Electrophysiological study of the effects of side products of RuBi-GABA uncaging on GABA_A receptors in cerebellar granule cells

Introduction

The GABA_A is the major inhibitory neurotransmitter receptor in the mammalian brain; it comprises five subunit proteins from different subfamilies, forming a selective chloride channel. The existence of various homologous subunits gives rise to many GABA_A receptor subtypes. Depending on the subunit composition and arrangement, these receptors exhibit different physiology and pharmacology together with a distinct regional, cellular, and subcellular distribution [1,2,3,4]; each receptor subtype also contributes to the modulation of distinct functions of the brain [5,6,7,8]. Thus, studying the distribution of these receptors and their subunits in specific neuronal domains (e.g., cell body, dendrites, and nerve endings) is of paramount importance to shed light on their role in the central nervous system (CNS).

Moreover, the distinct cellular and subcellular locations of individual receptor subtypes suggest that they exhibit specific functions in the brain that can be selectively modulated by subtype-specific drugs [9].

To achieve this ambitious goal, it is necessary to have the right tools. Since the 1970s, a new class of compounds, called caged compounds [10], has emerged as powerful tools for stimulating a selected biological target in space or time [11]. The idea behind the caging technique is that a molecule of interest is biologically inert (i.e., caged) until the chemical bond with the cage, which works as protecting group, is removed, for example, by photoexcitation with a laser beam. Illumination results in a concentration jump of the biologically active molecule that can bind to its cellular receptors, switching on (or off) the targeted process [11].

In particular, caged compounds represent powerful tools to stimulate the cell locally, especially when two-photon excitation is used, as it may allow focal bursts of release with a volume of much less than 1 fL [11,12]. The potentialities of this approach have boosted the development of many compounds so much that every kind of signaling molecule has been caged [13]. As reported by Mayer and Heckel, the spatially well-defined and rapid change in the concentration of caged agonists or antagonists of neuronal receptors, induced by flash photolysis, is of great value for investigating the kinetic and mechanistic aspects of receptors, transporters, and ion channels. Therefore, caging technology was applied to various neurotransmitters, including γ-aminobutyric acid (GABA), and different strategies and protecting groups, such as α-carboxy-2-nitrobenzyl (CNB)-, 4-carboxymethoxy-5,7-dinitroindolinyl (CDNI)-, [1,3-bis(dihydroxyphosphoryloxy)propan-2-yloxy]-7-nitroindoline (DPNI)-, 4-methoxy-5,7-dinitroindolinyl (MDNI)- [14] have been used for caging GABA. Among these groups, ruthenium-bipyridine-triphenylphosphine- (RuBi) has been used to cage GABA, resulting in a compound called RuBi-GABA, which has high one-photon efficiency of uncaging upon irradiation at a visible wavelength [15], a characteristic directly related to the product of the extinction coefficient and the quantum yield of photolysis. The extensive use of RuBi-GABA promoted a thorough investigation of its side effects. In particular, RuBi-GABA’s antagonistic effect on GABA_A receptors was studied by different researchers: Rial Verde et al. [16] stated that they observed antagonistic effects at millimolar concentrations, while for concentrations below 20 μM, no anomalies in membrane resistance and electrophysiological behavior were observed, Matsuzaki et al. [15] reported that no antagonism occurs below 20 µM, while more recently Ellis-Davies et al. [12,17] reported that RuBi-GABA has an IC₅₀ of 4.4 µM and supposed that the origin of the antagonism is due to the fact that a carboxylate of each probe can enter the GABA binding cleft and act as a competitive antagonist to GABA. Recently, we investigated how physical parameters, like distance, power, exposure time, etc., affect the electrophysiological behavior of cerebellar granule cells (CGC) in response to RuBi-GABA uncaging with one- and two-photon excitation [18]. In order to push further our investigation, we performed new experiments to analyze the effects of the side products obtained after one-photon (1P) uncaging process. Here we report the procedure we used to separate the cage from GABA after uncaging with a laser source, and we compare the electrophysiological measurements acquired with and without dispending the cage molecules in the perfusing bath. Moreover, to evaluate the role of the caging molecule both near and far from the cell soma, we compare experiments performed by changing the distance of the uncaging point from the cell.

