U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Chattopadhyay A, editor. Serotonin Receptors in Neurobiology. Boca Raton (FL): CRC Press/Taylor & Francis; 2007.

Cover of Serotonin Receptors in Neurobiology

Serotonin Receptors in Neurobiology.

Show details

Chapter 1Quantitative Imaging of Serotonin Autofluorescence with Multiphoton Microscopy

and .


In this article we describe a technique to directly image vesicular serotonin in live neurons. In the context of serotonergic signaling in the brain, it is important to map out serotonergic storage and release sites, together with the distribution of serotonin receptors. Receptor mapping is possible by using endogenously labeled receptors, but serotonin vesicle mapping is more difficult. Following the early success in mapping serotonin in mast cells with three-photon excitation microscopy [1], we have recently demonstrated that such mapping is possible also in live neurons [2]. We describe here the technique of serotonin imaging in detail, together with a few application examples. This three-photon excitation based imaging technique should also be useful in general for live cell microscopy of UV chromophores.

What we present here is intended to be a practical guide and not a review of three-photon microscopy. For the technical details described here, we have strictly relied on our own hands-on experience. The description is divided into three main sections. “Multiphoton Microscopy” presents a brief summary of the concepts of multiphoton excitation. This is followed by a description of the optical technique of serotonin imaging in “Optical Setup.” Finally, “Imaging Serotonin and Other Monoamines in Live Neurons” describes a few examples of serotonin imaging in neuronal cells and also describes how this technology can possibly be extended to image other monoamines.


Multiphoton Excitation: Basic Concepts

Multiphoton excitation is a novel method for exciting monoamines and UV chromophores in general, with relatively benign infrared light. Multiphoton excitation depends on a nonlinear quantum effect first predicted by Maria Goppert–Meyer in 1931 [3], observed in practice by Singh and Bradley in 1964 [4], and put to use in biological microscopy by Webb and coworkers in 1990 [5]. It allows a molecule to be photo-excited by light that has much lower energy (and hence a longer wavelength) compared to what would be predicted from its conventional (one-photon) excitation spectrum. Simply put, when the intensity of the excitation light is very high, a fluorophore molecule can absorb multiple low-energy photons simultaneously, whose total energy equals or exceeds its one-photon excitation energy.

n-Photon excitation depends on the nth-power of the intensity of the excitation light. This intensity dependence automatically confers three-dimensional (3-D) localization of the excitation (as described later), which is exploited to obtain high resolution 3-D. Also, this process requires a longer excitation wavelength (~nλ) compared to the conventional one-photon excitation wavelength of λ. This factor helps us in lowering both photodamage to the sample and photobleaching of the chromophore. This is also the factor that allows us to perform live cell microscopy of UV molecules using infrared light.

There are excellent reviews that describe the principle and practice of two-photon microscopy in general [6–8]. The conceptual framework of two-photon microscopy can be simply extended to describe three-photon microscopy, so we will discuss the principles only briefly. For practice, it is useful to understand how the signal and the resolution vary with the nature of the excitation light, and these aspects would be highlighted here. The practical excitation source in multiphoton microscopy is a pulsed laser, usually tunable in the near-infrared region. The repetition rate of this laser, the pulse width, and the pulse shape in time, and the size of the collimated laser beam at the back aperture of the microscope objective lens are the key parameters that determine the efficiency of excitation.

The Requirement for a Pulsed Laser

The rate of excitation of a given molecular species depends on the repetition rate R, the pulse width τ, the average power Pav, and the beam waist at the focus ω0 of the excitation beam. The molecular property contributes to the n-photon excitation process through the n-photon absorption cross section σn. The average rate of fluorescence emission (per molecule) is proportional to the average rate of absorption, which in turn is proportional to


For CW (continuous wave, steady) excitation, Rτ = 1. As the order of excitation increases excitation cross section σn goes down, so for the same pulse width, the required Pav requirement goes up. For a typical value of σ3 = 3.2 × 10−96 m6 s2 photon–2 (which is the value for serotonin at 740 nm [1]) and that of the beam waist of 300 nm, the Pav required to get a near-saturation (i.e., near maximal) signal from a CW laser is about 400 W. This is a tremendous amount of power, and therefore only short pulsed lasers with high peak powers (but much lower average powers) are practical for this kind of microscopy.

