Oxygen Consumption and Luminescence of Maurolicus Photophores Stimulated by Potassium Cyanide

The metabolic inhibitor potassium cyanide (KCN) triggers glowing and enhances the oxygen consumption of isolated photophores of the epipelagic luminescent fish Porichthys (Mallefet & Baguet, 1984). To explain this paradoxical effect of KCN, it has been suggested that the increase in oxygen consumption during the light response corresponds to that involved in the luciferin-luciferase light reaction (Cormier, Crane & Nakano, 1967). The aim of the present work was to investigate, in isolated photophores of a mesopelagic fish, whether KCN can also evoke a light emission associated with a simultaneous increase in oxygen consumption. The results show that KCN evokes a light response but blocks the oxygen consumption of the light organ. It is suggested that oxygen has different roles in the control of light production in Maurolicus and Porichthys photophores.

Fresh specimens of Maurolicus muelleri were collected, as previously described (Baguet & Christophe, 1983), during 8 days between 05.00 and 06.00 h in the Strait of Messina and brought to the Istituto Talassografico where they were transferred to glass vessels in cooled sea water and stored at 7°C.

The two rows of abdominal photophores are partly fused, forming a common narrow central trunk (Fig. 1A). Each experiment was performed on a portion containing 12 photophores (Fig. IB). Portions were carefully excised under a binocular microscope using fine scissors and forceps (Dumont no. 5). The ends of the preparation were tied to a thin stainless-steel needle which was fixed to the lower part of the oxygen electrode by an O-ring (Fig. 1C), the preparation being about 3 mm from the plane of the cathode. The oxygen electrode and the photophores were immersed in a glass tube (50mm length; 14mm internal diameter) in a temperature-controlled chamber (20°C); the tube was filled with 4 ml saline and sealed to the electrode at the top with paraffin.

The light-emitting area of the preparation was 10 mm from the photomultiplier (PM 270C, International Light) the photosensitive area of which was placed on the transparent wall of the thermostatic chamber. The signal was displayed on a two-channel strip-chart recorder. The apparatus was calibrated using a tritium-irradiated phosphor (Betalight, Saunders Roe, Nuclear Enterprises, Ltd) emitting on an area of 2 mm², in the same location as the light organ.

The oxygen consumption of the preparation was estimated from the oxygen content of the surrounding saline during a given time period, using a polarographic oxygen electrode as previously described (Mallefet & Baguet, 1984). To measure the oxygen consumption of the preparation, the electric current of the electrode was reduced by 90% with a back-off system and the remaining signal amplified.

In each experiment, the oxygen electrode was calibrated in oxygen-free saline (prepared by adding sodium dithionite, Na₂S₂O₄, to reduce dissolved oxygen) and in fully aerated saline (20°C); the oxygen consumption of the electrode, without the light organ, was then measured for 10 min.

The preparation, attached to the electrode, was immersed in fully aerated saline to measure the oxygen consumption and the luminescence level for 10 min. After this, 0·4 ml KCN was injected into the saline, through a capillary fixed on the side of the electrode, to achieve dilutions of 5× 10⁻⁴, 5× 10⁻⁵ and 5× l0⁻⁶moll⁻¹. Oxygen consumption and luminescence were followed for 40 min. Afterwards, the preparation was blotted with a filter paper and weighed with an electrobalance (Metier H10, accuracy 10μg).

From the measurements in aerated saline and the known volume of saline containing the preparation, the rate of oxygen consumption was calculated in nmol min⁻¹. The results are expressed as mean values with standard errors (mean ± S.E.M.) and the number of preparations (N), together with the statistical significance difference between different mean values.

The resting respiration rate was measured on 50 preparations, the experiments being carried out from 2 to 13 h after fishing. The duration of storage of the specimens did not appear to affect the oxygen consumption of the isolated preparations. The mean oxygen consumption rate of preparations excised 2–5 h after capture of the fish (1·38 ± 0·08 nmol min⁻¹) did not differ significantly from the mean rates of preparations excised 5–9 h (1·24 ± 0·11 nmol min⁻¹) and 9–13 h (0·91 ±0·23 nmol min⁻¹) after capture (Table 1).

The mean value for 50 preparations was 1·25 ± 0·07 nmol O₂ min⁻¹ for an average mass of 10·24 ± 0·38 mg. Regression analysis of the resting oxygen consumption, plotted as a function of the mass of the preparations, shows a significant correlation of the two parameters (r=0·69; N=50) which is described by the equation O₂ = 0·126 ± 0·019P–0·037, where O₂ is the oxygen consumption in nmol min⁻¹ and P the fresh mass in mg (Fig. 2).

