- Email: [email protected]
Monitoring neuronal calcium signalling using a new method for ratiometric confocal calcium imaging
Cell Calcium 34 (2003) 295–303
Instruments and techniques
Monitoring neuronal calcium signalling using a new method for ratiometric confocal calcium imaging Christian Lohr Abteilung für Allgemeine Zoologie, Universität Kaiserslautern, Postfach 3049, Kaiserslautern 67653, Germany Received 4 February 2003; received in revised form 2 April 2003; accepted 14 April 2003
Abstract Ca2+ signalling influences many processes in the adult and developing nervous system like exocytosis, synaptic plasticity, and growth cone motility. Optical techniques in combination with fluorescent Ca2+ indicators are the most frequently used methods to measure Ca2+ signalling in cells. In the present study, a new method for ratiometric confocal Ca2+ imaging was developed, and the usefulness of the system was tested with two different neuronal preparations. Developing Manduca sexta antennal lobe neurons were loaded with the Ca2+ -sensitive dye Fura Red-AM, and the ratio of fluorescence excited at 457 and 488 nm was measured with a confocal laser scanning microscope. During pupal stages 4–12, the antennal lobe neuropil is restructured which includes the ingrowth of olfactory receptor axons, dendritic outgrowth of antennal lobe neurons, and synaptogenesis. In antennal lobe neurons, application of the AChR agonist carbachol induced Ca2+ oscillations the amplitude and frequency of which changed during stages 4–9, while at the end of synaptogenesis, at stages 11 and 12, only single Ca2+ transients were elicited. The Ca2+ oscillations were blocked by d-tubocurarine and Cd2+ , indicating that they were due to Ca2+ influx through voltage-gated Ca2+ channels, activated by nAChR-mediated membrane depolarization. To test whether single action potentials can induce Ca2+ transients detectable by Fura Red, individual leech Retzius neurons were injected iontophoretically with the Ca2+ indicator, and the membrane potential was recorded during Ca2+ imaging. Single action potentials induced transient increases in the Fura Red ratio measured in the axon, while trains of action potentials elicited Ca2+ transients that could also be recorded in the cell body and the nucleus. The results show that Fura Red can be used as a ratiometric Ca2+ indicator for confocal imaging. © 2003 Elsevier Science Ltd. All rights reserved. Keywords: Ca2+ signalling; Calcium imaging; Ratiometric indicator
1. Introduction Calcium serves as an important second messenger in all kinds of animal cells (for reviews see Refs. [1–3]). To study the dynamics of Ca2+ in the cytosol, optical techniques in combination with Ca2+ -sensitive fluorescent dyes have become the most frequent method [4]. A variety of fluorescent Ca2+ indicators are available, most of which simply increase their fluorescence intensity among binding of Ca2+ . A few Ca2+ -sensitive dyes allow ratiometric Ca2+ imaging, by responding to binding of Ca2+ either with a shift in the excitation peak, like Fura-2, or with a shift in the emission peak, like Indo-1. Ratiometric Ca2+ imaging permits the calibration of the imaging system and the calculation of absolute Ca2+ concentrations in the cell. During the past decade, an increasing number of investigators have used confocal laser scanning microscopy to study Ca2+
E-mail address: [email protected] (C. Lohr).
signalling at high optical resolution [2,4]. For ratiometric Ca2+ imaging in conjunction with confocal microscopy, two methods have been employed. First, Lipp and Niggli [5] and Schild et al. [6] independently developed a method using a combination of two Ca2+ -sensitive dyes, Fluo-3 and Fura Red. This method, however, has the drawback that reproducible measurements can only be performed if both dyes load the cells at the same rate to achieve a reproducible ratio of dye concentrations, and if both dyes are equally affected by photobleaching, diffusion and transportation out of the cell or into organelles to maintain the ratio of dye concentrations. In particular with iontophoretical dye injection or during longer experiments, these assumptions are difficult to accomplish since different dyes tend to behave differently in the cellular environment [7]. Second, UV lasers and dyes like Indo-1 or Fura-2 are used for ratiometric Ca2+ measurements [8–10], but the high energy of UV light can cause photodamage of the tissue under investigation, impairing the viability of the cells. In addition, few confocal microscopes are equipped with a UV laser because of their high cost.
0143-4160/$ – see front matter © 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0143-4160(03)00105-2
296
C. Lohr / Cell Calcium 34 (2003) 295–303
The aim of the present study was to establish a new method for ratiometric confocal Ca2+ imaging, using visible light and a single dye. We used the dual excitation ratiometric Ca2+ indicator Fura Red and confocal microscopy based on visible light delivered by an argon laser for Ca2+ measurements and calibration. To test the usefulness of this new method, we investigated Ca2+ signalling in two different preparations, neurons in the developing antennal lobe of the sphinx moth Manduca sexta, and neurons in the central nervous system of the leech Hirudo medicinalis. Neuronal Ca2+ signalling has been shown to fulfil many functions like triggering exocytosis, modulating gene expression or mediating synaptic plasticity (for reviews see Refs. [11,12]). In the developing nervous system, Ca2+ signalling is involved in cell migration, growth cone motility and dendritic growth [13–15]. The present study shows that Fura Red may be a useful tool for ratiometric confocal Ca2+ imaging. The results revealed a changing pattern of Ca2+ oscillations in developing antennal lobe neurons of Manduca induced by activation of nicotinic acetylcholine receptors during metamorphosis. In addition, the method can be employed to measure local Ca2+ signals such as fast Ca2+ transients in axonal subcompartments induced by single action potentials, or Ca2+ signalling in the nucleus.
2. Material and methods 2.1. Animals and preparation M. sexta: pupae of M. sexta (Lepidoptera: Sphingidae) were taken from the departmental insect-rearing facility where they were reared from eggs on an artificial diet under a long-day photoperiod (17:00 h light:07:00 h dark) at 26 ◦ C. Metamorphosis is considered to start with the molt from larva to pupa and to end with the emergence of the adult moth. Metamorphic development can be divided into 18 stages, recognized by changes in structures that are visible through the pupal cuticle [16]. Before dissection, pupae were anaesthetized on ice for 5 min. The brains then were dissected quickly from the head and placed in cold physiological saline. H. medicinalis: adult leeches were obtained from Zaug (Biebertal, Germany) and maintained in tap water at 16–18 ◦ C. Before an experiment, the animals were adapted to room temperature (20–23 ◦ C) for 2–7 days. The dissection and experimental procedures were performed as described before [17,18]. Individual ganglia were removed from the ventral nerve cord and pinned ventral side upwards into a Sylgard-lined experimental chamber containing physiological saline. The ventral ganglionic capsule was removed. The cell bodies of the ventral neurons, except for the Retzius neurons, were sucked off by a fire-polished micropipette to enable optical access to the neuropil.
2.2. Solutions M. sexta: insect physiological saline consisted of 150 mM NaCl, 4 mM KCl, 2 mM CaCl2 , 1 mM MgCl2 , 10 mM HEPES buffer; pH was adjusted to 7.0 with NaOH, the osmolarity was adjusted to approximately 390 mOsm with mannitol and glucose. Carbachol (CCH; Sigma, Germany) and cadmium chloride were dissolved in physiological saline from stock solutions (100 mM, stored at 4 ◦ C for up to 1 month). Cyclopiazonic acid (CPA; Sigma, Germany) was dissolved in dimethylsulfoxide (DMSO) as a stock solution of 100 mM and kept frozen. Stock solutions were added to physiological saline directly before an experiment to achieve the final concentration. d-Tubocurarine (Sigma, Germany) was added directly to physiological saline at the final concentration. All drugs were bath-applied with the perfusion system. H. medicinalis: during the experiment, leech ganglia were continuously superfused with physiological saline containing NaCl, 85 mM; KCl, 4 mM; CaCl2 , 2 mM; MgCl2 , 1 mM; HEPES, 10 mM; pH adjusted to 7.4 with NaOH. 2.3. Ratiometric confocal calcium imaging M. sexta antennal lobe neurons were loaded with Fura Red-AM (Molecular Probes), the membrane-permeant form of Fura Red, which was dissolved in DMSO at a concentration of 3 mM and was kept frozen in aliquots of 1 l for up to 6 months. For dye-loading, 0.5 l of Pluronic F-127 (Molecular Probes) was added to one aliquot of Fura Red-AM which then was dissolved in 1 ml of physiological saline. Whole brains were incubated in the dye solution for 30 min, rinsed several times with dye-free physiological saline, and were then pinned in a Sylgard-lined experimental chamber (volume: 200 l). The experimental chamber was placed on the stage of an upright confocal microscope (LSM 510, Zeiss, Oberkochen) equipped with a water immersion “dipping” lens (Achroplan 40×, NA 0.75; Zeiss Oberkochen) with 2 mm working distance, and the preparation was continuously superfused with physiological saline. Single leech Retzius neurons were filled iontophoretically through a microelectrode as described by Beck et al. [19]. Fura Red pentapotassium salt was dissolved in 200 mM potassium acetate at a concentration of 5 mM, and a microelectrode pulled from borosilicate glass capillaries (GC150-F15, Clark Elektromedical Instruments, UK) was backfilled with the dye solution. For electrophysiological recordings and dye injection, an Axoclamp 1B was used (Axon Instruments). A Retzius neuron was impaled, and the dye was injected by 500-ms current pulses of −2 nA at 1 Hz for 2 min for Ca2+ measurements in the cell body, and for 10 min for Ca2+ measurements in the axon. For ratiometric calcium imaging, Fura Red was alternately excited with the 457 and 488-nm lines of the argon laser, using the acousto-optical tunable filter (AOTF) to switch between the wavelengths (Fig. 1). The AOTF is controlled
C. Lohr / Cell Calcium 34 (2003) 295–303
Fig. 1. Experimental setup. The AOTF was used to switch between the 457 and 488-nm line for illumination of the Fura Red-filled cells. A beam splitter (BS 488) separated excitation light from emission light which was collected by a PMT through a 585 longpass filter (LP 585). A microelectrode was used for injection of Fura Red and recording of the membrane potential (Vm ).