This relationship also shows that for a three-photon process, the signal will increase by a factor of eight if the power is increased by a factor of two, and it will decrease by a factor of four if the pulse width is increased by a factor of two (the pulse shape factor has been neglected in this description; the reader is referred to Xu et al. [9] for details). In practice, the pulse width used is frequently about ~100 fs, as shorter pulses get severely broadened by the microscope optics through group velocity dispersion and nonlinear effects [10].

We note here that the log–log plot of fluorescence vs Pav would give a straight line with a slope of n, which is the order of the excitation. This relation holds only in the absence of saturation and photo bleaching of the fluorophore. Near saturation there is no more fluorophore that can be excited, and fluorescence counts approach a plateau.

Figure 1.1 shows the change in excitation efficiency with τ for two- and three-photon excitation. We assume the excitation rates to be the same for some arbitrary pulse width. Both decrease as pulse width increases, but the decrease for three-photon excitation is much sharper.

FIGURE 1.1. The dependence of excitation efficiency on the pulse width for two- (solid) and three-photon (dash) excitation.


The dependence of excitation efficiency on the pulse width for two- (solid) and three-photon (dash) excitation. For τ = 1, excitation efficiency for both is normalized to unity. For three-photon excitation, the efficiency drops faster than two-photon (more...)

The Requirement for Tight Focusing

Now we focus on the spatial properties of the laser beam. For a focused beam that has a Gaussian spatial profile, the spatial dependence of the intensity I can be expressed as a function of z (the distance along the beam propagation direction, measured from the focus) and r (the distance perpendicular to the beam propagation direction, measured from the beam axis).


Where the dependence on z is implicit through the function ω(z), ω(z) is the radius of the beam (defined as the distance where the field intensity drops to 1/e2 of the intensity at the beam axis) at a distance z from the focus, and ω0 is the minimum radius (i.e., radius at the focus, z = 0). For a Gaussian beam of wavelength λ, the variation of the beam diameter with z can be expressed as [11]


Figure 1.2A shows the excitation intensity I(r,z) around the focus spot (the beam is assumed to remain Gaussian throughout). Figure 1.2B and Figure 1.2C show I [2](r,z) and I [3](r,z) distribution around the focus, which are relevant for two- and three-photon excitation, respectively. From the figures it is clear that in MPE, as the order of excitation increases, excitation also becomes more and more localized. This is the origin of the 3-D resolution inherent in MPE.

FIGURE 1.2. Multiphoton excitation is localized near the focus.


Multiphoton excitation is localized near the focus. For a normalized focused Gaussian excitation beam with intensity I (r,z), the plot of I (r,z), I2(r,z), and I3(r,z) near the focus are shown in parts A, B, and C, respectively. As the order of excitation (more...)

We note that a higher order MPE also requires longer wavelengths to excite the same chromophore. This would widen the excitation profile somewhat, compared to that shown in these figures. An estimate can be made by considering Equation 1.1 and Equation 1.3. It is suggested that for a two-photon process, the resolution would be worse by a factor of √2, and for a three-photon process, it would be worse by a factor of √3 (compared to a single-photon confocal detection). However, frequently the excitation peak for a multiphoton process occurs at a wavelength shorter than the expected value of nλ, and this then improves the resolution. Ultimately, the resolution of a single photon confocal microscope is typically comparable to that of a multiphoton microscope.

Equation 1.3 shows that for a particular z, as we increase the beam diameter, ω0 decreases. This implies that as one increases the beam diameter at the back aperture of the objective lens, MPE becomes more efficient with the higher intensity produced. For typical high resolution microscopy, the beam size is increased to a size slightly bigger than the back aperture of the objective lens to get the smallest possible beam waist. In this case, the beam no longer remains Gaussian at the focus. Fraunhoffer diffraction theory predicts what the light field will be in such cases [12–14], but the rules of thumb derived from the Gaussian description above remain valid. So for obtaining a higher signal in three-photon microscopy, and to obtain the highest possible resolution, the excitation beam should fill the back aperture of the objective lens.

In a mode locked fs-pulsed laser, the power is very high for the duration of the pulse. As it is focused with a high NA objective lens, the intensity is sufficient for MPE to occur at the focus spot. Hence, all the fluorescence originates from the focus spot only, which typically has a volume of ~10−13 cc (unless the intensity is increased beyond a limit, and the saturation effects start dominating). This focal volume can be characterized, if needed, using a technique called a fluorescence correlation spectroscopy [15–19].