Materials and methods

Results and discussion

The rationale of our experiments was investigating the effects of the side products of RuBi-GABA uncaging process on GABA_A receptors in CGCs, and studying the differences in electrophysiological response, due to these molecules, based on the distance from cell soma and on the polarity of the holding potentials. With this in mind, we first developed a strategy to separate the “RuBi-cage” from the γ-aminobutyric acid in a photolyzed RuBi-GABA solution. Then, we performed electrophysiological measurements, administering solutions of GABA, and solutions of GABA with the RuBi-cage. Finally, we compared patch-clamp experiments performed uncaging RuBi-GABA at different distances from the cell soma. In Figure 1a, we report the images of a pipette attached to a cerebellar granule during patch-clamp experiments with RuBi-GABA. The neuron soma and dendrites are highlighted by using biocytin, a fluorescent dye (Figure 1b), while two typical uncaging points, one close to (P1) and the other far from (P2) the cell soma, are shown in Figure 1c.

Figure 1

Bright-field image of the capillary tip sealed to a cerebellar granule during electrophysiological measurement using the patch-clamp technique in whole-cell configuration (a) Fluorescence image of the cerebellar granule, labeled with biocytin, attached to the glass tip (b). Merge of bright field and fluorescence images (c) P1 and P2 represent the uncaging positions near (∼2 µM) and far (∼30 µM) from the cell soma, respectively.

To retrieve the cage molecules from a RuBi-GABA solution exposed to laser irradiation, we exploited two analytical techniques: high-performance liquid chromatography (HPLC), a technique typically used to separate molecules from heterogeneous solutions, and mass spectrometry, which allows the identification of unknown compounds via molecular weight determination, and the quantification of known compounds, and the determination of the structure and chemical properties of molecules (e.g., mass to charge m/z ratio). Combining these techniques, we collected the cage molecules, i.e., ruthenium-bipyridine-triphenylphosphine, starting from a solution of RuBi-GABA, i.e. (bis(2,2′-bipyridine-N,N′)triphenylphosphine)-4-aminobutyric acid ruthenium hexafluorophosphate complex, and we determined the electrical charge of the cage, which is due to the positive charge of the [Ru(bpy)₂(PPh₃)(GABA)]⁺ (bpy = 2,2′-bipyridine, PPh₃ = triphenylphosphine) complex [21]. Here we describe, in detail, the procedure. First, we diluted RuBi-GABA, from stock solution, in Milli-Q water and analyzed the solution before compound photolysis. In Figure 2, the DIA analysis mass spectrum shows the presence of signals at m/z 924, these signals are compatible with a compound with the same raw formula of RuBi-GABA, as declared on the datasheet provided by Tocris-Bioscience, i.e., C₄₂H₃₉N₅PO₂RuPF₆, then, zooming the peak, it is clear that the isotopic distribution of molecular ions is compatible with the theoretical one, calculated using the Molecular Weight Calculator freeware software (ver 6.5), and reported in Figure S1 in Supplementary Information. Then, we processed RuBi-GABA by LC-MS; the mass spectrum at RT 16.4 min is shown in Figure 3. A chromatographic peak is present at an m/z of 778, which corresponds to a RuBi-GABA complex without PF₆; as we can see zooming the peak, the isotopic distribution of the molecular ions is compatible with the theoretical spectrum calculated considering the raw formula C₄₂H₃₈N₅PO₂Ru. Finally, we exposed a RuBi-GABA solution to irradiation with a visible wavelength (405 nm) portable laser for 5 s to photolyze RuBi-GABA molecules. After this treatment, we analyzed the solution in HPLC-ESI–MS. Two signals at RT 16.4 and 17.5 min are present (Figure 4), the former (m/z 778) corresponding to caged RuBi-GABA with no PF₆ complex, the latter (m/z 716) corresponding to uncaged RuBi-GABA, such that the m/z value is related to the addition of Na with one water molecule. In fact, a zoomed view of the mass spectrum, acquired at RT 17.5 min, shows an isotopic distribution of the molecular ions, which is compatible with the theoretical one calculated considering the expected raw formula C₄₂H₃₈N₅PO₂RuNaH₂O. We refer to this complex as “RuBi-cage.” The theoretical spectrum is shown in Figure S1 in Supplementary Information. In order to determine the exposure time necessary for complete photolysis of RuBi-GABA, we analyzed 5 samples exposed to laser irradiation for 3, 6, 9, 12, and 15 s. Results are summarized in Table 1, and show that, after a 6 s exposure time, RuBi-GABA is completely photolyzed.