297
by a digital signal processor which allows the change of the excitation wavelength within less than a millisecond. The output intensities of both wavelengths were measured at the nosepiece of the microscope with a photodiode and adjusted to equal values with the AOTF. This was achieved by setting the AOTF to 24% transmittance for the 457-nm line, and to 4% transmittance for the 488-nm line. The Fura Red fluorescence was measured by a photomultiplier tube (PMT) through a 585-nm longpass filter, resulting in a pair of images (Fig. 2). The same PMT was used to record both images of the 457/488 nm ratio image pair. Hence, changing the gain of the PMT affects both images equally and does not result in a shift of the image ratio. Therefore, adjusting the fluorescence values was only done by changing the PMT gain, while all other parameters, such as laser power and AOTF settings, were kept constant. The linewise multitrack mode was used to collect ratio image pairs. In that mode, the corresponding pixels of the two images of an image pair
Fig. 2. Cluster of approximately 15 cell bodies of Fura Red-filled Manduca antennal lobe neurons in situ in a low-calcium (A) and high-calcium (B) environment. The original images excited with 457 and 488 nm as well as the pseudo-coloured ratio images are shown. (C) Calibration of the Fura Red fluorescence ratio in antennal lobe neurons permeabilized by 4-bromo-A23187 and ionomycin. Note the decrease in fluorescence intensities of Fura Red at both wavelengths upon binding of Ca2+ , while the ratio of fluorescence F457 /F488 increases.
298
C. Lohr / Cell Calcium 34 (2003) 295–303
are temporally separated by a time range of less than 5 ms, resulting in a quasi-simultaneous recording of the image pair. For Ca2+ measurements in Manduca antennal lobe neurons, image pairs were recorded at a rate of 1 image pair/3 s. For recordings of activity-induced Ca2+ transients in leech Retzius neurons, the acquisition rate was set to 6–8 Hz.
stages 4 to 12, were investigated. The data from stages 6 and 7, and from stages 11 and 12 were pooled. Values are given as means±standard error of the mean (S.E.M.). Means were checked for significant differences using the Student’s t-test. Means were defined as statistically different when the null hypothesis was rejected at an error probability P < 0.05.
2.4. Calibration 3. Results The Fura Red fluorescence ratio was calibrated to permit calculation of the intracellular Ca2+ concentration ([Ca2+ ]i ) using the formula given by Grynkiewicz et al. [20]: [Ca
2+
Sf 2 R − Rmin ]i = Kd Sb2 Rmax − R
where R is the fluorescence ratio measured experimentally (see above), Rmin is the fluorescence ratio for Ca2+ -free and Rmax for Ca2+ -saturated conditions. Sf2 and Sb2 are the fluorescence values for Ca2+ -free and Ca2+ -bound dye excited at 488 nm. Rmin , Rmax , Sf2 , and Sb2 were determined experimentally using cells that were permeabilized by 20 M of the Ca2+ ionophore 4-bromo-A23187 and 50 M ionomycin (both Molecular Probes) to allow [Ca2+ ]i to equilibrate with [Ca2+ ]o (Fig. 2C). Rmin and Sf2 were determined in a Ca2+ -free solution containing 1 mM EGTA, and Rmax and Sb2 were determined in a solution containing 2 mM Ca2+ . The dissociation constant Kd of 140 nM was taken from the manufacturer [21]. 2.5. Data analysis Single cell bodies of antennal lobe neurons in situ could easily be identified (Fig. 2B). A region of interest (ROI), including either one cell body of a Manduca antennal lobe neuron or a subcellular compartment such as an axon segment or the nucleus of a leech Retzius neuron, was defined and the mean ratio value of the ROI was measured throughout the image sequence. From the ratio values, the Ca2+ concentration was calculated using the calibration data. Manduca neurons of different stages of metamorphic development, from
3.1. Calibration data The confocal calcium imaging system was calibrated using permeabilized neurons of Manduca and Hirudo. First, the cells were incubated for 10 min in a Ca2+ -free, EGTA-buffered saline to record Sf2 and the fluorescence ratio (Rmin ) in a virtually Ca2+ -free environment (Fig. 2C). Then, the saline was replaced by a saline containing 2 mM Ca2+ to achieve Sb2 and the ratio values (Rmax ) corresponding to the Ca2+ -saturated dye. In Manduca antennal lobe neurons, Rmin was 0.648, and Rmax was 1.497. Sf2 /Sb2 was determined as 5.46. In cell bodies of leech Retzius neurons, Rmin was 0.624, and Rmax was 1.656. Sf2 /Sb2 was calculated as 5.19 in this preparation. No attempt was made to calibrate the Fura Red fluorescence in axons of the Retzius neuron, because the axons are located deep in the tissue and are thus hardly accessible for the ionophore or Ca2+ . Hence, a sufficient permeabilization of the axonal membrane and a complete equilibration of the intraaxonal Ca2+ with the Ca2+ concentration in the bath can presumably not be achieved. 3.2. Effect of photobleaching and axial specimen shift The advantages of ratiometric Ca2+ imaging not only are the ability to calibrate the imaging system and hence the calculation of the Ca2+ concentration in cells, but also the reduction of certain artefacts. Photobleaching, for example, reduces the concentration of the Ca2+ indicator and causes a drop of the fluorescence during the experiment, as shown by
Fig. 3. Effect of photobleaching and shift of focal plane on the fluorescence ratio. (A) Steady illumination of a single Manduca neuron causes a decrease in fluorescence of both excitation wavelengths (upper traces), while the fluorescence ratio stays constant (lower trace). (B) Fura Red fluorescence and fluorescence ratio measured in a single Manduca neuron. A shift in the focal plane, induced by moving the specimen along the z-axis by 8 m, results in a decrease of the fluorescence intensity of both excitation wavelengths (upper traces) without changing the fluorescence ratio (lower trace).
