Neural Excitation of the Larval Firefly Photocyte: Slow Depolarization Possibly Mediated by a Cyclic Nucleotide

J. exp. Biol. (1976), 65, 213-327 213 ~Vith 10 figures xrinUd in Great Britain

NEURAL EXCITATION OF THE LARVAL FIREFLY PHOTOCYTE: SLOW DEPOLARIZATION POSSIBLY MEDIATED BY A CYCLIC NUCLEOTIDE

BY DONATA OERTEL AND JAMES F. CASE Department of Biological Sciences University of California at Santa Barbara, Santa Barbara, California 93106

(Received 3 February 1976)

SUMMARY 1. In fireflylarvae , extracellular recordings from the light organ nerve show that a volley of action potentials elicits a glow similar to the glow of an intact animal. 2. Intracellular recordings from the photocytes show that they respond to nerve stimulation with a slow, graded depolarization which precedes light emission. The depolarization begins about 0-5 s after the nerve is stimulated; it peaks about 1 s after stimulation; and subsides about 2^5 s after the stimulus. The glow increases fastest when the photocyte depolarization is at its peak and lasts 5-15 s. 3. Photocyte depolarization is associated with a decrease in the input resistance of the cell. 4. Adrenergic receptors in the light organ are pharmacologically similar to vertebrate a-receptors. 5. Phosphodiesterase inhibitors, aminophylline and theophylline, cause the light organ to glow, suggesting that cyclic nucleotides may mediate the effect of the adrenergic nerve transmitter.

INTRODUCTION Two aspects of neural control of bioluminescence in the firefly larva have been in- vestigated using electrophysiological and pharamacological methods: (1) synaptic mechanisms of the nerve-photocyte synapse; (2) intracellular mechanisms controlling luminescence. Although the biochemical reactions leading to light emission in the firefly have been extensively studied (McElroy & De Luca, 1973; Bowie, Horak & De Luca, 1973; Bowie et al. 1973), little is known about the physiological role played by nerve impulses. Several lines of evidence indicate that light emission in adult light organs is indeed controlled neurally (Buck & Case, 1961; Case & Buck, 1963; Buck, Case & Hanson, 1963). In addition it has been shown that the innervation of adult and larval light organs is probably adrenergic (Smalley, 1965; Carlson, 1968 a, b). The morphology of the light organ of the larva is simpler than that of the adult (Peterson, 1970; Oertel, Linberg & Case, 1975; Smith, 1963). A mass of about 2000 photocytes, well tracheated and innervated, is covered by dorsal layer cells. The cell 214 DONATA OERTEL AND J. F. CASE layers are clearly distinguishable in vivo because the photocyte layer is transparent id contrast to the opaque, white, dorsal cell layer. Each light organ receives one nerve branch containing only two axons which branch to innervate the photocytes directly. The synapse between nerve and photocyte has been identified as an electron-dense, irregularly shaped, inward projection from the nerve cell membrane which is surrounded by light core vesicles (Oertel et al. 1975). Several features of the light organ of the firefly larva make it a useful physiological preparation, namely: (1) Light emission is easily measured; (2) the photocyte layer is homogeneous, containing only photocytes, small nerve terminals and small tracheoles; (3) photocytes are easily exposed for experimentation by removing a small patch of dorsal layer cells; (4) the synapse between nerve and photocyte can be studied both presynaptically by recording extracellularly from the nerve, and postsynaptically by recording intracellularly from the photocyte.

MATERIALS AND METHODS General Larvae of Photuris pennsylvanica DeGeer were collected in the fall near Bethesda, Maryland and flown to California where they were maintained at 11 °C in Petri dishes lined with moist filter paper with a light cycle of 16 h light/8 h dark. Periodi- cally they were brought to room temperature, fed and cleaned (McLean, Buck & Hanson, 1972). The preparations on which electrophysiological recordings and pharamacological tests were made were similar. Larvae were decapitated and pinned on to a clear, Sylgard resin-filled dish, ventral side down. All subsequent procedures were done at x 62-5 magnification. The tergites were removed from all segments to expose the gut and fat bodies, which were then excised leaving the sternites, ventral musculature, light organs, and central nervous system (Fig. 1). For intracellular recording and for pharamacological tests a small hole about 75 /tm in diameter was made in the white, opaque, dorsal cell layer with an electrolytically sharpened tungsten pin, to expose the dorsal surface of the transparent mass of photocytes. If this is done carefully there is no background glow, as is characteristic of injured firefly light organs, and neurally excited glows do not change in shape or amplitude after the dissection. Extracellular recordings from the light organ nerve were made from the small branch just proximal to the site where the nerve enters the dorsal cell layer, a point which is conspicuous as a break in the dorsal cell layer. The preparations were bathed in physiological saline containing 90 mM-NaCl, 7 mM-KCl, 7 mM-CaCl2, 66 mM glucose, 1 mM Tris, pH 7-4. The proper concentrations of Na and K were determined by flame photometry of larval haemolymph. Blood was drawn into haematocrit capillary tubes and centrifuged. The serum was diluted in 116 mM Tris and the Na and K concentrations were then measured on an Eppendorf Flame Photometer. The optimal concentration of Ca2"1" was determined empirically and glucose was added to increase the osmolarity to 275 m-osmoles which was found to be optimal in electron microscopical fixation (Oertel et al. 1975). Neural excitation of the larval firefly photocyte 215

