Functional physiology of lantern shark (Etmopterus spinax) luminescent pattern: differential hormonal regulation of luminous zones

The vast majority of deep-sea organisms rely on luminescence for numerous functions including predation avoidance, predation and intraspecific communication (Buck, 1978; Wilson and Hastings, 1998; Herring, 2007). The possession of photogenic organs (photophores) is a common feature in many of these organisms, including crustaceans, molluscs and fishes (Haddock et al., 2010). These photophores are divided into two types: extrinsic photophores, which shelter symbiotic luminous bacteria, and intrinsic photophores, which are self-luminescent, containing an endogenous luminescent system (Herring, 1982). To bring an adaptive advantage to their owner, photophores should be precisely controlled as the efficiency of luminescent behaviours depends on the timing and the physical characteristics of luminescence (Warner et al., 1979; Denton et al., 1985; Branham and Greenfield, 1996; Harper and Case, 1999; Demary et al., 2006; Rivers and Morin, 2008).

A variety of luminescence control mechanisms exist in fishes (for reviews, see Herring and Morin, 1978; Herring, 1982; Claes and Mallefet, 2009a), which are the only luminous vertebrates. Photogenic organs with bacterial symbionts in bony fishes are controlled mechanically via mobile dark shutters, rotation into a dark pocket or chromatophores, and/or physiologically, by controlling the physicochemical characteristics (osmolarity, glucose, O₂ availability, etc.) of the bacterial growth medium (Bertelsen, 1951; Herring and Morin, 1978; Herring, 1982; Haygood, 1993; Munk, 1999) (Fig. 1A). In contrast, intrinsic photophores are under neural control (mostly by adrenergic nerves) in bony fishes but appear to be under hormonal control in Chondrichthyes (Baguet, 1975; Baguet and Marechal, 1978; Baguet and Christophe, 1983; Claes and Mallefet, 2009a) (Fig. 1B,C). In addition to these ‘luminescence switches’, accessory optical structures (including reflectors, light guides, optical filters and lenses) are often present and allow a physical modulation of the light emitted by photophores (Herring and Morin, 1978; Munk and Bertelsen, 1980; Herring, 1982; Denton et al., 1985).

The velvet belly lantern shark, Etmopterus spinax (Linnaeus 1758), is a small deep-sea shark endowed with numerous photophores forming a complex luminous pattern organised in nine different types of luminous zones (Claes and Mallefet, 2008) (Fig. 2A,B). Recent studies suggest that this pattern is involved in camouflage by counterillumination as well as in intraspecific behaviours such as cooperative swimming/hunting and even sexual signalling (Claes and Mallefet, 2008; Claes and Mallefet, 2009b).

Luminescence from the ventral luminous zone of E. spinax is controlled by hormones involved in the physiological control mechanism of Elasmobranch (sharks and rays) skin pigmentation: melatonin (MT) and prolactin (PRL), which trigger light emission, as well as α-melanocyte stimulating hormone (α-MSH), the application of which prevents MT- and PRL-induced luminescence (Claes and Mallefet, 2009a). As the E. spinax luminescent pattern is believed to be involved in more than one function (Claes and Mallefet, 2008; Claes and Mallefet, 2009b), the question arises whether functionally different luminous zones present different control mechanisms.

Through MT and PRL application on the different luminous zones that produce the luminescent pattern of the velvet belly lantern shark, E. spinax, the specific goals of this study were (i) to determine whether all the luminous zones respond to hormonal application, and if this is the case then (ii) to show whether different hormonal luminescent responses are observable among the different luminous zones, (iii) to investigate a potential sexual dimorphism in the luminescent responses to hormones, which would be a clue for the existence of sexual signalling, (iv) to investigate the effect of photophore density (P_D) on the luminescence kinetic parameters, and (v) to discuss the results in relation to the functional ecology of this shark species.

Twelve adult female and eight adult male living specimens of the velvet belly lantern shark E. spinax (30–51 cm total length) were caught by longlines lowered in a deep (≥200 m) area of the Raunefjord (Norway) during two field sessions (February and June 2009), and transferred to Espeland Marine Station (Espegrend, Norway) where they were kept in two 1 m×1 m×1 m tanks placed in a dark cold room (6°C) until use. Captive animals were killed by a blow to the head before the start of experimentation, following the local rules for experimental fish care.

In this work, photophore preparations consisted of standard (0.55 cm diameter) circular skin patches dissected out from E. spinax specimens [following the method of Claes and Mallefet (Claes and Mallefet, 2009a)]. Since we wanted to test the effects of MT and PRL separately, we selected all the luminous zones that were large enough to give two separate but similar patches, which correspond to six out of the nine zones described previously (Claes and Mallefet, 2008) (Fig. 2B).