C. Lohr / Cell Calcium 34 (2003) 295–303
the original fluorescence traces of a single Fura Red-filled antennal lobe neuron of M. sexta in Fig. 3A. In experiments using a single-wavelength Ca2+ indicator this leads to a continuous shift of the fluorescence baseline, which renders it difficult to quantify and compare Ca2+ transients of different parts of one experiment, or of different experiments. The fluorescence ratio, however, is not affected by photobleaching, because the fluorescence values of both excitation wavelengths are reduced at the same rate (Fig. 3A). A similar result is achieved when the focus plane is shifted along the z-axis by a few microns. Such a shift in the focal plane can occur during longer experiments for example by changes in the room temperature, which may affect the position of the microscope stage or the refraction properties of components in the optic path, or by movement of the specimen. The traces in Fig. 3B show the fluorescence values and the corresponding fluorescence ratio of a Manduca neuron during a recording in which the focal plane was artificially shifted by 8 m using the motorized stage. This shift led to a significant drop of the fluorescence of both the 457 and 488-nm excitation, while the fluorescence ratio stayed constant. 3.3. Ca2+ oscillations in Manduca antennal lobe neurons To test the usefulness of the Fura Red-based confocal calcium imaging system, the Ca2+ concentration was measured in Fura Red-filled antennal lobe neurons of M. sexta. Antennal lobe neurons are packed in cell groups in the periphery of the antennal lobe and are thus easily identifiable by their position. Only cells from the lateral cell group, containing both local interneurons and output neurons, were taken for Ca2+ measurements. Application of the acetylcholine receptor agonist carbachol (CCH, 100 M) for 5 min induced an increase in the Ca2+ concentration in more than 90% of the antennal lobe neurons in all developmental stages investigated (Fig. 4). Two principle response patterns were observed, single Ca2+ transients (Fig. 4A and F) or Ca2+ oscillations (Fig. 4B–E). Ca2+ oscillations were detected until stage 9, while single Ca2+ transients were still present at later stages. The shape of the Ca2+ oscillations, however, differed among the stages. The single Ca2+ transients of the Ca2+ oscillations became shorter, from plateau-like Ca2+ transients at stage 4 to spike-like Ca2+ transients at stage 9 (Fig. 4B–E). In addition, the regular pattern of Ca2+ oscillations at stage 4 changed to a rather irregular succession of Ca2+ transients at stage 9 when doublets of Ca2+ transients often occurred (Fig. 4E). Fig. 4G–I show the statistical summary of the Ca2+ responses induced by CCH. The fraction of neurons responding with Ca2+ oscillations to CCH increased from 23% at stage 4 to 51% at mid-stage 5, and 65% at stages 6 and 7 (Fig. 4G). At stage 9, 50% of the cells responded with Ca2+ oscillations, while at stages 11 and 12 Ca2+ oscillations could not be measured any more. The amplitude of the initial Ca2+ transient was evaluated, including the Ca2+ transients in those neurons in which
299
only a single Ca2+ transient could be measured (Fig. 4H). In stage 4 and mid-stage 5 neurons, the mean amplitudes of the Ca2+ transients were 370 ± 32 nM (n = 104) and 394 ± 14 nM (n = 69), respectively. There was a significant increase in the amplitude to 986 ± 67 nM (n = 184, P < 0.005) at stages 6 and 7. The mean amplitude significantly decreased to 688 ± 95 nM (n = 47, P < 0.005) at stage 9 although still larger than the amplitudes at stage 4 and mid-stage 5 (P < 0.01). At stages 11 and 12, the amplitude averaged 492 ± 71 nM (n = 40), which was not statistically different from stage 4, mid-stage 5 and stage 9, but was significantly smaller than at stages 6 and 7 (P < 0.005). The mean frequency of the Ca2+ oscillations significantly increased from 0.59 ± 0.04 events/min (n = 28) at stage 4 to 0.96 ± 0.05 events/min (n = 36) at mid-stage 5, and further to 1.31 ± 0.06 events/min (n = 141) at stages 6 and 7 and 1.67 ± 0.1 events/min (n = 26) at stage 9 (P < 0.005, Fig. 4I). Carbachol is an agonist of both nicotinic (nAChR) and muscarinic (mAChR) acetylcholine receptors. In order to identify the receptor type, the effect of the nAChR-specific blocker d-tubocurarine was investigated (Fig. 5A). 500 M d-tubocurarine completely blocked the CCH-induced Ca2+ oscillations (n = 29), indicating that they are mediated by nAChRs. The blocking effect of d-tubocurarine was irreversible. In M. sexta, nAChRs expressed in neurons of the abdominal ganglion have been shown to be Ca2+ permeable, and ACh-induced Ca2+ transients in these cells persisted in the presence of voltage-gated Ca2+ channel blockers like divalent cations [22]. In the present study of antennal lobe neurons, blocking voltage-gated Ca2+ channels with 500 M Cd2+ completely inhibited the CCH-induced Ca2+ oscillations (n = 15, Fig. 5B), indicating that the Ca2+ oscillations are exclusively mediated by voltage-gated Ca2+ channels, and hence that the nAChRs in the antennal lobe neurons are not Ca2+ permeable. In 33% of the cells the effect of Cd2+ was reversible (Fig. 5B). Preincubation of the preparation with CPA, which leads to the depletion of intracellular Ca2+ stores in Manduca motoneurons [23], had no effect on the Ca2+ oscillations (not shown), indicating that release of Ca2+ from intracellular stores does not contribute to the Ca2+ oscillations. 3.4. Ca2+ transients induced by action potentials in leech Retzius neurons Beck et al. [18] have shown that single action potentials can induce large Ca2+ transients in axons and dendrites of the leech Retzius neuron, using confocal microscopy and the non-ratiometric Ca2+ indicator Oregon Green BAPTA-1. In the present study, I employed this preparation to test whether Fura Red can be used to detect fast Ca2+ signals like those induced by a single action potential, and simultaneously perform electrophysiological recordings. A spontaneously generated action potential induced a clearly
300
C. Lohr / Cell Calcium 34 (2003) 295–303
Fig. 4. Typical recordings of Ca2+ transients elicited by 100 M carbachol (CCH, 5 min). Either single Ca2+ transients or Ca2+ oscillations were induced. Application bar and scale bars in (A) also apply for traces (B)–(F). (A) Single Ca2+ transient in a stage 4 antennal lobe neuron. (B) Ca2+ oscillation in a stage 4 neuron. (C) Ca2+ oscillation in a mid-stage 5 neuron. (D) Ca2+ oscillation in a stage 7 neuron. (E) Ca2+ oscillation in a stage 9 neuron. Note the appearance of doublets of Ca2+ transients (∗ ). (F) At stage 12, only single Ca2+ transients could be induced. (G) Fraction of cells responding to CCH with Ca2+ oscillations. The numbers on top of the bars reflect the number of responding cells vs. the number of all cells tested. (H) Amplitude of the initial CCH-induced Ca2+ transient including responses from cells with a single Ca2+ transient. The amplitude at stages 6 and 7 (pooled data) was significantly larger than at all other stages (∗∗∗ P < 0.005). The amplitude at stage 9 was significantly larger than at stage 4 and mid-stage 5 (∗∗ P < 0.01). (I) The frequency of the Ca2+ oscillations significantly increased between stages 4 and 9. Each stage was significantly different from all other stages (∗∗∗ P < 0.005). Abbreviations: st4, stage 4; m-st5, mid-stage 5; st6/7, pooled data of stages 6 and 7; st9, stage 9; and st11/12, pooled data of stages 11 and 12.
detectable increase in the fluorescence ratio measured in an axon segment with a signal-to-noise ratio of 5:1 (Fig. 6A and B), indicating that the response properties and the dynamic range of Fura Red are sufficient to resolve fast and relatively small Ca2+ signals. In the cell body of a Retzius neuron, Fura Red can be used to monitor Ca2+ changes in subcompartments such as the
nucleus (Fig. 6D). Like other Ca2+ indicators [17,18], the free anion of Fura Red accumulates in the nucleus, allowing to identify this organelle (Fig. 6D). During short trains of up to 10 action potentials at approximately 3 Hz induced by current injection into the cell body, Ca2+ increased in the cytoplasm but not in the nucleus. If the stimulus lasted for more than 3 s, the cytoplasmic Ca2+ signal was transmitted
C. Lohr / Cell Calcium 34 (2003) 295–303
301
Fig. 5. Pharmacological characterization of the Ca2+ oscillations. (A) d-Tubocurarine (500 M) irreversibly inhibited the CCH-induced Ca2+ oscillations. (B) Cd2+ (500 M) inhibited the CCH-induced Ca2+ oscillations. The effect of Cd2+ was reversible in 33% of the cells.
Fig. 6. Activity-induced Ca2+ transients in leech Retzius neurons. (A) Ca2+ transient induced by a single action potential (upper trace) measured in an axon segment as shown in (B). (C) Ca2+ transients induced by trains of action potentials during continuous current injection (1 nA, upper trace) measured in the cytoplasm (c) and the nucleus (N) as shown in (D).
into the nucleus, and a nuclear Ca2+ increase was detectable. The amplitude of the nuclear Ca2+ increase during a 15-s depolarization was 45.2 ± 8.7 nM (n = 7) compared to 68.1 ± 12.7 nM (n = 7) in the cytoplasm.