Int. Ext.

VIII abd. seg.

Tail

Fig. i. A diagram of the dissected firefly larva illustrating the intracellular recording arrangement from photocytes on the left and extracellular recording arrangement from the light organ nerve on the right. The light organs lie on the sternite of the eighth abdominal segment just anterior to the tail. The last of the chain of ganglia has a pair of nerves which leave the ganglion ventrolaterally, recognizable by their association with tracheae, and run posterior to the mid line of the eighth abdominal segment. One branch of this pair of nerves innervates each light organ. Its point of entry into the light organ can be seen as a break in the dorsal cell layer. To record intracellularly from photocytes, a patch of dorsal layer cells was cleared from the underlying photocyte layer so that the electrode approached the mass of photocytes dorsally, as shown on the left part of the diagram. A suction electrode was used to record from the light organ nerve just proximally to its entry into the light organ as shown on the right. In either recording condition, suction electrodes were used to stimulate the light organ through the ipsilateral segmental nerve (S). 216 DONATA OERTEL AND J. F. CASE

Photometric measurements Light emission from the preparation was monitored with an EMI Type ' S' photomultiplier tube powered by a Fluke model 409 A high voltage d.c. power supply.

Electrophysiology Extracellular recordings were made from the light organ nerve using glass suction electrodes of about 20 fim tip diameter, and a Grass P 5 a.c. coupled preamplifier. A second suction electrode was used for stimulation. Eserine, IO~3M, made up in physiological saline and applied to the eighth abdominal ganglion for short periods, was used in some experiments to cause the nerve to fire spontaneously (Case & Buck, 1963; Smalley, 1965). Intracellular recordings from photocytes were made using standard techniques with a WP Instrument, model M-4 A Electrometer. Microelectrodes were filled with 3 M-KC1 and had resistances of 60 MD. or more.

Pharmacology For pharmacological studies the eighth abdominal segment was cut at the mid line so that only one light organ of each pair was used, to avoid recording amgibuities. The dorsal cell layer was pierced and a few photocytes were exposed in order to reduce complications due to diffusion problems. Drugs were dissolved in larval firefly saline and were delivered to the light organ through fine polyethylene tubes with syringes. The chamber had a volume of 0-3 ml. To test relative effectiveness of agonists, the sensitivity of a single light organ to two or three drugs at equal concentrations was compared. Saline containing the dissolved drugs was added to the preparation in the dark. The solution was removed completely and the preparation was rinsed one or more times before the following test. The system allowed solution changes to be completed in 10 s. Since the sensitivity of the light organ changed with time, tests and controls were repeated several times during an experiment. To test a putative blocking agent, its effectiveness in reducing a neurally stimulated glow was measured using a preparation similar to that just described, but with the segmental nerve severed from the ganglion. The distal nerve stump was stimulated through a suction electrode with pulses of 5-15 V, 50 ms duration, at constant voltage and at regular intervals of about 45 s. The effectiveness of some agonists was tested by local application to the dorsal surface of the mass of photocytes. One microlitre of saline containing a drug was delivered over about 10 s to the exposed photocytes through a glass pipette with a tip diameter of 15-25/an, positioned with a micromanipulator. A microinjection apparatus was used to infuse the test solution slowly to avoid mechanical disturbance. The following chemicals were used in this study: adenosine 3':5'-cyclic mono- phosphoric acid, L-epinephrine, DL-epinephrine, guanosine 3':5'-cyclic mono- phosphoric acid, DL-isopropylnorepinephrine (DL-isoproterenol), L-norepinephrine (L-arterenol), DL-norepinephrine HC1 (DL-arterenol HC1), DL-octopamine, L-phenylephrine, DL-synephrine (Sigma); aminophylline (Schwarz/Mann); D-amphetamine Neural excitation of the larval firefly photocyte 217

5s Fig. 2. Extracellular recordings from light organ nerve showing that several impulses in close succession precede glowing with a constant latency. These recordings are from a single preparation which had been treated with eserine. (A) A weak stimulating pulse causes a volley of nerve impulses which precedes light emission. (B) A spontaneous volley of impulses and the associated glow. These recordings show only one nerve cell firing.

sulphate (U.S.P.); eserine sulphate, theophylline (Calbiochem); phentolamine, propranalol (tablets with lactose, CIBA Pharmaceutical Co.).

