Introduction

Cell life is possible only if intracellular pH (pHi) oscillations are kept within a very narrow range (Lagadic-Gossmann et al. 2004). The intracellular pH is a permissive facilitator of growth and development in normal tissue and in tumors (Pouysségur et al. 1984; Chiche et al. 2010; Boron 2004) and a intracellular H+ load is a indication of high metabolic activity such as respiration.

Manipulating the pHi in a controlled way implies controlling intracellular processes and activity such as cellular osmotic response (Bourque 2008) and the membrane polarization (Obara et al. 2008) nutrients transport (Thwaites and Anderson 2007).

Cells put in place several mechanisms to regulate pH. High H+ loads are extruded by specialized proteins in the cell membrane and intracellular vesicles through several mechanisms. The Na+/H+ exchanger (NHE) is an integral plasma membrane protein that mediates the electroneutral exchange of Na+ and H+ in a one-to-one stoichiometry (Orlowski et al. 1992; Harvey and Eherenfeld 1988). Vacuolar-type ATPase is present in the membranes of many organelles, where it exerts a proton or proton-potassium pumping action (Pedersen and Carafoli 1987; Finbow and Harrison 1997). Acid extrusion can also be produced by membrane transporters that load cells with HCO3 − (or CO3 2−) ions (Boron and Boulpaep 1983). These HCO3 − transport proteins include electroneutral Na+-HCO3 − cotransporters (NBC) (Deitmer and Schlue 1989; Choi et al. 2000; Romero et al. 1997) and Na+-dependent and independent Cl−/HCO3 − exchangers (Wang et al. 2000; Romero et al. 2000; Wong et al. 2002). Extracellular CO2/HCO3 − buffer facilitates these membrane-transport processes (Hulikova et al. 2011).

The intracellular pH may also be regulated via exogenous compounds, for instance by photoactivated proton caged compounds (PCC), as shown in the recent paper (Carbone et al. 2013). This allowed a high pH jump, which can be easily monitored by infrared spectroscopy. In particular, a long-chained proton caged compound, the 1-(2-nitrophenyl)- ethylhexadecyl sulfonate (HDNS), was purposely synthesized and dosed into 3T3-NIH cells. The dosed cells were then probed by infrared spectroscopy and irradiated by filtered UV light in order to induce intracellular HDNS photolysis. Upon each irradiation, the released protons react with the cytosolic HCO3 − and release CO2 according to the reactions:

$${\text{H}}^{ + } + {\text{ HCO}}_{ 3}^{ - } {\; \leftrightarrows \;} {\text{H}}_{ 2} {\text{CO}}_{ 3} {\; \leftrightarrows \;} {\text{H}}_{ 2} {\text{O}} + {\text{CO}}_{ 2}$$
(1)

Both HCO3 − and CO2 peaks may be monitored, however, although the latter appears at 2343 cm−1 in a clear region of the cell spectrum and can be clearly singled out, the former is broader and appears around 1360 cm−1 in a crowded region of the cell spectrum. Therefore, the detection of the intracellular pH can be achieved by probing the intensity of the CO2 peak. This implies a sufficiently long-lived intracellular CO2 pressure for detecting purposes. A better insight into the CO2 monitoring as outcome of intracellular acidification would improve the application of PCC as exogenous acidification tools. In general, the intracellular level of CO2 is a balance between CO2 influx/efflux and CO2 internal metabolization. In spite of the long-standing assumption that all cell membranes let gases pass extremely easily (Overton 1895) as effect of the high solubility of lipophilic gases in the membranes lipid phase, the CO2 flux is regulated both at the cholesterol level of the membranes (Itel et al. 2012) and by gas-channel proteins (Itel et al. 2012; Kustu and Inwood 2006; Endeward et al. 2008). As for the intracellular metabolization, carbonic anhydrase (CA) is perhaps alone among enzymes in being able to bind CO2 (Dodgson et al. 1980; Chegwidden et al. 2000).

In the current paper, we evaluate the modulation of intracellular pH as a function of the HDNS incubation time on the HEK-293 cells in order to get better insight into the overall PCC-induced intracellular acidification process. The cells are irradiated with UV light and monitored by infrared spectroscopy. We found out that the CO2 produced is similar for all of the samples. The time course of the CO2 integrated peak, however, depends on the incubation time and displays a combination of linear and steep increases followed by an exponential decay. This finding is interpreted in terms of HDNS confinement into intracellular vesicles of growing size depending on the incubation time and is sensitive to the UV light when reaching a critical average size. Intracellular vesicle explosion frees protons, which generate CO2 according to the equilibrium (1), which is then metabolized by CA/and or gas-channel proteins.