4. Discussion Calcium imaging has become a major technique to study cell physiology, and ratiometric calcium measurements are
used to quantify the Ca2+ concentration in cells. Only a few investigators, however, have applied ratiometric calcium imaging to confocal laser scanning microscopy, either using UV-excitable dyes like Indo-1 or Fura-2 [8–10], or using a combination of two dyes like Fluo-3 and Fura Red [5,6]. In the present study, a method for confocal Ca2+ imaging using visible light and a single dye was developed. In comparison with UV light, visible light causes less photodamage of the tissue, increasing the viability of the cells during the experiment. The use of a single dye avoids some assumptions
302
C. Lohr / Cell Calcium 34 (2003) 295–303
required when a combination of two dyes is used: both dyes must load the cells at the same rate to achieve a reproducible ratio of dye concentrations, and both dyes must be maintained equally inside the cell during the experiment. Thus, I believe that the method described here might be useful for ratiometric confocal Ca2+ imaging; it minimizes photodamage of the tissue, and only one dye is used to quantify Ca2+ measurements. The low molar extinction coefficient and quantum yield of Fura Red, however, indicate that higher laser power is required for excitation compared to other Ca2+ -sensitive probes used for confocal microscopy, such as Fluo-3, Fluo-4 or Calcium Green-1 [21]. In addition, the small dynamic range of the Fura Red ratio decreases the signal-to-noise ratio as shown by measurements of Ca2+ transients induced by single action potentials. Confocal imaging of Ca2+ -dependent changes of the Oregon Green BAPTA-1 fluorescence in axons of leech Retzius neurons yielded an increase in fluorescence of roughly 100% induced by a single action potential, at a noise level of less than 5% [18]. The resulting signal-to-noise ratio of 20 is much higher than the value of 5 measured with Fura Red here. However, at low noise levels and larger Ca2+ signals the dynamic range of the Fura Red ratio has been proven sufficient to perform Ca2+ imaging at high resolution by the present results. The new method was used in the current study to show a changing pattern of Ca2+ oscillations induced by application of CCH in antennal lobe neurons of M. sexta during metamorphic development. The stages chosen for this study cover the critical time period of neuropil development in the antennal lobe, including ingrowth of receptor axons, dendritic outgrowth of interneurons and projection neurons, and synaptogenesis [24] when Ca2+ might be expected to have a functional role. Ca2+ signalling has been shown to affect neuronal development in many systems. In the vertebrate retina, for example, calcium regulates dendritic development [15], and axonal outgrowth is dependent on Ca2+ transients in growth cones in frog and mouse [25,26]. In the present study, Ca2+ oscillations in M. sexta antennal lobe neurons were induced by the activation of nAChRs. Acetylcholine is the excitatory neurotransmitter released from the olfactory receptor axons in the antennal lobe [27–29]. Mature antennal lobe neurons respond to applied acetylcholine with changes in the membrane potential and the membrane conductance [29]. Spontaneous activity in receptor axons and synaptic transmission between the receptor axons and antennal lobe neurons, however, cannot be recorded before stages 7 and 8 [30,31]. Hence, activity-dependent acetylcholine release-mediating Ca2+ signalling in the antennal lobe neurons appears not to be involved in dendritic outgrowth which starts 2–3 days earlier (see Ref. [24]). It cannot be excluded, though, that the release of acetylcholine from ingrowing receptor axons occurs independent of action potentials. For example, the activation of cell adhesion molecules can induce Ca2+ influx [14] which may elicit Ca2+ -dependent acetylcholine release. Since M. sexta
antennal lobe neurons express functional nAChRs even at early developmental stages as shown by the Ca2+ measurements made here, this pathway to acetylcholine release from receptor neurons could induce Ca2+ signals important for dendritic growth and synaptogenesis. The Ca2+ oscillations induced by CCH could be blocked by d-tubocurarine, indicating that they are mediated by nicotinic AChRs. In M. sexta neurons of the abdominal ganglion, Ca2+ transients due to nAChR activation were not completely inhibited by blocking voltage-gated Ca2+ channels with 5 mM Ni2+ , indicating some Ca2+ permeability of the nAChR channel [22]. In contrast, the Ca2+ oscillations described in the present study were completely suppressed by Cd2+ , which is known to block voltage-gated Ca2+ channels in neurons and glial cells of Manduca [32–34]. Hence, in antennal lobe neurons the nAChRs appear not to be Ca2+ permeable, and the CCH-induced Ca2+ responses are presumably due to activation of voltage-gated Ca2+ channels induced by nAChR-mediated membrane depolarization. Acknowledgements I thank Drs. Joachim W. Deitmer and Lynne A. Oland for many fruitful discussions, and for their comments on a previous version of the manuscript. Jan E. Heil is gratefully acknowledged for managing the insect-rearing facility. This work was financially supported by the Deutsche Forschungsgemeinschaft (LO 779/2-1). References [1] J.W. Deitmer, A.J. Verkhratsky, C. Lohr, Calcium signalling in glial cells, Cell Calcium 24 (1999) 405–416. [2] M.J. Berridge, P. Lipp, M.D. Bootman, The versatility and universality of calcium signalling, Nat. Rev. Mol. Cell Biol. 1 (2000) 11– 21. [3] H. Tinel, E. Kinne-Saffran, R.K. Kinne, Calcium signalling during RVD of kidney cells, Cell Physiol. Biochem. 10 (2000) 297– 302. [4] A. Takahashi, P. Camacho, J.D. Lechleiter, B. Herman, Measurement of intracellular calcium, Physiol. Rev. 79 (1999) 1089–1125. [5] P. Lipp, E. Niggli, Ratiometric confocal Ca2+ -measurements with visible wavelength indicators in isolated cardiac myocytes, Cell Calcium 14 (1993) 359–372. [6] D. Schild, A. Jung, H.A. Schultens, Localization of calcium entry through calcium channels in olfactory receptor neurones using a laser scanning microscope and the calcium indicator dyes Fluo-3 and Fura-Red, Cell Calcium 15 (1994) 341–348. [7] D. Thomas, S.C. Tovey, T.J. Collins, M.D. Bootman, M.J. Berridge, P. Lipp, A comparison of fluorescent Ca2+ indicator properties and their use in measuring elementary and global Ca2+ signals, Cell Calcium 28 (2000) 213–223. [8] K. Kuba, S.Y. Hua, M. Nohmi, Spatial and dynamic changes in intracellular Ca2+ measured by confocal laser-scanning microscopy in bullfrog sympathetic ganglion cells, Neurosci. Res. 10 (1991) 245–259. [9] D.A. Przywara, S.V. Bhave, A. Bhave, T.D. Wakade, A.R. Wakade, Stimulated rise in neuronal calcium is faster and greater in the nucleus than in the cytosol, FASEB J. 5 (1991) 217–222.
C. Lohr / Cell Calcium 34 (2003) 295–303 [10] R. Nitschke, S. Wilhelm, R. Borlinghaus, J. Leipziger, R. Bindels, R. Greger, A modified confocal laser scanning microscope allows fast ultraviolet ratio imaging of intracellular Ca2+ activity using Fura-2, Pflugers Arch. 433 (1997) 653–663. [11] A.J. Verkhratsky, The endoplasmic reticulum and neuronal calcium signalling, Cell Calcium 32 (2002) 393–404. [12] M.J. Berridge, The endoplasmic reticulum: a multifunctional signaling organelle, Cell Calcium 32 (2002) 235–249. [13] H. Komuro, P. Rakic, Orchestration of neuronal migration by activity of ion channels, neurotransmitter receptors, and intracellular Ca2+ fluctuations, J. Neurobiol. 37 (1998) 110–130. [14] D.J. Dunican, P. Doherty, The generation of localized calcium rises mediated by cell adhesion molecules and their role in neuronal growth cone motility, Mol. Cell Biol. Res. Commun. 3 (2000) 255–263. [15] R.O. Wong, A. Ghosh, Activity-dependent regulation of dendritic growth and patterning, Nat. Rev. Neurosci. 3 (2002) 803–812. [16] L.A. Oland, L.P. Tolbert, Glial patterns during early development of antennal lobes of Manduca sexta: a comparison between normal lobes and lobes deprived of antennal axons, J. Comp. Neurol. 255 (1987) 196–207. [17] C. Lohr, J.W. Deitmer, Dendritic calcium transients in the leech giant glial cell in situ, Glia 26 (1999) 109–118. [18] A. Beck, C. Lohr, D.W. Deitmer, Calcium transients in subcompartments of the leech Retzius neuron as induced by single action potentials, J. Neurobiol. (2001) 1–18. [19] A. Beck, C. Lohr, H. Berthold, J.W. Deitmer, Calcium influx into dendrites of the leech Retzius neuron evoked by 5-hydroxytryptamine, Cell Calcium (2002) 137–149. [20] G. Grynkiewicz, M. Poenie, R.Y. Tsien, A new generation of Ca2+ indicators with greatly improved fluorescence properties, J. Biol. Chem. 260 (1985) 3440–3450. [21] R.P. Haugland, Handbook of Fluorescent Probes and Research Products, Molecular Probes, Eugene, 2002, p. 793. [22] R.M. Zayas, S. Qazi, D.B. Morton, B.A. Trimmer, Nicotinic-acetylcholine receptors are functionally coupled to the nitric oxide/cGMPpathway in insect neurons, J. Neurochem. 83 (2002) 421–431. [23] C. Duch, R.B. Levine, Changes in calcium signaling during postembryonic dendritic growth in Manduca sexta, J. Neurophysiol. 87 (2002) 1415–1425.
303
[24] L.A. Oland, L.P. Tolbert, Multiple factors shape development of olfactory glomeruli: insights from an insect model system, J. Neurobiol. 30 (1996) 92–109. [25] N.C. Spitzer, Activity-dependent neuronal differentiation prior to synapse formation: the functions of calcium transients, J. Physiol. Paris 96 (2002) 73–80. [26] F. Ciccolini, T.J. Collins, J. Sudhoelter, P. Lipp, M.J. Berridge, M.D. Bootman, Local and global spontaneous calcium events regulate neurite outgrowth and onset of GABAergic phenotype during neural precursor differentiation, J. Neurosci. 23 (2003) 103–111. [27] J.R. Sanes, J.G. Hildebrand, Acetylcholine and its metabolic enzymes in developing antennae of the moth, Manduca sexta, Dev. Biol. 52 (1976) 105–120. [28] J.R. Sanes, D.J. Prescott, J.G. Hildebrand, Cholinergic neurochemical development of normal and deafferented antennal lobes during metamorphosis of the moth, Manduca sexta, Brain Res. 119 (1977) 389–402. [29] B. Waldrop, J.G. Hildebrand, Physiology and pharmacology of acetylcholinergic responses of interneurons in the antennal lobes of the moth Manduca sexta, J. Comp. Physiol. A 164 (1989) 433– 441. [30] L.P. Tolbert, S.G. Matsumoto, J.G. Hildebrand, Development of synapses in the antennal lobes of the moth Manduca sexta during metamorphosis, J. Neurosci. 3 (1983) 1158–1175. [31] L.A. Oland, W.M. Pott, G. Bukhman, X.J. Sun, L.P. Tolbert, Activity blockadge does not prevent the construction of olfactory glomeruli in the moth Manduca sexta, Int. J. Dev. Neurosci. 14 (1996) 983– 996. [32] J.H. Hayashi, J.G. Hildebrand, Insect olfactory neurons in vitro: morphological and physiological characterization of cells from the developing antennal lobes of Manduca sexta, J. Neurosci. 10 (1990) 848–859. [33] C. Duch, R.B. Levine, Remodeling of membrane properties and dendritic architecture accompanies the postembryonic conversion of a slow into a fast motoneuron, J. Neurosci. 20 (2000) 6950–6961. [34] C. Lohr, E. Tucker, L.A. Oland, L.P. Tolbert, Development of depolarization-induced calcium transients in insect glial cells is dependent on the presence of afferent axons, J. Neurobiol. 52 (2002) 85–98.