Choosing the Right Excitation Wavelength

In MPE, instead of one photon of wavelength λ, a molecule absorbs n photons [9] of wavelength ~nλ. However, the emission spectrum remains the same for almost all the molecules [6] (Figure 1.3). A fluorophore whose excitation peak is at wavelength λ often has a multiphoton excitation peak at a wave length smaller than nλ (n = 2, 3, …) as mentioned previously. For a chromophore with an unknown excitation spectrum, one has to scan the excitation wavelength around nλ and measure the fluorescence obtained at each λex to find the peak of the excitation spectrum.

FIGURE 1.3. The energy scheme for multiphoton excitation.


The energy scheme for multiphoton excitation. In conventional one-photon excitation, a molecule absorbs a single photon of energy E = hc/λ, which matches the energy difference between the excited and the ground state. In two-photon excitation, (more...)

Advantages of Three-Photon Microscopy

  1. Less damaging for cells: With three-photon excitation, serotonin can be excited using an infrared light source (e.g., a mode-locked fs laser) instead of using ultraviolet light (~270 nm). For imaging monoamines in live biological samples, this is a big advantage over one-photon confocal imaging, as infrared is far less damaging than the UV.
  2. Quantitative imaging: As the image is a representation of the auto fluorescence intensity distribution of the molecule itself in the sample, MPM is a very quantitative method. Concentration of the fluorophore in situ can be estimated by calibrating with a known concentration of the same fluorophore (after taking into account all photo-physical and photo-chemical effects).
  3. Enables imaging deeper in the thicker sample: With MPE, imaging deeper inside the biological samples (tissue) is possible with little distortion in the resolution compared to one-photon confocal imaging. In multiphoton excitation longer wavelengths are used for excitation so scattering of the excitation is much less
    (scattering of light1λ4).
    The tissue absorption is also much less in the near infrared. This results in good focusing and resolution to a considerable depth (~70 μm).
  4. Exciting more than one fluorophore simultaneously with the same excitation: MPE spectra are typically broad, which ensures that there is a considerable possibility of finding a good wavelength for simultaneously exciting multiple types of fluorophores. As only a single beam is used, no colocalization errors can creep into these measurements.


Point Scanning

So far we have discussed the fluorescence obtainable from the MPE of a small stationary volume. To know the distribution of fluorophore in an extended object like a neuron or a piece of animal tissue, one has to collect fluorescence from all the points with the resolution (limited by Rayleigh criterion) of the excitation volume. To take a fluorescence image, a focused excitation spot scans the extended object, and the fluorescence emitted at each point is collected through a photo detector. The plot of the fluorescence intensity vs. the position then produces the fluorescence image. This scheme is known as point scanning.


For setting up an MPM one needs (1) a good microscope (inverted or upright, depending on the demands of the experiment) with adequate flexibility to put external detectors, (2) scanning optics which make excitation beam scan in 2-D and which have reasonable infrared throughput (a conventional confocal scanner or simply a set of galvanometric mirrors with proper software is adequate), (3) a fs pulsed laser with high repetition rate, and (4) some ultrafast optics. The following steps are involved:

Coupling the Scanning Optics with the Microscope

  1. Scanning optics should be coupled to the microscope such that the excitation beam entering the objective should be collimated or left a little diverging (over a 3m distance; 10% increase in the diameter is good). If the beam is converging, with a high NA, low-working-distance objective lens, the beam may get focused within the objective itself and damage it.
  2. The pivot point of scanning (while scanning, the collinear beam makes an angular movement about a point, and the center of the beam always passes through it) should lie in the back focal plane of the objective lens and remain in the optical axis of the objective. This ensures homogeneous intensity distribution during spatial scanning. Otherwise as the beam moves away from the center, excitation intensity and spot size may change considerably. For achieving steps 1 and 2, one may have to use a 10× eye piece between the scanner and the microscope.
  3. For good resolution, focus spot size should be the smallest possible, which is achieved by adjusting the beam diameter at the back aperture of the objective lens (beam diameter should be equal to or slightly bigger than the back aperture of the objective).
  4. Beam size on the scanning mirrors should be small to avoid distortion because of the rotation about their own axes.


