1. Introduction

Fluorescence is an important and well-established probe in biological research, and various fluorescent dyes have been developed and utilized for a variety of biological targets [1-5]. Of them, the voltage-sensitive dyes play an increasingly important role in neuron studies in that they can non-invasively detect membrane potential change [2-5]. Traditionally, membrane potential and electrical current within a neuronal network are measured using the patch-clamp technique [1]. Unfortunately, the invasive electrodes may disturb neurons and their networks, and, thereby, interfere with their signal propagation [2-4]. Moreover, it is very difficult to probe neurons of micron scale. Such disadvantages can be overcome by fluorescence probing. Thanks to the efforts of a number of groups [2-6], many voltage-sensitive dyes have been developed. Basically, they can be classified into fast- and slow-response groups. While a slow dye responds to a potential change with a more pronounced fluorescence change, a fast dye responds sufficiently fast to detect action potentials in neurons. Of the fast-response group, the ANEP (amino-naphthyl ethenyl-pyridinium) series are the most reliable, sensitive and consistent for detection of sub-millisecond membrane potential changes in a variety of tissue, cell and model membrane systems [5-15]. For example, di-3-ANEPPDHQ has been spectrally characterized and used to probe action potentials from mammalian nerve terminals in situ [8].

Of the ANEP series, di-4-ANEPPDHQ (C₃₂H₄₇Br₂N₃O₂) is among the most successful [9-15]. Its chimeric structure (Fig. 1) [9] , different from di-3-ANEPPDHQ only in that two butyl groups replace the two propyl groups of the di-3-ANEPPDHQ [8], consists of the napthylstyryl-pyridinium chromophore from di-8-ANEPPS, a better known member of the ANEP series, and the progenitor (quaternary ammonium headgroup, DHQ) from the dienylstyryl-pyridium dye RH795 [9]. While the DHQ progenitor makes di-4-ANEPPDHQ water-soluble and less phototoxic than di-8-ANEPPS, the napthylstyryl-pyridinium chromophore renders its fluorescence as sensitive to potential change as di-8-ANEPPS, and yet the dye does not internalize as easily as RH795 [9]. This dye has been used to image the submucousal plexus of a guinea-pig in vitro [9, 10], live tobacco BY-2 cells [11], and live human embryonic kidney 293 cells [12, 13], and to probe the plasma membrane potential change in the submucousal plexus [9, 10]. Its emission spectrum in live BY-2 cells has also been measured [11]. Additionally, the di-4-ANEPPDHQ fluorescence is also sensitive to the degree of order of lipid solutions in water, and this dye has been used to visualize model giant unilamellar lipid vesicles and rafts (liquid-ordered domains) on the plasma membrane of live neutrophils. Its excitation and emission spectra in large unilamellar lipid vesicles (LUVs), and its red/green emission intensity ratio in neutrophils, have also been measured [14, 15].

 

Fig. 1. Chemical structure of the dye di-4-ANEPPDHQ [9]

Download Full Size | PDF

Although di-3-ANEPPDHQ has been characterized and used to probe action potentials from a mouse neurohypophysis in situ [8], the spectral characteristics of di-4-ANEPPDHQ in a living cell environment have not, to our knowledge, been systematically characterized. Such information is a prerequisite for a fluorescence probe and its absence will limit the applications of the probe. This work, therefore, measures absorption, excitation and emission spectra of the dye in its stock solution and staining solution, and bound to the plasma membrane of in vitro primary and immortalized neurons. Usually primary neurons do not divide and have lower viability and shorter lifespan than immortalized cells, which divide readily. The primary and immortalized neurons we chose were rat hippocampal neurons and mouse hypothalamic neurons (GT1-7), respectively. The former have a differentiated morphology, with clearly distinguishable axons and pronounced dendrites, play a very important role in memory, and yet the related mechanism of their role has been widely debated [16]. The latter are a sub-line of GT1, the first neuroendocrine clonal cell line derived from the central nervous system [17], have a large body, tend to grow into a network-like colony, anchor firmly to untreated dishes or coverslips, and are widely used in neuronal research [17, 18].

2. Experimental

Commercially available di-4-ANEPPDHQ powder (Invitrogen, CA) was first dissolved with the inorganic solvent, dimethyl sulfoxide (DMSO), yielding a stock solution of 1.50mM. The stock solution was further diluted using a commercially available aqueous, phenol red-free, Hank′s buffered salt solution (HBSS) (Invitrogen, CA) to obtain a staining solution of 1.50μM.