Instruments and techniques
Monitoring neuronal calcium signalling using a new method for ratiometric confocal calcium imaging Christian Lohr Abteilung für Allgemeine Zoologie, Universität Kaiserslautern, Postfach 3049, Kaiserslautern 67653, Germany Received 4 February 2003; received in revised form 2 April 2003; accepted 14 April 2003
Abstract Ca2+ signalling influences many processes in the adult and developing nervous system like exocytosis, synaptic plasticity, and growth cone motility. Optical techniques in combination with fluorescent Ca2+ indicators are the most frequently used methods to measure Ca2+ signalling in cells. In the present study, a new method for ratiometric confocal Ca2+ imaging was developed, and the usefulness of the system was tested with two different neuronal preparations. Developing Manduca sexta antennal lobe neurons were loaded with the Ca2+ -sensitive dye Fura Red-AM, and the ratio of fluorescence excited at 457 and 488 nm was measured with a confocal laser scanning microscope. During pupal stages 4–12, the antennal lobe neuropil is restructured which includes the ingrowth of olfactory receptor axons, dendritic outgrowth of antennal lobe neurons, and synaptogenesis. In antennal lobe neurons, application of the AChR agonist carbachol induced Ca2+ oscillations the amplitude and frequency of which changed during stages 4–9, while at the end of synaptogenesis, at stages 11 and 12, only single Ca2+ transients were elicited. The Ca2+ oscillations were blocked by d-tubocurarine and Cd2+ , indicating that they were due to Ca2+ influx through voltage-gated Ca2+ channels, activated by nAChR-mediated membrane depolarization. To test whether single action potentials can induce Ca2+ transients detectable by Fura Red, individual leech Retzius neurons were injected iontophoretically with the Ca2+ indicator, and the membrane potential was recorded during Ca2+ imaging. Single action potentials induced transient increases in the Fura Red ratio measured in the axon, while trains of action potentials elicited Ca2+ transients that could also be recorded in the cell body and the nucleus. The results show that Fura Red can be used as a ratiometric Ca2+ indicator for confocal imaging. © 2003 Elsevier Science Ltd. All rights reserved. Keywords: Ca2+ signalling; Calcium imaging; Ratiometric indicator
1. Introduction Calcium serves as an important second messenger in all kinds of animal cells (for reviews see Refs. [1–3]). To study the dynamics of Ca2+ in the cytosol, optical techniques in combination with Ca2+ -sensitive fluorescent dyes have become the most frequent method [4]. A variety of fluorescent Ca2+ indicators are available, most of which simply increase their fluorescence intensity among binding of Ca2+ . A few Ca2+ -sensitive dyes allow ratiometric Ca2+ imaging, by responding to binding of Ca2+ either with a shift in the excitation peak, like Fura-2, or with a shift in the emission peak, like Indo-1. Ratiometric Ca2+ imaging permits the calibration of the imaging system and the calculation of absolute Ca2+ concentrations in the cell. During the past decade, an increasing number of investigators have used confocal laser scanning microscopy to study Ca2+
E-mail address: [email protected] (C. Lohr).
signalling at high optical resolution [2,4]. For ratiometric Ca2+ imaging in conjunction with confocal microscopy, two methods have been employed. First, Lipp and Niggli [5] and Schild et al. [6] independently developed a method using a combination of two Ca2+ -sensitive dyes, Fluo-3 and Fura Red. This method, however, has the drawback that reproducible measurements can only be performed if both dyes load the cells at the same rate to achieve a reproducible ratio of dye concentrations, and if both dyes are equally affected by photobleaching, diffusion and transportation out of the cell or into organelles to maintain the ratio of dye concentrations. In particular with iontophoretical dye injection or during longer experiments, these assumptions are difficult to accomplish since different dyes tend to behave differently in the cellular environment [7]. Second, UV lasers and dyes like Indo-1 or Fura-2 are used for ratiometric Ca2+ measurements [8–10], but the high energy of UV light can cause photodamage of the tissue under investigation, impairing the viability of the cells. In addition, few confocal microscopes are equipped with a UV laser because of their high cost.
0143-4160/$ – see front matter © 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0143-4160(03)00105-2
296
C. Lohr / Cell Calcium 34 (2003) 295–303
The aim of the present study was to establish a new method for ratiometric confocal Ca2+ imaging, using visible light and a single dye. We used the dual excitation ratiometric Ca2+ indicator Fura Red and confocal microscopy based on visible light delivered by an argon laser for Ca2+ measurements and calibration. To test the usefulness of this new method, we investigated Ca2+ signalling in two different preparations, neurons in the developing antennal lobe of the sphinx moth Manduca sexta, and neurons in the central nervous system of the leech Hirudo medicinalis. Neuronal Ca2+ signalling has been shown to fulfil many functions like triggering exocytosis, modulating gene expression or mediating synaptic plasticity (for reviews see Refs. [11,12]). In the developing nervous system, Ca2+ signalling is involved in cell migration, growth cone motility and dendritic growth [13–15]. The present study shows that Fura Red may be a useful tool for ratiometric confocal Ca2+ imaging. The results revealed a changing pattern of Ca2+ oscillations in developing antennal lobe neurons of Manduca induced by activation of nicotinic acetylcholine receptors during metamorphosis. In addition, the method can be employed to measure local Ca2+ signals such as fast Ca2+ transients in axonal subcompartments induced by single action potentials, or Ca2+ signalling in the nucleus.
2. Material and methods 2.1. Animals and preparation M. sexta: pupae of M. sexta (Lepidoptera: Sphingidae) were taken from the departmental insect-rearing facility where they were reared from eggs on an artificial diet under a long-day photoperiod (17:00 h light:07:00 h dark) at 26 ◦ C. Metamorphosis is considered to start with the molt from larva to pupa and to end with the emergence of the adult moth. Metamorphic development can be divided into 18 stages, recognized by changes in structures that are visible through the pupal cuticle [16]. Before dissection, pupae were anaesthetized on ice for 5 min. The brains then were dissected quickly from the head and placed in cold physiological saline. H. medicinalis: adult leeches were obtained from Zaug (Biebertal, Germany) and maintained in tap water at 16–18 ◦ C. Before an experiment, the animals were adapted to room temperature (20–23 ◦ C) for 2–7 days. The dissection and experimental procedures were performed as described before [17,18]. Individual ganglia were removed from the ventral nerve cord and pinned ventral side upwards into a Sylgard-lined experimental chamber containing physiological saline. The ventral ganglionic capsule was removed. The cell bodies of the ventral neurons, except for the Retzius neurons, were sucked off by a fire-polished micropipette to enable optical access to the neuropil.
2.2. Solutions M. sexta: insect physiological saline consisted of 150 mM NaCl, 4 mM KCl, 2 mM CaCl2 , 1 mM MgCl2 , 10 mM HEPES buffer; pH was adjusted to 7.0 with NaOH, the osmolarity was adjusted to approximately 390 mOsm with mannitol and glucose. Carbachol (CCH; Sigma, Germany) and cadmium chloride were dissolved in physiological saline from stock solutions (100 mM, stored at 4 ◦ C for up to 1 month). Cyclopiazonic acid (CPA; Sigma, Germany) was dissolved in dimethylsulfoxide (DMSO) as a stock solution of 100 mM and kept frozen. Stock solutions were added to physiological saline directly before an experiment to achieve the final concentration. d-Tubocurarine (Sigma, Germany) was added directly to physiological saline at the final concentration. All drugs were bath-applied with the perfusion system. H. medicinalis: during the experiment, leech ganglia were continuously superfused with physiological saline containing NaCl, 85 mM; KCl, 4 mM; CaCl2 , 2 mM; MgCl2 , 1 mM; HEPES, 10 mM; pH adjusted to 7.4 with NaOH. 2.3. Ratiometric confocal calcium imaging M. sexta antennal lobe neurons were loaded with Fura Red-AM (Molecular Probes), the membrane-permeant form of Fura Red, which was dissolved in DMSO at a concentration of 3 mM and was kept frozen in aliquots of 1 l for up to 6 months. For dye-loading, 0.5 l of Pluronic F-127 (Molecular Probes) was added to one aliquot of Fura Red-AM which then was dissolved in 1 ml of physiological saline. Whole brains were incubated in the dye solution for 30 min, rinsed several times with dye-free physiological saline, and were then pinned in a Sylgard-lined experimental chamber (volume: 200 l). The experimental chamber was placed on the stage of an upright confocal microscope (LSM 510, Zeiss, Oberkochen) equipped with a water immersion “dipping” lens (Achroplan 40×, NA 0.75; Zeiss Oberkochen) with 2 mm working distance, and the preparation was continuously superfused with physiological saline. Single leech Retzius neurons were filled iontophoretically through a microelectrode as described by Beck et al. [19]. Fura Red pentapotassium salt was dissolved in 200 mM potassium acetate at a concentration of 5 mM, and a microelectrode pulled from borosilicate glass capillaries (GC150-F15, Clark Elektromedical Instruments, UK) was backfilled with the dye solution. For electrophysiological recordings and dye injection, an Axoclamp 1B was used (Axon Instruments). A Retzius neuron was impaled, and the dye was injected by 500-ms current pulses of −2 nA at 1 Hz for 2 min for Ca2+ measurements in the cell body, and for 10 min for Ca2+ measurements in the axon. For ratiometric calcium imaging, Fura Red was alternately excited with the 457 and 488-nm lines of the argon laser, using the acousto-optical tunable filter (AOTF) to switch between the wavelengths (Fig. 1). The AOTF is controlled
C. Lohr / Cell Calcium 34 (2003) 295–303
Fig. 1. Experimental setup. The AOTF was used to switch between the 457 and 488-nm line for illumination of the Fura Red-filled cells. A beam splitter (BS 488) separated excitation light from emission light which was collected by a PMT through a 585 longpass filter (LP 585). A microelectrode was used for injection of Fura Red and recording of the membrane potential (Vm ).