Materials and methods

Synthesis of HDNS

HDNS was prepared following the procedure described by Kabalka et al. (1986), with some variations. The synthetic route of the HDNS is reported in the Supplemental Information along with the spectroscopic characterization.

Cell culture

Human embryonic kidney 293 (HEK-293) cells were cultured directly on UV and infrared-transparent CaF2 windows, in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 2 mM glutamine, 10 % FBS, maintained in a humidified incubator at 10 % CO2, and periodically tested to ensure the absence of mycoplasma contamination. Afterwards, the cells were incubated with 2.2 mM HDNS in DMEM for different times (0.5, 1, 2, and 3 h). The cell-coated windows were then transferred to a custom-built sample holder.

Infrared spectroscopy

FTIR (Fourier-transform infrared) spectroscopy measurements were performed in transmission mode, with a Bruker IFS66/VS interferometer, under vacuum (~10−3 mbar), to avoid interference from atmospheric CO2. The cells were housed in a sandwich holder and a 12-μm Mylar spacer was used to control the sample thickness. The signal-to-noise ratio in the 2300–2400 cm−1 was 0.5 ÷ 1 %.

Photolysis and CO2 monitoring

UV radiation (275–375 nm) was provided by a deuterium discharge lamp (Acton Research Corporation) equipped with a band-pass FGUV11 filter (Thorlabs). The power density delivered to the sample is of the order of 0.1 W/mm2. In the photolysis experiment, an area of ca. 7 mm2 was illuminated by UV light. FTIR spectra have been collected with a spectral resolution of 2 cm−1. The photoreactivity of HDNS is reported in Fig. 1S of the Supplemental Information. The cells were monitored for a few minutes by acquiring sequential FTIR spectra in the dark. After confirming stability of the spectra, measurements were interrupted and the cells were irradiated for 1 min with near-UV light. FTIR measurements were then resumed after the irradiated period and repeated at least 20 times before a subsequent irradiation. The spectra are recorded at room temperature and normalized for the first spectrum after the irradiation. The control irradiation of non-dosed cells is reported in the Supplemental Information Fig. 2S.

Results and discussion

The incubation time of HEK-293 cells with HDNS has a large effect on the modes of CO2 production upon photolysis. This is shown in Fig. 1, where the integrated peak area of CO2 is reported as a function of 0.5, 1-, 2-, and 3-h incubation time, with arrows indicating the irradiation events [panels (a) through (d)]. The irradiation times were chosen after some preliminary tests, so that different kinetics of CO2 production and metabolization as well as mixed kinetic behaviors can be evidenced, in measurements taken in the shortest possible time span. The response of HDNS in DMEM is reported in Fig. 3S. Unlike the cells, the spectral region of the CO2 band appears to be unaffected by the irradiation.

Fig. 1
figure 1

Time course of the integrated CO2 peak area after UV-irradiation of the HEK-293 cells dosed with 2.2 mM HDNS at different incubation times: a 0.5 h, b 1 h, c 2 h, and d 3 h. The arrows indicate the irradiation events. e The CO2 integrated peak area of the 1-h incubation samples that underwent a single irradiation event. I incubation time. The vertical dotted-dashed lines indicate the limits of linear fitting. For the 1-h-1 irradiation, the sample has been fitted with the lines in the two intervals marked with 1 and 2 in e. The inset in each panel reports the curves of CO2 with the associated fits. The horizontal lines in b and c correspond to the CO2 to be added to the final peak to estimate the total amount produced

Full size image

This is interpreted as due to rapid diffusion of photolysis products through the solution, which leads to a too little change of acidity to be detected.

The general behavior of the CO2 integrated intensity after an irradiation event can be described as the combination of a linear increase (CO2-l) and a steep increase followed by a decay (CO2-d), that in some cases follows an exponential behavior. Linear sections and exponential decays were fitted (where possible) and the parameters used are reported in Table 1 along with the associated errors. Furthermore, in Fig. 1, the vertical dashed-dotted lines indicate the ranges taken for the linear fittings.

Table 1 Fitting parameters of the CO2 production curves for the different incubation times and irradiation modes
Full size table

The 0.5-h sample reaches the maximum level of CO2 production after a few minutes, followed by an exponential decay nearly at the starting level. The time constant of the exponential decay is 242 ± 8 s.

The first irradiation of the cells incubated for 1 h causes a linear increase of the CO2 peak area. A second irradiation after 25 min causes CO2 production to reach its peak level followed by a decay, then by a short linear increase again to a level that remains constant also after the third irradiation event. The decay cannot be immediately singled out as exponential, though it can be fitted (with a large error) by the superposition of an exponential decay and a linear increase with the same slope as the first section.