RESULTS Extracellular recordings from nerve Electrical stimulation of the segmental nerve to the light organ produces a glow with a distribution and duration similar to that resulting from disturbing a healthy animal. Thus a stimulus of 5-15 V of 50 ms duration applied to a small loop of nerve drawn into a suction electrode (Fig. 1) causes a glow about 1*75 s later. Glow amplitude increases with stimulus intensity and duration. The duration of the glow is between 5 and 10 s, except at very high stimulus intensities when the glow duration increases (Fig. 5)- Recordings from the branch of the segmental nerve which innervates the light organ show that electrical stimulation produces a volley of nerve impulses, whose frequency increases with stimulus strength. The light organ nerve is inactive when undisturbed. When the nerve is stimulated with pulses of short duration, the stimulus artifacts obscure recorded action potentials. To reduce this problem, extracellular recordings were done under one of two special conditions: (1) the preparation was treated briefly with io~3 M eserine to make the nerve fire spontaneously and to lower the threshold for electrical stimulation (Figs. 2, 3); or (2) the stimulating pulses were of very long duration (o- 5-1 s) (Fig. 4). Eserine causes the light organ to glow sporadically (Figs. 2, 3). This effect is observed only when the segmental nerve to the light organ is left unsevered from the ganglion in larvae as well as in adults (Case & Buck, 1963; Smalley, 1965), indicating that the site of action is central rather than peripheral. Simultaneous recording of light emission and extracellular action potentials from 2l8 DONATA OERTEL AND J. F. CASE

2s Fig. 3. An extracellular recording from the light organ nerve showing at least two cells firing. The upper trace records nerve activity and the lower records light emitted from the light organ. In this eserine-treated preparation the nerve cells fire continually while the light organ glows continually, brightening occasionally; throughout the experiment an increase in firing frequency preceded a brightening of the glow.

-. I

•3. 2 e

10 20 30 40 50 60 •^00 Number of nerve impulses Fig. 4. Relationship between nerve impulses and glow amplitude. An stimulating pulse causes the light organ nerve to fire for the duration of the pulse. The firing rate is voltage dependent. Plotting number of impulses against glow amplitude gives a sigmoid curve. Error bars showing the standard deviation of the mean are shown where more than one measurement was made. Neural excitation of the larval firefly photocyte 219 mV Or

-6OU

10 15 20 25 30 Time (s)

Fig. 5. Intracellular record from photocyte. The light, lower trace shows intracellular potential and the heavy, upper trace shows light emission. A strong stimulus was given at time o which caused a prolonged glow. The depolarization clearly precedes the glow. the nerve of an eserine-treated preparation shows that when the light organ glows spontaneously, a volley of action potentials precedes each glow. When the light organ glows continually, with occasional brightening, the nerve fires continually with an increase in firing frequency preceding each brightening. The lag between electrical activity in the nerve and light emission from the light organ is consistently about 1-75 s. This is seen in glows elicited by spontaneous nerve impulses (Fig. 2) and by electrically stimulated nerve impulses (Figs. 2, 3, 5, 7 and 8). Comparison of the latencies between nerve activity and light emission shows that nerve impulses at higher frequencies are much more effective in eliciting a glow than nerve impulses at lower frequencies (Figs. 2, 3). The recording shown in Fig. 3 is from an eserine-treated preparation. It shows at least two cells firing, a result consistent with the morphological observation that this nerve has two axons (Oertel et al. 1975). In a further examination of the relationship between nerve impulses and glow amplitude, long (1 s) pulses were used to stimulate the nerve. Under these conditions firing frequency during the pulse is voltage dependent with results as shown in Fig. 4. When the number of impulses per 1 s stimulus are plotted against glow amplitude, the resulting curve is clearly sigmoid, confirming the earlier observation that several impulses in close succession are more effective in eliciting a glow than is a single nerve impulse. Increasing the firing frequency of the nerve beyond about 30/s does not cause a further increase in light emission.