In the presence of 5× 10⁻⁶mol l⁻¹ KCN, two preparations produced a dim luminescence, commencing 303 ± 33 s after the application of cyanide and reaching a constant value of 44·87 ± 3·97Mquantas⁻¹ after 1554 ± 150s. The two glowing and seven non-glowing preparations showed similar time courses of oxygen consumption (Fig. 3A). The oxygen uptake decreased immediately after the application of cyanide. After 3min the mean respiration rate was 0·85 ± 0·28 nmol min⁻¹, which corresponds to about 70% of the previous resting level. Subsequently the oxygen uptake declined, at a slower rate, to reach a constant level after 27 min corresponding to about 50% of the previous resting level.

In 5× 10⁻⁵moll⁻¹ KCN, all the preparations (N = 9) produced a slow, low light emission, beginning 341·4±52·6s after the application of cyanide and reaching a plateau (49·73 ±21·18Mquantas⁻¹) after 1228±188·8s. The mean oxygen consumption rate seemed to decrease in two steps (Fig. 3B): a rapid fall of 40% after 3 min, followed by a slower decrease for the next 30 min, when the oxygen uptake reached a steady level corresponding to about 23% of the initial resting level.

The time course of the mean oxygen consumption and the mean production of light in 5× l0⁻⁴moll⁻¹ KCN are shown in Fig. 3C. All 24 preparations responded with a long-lasting luminescence, commencing 102·6 ± 13·9 s after addition of KCN and reaching a maximal value (826·8 ± 174·9 Mquantas⁻¹) after 1466·1 ± 95·2s. 16 min later, at the end of the experimental period, the light emission level dropped to 42% of the peak value. The oxygen uptake dropped dramatically on addition of KCN, before the preparation had begun to luminesce, and reached a mean value of 0·45 ± 0·08 nmol min⁻¹ after 3 min, which corresponded to 35 % of the initial resting level. Subsequently, there was a further, slower decrease: after 15 min oxygen consumption had decreased by 92% and it remained at this steady level for the next 25 min. It is noteworthy that most of the light was produced during this period, i.e. after the oxygen consumption had been virtually blocked (Fig. 3C). Although these results indicate that light production is independent of oxygen uptake, it seems that the ability of an isolated preparation to emit light is dependent on the intensity of its resting oxygen consumption (Fig. 4). The data in Fig. 4 show that light production is not significantly different from zero when there is no oxygen consumption; the slope, b, of the calculated regression line indicates that for 1 nmol oxygen consumption per minute at rest, a preparation produces 1350 Mquantas⁻¹ in response to 5× 10⁻⁴ moll⁻¹ KCN.

The sensitivity of isolated photophores of Maurolicus to adrenaline and noradrenaline, and the significant decrease of the light response in the presence of α-and β-adrenergic antagonists (Baguet & Christophe, 1983), suggest that they could be innervated by cathecholaminergic neurones. It could be that KCN causes release of a neural transmitter, thus inducing luminescence of the light organ. We tested this hypothesis by comparing the light emission evoked by KCN in preparations which had been pretreated with phentolamine (α-adrenergic antagonist) and propranolol (β-adrenergic antagonist).

To do this, two preparations were isolated from each fish and immersed either in normal saline or in saline containing phentolamine or propranolol (5× l0⁻⁴moll⁻¹) for 10min; KCN was then added, to achieve a concentration of 5× 10⁻⁴moll⁻¹. The mean differences from controls of the light emission responses were 49·7 ± 41·1 Mquantas⁻¹(N = 11) in phentolamine-saline and 31·3 ± 34·8Mquantas⁻¹ (N = 11) in propranolol-saline.

Thus neither phentolamine nor propranolol significantly affected the intensity of the light emitted by Maurolicus photophores in the presence of 5× 10⁻⁴ mol l⁻¹ KCN. Similarly, no change occurred in the time course of the oxygen consumption in the presence of either adrenergic antagonist.

Previous research (Baguet & Christophe, 1983) has shown that isolated photophores from freshly collected specimens remain chemically excitable and luminesce in the presence of adrenaline and noradrenaline. The present work shows that the abdominal photophores exhibit a stable resting respiratory rate even when stored for 13 h and are therefore in a good physiological condition. The resting respiratory rate of 1 g of isolated photophores of Maurolicus is 126·5 ± 19·1 nmol O₂min⁻¹ (20°C; N =50). This value is not significantly different from those of 87·0 ± 15·6nmol O₂g⁻¹ min⁻¹(N = 18) for the isolated ventral light organ of the meso-pelagic fish Argyropelecus hemigymnus (Mallefet & Baguet, 1985) and 102·6 ± 21·4 nmol O₂g⁻¹ min⁻¹(N= 12) for the isolated photophores of the mesopelagic fish Myctophum punctatum. However, the resting respiration rate (335 ± 24·4 nmol O₂g⁻¹ min⁻¹) of the isolated photophores of the epipelagic fish Porichthys is much higher (J. Mallefet & F. Baguet, unpublished observation).