Detecting the Fluorescence

When a fluorophore is excited in an isotropic specimen it also emits isotropically. While scanning the sample, depending on the emission spectra and the demand of the experimental setup, the fluorescence can be collected mainly in three ways.

  1. Epifluorescence collection
  2. Non-epi collection [20]
  3. 4π collection [21] (i.e., both epi and non-epi)

Epi Collection

In an epifluorescence setup (Figure 1.4), excitation is focused with an objective and the part of the fluorescence emitted in the backward direction is collected with high efficiency by the same objective. The collected fluorescence counter propagates along the excitation path and is separated from the excitation by a dichroic mirror placed next to the objective at an angle of 45° to the beam direction.

FIGURE 1.4. Schematic diagram of the setup for epi collection.


Schematic diagram of the setup for epi collection. (Adapted from Kaushalya, S.K., Balaji, J., Garai, K., and Maiti, S., Fluorescence correlation microscopy with real-time alignment readout, Appl Opt, 44, 3262–3265, 2005. With permission from the (more...)

The fluorescence reflected by the dichroic passes through appropriate filters and finally falls on a wide area photo detector (usually a photo multiplier tube, PMT). A wide-area detector is required because the dichroic-reflected fluorescence is not stationary in space: it moves as the excitation beam scans the sample. No confocal pinhole is needed in front of the detector because all the emission is coming from a diffraction-limited focus spot where intensity is enough to produce MPE. This scheme is known as nondescanned detection and is usually the preferred (and more sensitive) way of collecting the multiphoton fluorescence signal.

In the epi setup, collection will be limited for fluorophores that emit at UV wavelengths which do not pass through the objective. Typically objectives are transparent above 350 nm (see Figure 1.6). So, for imaging a fluorophore which emits in the UV region, one should carefully check the transmission of the objective.

FIGURE 1.5. Schematic diagram of the setup for non-epi collection.


Schematic diagram of the setup for non-epi collection. Here fluorescence does not pass through any optics except for filters which block the excitation but transmit the fluorescence. (Adapted from Balaji et al. [20] With permission from the Optical Society (more...)

Non-Epi Collection (without a Lens)

In the non-epi detection scheme, maximum possible fraction of the fluorescence emitted in the forward direction is collected onto a wide area detector placed at the nearest possible separation from the sample with suitable filters placed before it (Figure 1.5).

FIGURE 1.6. Spectral characteristics of serotonin and optical elements.


Spectral characteristics of serotonin and optical elements. Peak normalized fluorescence spectra of 400 mM serotonin (filled circles) and NADH (filled triangles), transmission characteristics of CuSO4 (inverted empty triangles), Cu(NH3)4SO4 (empty squares) (more...)

Non-epi detection is mainly useful for the fluorophores whose emission is deeper in the UV region and which will not pass through the objective lens. The main limitation of this scheme of detection is in blocking the excitation properly. Because excitation intensity is 1012 ~ 1013 times higher than the fluorescence, and it falls on the detector almost unattenuated, so excitation shielding is very crucial for obtaining

Specific Optical Setup for Serotonin Imaging

Serotonin is one of the monoamines that is suitable for MPM with epi-collection because of a good overlap of its fluorescence spectra (Figure 1.6) with the detection optics. It is three-photon excitable at 740 nm, which is far less photo toxic to cells than the one-photon excitation wavelength of 270 nm. In the neuronal cells, serotonin is packed in vesicles of size ~100 nm in reasonably high concentration (~400 mM), which makes it easy to detect and image.

For serotonin imaging we use RN46A cells or primary culture neurons from the raphe nucleus of the 1~2-day-old rat or raphe tissue itself. For excitation, a mode-locked 100 fs, 76 MHz pulsed laser (Mira-9000, Coherent, U.S.) at 740 nm is focused with a water immersion 1.2 NA, 60× objective lens (Nikon, Japan).

A dichroic mirror (675 DCXRUV, Chroma, U.S.) is used just below the objective which reflects the emission. Emission is passed through the saturated CuSO4 filter (custom-made 1-cm path length). CuSO4 solution nicely serves the purpose of a filter because it blocks all the excitation and passes the emission with a wide transparent window (Figure 1.6) which allows the serotonin emission to pass through. (We note that multiphoton excitation of serotonin at 740 nm results in a second emission peak at 500 nm.) For detection, a 30 mm head-on single-photon-counting PMT (Electron Tubes, U.K.) is used. Separation between the dichroic and the filter with PMT is kept to a minimum, so that scattered fluorescence light can also be collected.