Intracellular recording from photocytes It is relatively easy to dissect away the dorsal cell layer, exposing the photocyte for electrode penetration. Photocyte resting potentials were between 50 and 65 mV. When the nerve is stimulated, causing the light organ to glow, photocytes depolarize slowly. Depolarization begins about 0-5 s after the stimulus, reaches a peak after about 1 s, and then decreases to the resting level (Figs. 5-8). This depolarization is graded as a function of stimulus strength (Fig. 6). Stimuli given more frequently than about 1 per min generate progressively smaller depolarizations and associated glows (Fig. 7)- 220 DONATA OERTEL AND J. F. CASE mV -30

-40

-50

-66*-

0 12 3 4 5 Time (s)

Fig. 6. Three superimposed sweeps show the graded photocyte depolarization in response to nerve stimulation. Stimulating pulses of 50 mi duration and various strengths were delivered at o s to the segmental nerve; the strongest pulse gives the strongest depolarization.

-701-

20 s Fig. 7. Simultaneous recording of intracellular membrane potential from photocytes (heavy lower trace) and light emission (light upper trace) show photocyte depolarizations and associated light emission after three successive stimulating pulses of equal strength were given to the segmental nerve. A larger depolarization is associated with a larger glow than with a smaller glow. Note: light emission measured is that of the entire light organ.

Light emission from an entire light organ was measured simultaneously with the intracellular potential. Depolarization of the photocyte in response to stimulation of the nerve precedes light emission; the depolarization reaches a peak while light emission is increasing most rapidly (Fig. 5) and then returns to rest level long before the light organ stops glowing (Figs. 5, 7 and 8). Intracellular recordings could be made from only one photocyte at a time, whereas the light emission measured was necessarily that of a whole light organ, about 2000 photocytes. Therefore it might be argued that photocyte depolarization may not always precede the glow but that some might depolarize later. This is unlikely, however, since photocyte depolarization preceded light emission in all measurements without exception. Depolarization of the photocyte membrane is associated with an increase in the membrane conductance. Fig. 8 shows the result of an experiment in which small pulses of current are injected into the cell at regular intervals through the recording electrode. The resultant voltage drop across the microelectrode resistance and input resistance of the cell is balanced prior to stimulation of the nerve. After stimulation, the balance shifts rapidly, reflecting a decrease in the input resistance of the photocyte, the direction of the shift depending on the direction of the injected current. The time course of the change in membrane conductance is the same as the time course of the Neural excitation of the larval firefly photocyte 221

n Or

-40

t t . _ * \ \

-801-

Q 4 8 12 16 20 Time (s)

Fig. 8. Photocyte depolarization is associated with a conductance increase of the photocyte membrane. The upper trace shows light emission, the lower trace, intracellular potential. In this experiment small pulses of current (less than io~10 A) were delivered through the recording electrode. The voltage drop across the microelectrode and across the input resistance of the cell is balanced before the nerve is stimulated. After the nerve is stimulated at o e the balance shifts rapidly as the photocyte depolarizes, then returns to the original position. This reflects a transient decrease in the input resistance of the photocyte after nerve stimulation.

depolarization. It begin9 before the light organs starts to glow, reaches a peak when light emission from the light organ is increasing most rapidly, and then returns to its original level before cessation of light emission. Application of norepinephrine, epinephrine and synephrine in concentrations ranging from io~4 to io~a M while recording intracellularly from photocytes resulted in light emission but not in depolarizations. The membrane potential and membrane conductance varied slightly and irregularly but there was not a consistent large depolarization nor a large conductance increase.

Pharmacology Our results confirm that the photocytes of the firefly larva are demonstrably sensitive to adrenergic drugs and insensitive to other putative transmitters (Carlson, 1968 a, b). Local application of putative transmitters directly on to photocytes shows that photocytes are sensitive to adrenergic drugs at relatively low concentrations (Fig. 9). Threshold sensitivities for epinephrine and norepinephrine measured by local application were similar to sensitivities measured by changing solutions in the entire bath (Figs. 9, 10). They were between io~6 and io~5 M, varying somewhat from preparation to preparation. The only other drugs tested to which the light organ was equally sensitive were synephrine (io~6 M) and octopamine (io~B M). Most pharmacological tests were done by removing the solution in the bath around the preparation with one syringe and adding the following test solution through another. Fig. 10 shows the results of an experiment comparing the effectiveness of different concentrations of epinephrine. When the light organ was not glowing, removal of the solution around it had little effect; however, if a solution was removed while the light organ glowed, the glow brightened. This brightening is most likely caused by variation in oxygenation of the light organ and suggests that while glowing, the light organ is sensitive to oxygen tension. 222 DONATA OERTEL AND J. F. CASE

2X10"3M

IS 3 Fig. 9. Local application of epinephrine. One microlitre of saline containing epinephrine at various concentrations was delivered on to photocyte surfaces of one preparation during the periods marked by bars. The sensitivity of the records varies; calibration bars on the left mark equal absolute magnitudes of glow.

