Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2009 Oct;76(4):754-65.
doi: 10.1124/mol.109.057687. Epub 2009 Jul 13.

Photodynamic Effects of Steroid-Conjugated Fluorophores on GABAA Receptors

Affiliations
Free PMC article

Photodynamic Effects of Steroid-Conjugated Fluorophores on GABAA Receptors

Hong-Jin Shu et al. Mol Pharmacol. .
Free PMC article

Abstract

We have shown that fluorescent, 7-nitro-2,1,3-benzoxadiazol-4-yl amino (NBD)-conjugated neurosteroid analogs photopotentiate GABA(A) receptor function. These compounds seem to photosensitize a modification of receptor function, resulting in long-lived increases in responses to exogenous or synaptic GABA. Here we extend this work to examine the effectiveness of different fluorophore positions, conjugations, steroid structures, and fluorophores. Our results are generally in agreement with the idea that steroids with activity at GABA(A) receptors are the most potent photopotentiators. In particular, we find that an unnatural enantiomer of an effective photopotentiating steroid is relatively weak, excluding the idea that membrane solubility alone, which is identical for enantiomer pairs, is solely responsible for potent photopotentiation. Furthermore, there is a significant correlation between baseline GABA(A) receptor activity and photopotentiation. Curiously, both sulfated steroids, which bind a presumed external neurosteroid antagonist site, and hydroxysteroids, which bind an independent site, are effective. We also find that a rhodamine dye conjugated to a 5beta-reduced 3alpha-hydroxy steroid is a particularly potent and effective photopotentiator, with minimal baseline receptor activity up to 10 muM. Steroid conjugated fluorescein and Alexa Fluor 546 also supported photopotentiation, although the Alexa Fluor conjugate was weaker and required 10-fold higher concentration to achieve similar potentiation to the best NBD and rhodamine conjugates. Filling cells with steroid-conjugated or free fluorophores via whole-cell patch pipette did not support photopotentiation. FM1-43, another membrane-targeted, structurally unrelated fluorophore, also produced photopotentiation at micromolar concentrations. We conclude that further optimization of fluorophore and carrier could produce an effective, selective, light-sensitive GABA(A) receptor modulator.