Checklist before Imaging

  1. To reduce optical noise and background, the whole setup, including the microscope stage and the PMT, should be made light proof. Before starting any experiment, the background leakage should be checked and minimized.
  2. For calibration purposes, every time a known concentration (~100 mM) of serotonin solution in buffer (with a prefixed average excitation power and pulse width) should be taken, and the signal obtained should be checked with previous records.
  3. In an experiment in which different excitation powers have to be used, one should be careful about the change in the pulse width, because typical means used to change the power (such as an neutral density filter) may affect the pulse width differently. One of the best ways to control power is to use a half-wave plate and a polarizer combination.
  4. Polarization of the excitation beam should be fixed, as the throughput of the optical system may be strongly polarization dependent.
  5. Before imaging a biological sample, one should be aware of the distance calibration of the MPI setup. From that one can easily calculate the length and the width of a pixel. For best resolved images, the settings should be such that each pixel corresponds to about half of the optical resolution in length and width. Typical pixel dimensions for good resolution would be about 120 nm in the x–y and 400 nm in the z direction.


Imaging Protocol

  1. Raphe neurons are isolated using standard primary culture protocols [22]. Cells cultured on homemade cover-slip-bottomed petri dishes (cover slips glued to the 15mm bottom-drilled petri dishes with Canada balsam which we find to be nontoxic) are placed in the imaging buffer (NaCl 146 mM, KCl 5.4 mM, CaCl2 1.8 mM, MgSO4 0.8 mM, KH2PO4 0.4 mM, Na2HPO4 0.3 mM, d-Glucose 5 mM, Na-HEPES 20 mM) after washing twice with the same buffer.
  2. An appropriate field is chosen while viewing the specimen under phase contrast.
  3. Cells should be viewed in transmission with a high NA water immersion objective, which is used for MPM (with the aperture before the condenser partially closed, thereby increasing the contrast by reducing the depth of field). This ensures that the object to be imaged is in the focal plane of the objective lens and reduces the time required to search for the appropriate z-plane once scanning is started.
  4. The transmission lamp is switched off, and the instrument is covered so that no stray light couples to the detection system.

Establishing the Source of the Signal

In Figure 1.7A, serotonin is seen as bright-punctate-distributed spots of varying size in the cell. What we see as the bright punctate structure is vesicular serotonin. When calibrated with known concentration of serotonin in solution, the average concentration inside the vesicle comes out to be about 400 mM. However, multiple components of a cell can emit autofluorescence [23], and some of these may have spectral overlap with serotonin (this is especially true for NADH). So for any type of cell that has not been characterized previously, one has to run a battery of tests before assigning the fluorescence to serotonin. Some of the possible tests are: immunohistochemistry, inhibition of serotonin synthesis, up-regulation of serotonin synthesis, mass spectrometry of the cell extract, checking the order of excitation for its three-photon nature (many fluorescent components such as NADH would be two-photon fluorescent at 740 nm), labeling other cellular objects such as mitochondria to identify the source of the fluorescence (causing serotonin exocytosis by depolarization of the membrane), and inducing serotonin release with amphetamines. Many of these tests have been described elsewhere [2], and we will only describe a few of these here. In the context of the applications of serotonin imaging, we will describe a few of the other ones in the subsequent section.

FIGURE 1.7. Imaging serotonin in cells.


Imaging serotonin in cells. (A) a three-photon image of serotonergic cell line RN46-A. (B) image of the same field after exposure to 75 mM K+. Serotonin vesicles are seen as bright punctate structures. Bar = 10 μm. (Adapted from Balaji et al. (more...)

Test for the Order of Excitation

At 740 nm the serotonin solution is three-photon excitable. This suggests that if the bright structures are serotonin they should also follow the same excitation order. To check this, the same field of cells was imaged with different excitation power, and the order of excitation was calculated from the fluorescence intensity of the bright structures at different excitation power. It came out to be (2.81 ± 0.29) (Figure 1.8).

FIGURE 1.8. The proof for three-photon excitation.


The proof for three-photon excitation. Here, log of the fluorescence signal from the bright cellular structures is plotted against the log of the corresponding excitation powers. Slope of the straight line fit is 2.81 ± 0.29, from which we conclude (more...)