Fig. 10. Application of epinephrine by changing bath. These two traces show the course of one continuous experiment. Epinephrine was first added at a low concentration, this solution was removed then saline was used to wash the preparation twice; finally epinephrine was added at a higher concentration and removed. Neural excitation of the larval firefly photocyte 223

Table 1. Relative sensitivity of light organ to drugs (This table shows the relative effectiveness of drugs on single light organ preparations with part of the dorsal cell layer dissected away. Each experiment (I, II and III) consists of tests of the relative effectiveness of several drugs at equal concentrations. The test solution was introduced, causing the light organ to glow, and was removed when the glow reached a fairly steady level (after 2-4 min); then the preparation was washed and another test drug was applied. Each test was repeated 2-4 times. The numbers represent the glow amplitudes measured from oscillograph records of photomultiplier output during the experiment. The table gives the mean relative glow amplitudes of repeated tests of the same solution ± standard deviation of the mean. Numbers are comparable only within one experiment for tests done on a single light organ.) I II III

5 x io~* M 1-5 x IO-*J I-OX IO"' M L-Epinephrine 37-3 ±3-8 L-Norepinephrine 575 ±25 L-Phenylephrine HCt o o 225 ±143 DL-l80propylnorepinephrine o o S-o±i-4

In order to compare the photocyte adrenergic receptor with other adrenergic receptors, the relative effectiveness of epinephrine, norepinephrine, phenylephrine and isopropylnorepinephrine was determined. The results of three experiments establishing the order of sensitivity are set forth in Table 1. In order of decreasing sensitivity these are: L-epinephrine > L-norepinephrine > phenylephrine > DL- isopropylnorepinephrine. The effects of certain adrenergic blockers on neural induction of luminescence were also tested. Nerve stimulation repeated at regular intervals excites a glow which is only slightly variable. The effects of possible blockers on the amplitude of these neurally excited glows were measured. The results from experiments with propranalol, a specific yff-adrenergic blocker, and phentolamine, an a-adrenergic blocker, were suggestive. At comparable concentrations propranalol caused a slower and smaller reduction in glow magnitude than phentolamine. In a further comparison of vertebrate adrenergic receptors and the photocyte adrenergic receptor, the effects of phosphodiesterase inhibitors, theophyUine and aminophylline, were tested. Both theophylline and aminophylline at io~* M cause the light organ to glow. The phosphodiesterase inhibitors, theophylline and aminophylline, are known to raise the levels of cyclic nucleotides intracellularly by specifically inhibiting their breakdown (Butcher & Sutherland, 1962). Therefore, a series of experiments was done testing the effects of 3':s'-cyclic AMP and 3':5'-cyclic GMP. Neither 5 rnM cAMP nor 5 mM cGMP had a measurable effect on the light organ; nor was it possible to measure a synergistic effect of aminophylline and either of these cyclic nucleotides.