297
by a digital signal processor which allows the change of the excitation wavelength within less than a millisecond. The output intensities of both wavelengths were measured at the nosepiece of the microscope with a photodiode and adjusted to equal values with the AOTF. This was achieved by setting the AOTF to 24% transmittance for the 457-nm line, and to 4% transmittance for the 488-nm line. The Fura Red fluorescence was measured by a photomultiplier tube (PMT) through a 585-nm longpass filter, resulting in a pair of images (Fig. 2). The same PMT was used to record both images of the 457/488 nm ratio image pair. Hence, changing the gain of the PMT affects both images equally and does not result in a shift of the image ratio. Therefore, adjusting the fluorescence values was only done by changing the PMT gain, while all other parameters, such as laser power and AOTF settings, were kept constant. The linewise multitrack mode was used to collect ratio image pairs. In that mode, the corresponding pixels of the two images of an image pair
Fig. 2. Cluster of approximately 15 cell bodies of Fura Red-filled Manduca antennal lobe neurons in situ in a low-calcium (A) and high-calcium (B) environment. The original images excited with 457 and 488 nm as well as the pseudo-coloured ratio images are shown. (C) Calibration of the Fura Red fluorescence ratio in antennal lobe neurons permeabilized by 4-bromo-A23187 and ionomycin. Note the decrease in fluorescence intensities of Fura Red at both wavelengths upon binding of Ca2+ , while the ratio of fluorescence F457 /F488 increases.
298
C. Lohr / Cell Calcium 34 (2003) 295–303
are temporally separated by a time range of less than 5 ms, resulting in a quasi-simultaneous recording of the image pair. For Ca2+ measurements in Manduca antennal lobe neurons, image pairs were recorded at a rate of 1 image pair/3 s. For recordings of activity-induced Ca2+ transients in leech Retzius neurons, the acquisition rate was set to 6–8 Hz.
stages 4 to 12, were investigated. The data from stages 6 and 7, and from stages 11 and 12 were pooled. Values are given as means±standard error of the mean (S.E.M.). Means were checked for significant differences using the Student’s t-test. Means were defined as statistically different when the null hypothesis was rejected at an error probability P < 0.05.
2.4. Calibration 3. Results The Fura Red fluorescence ratio was calibrated to permit calculation of the intracellular Ca2+ concentration ([Ca2+ ]i ) using the formula given by Grynkiewicz et al. [20]: [Ca
2+
Sf 2 R − Rmin ]i = Kd Sb2 Rmax − R
where R is the fluorescence ratio measured experimentally (see above), Rmin is the fluorescence ratio for Ca2+ -free and Rmax for Ca2+ -saturated conditions. Sf2 and Sb2 are the fluorescence values for Ca2+ -free and Ca2+ -bound dye excited at 488 nm. Rmin , Rmax , Sf2 , and Sb2 were determined experimentally using cells that were permeabilized by 20 M of the Ca2+ ionophore 4-bromo-A23187 and 50 M ionomycin (both Molecular Probes) to allow [Ca2+ ]i to equilibrate with [Ca2+ ]o (Fig. 2C). Rmin and Sf2 were determined in a Ca2+ -free solution containing 1 mM EGTA, and Rmax and Sb2 were determined in a solution containing 2 mM Ca2+ . The dissociation constant Kd of 140 nM was taken from the manufacturer [21]. 2.5. Data analysis Single cell bodies of antennal lobe neurons in situ could easily be identified (Fig. 2B). A region of interest (ROI), including either one cell body of a Manduca antennal lobe neuron or a subcellular compartment such as an axon segment or the nucleus of a leech Retzius neuron, was defined and the mean ratio value of the ROI was measured throughout the image sequence. From the ratio values, the Ca2+ concentration was calculated using the calibration data. Manduca neurons of different stages of metamorphic development, from
3.1. Calibration data The confocal calcium imaging system was calibrated using permeabilized neurons of Manduca and Hirudo. First, the cells were incubated for 10 min in a Ca2+ -free, EGTA-buffered saline to record Sf2 and the fluorescence ratio (Rmin ) in a virtually Ca2+ -free environment (Fig. 2C). Then, the saline was replaced by a saline containing 2 mM Ca2+ to achieve Sb2 and the ratio values (Rmax ) corresponding to the Ca2+ -saturated dye. In Manduca antennal lobe neurons, Rmin was 0.648, and Rmax was 1.497. Sf2 /Sb2 was determined as 5.46. In cell bodies of leech Retzius neurons, Rmin was 0.624, and Rmax was 1.656. Sf2 /Sb2 was calculated as 5.19 in this preparation. No attempt was made to calibrate the Fura Red fluorescence in axons of the Retzius neuron, because the axons are located deep in the tissue and are thus hardly accessible for the ionophore or Ca2+ . Hence, a sufficient permeabilization of the axonal membrane and a complete equilibration of the intraaxonal Ca2+ with the Ca2+ concentration in the bath can presumably not be achieved. 3.2. Effect of photobleaching and axial specimen shift The advantages of ratiometric Ca2+ imaging not only are the ability to calibrate the imaging system and hence the calculation of the Ca2+ concentration in cells, but also the reduction of certain artefacts. Photobleaching, for example, reduces the concentration of the Ca2+ indicator and causes a drop of the fluorescence during the experiment, as shown by
Fig. 3. Effect of photobleaching and shift of focal plane on the fluorescence ratio. (A) Steady illumination of a single Manduca neuron causes a decrease in fluorescence of both excitation wavelengths (upper traces), while the fluorescence ratio stays constant (lower trace). (B) Fura Red fluorescence and fluorescence ratio measured in a single Manduca neuron. A shift in the focal plane, induced by moving the specimen along the z-axis by 8 m, results in a decrease of the fluorescence intensity of both excitation wavelengths (upper traces) without changing the fluorescence ratio (lower trace).
C. Lohr / Cell Calcium 34 (2003) 295–303
the original fluorescence traces of a single Fura Red-filled antennal lobe neuron of M. sexta in Fig. 3A. In experiments using a single-wavelength Ca2+ indicator this leads to a continuous shift of the fluorescence baseline, which renders it difficult to quantify and compare Ca2+ transients of different parts of one experiment, or of different experiments. The fluorescence ratio, however, is not affected by photobleaching, because the fluorescence values of both excitation wavelengths are reduced at the same rate (Fig. 3A). A similar result is achieved when the focus plane is shifted along the z-axis by a few microns. Such a shift in the focal plane can occur during longer experiments for example by changes in the room temperature, which may affect the position of the microscope stage or the refraction properties of components in the optic path, or by movement of the specimen. The traces in Fig. 3B show the fluorescence values and the corresponding fluorescence ratio of a Manduca neuron during a recording in which the focal plane was artificially shifted by 8 m using the motorized stage. This shift led to a significant drop of the fluorescence of both the 457 and 488-nm excitation, while the fluorescence ratio stayed constant. 