Checking Co-Localization of Bright Structures with NADH

In the cells NADH is stored inside mitochondria, and it is two-photon excitable at 740 nm [24,25]. Its fluorescence spectrum has good overlap with the serotonin spectra and the transmission window of a CuSO4 filter. To check whether the bright punctuate structures in the image are serotonin or NADH, one has to simultaneously image mitochondria and serotonin in two separate channels and check for colocalization of the bright structures.

In a live neuron, mitochondria can be labeled by incubating cells with 10 μM rhodamine 1-2-3 (a fluorescent mitochondrial marker, which specifically gets packaged inside the mitochondria). To check for colocalization, simultaneous imaging for rhodamine 1-2-3 and serotonin is required. A slight modification on the detector side allows us to simultaneously image serotonin as well as rhodamine 1-2-3. The fluorescence after reflection from the dichroic (Figure 1.4) is split into two parts, using a 70/30 beam splitter; 30% of the beam is reflected for rhodamine 1-2-3 detection, whereas 70% goes straight to the serotonin channel. Rhodamine 1-2-3 fluorescence is selectively detected through a 555/50 filter (Chroma, Inc.) placed directly in front of the PMT. The serotonin channel has a 0.3 cm thick filter of saturated Cu(NH3)4SO4 solution, which blocks the rhodamine 1-2-3 fluorescence. Filters used in either channel don’t allow any cross talk and block residual infrared also. In the rhodamine 1-2-3 channel, elongated cylindrical mitochondrial structures are seen. In the serotonin channel the observed structures are mostly spherical. These two types of structures are also seen in different planes (Figure 1.9). So structures seen in the two channels do not colocalize, and the fluorescence therefore cannot originate from NADH in the mitochondria.

FIGURE 1.9. (Color figure follows p. 110.) Serotonin vesicles and the mitochondria.


(Color figure follows p. 110.) Serotonin vesicles and the mitochondria. (A) a pseudo-colored merged image of primary cultured cell simultaneously imaged at 6 μm above the cover slip, for serotonin (green, auto fluorescence) and mitochondria (red, (more...)

Applications of Serotonin Imaging

The ability to image serotonin directly and quantitatively provides a great assay for multiple types of studies involving serotonergic signaling. We describe a few of these here: (1) depolarization induced dynamics of serotonin vesicles, (2) differentiation-induced up-regulation of serotonin synthesis, and (3) amphetamine induced nonexocytotic release of serotonin from raphe neurons.

Exocytosis with K+ Depolarization

Physiologically, serotonergic vesicles in a neuron are exocytosed when the membrane is depolarized. A large number of studies have relied on electrophysiological techniques to extracellularly monitor exocytosis. Alternatively, optical techniques have also been used to monitor vesicle recycling during exocytosis [26,27] (after labeling the membrane with a fluorescent dye). However, three-photon imaging of serotonin, for the first time, allows us to directly monitor the neurotransmitter during exocytosis [1]. We provide an example here.

We use differentiated cells from a serotonergic cell line (RN46A) [2]. In Figure 1.7A, we see the cells prior to any treatment. The serotonin vesicles (or vesicular clusters) are clearly visible as bright punctate structures. The cells are then treated with 50 mM KCl solution, with KCl replacing an equivalent amount of NaCl in the solution. The cells are then imaged repeatedly at 5-min time intervals. Figure 1.7B shows the cells after 15 min. A large number of vesicles have disappeared in the intervening time, presumably due to exocytosis. We note that most of the vesicles that remain approximately retain their original intensity, showing that the effects of photobleaching, etc., are negligible.

Monitoring the Differentiation of Serotonergic Cells

The ability to directly monitor serotonin enables us to monitor the upregulation of serotonin synthesis and its packaging, in neurons. An interesting possibility is to use this technique as an assay for the emergence of serotonergic traits in pluripotent cells. We have followed the differentiation of a temperature-sensitive mutant cell line RN46A as the temperature is elevated from 33°C to 39°C [2,28]. The images of the cells in 33°C (incompletely differentiated) and those from the cells kept for 5 days at 39°C (well differentiated) are shown in Figure 1.10B and Figure 1.10A, respectively. Quantitative measurement shows that the total amount of serotonin per cell has increased by a factor of two [2]. In addition, the packaging of serotonin becomes much more prominent, and the number of serotonergic vesicles increases by a factor of four.