DISCUSSION The light organ of the firefly larva is structurally one of the least complex of neurally regulated light organs (Peterson, 1970; Oertel et al. 1975). It contains only about 2000 transparent photocytes, which are small (12-20/on in diameter) polyhedral cells, packed together with highly interdigitating surfaces. In P. petmsylvanica, the organ is 224 DONATA OERTEL AND J. F. CASE innervated by only two axons which branch profusely so that many nerve terminal can be seen surrounding each photocyte. Membrane specializations presumed to be synapses between nerve and photocyte can be seen in the nerve terminals as electron- dense, irregularly shaped structures surrounded by small (40-65 rim), light core vesicles. Serial sections show that a single nerve terminal has many synapses (Oertel et al. 1975). Extracellular recordings from the light organ nerve show that it is usually inactive and that upon becoming active a volley of impulses excites a glow in the light organ that is similar to the glow of the disturbed animal. A single nerve impulse can pro- duce a weak glow and a series of impulses in close succession produces a facilitating luminescent response. Maximum light intensity is attained at less than the maximum attainable impulse frequency, suggesting either synaptic limitation or limitation associated with the luminescence mechanism. The synaptic mechanisms of the nerve-photocyte synapse are different from classi- cal nerve-nerve or nerve-muscle synapses where the transmitter directly causes a postsynaptic conductance increase. The photocyte depolarization is about one hun- dred times slower than the end-plate potential in vertebrate striated muscle (Katz, 1966) or the postsynaptic potential in secondary neurones of the dragonfly ocellus (Dowling & Chappell, 1972; Chappell & Dowling, 1972). Also, the application of putative transmitters does not cause a large depolarization and conductance increase even though the light organ responds by glowing. Since the synaptic gap is not particu- larly wide in the light organ (see fig. 13 in Oertel et al. 1975), the slowness of the depolarization must be due to causes other than slow diffusion across the gap. The nature of these remains to be examined. One possibility is that the depolarization of the photocyte is an indirect consequence of transmitter action postsynaptically. If, for example, the transmitter activated a nucleotide cyclase, the cyclic nucleotide might affect membrane permeability as well as activating the light-emitting reaction. Such a mechanism would be consistent with what is known about the action of cyclic nucleotides (Rasmussen, Goodman & Tenenhouse, 1972). The findings of Smalley (1965) and Carlson (1968a, b), suggesting that innervation of the fireflyligh t organ is adrenergic, raised the possibility that catecholamines might act indirectly, by way of cyclic AMP. Vertebrate adrenergically controlled effector systems fall into two general classes, the a- and /?-adrenergic receptors (Furchgott, 1972). This division is not strictly limited to vertebrates. For example, the development of sea urchins is regulated by a /?-adrenergic system (McMahon, 1974). The results presented here indicate that the adrenergic receptor of the firefly light organ is similar to the a-adrenergic receptor according to the definition given by Furchgott (1972, p. 286). Not all the specific blockers of the Furchgott characterization were tested, because dibenamine and phenoxybenzamine are not soluble in aqueous solutions, and pronethanol was not available. However, the pharmacological charac- teristics of this adrenergic receptor are consistently those of a-adrenergic receptors. And although a-adrenergic systems are not generally mediated through cyclic AMP, the phosphodiesterase inhibitors, aminophylline and theophylline, cause the light organ to glow. It has long been known that methyl xanthines inhibit phosphodiesterases (Suther- land & Rail, 1958). Theophylline and aminophylline have since then regularly been Neural excitation of the larval firefly photocyte 225 sed specifically to inhibit phosphodiesterases in vitro (Butcher & Sutherland, 1962; Kebabian, Kuo & Greengard, 1972). The effects of applying these phosphodiesterase inhibitors in vivo are also consistent with their hypothesized function as specific inhibitors of phosphodiesterases (Hess & Haugaard, 1958; Rail & West, 1963). A study done on the rat brain showed that theophylline competitively inhibits phosphodiesterases specific for cyclic AMP as well as those more specific for cyclic GMP (Goldberg et al. 1970; O'Dea, Haddox & Goldberg, 1970). Therefore it is highly probable that these phosphodiesterase inhibitors cause the light organ to glow by inhibiting the breakdown of cyclic nucleotides, thereby increasing their intracellular levels. The catecholamines and phosphodiesterase inhibitors must be acting on at least one of the three cell types found in the light organ: the photocytes, the nerve terminals, and the tracheolar epithelium. Although Smith (1963) found that nerves in the adult are associated with tracheal end cells, suggesting that the adult flash is somehow controlled by oxygen, there is no reason to believe that the tracheoles in the larval light organ have an active role in controlling light emission. There are no specialized junctions between tracheolar cells and photocytes or between tracheolar cells and nerve terminals (Oertel et al. 1975). Therefore these drugs act probably on the nerve terminals or on the photocytes. In many systems cyclic nucleotides act on the target cells rather than at the site of release of extracellular messengers: in the fly salivary gland, cyclic AMP mediates the effect of 5-hydroxytryptamine (Berridge & Patel, 1968; Berridge, 1970; Prince, Berridge & Rasmussen, 1972), in /?-adrenergic systems it mediates the effects of catecholamines (Sutherland, Robison & Butcher, 1968), and cyclic GMP mediates the effects of cholinergic drugs in the heart (George et al. 1970) and in the bullfrog sympathetic ganglion (Weight, Petzold & Greengard, 1974). It seems likely, therefore, that phosphodiesterase inhibitors act on photocytes and not on nerve terminals. The finding that neither cyclic AMP nor cyclic GMP caused a glow in the light organ does not exclude their possible action in the control system of larval bioluminescence. Even when the dorsal cell layer is dissected away, diffusion through the con- voluted extracellular space (Oertel et al. 1975) around the photocytes is slow and the concentration of cyclic nucleotides must be lower between the photocytes than in the bath. This means that the concentration gradient of cyclic nucleotides across photocyte membranes may not be great enough to raise the intracellular level of the cyclic nucleotide significantly because cyclic nucleotides do not readily cross cell membranes. Another problem might arise from the presence of Ca*+ in the saline. Aminophylline does not cause the light organ to glow in the absence of Ca2"1" so that cyclic nucleotides would also not be expected to affect the light organ without Ca2"1". However, the Ca2"1" in the saline might interfere with the action of cyclic nucleotides by forming com- plexes with the phosphate groups of the cyclic nucleotides (Daniel Morse, personal communication). A recent report by Robertson (1975) that the adult firefly light organ may contain octopamine raises the possibility that octopamine might be the transmitter contained in the nerve terminals of larval light organs. The light organ is very sensitive to octopamine (Carlson, 19686). Octopamine has been found in nervous systems of arthropods (Robertson & Steel, 1972, 1973, 1974; Wallace et al. 1974). Robertson & IJ IXB 65 226 DONATA OERTEL AND J. F. CASE Steel (1972, 1973, 1974) found that octopamine acts in an analogous way to epineJ phrine and norepinephrine, in that octopamine activates a phosphorylase in the cockroach nerve cord indirectly mediated by cyclic AMP. Robertson & Steel (1974) also speculate that octopamine may act as a transmitter in the cockroach. This might explain the result of an experiment in which we attempted histochemically to localize epinephrine and norepinephrine with potassium dichromate at acid and neutral pH (unpublished observation). The result was not positive for either epinephrine or norepinephrine. The biochemical mechanisms elucidated by McELroy and his colleagues (McElroy & De Luca, 1973) do not immediately suggest where a cyclic nucleotide might act to control the reaction.