3.3. Ca2+ oscillations in Manduca antennal lobe neurons To test the usefulness of the Fura Red-based confocal calcium imaging system, the Ca2+ concentration was measured in Fura Red-filled antennal lobe neurons of M. sexta. Antennal lobe neurons are packed in cell groups in the periphery of the antennal lobe and are thus easily identifiable by their position. Only cells from the lateral cell group, containing both local interneurons and output neurons, were taken for Ca2+ measurements. Application of the acetylcholine receptor agonist carbachol (CCH, 100 M) for 5 min induced an increase in the Ca2+ concentration in more than 90% of the antennal lobe neurons in all developmental stages investigated (Fig. 4). Two principle response patterns were observed, single Ca2+ transients (Fig. 4A and F) or Ca2+ oscillations (Fig. 4B–E). Ca2+ oscillations were detected until stage 9, while single Ca2+ transients were still present at later stages. The shape of the Ca2+ oscillations, however, differed among the stages. The single Ca2+ transients of the Ca2+ oscillations became shorter, from plateau-like Ca2+ transients at stage 4 to spike-like Ca2+ transients at stage 9 (Fig. 4B–E). In addition, the regular pattern of Ca2+ oscillations at stage 4 changed to a rather irregular succession of Ca2+ transients at stage 9 when doublets of Ca2+ transients often occurred (Fig. 4E). Fig. 4G–I show the statistical summary of the Ca2+ responses induced by CCH. The fraction of neurons responding with Ca2+ oscillations to CCH increased from 23% at stage 4 to 51% at mid-stage 5, and 65% at stages 6 and 7 (Fig. 4G). At stage 9, 50% of the cells responded with Ca2+ oscillations, while at stages 11 and 12 Ca2+ oscillations could not be measured any more. The amplitude of the initial Ca2+ transient was evaluated, including the Ca2+ transients in those neurons in which
299
only a single Ca2+ transient could be measured (Fig. 4H). In stage 4 and mid-stage 5 neurons, the mean amplitudes of the Ca2+ transients were 370 ± 32 nM (n = 104) and 394 ± 14 nM (n = 69), respectively. There was a significant increase in the amplitude to 986 ± 67 nM (n = 184, P < 0.005) at stages 6 and 7. The mean amplitude significantly decreased to 688 ± 95 nM (n = 47, P < 0.005) at stage 9 although still larger than the amplitudes at stage 4 and mid-stage 5 (P < 0.01). At stages 11 and 12, the amplitude averaged 492 ± 71 nM (n = 40), which was not statistically different from stage 4, mid-stage 5 and stage 9, but was significantly smaller than at stages 6 and 7 (P < 0.005). The mean frequency of the Ca2+ oscillations significantly increased from 0.59 ± 0.04 events/min (n = 28) at stage 4 to 0.96 ± 0.05 events/min (n = 36) at mid-stage 5, and further to 1.31 ± 0.06 events/min (n = 141) at stages 6 and 7 and 1.67 ± 0.1 events/min (n = 26) at stage 9 (P < 0.005, Fig. 4I). Carbachol is an agonist of both nicotinic (nAChR) and muscarinic (mAChR) acetylcholine receptors. In order to identify the receptor type, the effect of the nAChR-specific blocker d-tubocurarine was investigated (Fig. 5A). 500 M d-tubocurarine completely blocked the CCH-induced Ca2+ oscillations (n = 29), indicating that they are mediated by nAChRs. The blocking effect of d-tubocurarine was irreversible. In M. sexta, nAChRs expressed in neurons of the abdominal ganglion have been shown to be Ca2+ permeable, and ACh-induced Ca2+ transients in these cells persisted in the presence of voltage-gated Ca2+ channel blockers like divalent cations [22]. In the present study of antennal lobe neurons, blocking voltage-gated Ca2+ channels with 500 M Cd2+ completely inhibited the CCH-induced Ca2+ oscillations (n = 15, Fig. 5B), indicating that the Ca2+ oscillations are exclusively mediated by voltage-gated Ca2+ channels, and hence that the nAChRs in the antennal lobe neurons are not Ca2+ permeable. In 33% of the cells the effect of Cd2+ was reversible (Fig. 5B). Preincubation of the preparation with CPA, which leads to the depletion of intracellular Ca2+ stores in Manduca motoneurons [23], had no effect on the Ca2+ oscillations (not shown), indicating that release of Ca2+ from intracellular stores does not contribute to the Ca2+ oscillations. 3.4. Ca2+ transients induced by action potentials in leech Retzius neurons Beck et al. [18] have shown that single action potentials can induce large Ca2+ transients in axons and dendrites of the leech Retzius neuron, using confocal microscopy and the non-ratiometric Ca2+ indicator Oregon Green BAPTA-1. In the present study, I employed this preparation to test whether Fura Red can be used to detect fast Ca2+ signals like those induced by a single action potential, and simultaneously perform electrophysiological recordings. A spontaneously generated action potential induced a clearly
300
C. Lohr / Cell Calcium 34 (2003) 295–303
Fig. 4. Typical recordings of Ca2+ transients elicited by 100 M carbachol (CCH, 5 min). Either single Ca2+ transients or Ca2+ oscillations were induced. Application bar and scale bars in (A) also apply for traces (B)–(F). (A) Single Ca2+ transient in a stage 4 antennal lobe neuron. (B) Ca2+ oscillation in a stage 4 neuron. (C) Ca2+ oscillation in a mid-stage 5 neuron. (D) Ca2+ oscillation in a stage 7 neuron. (E) Ca2+ oscillation in a stage 9 neuron. Note the appearance of doublets of Ca2+ transients (∗ ). (F) At stage 12, only single Ca2+ transients could be induced. (G) Fraction of cells responding to CCH with Ca2+ oscillations. The numbers on top of the bars reflect the number of responding cells vs. the number of all cells tested. (H) Amplitude of the initial CCH-induced Ca2+ transient including responses from cells with a single Ca2+ transient. The amplitude at stages 6 and 7 (pooled data) was significantly larger than at all other stages (∗∗∗ P < 0.005). The amplitude at stage 9 was significantly larger than at stage 4 and mid-stage 5 (∗∗ P < 0.01). (I) The frequency of the Ca2+ oscillations significantly increased between stages 4 and 9. Each stage was significantly different from all other stages (∗∗∗ P < 0.005). Abbreviations: st4, stage 4; m-st5, mid-stage 5; st6/7, pooled data of stages 6 and 7; st9, stage 9; and st11/12, pooled data of stages 11 and 12.
detectable increase in the fluorescence ratio measured in an axon segment with a signal-to-noise ratio of 5:1 (Fig. 6A and B), indicating that the response properties and the dynamic range of Fura Red are sufficient to resolve fast and relatively small Ca2+ signals. In the cell body of a Retzius neuron, Fura Red can be used to monitor Ca2+ changes in subcompartments such as the
nucleus (Fig. 6D). Like other Ca2+ indicators [17,18], the free anion of Fura Red accumulates in the nucleus, allowing to identify this organelle (Fig. 6D). During short trains of up to 10 action potentials at approximately 3 Hz induced by current injection into the cell body, Ca2+ increased in the cytoplasm but not in the nucleus. If the stimulus lasted for more than 3 s, the cytoplasmic Ca2+ signal was transmitted
C. Lohr / Cell Calcium 34 (2003) 295–303
301
Fig. 5. Pharmacological characterization of the Ca2+ oscillations. (A) d-Tubocurarine (500 M) irreversibly inhibited the CCH-induced Ca2+ oscillations. (B) Cd2+ (500 M) inhibited the CCH-induced Ca2+ oscillations. The effect of Cd2+ was reversible in 33% of the cells.
Fig. 6. Activity-induced Ca2+ transients in leech Retzius neurons. (A) Ca2+ transient induced by a single action potential (upper trace) measured in an axon segment as shown in (B). (C) Ca2+ transients induced by trains of action potentials during continuous current injection (1 nA, upper trace) measured in the cytoplasm (c) and the nucleus (N) as shown in (D).
into the nucleus, and a nuclear Ca2+ increase was detectable. The amplitude of the nuclear Ca2+ increase during a 15-s depolarization was 45.2 ± 8.7 nM (n = 7) compared to 68.1 ± 12.7 nM (n = 7) in the cytoplasm.