FIGURE 1.10. Serotonin imaging as an assay for differentiation.


Serotonin imaging as an assay for differentiation. Differentiated RN46A cells contain more serotonin and many more vesicles (A) compared to their undifferentiated counterparts (B) Bar = 10 μm. (Adapted from Balaji et al. [2] With permission from (more...)

Amphetamine-Induced Serotonin Release

A number of psychoactive agents work through the monoaminergic system, and serotonin imaging can provide a direct assay for the action of these agents. One of the most well-known classes of these agents is the amphetamine family. Extracellular measurements indicate that amphetamines affect the central nervous system in three main steps. They induce vesicular monoamines (serotonin and dopamine) to leak out into the cytosol, then this cytosolic monoamine is nonexocytotically expelled from the cell through reverse transport by the plasma membrane transporters. The resultant massive increase of the monoamine concentration gives rise to psychedelic effects. We can now directly image the first two steps of this process. In sequential images, we resolve the expulsion of the serotonin from the vesicles into the cytoplasm, and then observe a slower decrease of the total serotonin content of the cell (data not shown).

Catecholamine Detection

It is important to explore whether this technique is capable of imaging other neurotransmitters. In general, if a neurotransmitter is intrinsically fluorescent, even if only under UV excitation, multiphoton microscopy should be able to image it. Neurotransmitters give us the unique opportunity of imaging them, as they are concentrated in small punctate structures in the cell and can stand out against even relatively strong but diffuse background. Catecholamines are of immediate interest in this context, as they are fluorescent in the UV, albeit at wavelengths (~300 nm) which are even shorter than that of serotonin. This is a wavelength range where the throughput of normal glass-based objective lenses and other optical elements is very poor. However, the non-epi detection scheme discussed earlier can be the perfect solution to this throughput problem. UV transparent filters do not adequately block the infrared (we need more than 1010× blocking of the infrared in the non-epi mode). So we need to employ a special type of fs laser, known as an optical parametric oscillator (e.g., MIRA-OPO from Coherent) for this purpose. This laser operates in the visible region and can excite the fluorophores through a two-photon route. Also, very good glass-based filters (such as UG11, Schott Glass, Germany) are available for blocking the visible wavelengths while passing through the UV. This makes the catecholamines accessible to imaging. For example, dopamine is two-photon excitable at a 550 nm wavelength. Using the scheme outlines here, we have been able to obtain sufficient signal from a dopamine solution to speculate that dopamine imaging in live neurons would also be possible in the near future, using the non-epi detection route [20].


J. Balaji and Radha Desai have been instrumental in procuring some of the data reported here. This research is supported by a Wellcome Trust Senior Overseas Research Fellowship (no. 05995/Z/99/Z/HH/KO) to SM.