Cells incubated for 2 h and irradiated at 0, 25, and 50 min display a linear increase after the first irradiation event. The second irradiation causes a small steep increase readily metabolized followed again by a linear increase. The third irradiation causes the CO2 to reach its peak level and is followed by a decay to a level slightly higher than prior to the irradiation. The level remains constant also upon the following irradiation. The decays after the steep increases are too sharp and too small for fitting. Therefore, only a linear fitting has been made up to 2600 s, though excluding the points corresponding to the first steep increase (and decay).

The 3-h incubated sample is characterized by a linear increase after each irradiation event until it reaches the saturation level and then remains constant (check was made for a few minutes).

It is noteworthy that only the fraction of CO2 that was generated by a steep increase is subject to the (exponential) decrease. The part of CO2 that is generated according to a linear increase remains constant. As a consequence, the steady intracellular content of CO2 upon HDNS dosing into 293-HEK cells and subsequent irradiations can be related to the incubation time. The slope of the linear sections is different for the different incubation times. A check was made to show that the CO2 variation is not infrared triggered by interrupting the spectroscopic monitoring and retrieving it after a few minutes (region marked with a circle in Fig. 1d). The spectra recorded afterwards were still on the same curve.

Estimate of total CO2 production

The estimate of total CO2 produced in each of the measurements is straightforward in the 0.5- and 3-h incubation samples. In the former case, it corresponds to the CO2 peak, and in the latter, to the saturation level. For the 1-h sample, it has been considered that the overall CO2 production comprises the peak level reached after the second flash as well as the CO2 generated by linear increase after the decay. This latter contribution is marked in Fig. 1b between two horizontal lines. Similarly, for the 2-h sample, the peak level reached after the third irradiation had been additioned by the fraction of CO2 produced and metabolized after the second irradiation (marked by two horizontal lined in Fig. 1c).

The errors associated are of the order of 4 %, estimated as an average error of 2 % in the CO2 peak integral and another 2 % error in determining the saturation level. The estimated values of total amounts of CO2 production are reported in Table 2.

Table 2 Estimate of the total CO2 production
Full size table

The total CO2 production is very similar for all samples and can be related to the amount of PCC absorbed by the cells and sensitive to the irradiation.

Steady or steep CO2 increase

The CO2 peak increases after an irradiation event either linearly or steeply. A short incubation time leads to a steep increase of CO2 subsequently metabolized. Longer incubation times correspond to an initial linear increase followed by a steep increase of lesser extent up to the saturation level.

The rationalization of the “double behavior” can be made according to a few considerations.

  1. 1.

    The total CO2 production, CO2-l + CO2-d (i.e., the amount obtained both by linear and by steep increase followed by a decay) is similar for all incubation times. This is an indication that the all samples absorbed the same amount of HDNS and that 0.5-h incubation time is sufficient to reach the maximum HDNS absorption by the cell. Longer incubation times affect the kinetics of the intracellular metabolism, but not the total amount absorbed.

  2. 2.

    In aqueous solutions, the flash photolysis of proton caged compounds is of the order of nanoseconds (Gutman et al. 1981) and the CO2 generation according to equilibrium (1) has a time scale of 10−2 to 10−5 s (Schulz et al. 2006). Of the observed processes, the steep increase occurs in a time scale comparable with the flash photolysis of PCCs and CO2 generation in aqueous solution, whereas the linear increase takes place on the minutes-to-hours timescale. A slow CO2 increase as a result of a partial proton release is to be excluded because the behavior of the 0.5-h incubation time sample. In this case, the first irradiation event caused a steep increase of the CO2 level to saturation, thus proving both that it is sufficient to convert the whole HDNS absorbed by the cells and that the photolysis of the PCC is fast event on the scale of the whole CO2 generation and metabolism.

  3. 3.

    CO2-d is ready metabolized to a level similar to the one before the steep increase. At variance with this, the CO2-l is not metabolized at all. The linear behavior as a result of a balance of CO2 released and metabolized is to be excluded, because this would result in different total CO2 production for the different incubation times.

The overall CO2 production upon irradiation of HEK-293 cells dosed with HDNS for different incubation times is compatible with an HDNS confinement into intracellular vesicles of growing average size with growing incubation time, and a minimum size of 100 nm that ensures their stability (Pornpattananangkul et al. 2010). The accumulation into intracellular vesicles is a common reaction of cells to isolate intracellular exogenous agents. This is typical, for instance, for exogenous amines on mammalian cells, for which time of confinement and type of vesicle depend on several parameters, such as lipophilicity/hydrophilicity ratio, pKa, type of amine, i.e., if primary, secondary of tertiary (Dean et al. 1984). Later, it was shown that the accumulation of polyamines in Chinese hamster ovary cells proceeds in two tightly connected steps, i.e., the initial transport of the substrate across the plasma membrane into the cytosol via a classical transporter that requires rapidly followed by its sequestration in the polyamine sequestering vesicles (Soulet et al. 2004). A similar model, in our case, would explain the CO2 linear increase, the lack of CO2-l metabolization, and why the linear section precedes the steep increase.