We thank Dr J. Buck, Mrs E. Buck, Dr F. Hanson, Dr and Mrs K. Friedman for keeping us supplied with firefly larvae. We also thank Dr J. Nicholls for his valuable comments on the manuscript and Dr R. Eckert for advice on the electrophysiology. Mr W. Baumann kindly helped us with some of the illustrations. This work was supported by NSF Grant GB3339X, Office of Naval Research Contract N0014-75-C- 00242 and University of California Faculty Research Funds.

REFERENCES BERRIDGE, M. J. (1970). The role of 5-hydroxytryptamine and cyclic AMP in the control of fluid secretion by isolated salivary glands. J. exp. Biol. 53, 171-86. BERRIDOE, M. J. & PATEL, N. G. (1068). Insect salivary glands: stimulation of fluid secretion by 5- hydroiytryptamine and adenosine 3',5'-monophosphate (cyclic AMP). Science, N.Y. 163, 462-3. BOWIE, L. J.F HORAK, V. & DE LUCA, M. (1973). Synthesis of a new substrate analog of firefly luciferin. An active-site probe. Biochemistry, N. Y. 13, 1845-52. BOWIE, L. J., IRWIN, R., LOKEN, M., DE LUCA, M. & BRAND, L. (1973). Excited-state proton transfer and the mechanism of action of firefly luciferase. Biociiemxstry, N. Y. la, 1851—7. BUCK, J. B. & CASE, J. F. (1961). Control of flashing in fireflies. I. The lantern as a neuroeffector organ. Biol. Bull. mar. biol. Lab., Woods Hole iai, 234-56. BUCK, J. B., CASE, J. F. & HANSON, F. (1963). Control of flashing in fireflies. III. Peripheral excitation. Biol. Bull. mar. biol. Lab., Woods Hole 135, 251-69. BUTCHER, R. W. & SUTHERLAND, E. W. (1962). Adenosine 3',5'-phosphate in biological materials. I. Purification and properties of cyclic 3',s'-nucleotide phosphodiesterases and use of this enzyme to characterize adenosine 3',s'-phosphate in human urine. J'. biol. Chem. 337, 1044-50. CARLSON, A. D. (1968 a). Effect of adrenergic drugs on the lantern of the larval Photuris firefly. J. exp. Biol. 48, 381-7. CARLSON, A. D. (19686). Effect of drugs on luminescence in larval fireflies. J. exp. Biol. 49, 195-9. CASE, J. F. & BUCK, J. B. (1963). Control of flashing in fireflies II. Role of the central nervous system. Biol. Bull. mar. biol. Lab., Woods Hole 135, 234-50. CHAPPELL, R. & DOWLING, J. (1972). Neural organization of the median ocellus of the dragonfly. I. Intracellular electrical activity. J. gen. Pltysiol. 60, 121-47. DOWLINO, J. & CHAPPELL, R. (1972). Neural organization of the median ocellus of the dragonfly. II. Synaptic structure. J. gen. Physiol. 60, 148—65. FURCHGOTT, R. F. (1972). The classification of adrenoceptors (adrenergic receptors). An evaluation from the standpoint of receptor theory. In Catecholamines. Handb. exp. Pharmak. 33. GEORGE, W. J., POLSON, J. B., O'TOOLE, A. G. & GOLDBERG, N. D. (1970). Elevation of guanosine 3',5'-cyclic phosphate in rat heart after perfusion with acetylcholine. Proc. natn. Acad. Set. U.S.A. 66, 398-403- GOLDBERG, N. D., LUST, W. D., O'DEA, R. F., WEI, S. & O'TOOLE, A. G. (1970). The role of cyclic nucleotides in brain metabolism. Adv. in Biochem. Psychopharm. 