mammalian Central Nervous System under Normal and Pathological Conditions: Facts and Hypotheses. Neurochem. Int. 2008, 52, 905– 919. (6) Anderson, M.; Moshnikova, A.; Engelman, D. M.; Reshetnyak, Y. K.; Andreev, O. A. Probe for the Measurement of Cell Surface pH in Vivo and Ex Vivo. Proc. Natl. Acad. Sci. 2016, 113, 8177-8181. (7) Silver, I. A.; Murrills, R. J.; Etherington, D. J. Microelectrode Studies on the Acid Microenvironment beneath Adherent Macrophages and Osteoclasts. Exp. Cell Res. 1988, 175, 266–276.

7 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

For TOC only:

8 ACS Paragon Plus Environment

Page 8 of 8