Figure 2

Mass spectrum of the compound collected after the insertion of RuBi-GABA. The isotopic distribution of the molecular ions is zoomed in the lower graph; it is compatible with the theoretical isotopic distribution of a compound with raw formula C₄₂H₃₉N₅PO₂RuPF₆, whose graph is reported in panel A of Figure S1 (Supplementary Information).

Figure 3

The compound mass spectrum with an m/z of 778 collected after 16.4 min. The isotopic distribution of the molecular ions is zoomed in the lower graph; it corresponds to the theoretical isotopic distribution of a compound with raw formula C₄₂H₃₈N₅PO₂Ru, whose graph is reported in panel B of Figure S1 (Supplementary Information).

Figure 4

Mass spectrum of the compound after photolysis. The isotopic distribution of the molecular ions, zoomed in the lower box, is compatible with the theoretical distribution of a compound with raw formula C₄₂H₃₈N₅PO₂RuNaH₂O, reported in panel C of Figure S1 (Supplementary Information).

In order to retrieve enough RuBi-cage molecules for electrophysiological experiments, we processed photolyzed RuBi-GABA solutions to collect the molecules corresponding to the peak of interest. Then, we dried the sample by using a speed vac. To evaluate possible interferences of the cage on GABA_A receptors of CGCs, we prepared solutions of RuBi-cage alone and RuBi-cage plus GABA, having the same molarity of RuBi-GABA used for patch-clamp experiments, i.e., 10 µM. We used these solutions to verify if the RuBi-cage elicited a response on GABA_A receptors in living neurons. For this purpose, we bath-applied the compound to cultured CGCs from P7 rats and recorded whole-cell patch-clamp current traces, maintaining the holding potential at −40 mV; no current pulses were evoked. Then, we performed patch-clamp experiments to compare current pulses, evoked by 10 µM GABA, with pulses evoked by 10 µM GABA and 10 µM RuBi-cage, both for negative and positive holding potentials (Figure 5). For both polarities, the solutions of GABA with RuBi-cage evoked peak currents with lower intensity (in absolute value), an effect probably related to the interaction of the cage with the GABA binding site of GABA_A receptors. In Figure 6, we report a scatter plot with the values of the current peak evoked, on the same cell, by administration of GABA (x-axis) and GABA with the RuBi-cage (y-axis) (same molarity). Fitting the data by using a linear curve, we obtain a slope of 0.79 ± 0.04, which confirms the role of the RuBi-cage in decreasing the peak current evoked by GABA. Then, we quantified the percentage decrease (D%) of the current peak intensity on N = 6 cells by using the formula:

Figure 5

Representative current traces, in whole-cell configuration, at a holding potential of −40 mV (trace a) and +40 mV (trace b) of the effect 10 µM RuBi-cage plus 10 µM GABA on the same cell. The current is evoked by 10 µM GABA (left), 10 µM RuBi-cage alone (center), and 10 µM RuBi-cage plus 10 µM GABA (right).

Figure 6

Scatter plot of the peak current intensity values, measured on the same cell (measures performed on the same cell are reported using same colors), during administration of a solution of GABA (x-axis) and GABA plus the RuBi-cage (y-axis) (holding potentials −40 mV and +40 mV). The slope of the linear fit is 0.79 ± 0.04. The value of the correlation coefficient 0.97 suggests a linear relationship between the two series of experimental data N = 6. Error bars are smaller than symbols’ size.