The culturing procedures for the primary hippocampal neurons and immortalized hypothalamic neurons were quite different and described in detail in [19] and [2], respectively. Briefly, hippocampal neurons were extracted from the hippocampal tissue of embryonic day 18 rats and delivered in Hibernate E solution from Brainbits LLC (Brainbits, IL). The hippocampal tissues were dissociated at 30°C for 30min in 2mg/ml papain (Worthington, NJ) in hibernate E-Calcium solution (Brainbits, IL), then triturated and plated at a density of 160 cells/mm² onto a glass coverslip coated with poly-D-lysine (Brainbits, IL). Plated neurons were maintained in neurobasal medium (Invitrogen, CA) containing B27 supplement (Invitrogen, CA), at 37°C in an incubator with humidified 8% CO2. After 4 days, one half of the medium was changed with Neurobasal/B27 medium containing 0.5 mM glutamine (Invitrogen, CA). By comparison, the immortalized neuron culture was simpler. The GT1-7 cells were maintained in 25cm² flasks (Fisher Scientific, GA) at 37°C in an incubator with humidified 8% CO². Cells were grown in Dulbecco’s modification of Eagle’s medium (Gibco, NY) supplemented with 1 mM sodium pyruvate, 10 mM Na₂CO₃, 2 mM L-glutamine, 10 mM Hepes buffer, and 10% fetal bovine serum (Gibco, NY).

Prior to staining, the culturing media was removed and cells were washed using Hank’s buffered salt solution. Cells were stained with 1.50μM di-4-ANEPPDHQ for 17 min, and washed again using HBSS. Stained neurons were measured directly on coverslips, although coverslips do not easily fit our spectrophotometers. To circumvent this problem, as well as to allow depolarization of membranes by KCl, additional samples of stained GT1-7 neurons were trypsinized, and dispersed in HBSS to obtain a suspension for spectral measurement. For comparison, and to obtain sufficient solution for spectral analysis, the dye stock solution was diluted in DMSO to the same concentration of the staining solution (1.50μM). The diluted dye stock solution, the staining solution, and the suspension of stained GT1-7 in HBSS were loaded into 10mm×10 mm×65mm quartz cuvettes for spectral characterization. While the use of a cuvette allows for easy addition of KCl solution to depolarize their plasma membrane, as described below, suspension cultures of hippocampal neurons could not be obtained because the stained primary neurons had low viability and did not survive trypsinization and dispersion.

The absorption and fluorescence spectra were measured using a Varian Cary 500 UV-VIS-NIR spectrophotometer and a Varian Cary Eclipse fluorescence spectrophotometer, successively. First, the absorption spectra were collected at a wavelength scan rate of 60nm/min and with a slit bandwidth of 2nm. Then, emission spectra were collected with the excitation slit fixed and centered at the absorption peak. Finally, excitation spectra were measured with the emission slit fixed and centered at the emission peak. The wavelength scan rate was 120nm/min and the band width of the excitation is 5nm, which corresponds to the resolution of the excitation and emission spectra. Since the potassium ion concentration controls the neuron membrane potential, addition of highly concentrated KCl solution can effectively depolarize the membrane potential [1], which, in turn, causes a change in the fluorescence spectrum. To observe the effect of chemical stimulation on fluorescence spectra, 1M KCl solution was added to the stained GT1-7 cells/HBSS suspension in a cuvette immediately after the spectra were measured, and fluorescence spectra were collected once more. Again, as KCl could not be added to coverslip cultures, we did not measure spectra of stained, depolarized hippocampal cells. Stained hippocampal and GT1-7 neurons on coverslips were imaged using an Olympus IX81 microscope in phase-contrast and fluorescence modes, respectively. The fluorescence filters were configured according to the excitation and emission spectra of the stained cells: 450-490nm (excitation), 560nm (longpass edge, dichromatic), and 575-625nm (emission).

3. Results and discussion

Figure 2 shows the absorption spectra of the diluted stock solution, the staining solution, and the stained primary and immortalized neurons. Obviously, both the dye and solvent contribute to absorption of the diluted stock solution and the staining solution. The dye absorption peaks around 503nm after it was dissolved in DMSO and the peak shifts to 485nm after the dye was dissolved in H₂O (in HBSS). The change of solvent from DMSO to H₂O also reduces the absorbance. After the staining, the absorbance of the dye bound to the plasma membrane of both neurons increases drastically and peaks around 470nm. In contrast, the absorption of mouse neurohypophysis stained with di-3-ANEPPDHQ peaks around 520nm [8].

 

Fig. 2. Absorption spectra of the diluted stock solution (dye/DMSO), the staining solution (dye/HBSS), and stained primary hippocampal neurons and immortalized hypothalamic neurons. For comparison, the absorption spectra of DMSO and HBSS are also presented.