4. Discussion Calcium imaging has become a major technique to study cell physiology, and ratiometric calcium measurements are
used to quantify the Ca2+ concentration in cells. Only a few investigators, however, have applied ratiometric calcium imaging to confocal laser scanning microscopy, either using UV-excitable dyes like Indo-1 or Fura-2 [8–10], or using a combination of two dyes like Fluo-3 and Fura Red [5,6]. In the present study, a method for confocal Ca2+ imaging using visible light and a single dye was developed. In comparison with UV light, visible light causes less photodamage of the tissue, increasing the viability of the cells during the experiment. The use of a single dye avoids some assumptions
302
C. Lohr / Cell Calcium 34 (2003) 295–303
required when a combination of two dyes is used: both dyes must load the cells at the same rate to achieve a reproducible ratio of dye concentrations, and both dyes must be maintained equally inside the cell during the experiment. Thus, I believe that the method described here might be useful for ratiometric confocal Ca2+ imaging; it minimizes photodamage of the tissue, and only one dye is used to quantify Ca2+ measurements. The low molar extinction coefficient and quantum yield of Fura Red, however, indicate that higher laser power is required for excitation compared to other Ca2+ -sensitive probes used for confocal microscopy, such as Fluo-3, Fluo-4 or Calcium Green-1 [21]. In addition, the small dynamic range of the Fura Red ratio decreases the signal-to-noise ratio as shown by measurements of Ca2+ transients induced by single action potentials. Confocal imaging of Ca2+ -dependent changes of the Oregon Green BAPTA-1 fluorescence in axons of leech Retzius neurons yielded an increase in fluorescence of roughly 100% induced by a single action potential, at a noise level of less than 5% [18]. The resulting signal-to-noise ratio of 20 is much higher than the value of 5 measured with Fura Red here. However, at low noise levels and larger Ca2+ signals the dynamic range of the Fura Red ratio has been proven sufficient to perform Ca2+ imaging at high resolution by the present results. The new method was used in the current study to show a changing pattern of Ca2+ oscillations induced by application of CCH in antennal lobe neurons of M. sexta during metamorphic development. The stages chosen for this study cover the critical time period of neuropil development in the antennal lobe, including ingrowth of receptor axons, dendritic outgrowth of interneurons and projection neurons, and synaptogenesis [24] when Ca2+ might be expected to have a functional role. Ca2+ signalling has been shown to affect neuronal development in many systems. In the vertebrate retina, for example, calcium regulates dendritic development [15], and axonal outgrowth is dependent on Ca2+ transients in growth cones in frog and mouse [25,26]. In the present study, Ca2+ oscillations in M. sexta antennal lobe neurons were induced by the activation of nAChRs. Acetylcholine is the excitatory neurotransmitter released from the olfactory receptor axons in the antennal lobe [27–29]. Mature antennal lobe neurons respond to applied acetylcholine with changes in the membrane potential and the membrane conductance [29]. Spontaneous activity in receptor axons and synaptic transmission between the receptor axons and antennal lobe neurons, however, cannot be recorded before stages 7 and 8 [30,31]. Hence, activity-dependent acetylcholine release-mediating Ca2+ signalling in the antennal lobe neurons appears not to be involved in dendritic outgrowth which starts 2–3 days earlier (see Ref. [24]). It cannot be excluded, though, that the release of acetylcholine from ingrowing receptor axons occurs independent of action potentials. For example, the activation of cell adhesion molecules can induce Ca2+ influx [14] which may elicit Ca2+ -dependent acetylcholine release. Since M. sexta
antennal lobe neurons express functional nAChRs even at early developmental stages as shown by the Ca2+ measurements made here, this pathway to acetylcholine release from receptor neurons could induce Ca2+ signals important for dendritic growth and synaptogenesis. The Ca2+ oscillations induced by CCH could be blocked by d-tubocurarine, indicating that they are mediated by nicotinic AChRs. In M. sexta neurons of the abdominal ganglion, Ca2+ transients due to nAChR activation were not completely inhibited by blocking voltage-gated Ca2+ channels with 5 mM Ni2+ , indicating some Ca2+ permeability of the nAChR channel [22]. In contrast, the Ca2+ oscillations described in the present study were completely suppressed by Cd2+ , which is known to block voltage-gated Ca2+ channels in neurons and glial cells of Manduca [32–34]. Hence, in antennal lobe neurons the nAChRs appear not to be Ca2+ permeable, and the CCH-induced Ca2+ responses are presumably due to activation of voltage-gated Ca2+ channels induced by nAChR-mediated membrane depolarization. Acknowledgements I thank Drs. Joachim W. Deitmer and Lynne A. Oland for many fruitful discussions, and for their comments on a previous version of the manuscript. Jan E. Heil is gratefully acknowledged for managing the insect-rearing facility. This work was financially supported by the Deutsche Forschungsgemeinschaft (LO 779/2-1). References [1] J.W. Deitmer, A.J. Verkhratsky, C. Lohr, Calcium signalling in glial cells, Cell Calcium 24 (1999) 405–416. [2] M.J. Berridge, P. Lipp, M.D. Bootman, The versatility and universality of calcium signalling, Nat. Rev. Mol. Cell Biol. 1 (2000) 11– 21. [3] H. Tinel, E. Kinne-Saffran, R.K. Kinne, Calcium signalling during RVD of kidney cells, Cell Physiol. Biochem. 10 (2000) 297– 302. [4] A. Takahashi, P. Camacho, J.D. Lechleiter, B. Herman, Measurement of intracellular calcium, Physiol. Rev. 79 (1999) 1089–1125. [5] P. Lipp, E. Niggli, Ratiometric confocal Ca2+ -measurements with visible wavelength indicators in isolated cardiac myocytes, Cell Calcium 14 (1993) 359–372. [6] D. Schild, A. Jung, H.A. Schultens, Localization of calcium entry through calcium channels in olfactory receptor neurones using a laser scanning microscope and the calcium indicator dyes Fluo-3 and Fura-Red, Cell Calcium 15 (1994) 341–348. [7] D. Thomas, S.C. Tovey, T.J. Collins, M.D. Bootman, M.J. Berridge, P. Lipp, A comparison of fluorescent Ca2+ indicator properties and their use in measuring elementary and global Ca2+ signals, Cell Calcium 28 (2000) 213–223. [8] K. Kuba, S.Y. Hua, M. Nohmi, Spatial and dynamic changes in intracellular Ca2+ measured by confocal laser-scanning microscopy in bullfrog sympathetic ganglion cells, Neurosci. Res. 10 (1991) 245–259. [9] D.A. Przywara, S.V. Bhave, A. Bhave, T.D. Wakade, A.R. Wakade, Stimulated rise in neuronal calcium is faster and greater in the nucleus than in the cytosol, FASEB J. 5 (1991) 217–222.
C. Lohr / Cell Calcium 34 (2003) 295–303 [10] R. Nitschke, S. Wilhelm, R. Borlinghaus, J. Leipziger, R. Bindels, R. Greger, A modified confocal laser scanning microscope allows fast ultraviolet ratio imaging of intracellular Ca2+ activity using Fura-2, Pflugers Arch. 433 (1997) 653–663. [11] A.J. Verkhratsky, The endoplasmic reticulum and neuronal calcium signalling, Cell Calcium 32 (2002) 393–404. [12] M.J. Berridge, The endoplasmic reticulum: a multifunctional signaling organelle, Cell Calcium 32 (2002) 235–249. [13] H. Komuro, P. Rakic, Orchestration of neuronal migration by activity of ion channels, neurotransmitter receptors, and intracellular Ca2+ fluctuations, J. Neurobiol. 37 (1998) 110–130. [14] D.J. Dunican, P. Doherty, The generation of localized calcium rises mediated by cell adhesion molecules and their role in neuronal growth cone motility, Mol. Cell Biol. Res. Commun. 3 (2000) 255–263. [15] R.O. Wong, A. Ghosh, Activity-dependent regulation of dendritic growth and patterning, Nat. Rev. Neurosci. 3 (2002) 803–812. [16] L.A. Oland, L.P. Tolbert, Glial patterns during early development of antennal lobes of Manduca sexta: a comparison between normal lobes and lobes deprived of antennal axons, J. Comp. Neurol. 255 (1987) 196–207. [17] C. Lohr, J.W. Deitmer, Dendritic calcium transients in the leech giant glial cell in situ, Glia 26 (1999) 109–118. [18] A. Beck, C. Lohr, D.W. Deitmer, Calcium transients in subcompartments of the leech Retzius neuron as induced by single action potentials, J. Neurobiol. (2001) 1–18. [19] A. Beck, C. Lohr, H. Berthold, J.W. Deitmer, Calcium influx into dendrites of the leech Retzius neuron evoked by 5-hydroxytryptamine, Cell Calcium (2002) 137–149. [20] G. Grynkiewicz, M. Poenie, R.Y. Tsien, A new generation of Ca2+ indicators with greatly improved fluorescence properties, J. Biol. Chem. 260 (1985) 3440–3450. [21] R.P. Haugland, Handbook of Fluorescent Probes and Research Products, Molecular Probes, Eugene, 2002, p. 793. [22] R.M. Zayas, S. Qazi, D.B. Morton, B.A. Trimmer, Nicotinic-acetylcholine receptors are functionally coupled to the nitric oxide/cGMPpathway in insect neurons, J. Neurochem. 83 (2002) 421–431. [23] C. Duch, R.B. Levine, Changes in calcium signaling during postembryonic dendritic growth in Manduca sexta, J. Neurophysiol. 87 (2002) 1415–1425.
303
[24] L.A. Oland, L.P. Tolbert, Multiple factors shape development of olfactory glomeruli: insights from an insect model system, J. Neurobiol. 30 (1996) 92–109. [25] N.C. Spitzer, Activity-dependent neuronal differentiation prior to synapse formation: the functions of calcium transients, J. Physiol. Paris 96 (2002) 73–80. [26] F. Ciccolini, T.J. Collins, J. Sudhoelter, P. Lipp, M.J. Berridge, M.D. Bootman, Local and global spontaneous calcium events regulate neurite outgrowth and onset of GABAergic phenotype during neural precursor differentiation, J. Neurosci. 23 (2003) 103–111. [27] J.R. Sanes, J.G. Hildebrand, Acetylcholine and its metabolic enzymes in developing antennae of the moth, Manduca sexta, Dev. Biol. 52 (1976) 105–120. [28] J.R. Sanes, D.J. Prescott, J.G. Hildebrand, Cholinergic neurochemical development of normal and deafferented antennal lobes during metamorphosis of the moth, Manduca sexta, Brain Res. 119 (1977) 389–402. [29] B. Waldrop, J.G. Hildebrand, Physiology and pharmacology of acetylcholinergic responses of interneurons in the antennal lobes of the moth Manduca sexta, J. Comp. Physiol. A 164 (1989) 433– 441. [30] L.P. Tolbert, S.G. Matsumoto, J.G. Hildebrand, Development of synapses in the antennal lobes of the moth Manduca sexta during metamorphosis, J. Neurosci. 3 (1983) 1158–1175. [31] L.A. Oland, W.M. Pott, G. Bukhman, X.J. Sun, L.P. Tolbert, Activity blockadge does not prevent the construction of olfactory glomeruli in the moth Manduca sexta, Int. J. Dev. Neurosci. 14 (1996) 983– 996. [32] J.H. Hayashi, J.G. Hildebrand, Insect olfactory neurons in vitro: morphological and physiological characterization of cells from the developing antennal lobes of Manduca sexta, J. Neurosci. 10 (1990) 848–859. [33] C. Duch, R.B. Levine, Remodeling of membrane properties and dendritic architecture accompanies the postembryonic conversion of a slow into a fast motoneuron, J. Neurosci. 20 (2000) 6950–6961. [34] C. Lohr, E. Tucker, L.A. Oland, L.P. Tolbert, Development of depolarization-induced calcium transients in insect glial cells is dependent on the presence of afferent axons, J. Neurobiol. 52 (2002) 85–98.