We summarize the results in Table 2.

For positive and negative holding potentials, the values of D% are close; therefore, the role of the cage seems not to be dependent on the current direction. Even if early studies by Matsuzaki and Rial Verde [15,16] show no antagonistic effects for RuBi-GABA molarity below 20 µM, our results confirm more recent studies, which report RuBi-GABA antagonism on GABA_A receptors for molarities as low as 4.4 µM [12,17].

Then, we acquired I–V curves (from −60 to +60 mV, step: +20 mV) uncaging RuBi-GABA taking advantage of the point-by-point confocal laser scanning architecture to investigate this aspect further. In Figure 7, we report scatter plots of the peak current rectification ratio ( ∣ I peak + / I peak − ∣ ), defined as the absolute value of the quotient between the peak current recorded for one voltage polarity, e.g., positive I peak + , and the peak current recorded for the same voltage at the opposite polarity, e.g., negative I peak − , versus the absolute value of the holding potential. In particular, the plots refer to I–V curves acquired by uncaging RuBi-GABA nearly 2 µm (left) and 30 µm (right) far from the cell soma. For all holding potentials, the rectification ratio is compatible with 1 (Figure S2 in Supplementary Information); thus, a clear asymmetry in the electrophysiological behavior, due to RuBi-GABA uncaging, was not observed.

Figure 7

Absolute value of the ratio of current peaks measured for positive I peak + and negative I peak − holding potential, respectively, i.e., same voltage but opposite polarity. The graph on the left refers to data recorded for uncaging positions near the cell soma (like, for example, position P1), the graph on the right refers to uncaging points 30 µm far from the cell soma, such as position P2. Data points of the same color refer to the same cell measured at different holding potentials. The graph on the left refers to measurements on 20 cells, on the right on 7 cells.

Nevertheless, the values of ∣ I peak + / I peak − ∣ measured in the cell soma's proximity are distributed on a broader range, i.e., have a significantly higher relative standard deviation (RSD), compared to data collected far from the cell soma (Table S1). We speculate this phenomenon could be attributed to the higher probability of interaction of RuBi-cage molecules with GABA_A receptors occurring when molecules are uncaged in the proximity of the soma. These interactions may cause instabilities in the electrophysiological behavior of neurons, an aspect that must be carefully considered when using the RuBi-GABA compound to study GABAergic mechanisms in neurons.

Conclusion

Caged compounds are powerful tools for many research fields in life sciences, as they allow to control, in space and time, the release of molecules of interest. This functionality is of paramount importance when studying neuroscience as it allows to stimulate locally a specific target, for example, receptors like GABA_A, and map their distribution on a cell or a tissue. Our study concentrated on a specific commercial GABA-caged compound called RuBi-GABA. Its extensive use motivated us to investigate the influence of the side products obtained after photolysis on the electrophysiological behavior of GABA_A receptors in CGCs due to the extensive experience of our group in this field. We focused on studying the effect of the RuBi-cage, i.e., the compound released after photolysis, whose raw formula is C₄₂H₃₈N₅PO₂RuNaH₂O. We developed an analytical procedure to completely photolyze a solution of RuBi-GABA and extract only the RuBi-cage. Then, we studied the effect of this side-product comparing electrophysiological currents evoked administering GABA and GABA with the RuBi-cage, i.e., avoiding in situ-uncaging. A percentage decrease (around 15–19%) of the peak current, independent of the current direction, suggests that the RuBi-cage may interact with GABA_A receptors with a mechanism not dependent on the RuBi-cage positive sign. Moreover, the rectification ratio of peak currents evoked by one-photon in-situ uncaging near and far from the cell soma shows that traces recorded near the soma are affected by higher variability in electrophysiological behavior. This phenomenon could be reasonably related to the higher probability of interaction of the RuBi-cage with the GABA_A receptor when uncaging occurs near the cell soma.