Main Content

Gordon Research Conference

Real-Time Imaging of Neuron Network Activity by Chemiluminescence

John R. Feick
Department of Biology
Saint Anselm College

Information on the performance of networks of living nerve cells would be enhanced if visible readouts of neuron activity were possible. A protocol has been developed for testing potential chemiluminescent indicators of the metabolic activity of nerve cells in cell culture, using lucigenin. Fluorescent microscopy confirms that lucigenin is concentrated by mitochondria. Active oxygen is an expected by-product of mitochondrial metabolism but is only weakly detected because cells are protected by active oxygen acceptors. Work by M. Trush demonstrates that this oxygen scaveng-ing system can be overloaded using very dilute KCN to derail the electron transport system. A modified scintillation counter is used to read the light output from the treated neurons. Photon output is low. The IPD low light imaging equipment at the MBL, Woods Hole, required computer processing to produce minimal information pictures. Lucigenin analogs, synthesized by Helix Research, are being surveyed as alternative photon producers. Current light output is not strong enough to support practical applications, but small improvements might make the method useful.

Twenty years ago Israel (1) reported that synaptic transmission could be detected as photon output, using luminol to react with active oxygen from the acetylcholine breakdown reaction. Israel's experiments suggested that imaging the activity of inter-connected neurons could be useful, and his work continues. About the same time Grinvald (2) employed electro-luminescent dyes to image action potentials and has used the technique to demonstrate results compatible with the data from many electrode-based neurophysiology experiments. Grinvald's work also continues, with a recent paper in Science (3).

After reading Israel's and Grinvald's papers, I began trying to image neuron activity using a chemiluminescent indicator (lucigenin) to detect the active oxygen produced "accidentally" by the cytochrome series in the final electron transport step that forms ATP.

On my sabbatical five years ago, I visited with M. Trush (4) at Johns Hopkins, who works with the chemiluminescent detection of mitochondrial activity in WBCs and macrophages, using lucigenin. Trush has demonstrated that a dilute KCN block derails energy transfer to ATP at the final electron transport stage, enhancing the production of active oxygen by 23%. Trush has also demonstrated that lucigenin acts as a fluorescent dye that concentrates in mitochondria. (Molecular Probes now distributes lucigenin as a mitochondrial stain, with results similar to Rhodamine 123.)

Trush's work has been at the center of a recent controversy because it has been suggested by I. Fridovich (5) that lucigenin at high concentration drives the chemilumi-nescent reaction, yielding a photon output that exceeds the metabolic production of reactive oxygen (redox cycling of lucigenin). Trush and others (6) have produced convincing evidence that lucigenin used at the concentration they employ does not create this artifact.

There is no difficulty recording the low-level photon output from nerve cells using a photomultiplier tube detection system. Imaging is more difficult. The photo-multiplier tube registers the integrated light output from a mass culture. Imaging requires detecting and localizing light from a single cell, a single axon, or better, an individual mitochondrion. The type of cooled CCD camera used by astronomers has been steadily improved to form images from very low photon-output sources, but these cameras have proven inadequate for this application.

A resistive anode imaging photon detector (IPD) imaging system at the Low Level Light lab at the MBL, Woods Hole, operated by Eric Karplus using computer processing produced a "proof-of-concept" image, but demonstrated that the level of light output currently obtained is unlikely to have practical application.

Just as Grinvald was able to improve the electroluminescent dye technique by becoming involved with the development of more efficient dyes, it is likely that a practical chemiluminescent demonstration of mitochondrial metabolism, sensing reactive oxygen output, will depend upon the development of a more suitable chemi-luminescent molecule. I have been testing lucigenin analogs synthesized by M. Kuhn of Helix Research. Three molecules tested thus far---N, N-dimethyl biacridanyl (Helix), lucigenin-HA (Helix), 9-cyano-10-methyl acridinum (Sigma library of rare chemicals)---have been both chemiluminescent and fluorescent, staining mitochondria specifically. With the protocols tested, these analogs showed no dramatic improvement in photon output.

There are interesting and useful chemiluminescence experiments that can be done using mass nerve cell cultures and a photomultiplier tube detector, but my efforts have been directed toward developing a protocol that would allow useful imaging. The most productive focus for future research seems to be the synthesis of promising new analog chemiluminescent dyes. This is an expensive approach (approximately $5,000/ analog) and my participation is dependent upon a co-operative testing agreement with Helix Research Company.

Further modification of the protocol by controlling the chemical environment of the cells under study may produce improved light output, however one cannot pro-ductively study nerve cell activity in an unnatural chemical soup. (Benzaldehyde produces mitochondrial fireworks!) My minimal requirement for the chemical system under study is that the cells used survive the experiment and can be cultured and reproduce when returned to cell culture medium.

Increased photon output is usually detected when the test saline solution is made hypotonic. This dilution may drive cell metabolism (pumping) or it may result from a reduction of chloride ion, which is a lucigenin inhibitor. The lucigenin reaction is pH sensitive, light output increases with pH.

George Reynolds told me (1993) that the minimal light output to support imaging was 200 photons/sq mm/min. Detectors have improved, and it has been possible to plot detected photon localizations over a conventional microscope image. Background noise, electronic and photon, interferes with the idealized concept of single photon detection. (The light levels discussed here are far below the sensitivity of the human eye.) Current light output is not strong enough to support practical applications, but small improvements may make the method useful.

The HT4 cells used in these experiments were a gift of Dan Koshland. The NIE115 cells were provided by Marshal Nirenberg. Lucigenin analogs were from M. Kuhn of Helix Research. The IPD trials were arranged by Eric Karplus, Science Wares, Woods Hole, Massachusetts and by Robert Creton of the Lionel Jaffee Lab. All support for this research has been provided by Saint Anselm College.


  1. Israel, M. and B. Lesbats. 1981. Chemiluminescent determination of Acetylcholine, and Continuous Detection of Its Release From Torpedo Electric Organ Synapses and Synaptosomes. Neurochemistry International, Vol. 1, No. 1, pp. 81-90.
  2. Grinvald, A., A. Manker and M. Segal. 1982. Visualization of the Spread of Electrical Activity in Rat Hippocampal Slices by Voltage-Sensitive Optical Probes. J. Physio., Vol. 333, pp. 269-291.
  3. Tsodyks, M., T. Kenet, A. Grinvald, and A. Arieli. 1999. Linking Spontaneous Activity of Single Cortical Neurons and the Underlying Functional Architecture. Science, Vol. 256, p. 1943.
  4. Esterline, Russell L. and Michael A. Trush. 1989. Lucigenin Chemiluminescence and Its Relationship to Mitochondrial Respiration in Phagocytic Cells. Biochemical and Biophysical Research Communications, Vol. 159, No. 2, pp. 584-591.
  5. Fridovich, Irwin. 1997. Superoxide Anion Radical (O2), Superoxide Dismutases, and Related Matters. J. Biol. Chem., Vol. 272, No. 30, pp. 18515-18517.
  6. Li Y, Zhu H, P. Kuppusamy, V. Roubaud, J. L. Zweier, and M. A. Trush. 1998. Validation of Lucigenin (bis-N-methylacridinium) as a Chemilumigenic Probe for Detecting Superioxide Anion Radical Production by Enzymatic and Cellular Systems. J. Biol. Chem., Vol 273, No. 4, pp. 2015-2023.