Figures

Fig. 1.
Fig. 1.
Examples of the data and protocols used for the evaluation of compound activity. A, responses from voltage clamped oocytes (−70 mV) expressing α1β2γ2L GABA receptor subunits and challenged with 2 μM GABA and GABA plus increasing concentrations of test compounds. Compound CW17 has been used in our previous publication on photopotentation (Eisenman et al., 2007) and therefore serves as our exemplar here. B, example of inhibition of [35S]TBPS binding by compound CW17. The IC50 is given in Table 1. C, photopotentiation assay in hippocampal neurons using CW17 as an example. The horizontal bars denote application times. Excitation light was 480 nm. The dotted line in this and subsequent traces indicates initial holding current. Note that at 0.1 μM compound CW17, no baseline GABAA receptor activity is detected. Note that light induced current is largely lost upon GABA removal, indicating that the outward light generated current is potentiated GABA current with some small directly gated residual outward current.
Fig. 2.
Fig. 2.
Compound fluorescence. A, in vitro fluorescence of the compounds (0.1 μM) tested in this work (ethanol solvent). Each mean and S.E.M. is calculated from three measurements. B, cellular fluorescence of the compounds, taken from images acquired during the photopotentiation assay. Note that the charged compounds CW38, CW28 (Alexa Fluor 546 conjugate), and AKB5 exhibit strong in vitro fluorescence but weak intracellular fluorescence, presumably because the negative charge on these compounds prevents good cellular penetration. Number of cells tested is the same as that given in summary graphs of functional effects shown in Figs. 3 to 5 and 7.
Fig. 3.
Fig. 3.
Comparisons among steroid analogs with fluorescent groups attached to steroid carbon 2. A1 and B1 show structures of the compounds. A2 and B2 show baseline activity on GABAA receptors expressed in oocytes at three concentrations (mean ± S.E.; n = 4 oocytes for each compound). The dotted line at 100% represents the response to GABA alone. Assays were performed as in Fig. 1A; A3 and B3 show summary data from hippocampal neurons challenged as in Fig. 1C with the indicated compounds at 0.1 μM (n = 3–6 cells per bar). Black bar, normalized baseline compound activity before light onset; dotted line, response to GABA alone before compound addition; white bar, normalized response in the presence of 480 nm light.
Fig. 4.
Fig. 4.
Compounds with NBD fluorophore attached by different linkers to steroid carbon 11. Data from compound CW17 are repeated here for comparison. Summary data are similar in format to Fig. 3. A, structures. B, oocyte responses (n = 4 oocytes each compound). C, baseline (black bars) and light-activated (white bars) activity on GABA responses in hippocampal neurons. The inset shows the baseline activity of 0.1 μM compound at higher resolution. Consistent with oocyte data in B, compounds CW23 and CW26 show most baseline activity and highest photopotentiation in this series.
Fig. 5.
Fig. 5.
Effects of enantiomeric steroid ent-CW25 on photopotentiation. A, structures of compound CW25 and its enantiomer ent-CW25. B, baseline GABAA receptor activity is not observed for ent-CW25 on oocytes at concentrations up to 10 μM, C, example traces from the photopotentiation protocol in hippocampal neurons (0.1 μM compound and 480-nm light wavelength). The traces come from different cells. D, summary of photopotentiation protocol in hippocampalneurons (n = 9 for 0.1 μM CW25; n = 10 for 0.1 μM ent-CW25 and n = 6 for ent-1 μM CW25). Black bars, baseline (before light); white bars, excitation with 480 nm light.
Fig. 6.
Fig. 6.
Correlation between GABAA receptor activity and photopotentiation. Ranked GABA receptor activity was established by taking potentiation values at 10 μM compound, or the reciprocal of the fractional response in the case of inhibitory actions, in the oocyte assay. This ranked order was correlated with photopotentiation values from hippocampal neurons. The correlation was significant (Spearman's ρ = 0.588, p < 0.05, n = 13 NBD/fluorescein-tagged compounds). Compound identifiers are used as symbols. Solid line is a linear regression.
Fig. 7.
Fig. 7.
Effects of 5β-reduced steroid carrier and of red-shifted fluorophores. A, structures. B, summary of oocyte effects (n = 4 oocytes for each compound). C, top traces are representative of responses to blue light of the three compounds (0.1 μM) in hippocampal neurons. The bottom traces are responses from cells evaluated with green light under the same conditions. Each trace is from a separate cell. D, summary of evaluations performed as in C (n = 3–5 for each bar). Note that optimal photopotentiation occurs with excitation wavelengths best matched to the fluorophore's excitation spectrum. Compound CW28 did not respond to green light until the concentration was raised to 1 μM (D, inset).
Fig. 8.
Fig. 8.
Intracellular loading of fluorophores is not effective. A, comparison of CW38 (sulfated steroid) photopotentiation with extracellular application (left) and intracellular application of 50-fold higher concentration (right). Effective loading was verified by intracellular fluorescence (insets). B, comparison of compound CW28 (negatively charged Alexa Fluor 546 fluorophore) extracellularly and intracellularly applied. C, summary of effects of the two charged compounds applied to either side of the membrane (n = 3 cells for each bar, except for compound CW38 outside, where n = 24).
Fig. 9.
Fig. 9.
FM1-43 photopotentiation. A, robust photopotentiation by 5 μM FM1-43. GABA concentration was 0.5 μM, A, the standard screening protocol is shown in the left trace. In the right trace, GABA alone was re-applied 1 min after the initial photopotentiation. The dashed line extending from A1 to A2 represents the approximate original GABA response, to highlight the residual potentiation after a 1 min wash. B, structure of FM1-43. C, fluorescence image of a cell pre-exposed to 5 μM FM1-43 for 20 s, at the onset of illumination.

Similar articles

See all similar articles

Cited by 2 articles

Publication types

LinkOut - more resources

Feedback