The confinement of HDNS in vesicles creates a barrier between the photo-generated protons and the required HCO3 − to give rise to the CO2. Therefore, although the first irradiation event is sufficient to free all protons from intracellular HDNS, the slow step of the CO2 production is due to the migration either of the protons outside or the HCO3 − inside the vesicles. The latter option is more likely to occur because the subsequent CO2 production would occur inside the vesicles, where it cannot be metabolized. An outward flow of protons would cause CO2 production in the cellular cytosol where it could be metabolized by the CA and/or gas-channel proteins.

In this scenario, the vesicles would have a size distribution that changes with the incubation time. Factors that influence the vesicles growth after the irradiation event are the HCO3 − influx and/or CO2 production, a process which may occur with the aid of the cytosolic lipid droplets to help extend the vesicle membranes. Possibly, also vesicle fusion contributes to vesicle growth, though without really playing a role in the kinetic trend. In general, vesicle fusion may occur with or without content mixing, or with a content loss, depending on the type of cargo and on the fusion mechanism (Marrink and Mark 2003). In our case, a payload loss upon fusion can be excluded, because this would imply a CO2 release into the cell, with a consequent fast metabolization and deviation from the linear behavior, which is not observed. In a fusion without mixing (which would ensure the CO2 confinement), the vesicles can still be treated as separate entities in kinetic terms, hence with no impact on the kinetic behavior. A fusion with mixing content would have marginal effects on the kinetic behavior because the slowing effects related to the larger size of the vesicles (i.e., longer pathway to reach the protons in the inner shells of the vesicles, after depletion of the outer shells) are compensated by the actual mixing, which partially removes the proton concentration gradients. In the latter case, only the CO2-d/CO2-l ratio would ultimately be influenced because a larger number of vesicles would reach the critical sizes, at a given moment, due to the fusion, and a larger amount of CO2 is freed in the cell, to be metabolized with an exponential behavior (CO2-d).

The vesicle growth continues until reaching a critical size range, sensitive to UV irradiation. A vesicle’s variable size distribution would also explain the decreasing slope of the linear sections with the incubation time. The larger the vesicle, the longer, on average, the pathway for the HCO3 − to reach the protons, and the slower the process. The steep increase is then likely related to a disruption of the vesicles of the critical sizes, provided they are sensitive to the UV irradiation.

Free protons can then be metabolized by the cell in a similar way as for the short incubation time (0.5 h). The existence of critical size vesicles may be the cause of the small exponential for the 2-h incubation sample at the second irradiation event. In this round, only a small fraction of the vesicles reached a critical size that can be freed upon irradiation.

In order to verify the vesicle hypothesis and determine whether the irradiation event can trigger the explosion of the critical-sized ones, a test was made on a 1-h incubation sample. This sample was only initially irradiated for 1 min with UV light and then monitored by infrared spectroscopy, without any further irradiation. The integrated CO2 peak area is reported in Fig. 1e and it only shows a monotonous increase, with the same slope as for the multi-irradiated one in the same time range. Afterwards, the slope changes to a value similar to the 3-h sample until it reaches the saturation level. Therefore, we conclude that the UV irradiation is not only the trigger for the HDNS photolysis but also for the critical-size vesicle disruption.

Conclusions

HDNS is a proton caged compound that may be used as an exogenous agent for intracellular acidification upon dosing to cells and photolysis for effect of near-UV light. This causes an increase of intracellular CO2 due to the generation and decomposition of carbonic acid, which can be probed by infrared spectroscopy. The integrated CO2 peak intensity displays a kinetic behavior that is a function of the incubation time, and is a combination of a linear increase and a steep increase followed by an exponential decay.

The total CO2 produced is similar for all the samples, regardless of the incubation time. This implies that it is only related to the initial HDNS concentration and can be used as a probe of the overall intracellular acidification. The time scale of the process, though, changes with the incubation time and is much faster (a few minutes) for the short incubation and it is over 1 h for the longest incubation time.

The CO2 peak intensity behavior is interpreted in terms of confinement of the HDNS in intracellular vesicles with an average size distribution that grows with the incubation time. The vesicles are subject to size variation upon HCO3 − influx and subsequent CO2 generation according to a linear kinetic. Vesicle fusion also possibly occurs, though with an estimated marginal impact on the kinetic trend. The UV irradiation of the critical-size vesicles causes their breaking and release of the protons, which readily react and yield CO2.