Maiti S, Shear JB, Williams Rebacca M, Zipfel WR, Webb WW. Measuring Serotonin Distribution in Live Cells with Three-Photon Excitation. Science. 1997;275:530–532. [PubMed: 8999797]
Balaji J, Desai R, Kaushalya SK, Eaton MJ, Maiti S. Quantitative measurement of serotonin synthesis and sequestration in individual live neuronal cells. J Neurochem. 2005;95:1217–1226. [PubMed: 16269014]
Maria Göppert-Mayer. Über Elementarakte mitzwei Quantensprüngen. Annalen der Physik. 1931;401:273–294.
Singh S, Bradley LT. Three-photon absorption in napthalene crystals by laser excitation. Phys Rev Lett. 1964;12:612–614.
Denk W, Strickler JH, Webb WW. Two-Photon Laser Scanning Fluorescence Microscopy. Science. 1990;248:73. [PubMed: 2321027]
Xu C, Zipfel W, Shear JB, Williams RM, Webb WW. Multiphoton fluorescence excitation: new spectral windows for biological nonlinear microscopy. Proc Natl Acad Sci USA. 1996;93:10763–10768. [PMC free article: PMC38229] [PubMed: 8855254]
Denk W, Piston DW, Webb WW. Handbook of Confocal Microscopy. Pawley J, editor. Plenum Press; New York: 1994.
Svoboda K, Yasuda R. Principles of two-photon excitation microscopy and its applications to neuroscience. Neuron. 2006;50:823–839. [PubMed: 16772166]
Xu C, Webb WW. Measurement of two photon excitation cross sections of molecular fluorophores with data from 690 to 1050 nm. J Opt Soc Am B. 1996;13:481.
Guild JB, Xu C, Webb WW. Measurement of group delay dispersion of high numerical aperture objective lenses using two-photon excited fluorescence. Appl Opt. 1997;36:397–401. [PubMed: 18250687]
Moore JH, Davis CC, Coplan MA. Building Scientific Apparatus: A Practical Guide to Design and Construction. Addison Wesley; Redwood City, Calif.: 1989.
Born M, Wolf E. Principles in Optics: Electromagnetic Theory of Propagation, Interference and Diffraction of Light. Cambridge University Press; 1980.
Sheppard CJR, Matthews HJ. Imaging in high-aperture optical systems. J Opt Soc Am A. 1987;4:1354–1360.
Wilson T, Sheppard CJR. Theory and Practice of Scanning Optical Microscopy. Academic; London: 1984.
Magde D, Elson E, Webb WW. Thermodynamic fluctuation in a reaction system — measurement by fluorescence correlation spectroscopy. Phys Rev Lett. 1972;29:705–708.
Elson E, Magde D. Fluorescence correlation spectroscopy: I Conceptual basics and theory. Biopolymers. 1974;13:1–27. [PubMed: 4818131]
Magde D, Elson E, Webb WW. Fluorescence correlation spectroscopy: II An experimental realization. Biopolymers. 1974;13:29–61. [PubMed: 4818131]
Maiti S, Haupts U, Webb WW. Fluorescence correlation spectroscopy: diagnostics for sparse molecules. Proc Natl Acad Sci USA. 1997;94:11753–11757. [PMC free article: PMC33774] [PubMed: 9342306]
Sengupta P, Balaji J, Maiti S. Measuring diffusion in cell membranes by fluorescence correlation spectroscopy. Methods. 2002;27:374–387. [PubMed: 12217654]
Balaji J, Reddy CS, Kaushalya SK, Maiti S. Microfluorometric detection of catecholamines with multiphoton-excited fluorescence. Appl Opt. 2004;43:2412–2417. [PubMed: 15119610]
Hell S, Stelzer EHK. Fundamental improvement of resolution with a 4Pi-confocal fluorescence microscope using two-photon excitation. Opt Commun. 1992;93:277–282.
Banker G, Goslin K. Culturing Nerve Cells. The MIT Press; Cambridge, Mass.: 1998.
Zipfel WR, Williams RM, Christie R, Nikitin AY, Hyman BT, Webb WW. Live tissue intrinsic emission microscopy using multiphoton-excited native fluorescence and second harmonic generation. PNAS. 2003;100:7075–7080. [PMC free article: PMC165832] [PubMed: 12756303]
Kasischke KA, Vishwasrao HD, Fisher PJ, Zipfel WR, Webb WW. Neural Activity Triggers Neuronal Oxidative Metabolism Followed by Astrocytic Glycolysis. Science. 2004;305:99–103. [PubMed: 15232110]
Vishwasrao HD, Heikal AA, Kasischke KA, Webb WW. Conformational dependence of intracellular NADH on metabolic state revealed by associated fluorescence anisotropy. J Biol Chem. 2005;280:25119–25126. [PubMed: 15863500]
Betz WJ, Bewick GS. Optical Analysis of Synaptic Vesicle Recycling at the Frog Neuromuscular Junction. Science. 1992;255:200–203. [PubMed: 1553547]
Rizzoli SO, Betz WJ. The Structural Organization of the Readily Releasable Pool of Synaptic Vesicles. Science. 2004;303:2037–2039. [PubMed: 15044806]
White LA, Eaton MJ, Castro MC, Klose KJ, Globus MY, Shaw G, Whittemore SR. Distinct regulatory pathways control neurofilament expression and neurotransmitter synthesis in immortalized serotonergic neurons. J Neurosci. 1994;14:6744–6753. [PMC free article: PMC6577239] [PubMed: 7965075]



multiphoton excitation


multiphoton microscopy




femto second

Copyright © 2007, Taylor & Francis Group, LLC.
Bookshelf ID: NBK5208PMID: 21204456


  • PubReader
  • Print View
  • Cite this Page

Other titles in this collection

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...