These skin patches were placed in a shark saline [292 mmol l⁻¹ NaCl, 3.2 mmol l⁻¹ KCl, 5 mmol l⁻¹ CaCl₂, 0.6 mmol l⁻¹ MgSO₄, 1.6 mmol l⁻¹ Na₂SO₄, 300 mmol l⁻¹ urea, 150 mmol l⁻¹ trimethylamine N-oxide, 10 mmol l⁻¹ glucose, 6 mmol l⁻¹ NaHCO₃; total osmolarity 1.080 mosmol l⁻¹; pH 7.7 (Bernal et al., 2005)]. For each shark, photophore densities present in the six investigated luminous zones was estimated using a binocular microscope (Leitz Diaplan, Oberkochen, Germany) by counting the number of photophores present in one of the two patches and by dividing this number by its surface area (0.2375 cm²). Skin patches were then stored in saline at 6°C until use (<2 h; maximum survival time of skin patches was more than 12 h).

The same procedure was applied for each luminous zone of the 20 E. spinax specimens. Skin patches were transferred to a 96-well microplate (each well containing 100 μl saline) from a luminometer (Berthold Orion, Pforzheim, Germany) calibrated with a standard 470 nm light source (Beta light, Saunders Technology, Hayes, UK). These patches were stimulated by a single application of either MT (Tocris Bioscience, Ellisville, MO, USA) or PRL (Sigma, St Louis, MO, USA) at a final concentration of 10⁻⁶ mol l⁻¹, which is the concentration eliciting the highest luminescent response in the ventral zone, i.e. the concentration to which the luminous tissue is most sensitive (Claes and Mallefet, 2009a). The experimental procedure was performed at room temperature (18°C).

The luminescent responses were recorded for 1 h after the stimulation and were characterised using different kinetic parameters (Fig. 2C): the maximum intensity of light emission (L_max, in megaquanta per second, Mq s⁻¹), the total quantity of light emitted during the experiment (L_tot, in teraquanta per hour, Tq h⁻¹) and the time to reach L_max from the stimulation time (TL_max, in min). Light parameters were standardised by skin surface area (in cm²). In order to compare the luminous response of six different luminous zones to the two hormones, a total of 12 skin patches were dissected out from each animal and stimulated simultaneously. Then, according to the number of experiments run at the same time in the luminometer, Berthold Simplicity software parameters were adjusted in order to obtain at least one point every minute.

Knowing that L_max and L_tot parameters of luminescent responses to MT and PRL vary according to the time of year (Claes and Mallefet, 2009a), and as our specimens were obtained at two different times (February and June), the results were standardised for each specimen by dividing the values obtained for these two parameters in the different luminous zones by the value of L_max and L_tot of the luminescent response from the ventral zone (which has the lowest P_D). In order to perform coherent comparisons, the P_D of the skin patches was also standardised using ventral density as the reference. No such standardisation was applied for TL_max, which was found to be constant (for both hormones) over the year (Claes and Mallefet, 2009a). Control Student's t-tests were performed between sexes for (absolute) luminescence parameters and P_D in the ventral luminous zone.

Two-way analyses of variance (ANOVA) were performed to determine the effect of sex and luminous zone on the three different luminescence parameters and the P_D. If a sexual difference was detected, each luminous zone was investigated for sexual differences using Student's t-test. In addition, a linear regression was performed to identify a potential link, across the different luminous zones, between the mean values of the light parameters of hormone-induced luminescence and the mean P_D.

Statistical analyses (ANOVA, Student's t-tests and linear regressions) were performed with the software SAS/STAT 1990 (SAS Institute Inc., Cary, NC, USA). Each mean value is expressed with its standard error (means ± s.e.m.) and N is the number of skin patches used for a specific treatment (which corresponds to the number of shark specimens tested as one skin patch from each individual was used in each treatment).

Control Student's t-tests showed that the ventral luminous zones of male and female E. spinax specimens used in this study are rather uniform in terms of their P_D and their MT-induced luminescence kinetic parameters (Table 1). Sexual differences were, however, observed for the luminescence induced by PRL; females showed higher L_max and L_tot than males (Table 1). This difference might reflect the fact that males and females were not tested at the same time of year (all females were tested in February while 6 out of the 8 males were tested in June), as seasonal variations in the L_max (and L_tot) of hormonally induced luminescence has already been detected in this species (Claes and Mallefet, 2009a). This seasonal variation in L_max (and L_tot) confirms the need, for these two light parameters, to use the standardisation detailed above in the following analyses. Two-way ANOVA revealed a significant difference in P_D among the different luminous zones but failed to show any effect of sex on this variable (Table 2).

Application of MT always evoked light emission in all the skin patches from the different luminous zones tested. Two-way ANOVA detected highly significant differences (P<0.001) among the light parameters of luminescence from these luminous zones (Table 2). These ANOVA failed, however, to detect any sexual effect on the MT-induced luminescence (Table 2).

Linear regression analysis revealed a significant relationship between P_D and the kinetic parameters L_max and L_tot of the MT-induced luminescence (Fig. 3A,B; Table 3). No relationship was found between P_D and TL_max (Fig. 3C; Table 3).

Application of PRL always evoked light emission in all the skin patches from the different luminous zones tested. Two-way ANOVA detected significant differences (P<0.05) among the luminescence kinetic parameters of these luminous zones as well as sexual differences in L_tot and TL_max (Table 2).