According to Torres, Belman & Childress (1979) the oxygen consumption of mesopelagic fish is about two orders of magnitude less than that of epipelagic species ; most of the difference appears to be due to lower aerobic metabolism in the muscle tissues (Siebenaller & Yancey, 1984). We therefore suggest that the amount of metabolically active tissue is higher in Porichthys than in Maurolicus photophores. Differences exist, but they are opposite to what could be expected: the protein content is greater in Maurolicus (312 ± 84mgg⁻¹) than in Porichthys (112 ± 33mgg⁻¹) photophores (D. Sempoux-Thinès & F. Baguet, unpublished observations) and the photocyte mass represents about 35 % of the overall mass of the photophore in Maurolicus as compared with only 2·4–4·2% in Porichthys (LaRivière & Anctil, 1984). The present results confirm our suggestion (Mallefet & Baguet, 1985) that the lower rate of aerobic metabolism of photophores of a mesopelagic fish might be an inherent characteristic of these deep-living luminescent species.

Our results suggest that KCN has a dual effect on isolated photophores of Maurolicus; (i) it induces occasional light production at 5× l0⁻⁶moll⁻¹ and always induces luminescence at 5× 10^−s and 5× 10⁻⁴moll⁻¹, and (ii) it decreases the oxygen consumption of the photophores whether they produce light or not, the inhibitory effect increasing from 5× l0⁻⁶ to 5× 10⁻⁴moll⁻¹. When KCN induces light production in Maurolicus photophores it decreases the oxygen consumption prior to the production of light. As there is good evidence that cyanide blocks the mitochondrial respiration of fishes in the same way as in mammals (Gumbmann & Tappel, 1962) it is likely that the inhibitory effect on the Maurolicus photophores can be explained by an inhibition of the resting respiration rate of the tissues. As 5× 10⁻⁴mol l⁻¹ KCN evokes a large and long-lasting luminescence, our results emphasize that mitochondrial respiration is not essential for the process of light generation. Moreover, the decrease in oxygen consumption suggests the inhibition of a mechanism that must prevent spontaneous luminescence of the photophore. By analogy with the photophores of the epipelagic luminescent fish Porichthys, which also luminesce in the presence of 10⁻⁵—10⁻³ mol l⁻¹ KCN (Mallefet & Baguet, 1984), it might be that, by inhibiting cellular respiration, the mechanism which sequesters intracellular calcium would be inhibited by KCN, so that the increase in free Ca²⁺ could stimulate the luminescent system.

Since histological and anatomical studies on the photophores of Maurolicus have demonstrated the presence of nerves (Oshima, 1911), it might be argued that KCN does not act directly on the photocytes but indirectly on the neural processes of the photophore. The absence of any significant effect of adrenergic inhibitors on the luminescence induced by KCN rules out the possibility that this metabolic inhibitor triggers release of an adrenergic neuromediator.

The biochemical mechanism for the luminescence in Porichthys photophores consists of a luciferase-catalysed oxidation of luciferin by oxygen (Tsuji et al. 1975). It has been suggested that the increase in oxygen consumption associated with the light emission evoked by KCN originates from the oxidation of luciferin by molecular oxygen (Mallefet & Baguet, 1984).

Since the production of light by Maurolicus photophores is not associated with increased oxygen consumption, it is likely that oxygen plays different roles in the control of light generation in Maurolicus and Porichthys photophores. All known bioluminescent systems require oxygen to produce light in vitro (Hastings, 1983) ; a bioluminescent system such as coelenterazine (coelenterate luciferin) can be induced to produce light in vitro simply by the addition of calcium, oxygen apparently not being required (Shimomura, Johnson & Saiga, 1962). However, Hastings & Morin (1969) have shown that coelenterazine is precharged with oxygen to form a stable peroxide. In this case, it would be expected that no oxygen would be consumed during light emission, since it is stored as a preoxygenated intermediate. Although we have not examined the biochemical nature of the bioluminescent system in Maurolicus photophores, it seems likely that some preoxygenated substance could also be involved in luminescence.

This work was made possible by grants from the Fonds National de la Recherche Scientifique of Belgium to JM, from the Fonds Léopold III and from the Belgian Foundation of Bioluminescence. We wish to acknowledge the help of Professor A. Cavalière, Director of the Istituto Sperimentale Talassografico who provided us with the necessary information and bench space.