Download Full Size | PDF

Figure 3 shows the excitation and emission spectra of the diluted stock and staining solutions, and the stained neurons before and after KCl addition. (As GT1-7 neurons on coverslips and in suspension produced the same spectra, only suspension spectra are displayed in Fig. 3 to allow direct comparison with KCl-depolarized GT1-7 cells.) The fluorescence intensity of the staining solution is significantly lower than that of the diluted stock solution, manifesting an obvious fluorescence-quenching effect of the aqueous solvent (Fig. 3(a)). The change in solvent from DMSO to water also shifts the excitation and emission peaks from 505nm and 728nm to 478nm and 660nm, respectively (Fig. 3(b)). Such changes are typically observed as the local chemical environment of dye molecules changes (solvatochromism) [3]. In our case, water is more polar than DMSO and of a higher dielectric constant.

After dye molecules bind to the neuron membrane, the fluorescence intensity increases greatly (Fig. 3(a)), and the excitation and emission peaks further shift to 479 nm and 570 nm for primary neurons, and to 476nm and 585nm for GT1-7 neurons (Fig. 3(b)). Such variations are attributed not only to the solvatochromism, as the chromophores of the dye molecules are detached from water molecules and inserted between the much less polar hydrocarbon chains of the lipids of the plasma membrane, but also to the intra-membrane electric field as it induces the chromophore’s excitable charge to shift from the pyridinium to the naphthystyryl anilino nitrogen [3, 5, 9]. Obviously, excitation and emission spectra of the stained hippocampal and GT1-7 neurons are different, and differ from those of LUVs and neutrophils stained with di-4-ANEPPDHQ [9-15] and other bio-targets stained with the other ANEP dyes [6-8] (Table 1). This strongly suggests that a specific spectral characterization is important if one wants to apply di-4-ANEPPDHQ to a new bio target or adopt a new staining protocol. Interestingly, the difference between excitation spectra is less than emission spectra.

Although the KCl addition increases the fluorescence intensity (F) of stained GT1-7 neurons (Fig. 3(a)), it does not change the normalized spectrum (Fig. 3(b)), or the relative change ∆F/F is positive and roughly constant. By comparison, upon membrane depolarization in bio-targets stained with di-4-ANEPPDHQ or other ANEP dyes, ∆F/F varies with the wavelength [4-15]. For example, electrical depolarization in hypothalamic nerve terminals stained with di-3-ANEPPDHQ, increases fluorescence intensity at 400nm excitation, but decreases it around 525nm [8]. The ∆F/F sign changes because the absorption spectrum of di-3-ANEPPDHQ is blue-shifted in response to depolarization [8]. The mechanism of the depolarization effect in our case may be different and needs further investigation.

 

Fig. 3. (a). Raw and (b) normalized excitation and emission spectra of the diluted stock and staining solutions, the stained primary hippocampal and immortalized hypothalamic neurons in HBSS before and after the addition of 1M KCl solution. (a) and (b) have the same legends.

Download Full Size | PDF

Table 1. Spectral characteristics of different solutions or targets tagged with ANEP dyes

View Table

Figure 4 shows phase-contrast and fluorescence micrographs of the primary hippocampal neurons and immortalized hypothalamic neurons. As had been expected, the hippocampal cells have more differentiated morphology, with clearly distinguishable axons and pronounced dendrites. However, their density is lower than that of the GT1-7 neurons because they, unlike the latter, can not divide. The occasional spherical cells in Figs. 4(b) and 4(d) are detached from the Petri dish bottom, floating in HBSS. Still other cells appear substantially brighter (Fig. 4(b)). They might have internalized the dye as the micrographs were taken some time after staining. The internalization, if it really occurred, implies that considerable care is still needed even for a dye that does not internalize readily. However, the elevated brightness may also be due to, or enhanced by, the absence of optical sectioning in our wide field microscope. The fluorescence images in Figs. 4(c) and 4(d) were obtained using a fluorescence filter set especially tailored to match the excitation and emission spectra of GT1-7 cells, and to ensure optimal visualization of the stained cells.

 

Fig. 4. phase-contrast (a, b) and fluorescence (c, d) images of stained primary hippocampal neurons (a, c) and immortalized hypothalamic neurons (b, d).

Download Full Size | PDF

4. Conclusions

Absorption, excitation and emission spectra of a recently introduced, fast voltage-sensitive dye, di-4-ANEPPDHQ, were measured in its stock solution with dimethyl sulfoxide as the solvent, in its staining solution with water as the main solvent, and bound to the plasma membrane of in vitro primary rat hippocampal neurons and immortalized mouse hypothalamic neurons. The measured spectral characteristics are sensitive to the local chemical and electrical environment of the dye molecules. The excitation and emission peaks shift from 505nm and 728nm for the stock solution to 476nm and 585nm for the stained primary cells, and to 479nm and 570nm for the stained immortalized cells. The spectra of the stained hippocampal and hypothalamic neurons are different, and differ from those of cells or other bio-targets stained with di-4-ANEPPDHQ or the other ANEP dyes such as di-3-ANEPPDHQ. A specific spectral characterization is, therefore, necessary for any application of the dye to a new target. Interestingly, the difference between excitation spectra is smaller than between emission spectra.