3, 67-87. HESS, M. E. & HAUOAARD, N. (1958). The effect of epinephrine and aminophylline on the phosphorylase activity of perfused contracting heart muscle. J. Pharmac. exp. Ther. 133, 169-75. KATZ, B. (1966). Nerve, Muscle and Synapse. New York: McGraw-Hill. Neural excitation of the larval firefly photocyte 227 , J. W., Kuo, J.F. & GREENCARD, P. (1972). Determination of relative levels of cyclic AMP in tissues or cells prelabeled with radioactive adenine. Adv. Nuc. Res. 3, 131-8. MCELROY, W. D. & DE LUCA, M. (1973). Chemical and enzymatic mechanisms of firefly luminescence. In Chemiluminescence and Bioluminescence (ed. Cormier, Hercules and Lee.). New York: Plenum Press. MCLEAN, M., BUCK, J. B. & HANSON, F. (1972). Culture and larval behavior of Photurid fireflies. Am. Midi. Natur. 87, 133-45. MCMAHON, D. (1974). Chemical messengers in development. A hypothesis. Science, N.Y. 185, 1012- 21. O'DBA, R. F., HADDOX, M. K. & GOLDBERG, N. D. (1970). Kinetic analysis of a soluble rat brain cyclic nucleotide phosphodiesterase. Fedn Proc Fedn Am. Socs exp. Biol. 29, 473 abs. OERTEL, D., LINBERG, K. A. & CASE, J. F. (1975). Ultrastructure of the larval firefly light organ as related to control of light emission. Cell Tits. Res. 164, 27-44. PETERSON, M. K. (1970). The fine structure of the larval firefly light organ. J. Morph. 131, 103-16. PRINCE, W. T., BEHRIDOE, M. J. & RASMUSSEN, H. (1972). The role of calcium and cyclic AMP in the secretory response of the blowfly salivary gland to 5-hydroxytryptamine. Proc. natn. Acad. Set. U.S.A. 69. 553-7- RALL, T. W. & WEST, T. C. (1963). The potentiation of cardiac inotropic responses to norepinephrine by theophylline. J. Pharmac. exp. Ther. 139, 269-74. RASMUSSEN, H., GOODMAN, D. B. P. & TENENHOUSB, A. (1972). The role of cyclic AMP and calcium in cell activation. Critical review in Biochemistry Feb. 1972. ROBERTSON, H. A. (1975). Octopamine: presence in firefly lantern suggests transmitter role. Proc. Cent. Fed. Biol. Soc. 18, 66. ROBERTSON, H. A. & STBELB, J. E. (1972). Activation of insect nerve cord phosphorylase by octopamine and adenosine 3',5'-monophosphate. J. Neurochem. 19, 1603-6. ROBERTSON, H. A. & STEBLE, J. E. (1973). Effect of monophenolic amines on glycogen metabolism in the nerve-cord of the American cockroach, Periplaneta americana. Insect Biochem. 3, 53—9. ROBERTSON, H. A. & STEELB, J. E. (1974). Octopamine in the insect central nervous system: distribution, biosynthesis and possible physiological role. J. Physiol. Lond. 337, 34P—35P SMALLEY, K. N. (1965). Adrenergic transmission in the light organ of the firefly Photimu pyralis. Comp. Biocliem. Physiol. 16, 467-77. SMITH, D. S. (1963). The organization and innervation of the luminescent organ in a firefly, Photuris pennsylvanica (Coleoptera). J. Cell Biol. 16, 323-59. SUTHERLAND, E. W. & RALL, T. W. (1958). Fractionation and characterization of a cyclic adenine nucleotide formed by tissue particles. J. biol. Chem. 333, 1077-91. SUTHERLAND, E. W., ROBISON, G. A. & Butcher, R. W. (1968). Some aspects of the biological role of adenosine 3',5'-monophosphate (cyclic AMP). Circulation Res. 37, 279—306. WALLACE, B. G., TALAMO, B. R., Evans, P. D. & KRAVITZ, E. A. (1974). Octopamine: selective association with specific neurons in the lobster nervous system. Brain Res. 74, 349—55. WEIGHT, F. F., PETZOLD, G. & GREENOARD, P. (1974). Guanosine 3',5'-monophosphate in sympathetic ganglia: increase associated with synaptic transmission. Science, N.Y. 186, 942-4.

13-2