Student's t-tests confirmed sexual differences in L_tot in lateral (t_12,3=4.89, P=0.001), pelvic (t_11,3=3.90, P<0.01) and infra-pelvic (t_11,5=3.39, P<0.01) luminous zones, and sexual differences in TL_max in lateral (t_17,3=3.31, P<0.01) and pelvic (t_12,8=2.82, P<0.05) luminous zones. Linear regression analysis revealed a significant relationship between P_D and the kinetic parameters L_max and L_tot of the PRL-induced luminescence (Fig. 4A,B; Table 3). No relationship was found between P_D and TL_max (Fig. 4C; Table 3).

Standardised mean profiles of MT- and PRL-induced luminescence time courses from the different luminous zones investigated in this study (Fig. 5) clearly demonstrate that luminous zones respond differently to PRL and MT. While no difference was detected between male and female E. spinax specimens in response to MT, some zones (lateral, pelvic and infra-pelvic zones) showed sexual differences in response to PRL, with male sharks showing proportionally quicker light responses than females for these zones.

The luminescent pattern of the lantern shark E. spinax is complex and appears to be involved in different functions (Reif, 1985; Claes and Mallefet, 2008; Claes and Mallefet, 2009a; Claes and Mallefet, 2009b). As bioluminescent behaviours generally imply kinetic differences in light emission or even do not occur at the same time, we asked whether the different functional areas of E. spinax have similar or different luminescence control mechanisms. A recent study showed that luminescence from the ventral luminous zone of this shark may be triggered by MT and PRL (Claes and Mallefet, 2009a). If other luminous zones of this shark are controlled by hormones as well, another question arises: how does a signalling molecule that diffuses in the blood throughout all parts of the organism efficiently control a multifunctional luminescent pattern such as the one of E. spinax?

In this study we showed that all luminous zones investigated so far respond to application of either MT or PRL by glowing for variable time courses (all other zones of the luminous pattern also responded to both hormones but their small size prevented them from being tested quantitatively; J.M.C. and J.M., unpublished data). The observed difference in the light parameters L_max and L_tot of the different luminous zones certainly reflects the difference in P_D present in this zone. Not surprisingly, zones with higher photophore densities emitted more light than others after hormone application. This relationship is quite simple to understand: when MT- or PRL-containing blood reaches a specific luminous zone, all the photophores present in this zone are stimulated to glow. Direct observation of spontaneous luminescence from living specimens of E. spinax confirmed that brighter zones are the zones displaying the highest photophore densities (Claes and Mallefet, 2009b). A direct relationship was found between chemically induced luminescence intensity (via oxygen peroxide, H₂O₂) and P_D accross E. spinax's luminous zones (Claes and Mallefet, 2009b). In this case, however, the relationship mainly reflected the fact that the amount of luminous substrate available for the chemiluminescent reaction in a given luminous zone is linked to the number of photophores in a given surface area (Claes and Mallefet, 2009b), suggesting that each photophore might contain a similar amount of luminous substrate. However, as the variation in TL_max between the different luminous zones of E. spinax does not appear to be linked to a difference in P_D, the hormonal luminescence control present in this shark may act differently according to the part of the luminescent pattern that is stimulated. This heterogeneity in the timing of the light emission reflects a difference in the sensitivity of the luminous zones to both hormones, indicating that additional control mechanisms might be present. The observed sexual differences in response to PRL, which also do not seem to be linked to a difference in P_D, is another indication that the hormonal luminescence control mechanism of E. spinax is more complex than previously thought, and approaches in selectivity the neural luminescence control mechanism of bony fishes.

The luminous zones of E. spinax do not appear at the same time during embryogenesis (Claes and Mallefet, 2008) and present different surface area and P_D scaling patterns throughout ontogeny. These differences lead to the establishment of a complex heterogeneous luminescent pattern in mainly adult but also subadult individuals (Claes and Mallefet, 2009b). This luminous pattern heterogeneity appears to actually represent a trade-off between the continuously useful function of camouflage and periodic intraspecific functions such as cohesive swimming/hunting and sexual signalling which would only be useful in larger sharks (Reif, 1985; Claes and Mallefet, 2009b).

The results of this study (which focused on adult individuals of E. spinax) agree well with these hypotheses as they show that hormones controlling the luminescence from photophores of the E. spinax ventral luminous zone are able to not only trigger light in other luminous zones but also, when applied at the same concentration, make them luminesce proportionally to their P_D, thus conserving the heterogeneity of the luminescent pattern. Moreover, this study adds a new step in the complexity of the luminescent pattern as it shows that the kinetics of the luminescent response to PRL and MT also varies independently from the P_D of the zone. The rate of light emission is very different from one zone to another, thus reinforcing the idea of the multifunctionality of the luminous pattern. Similar versatility in the luminescence emission has been found in lanternfishes (myctophids) which feature a caudal organ that flashes rapidly and is believed to be used in intraspecific communication, and slow glowing ventral photophores involved in camouflage (Young et al., 1980; Herring, 2007). In addition, the data presented here bring, for the first time, evidence of a clear dimorphism in the response to PRL between males and females of E. spinax. Interestingly, this dimorphism is only apparent in the pelvic region, where the sexually dimorphic organs of sharks, i.e. male's claspers and the cloaca of the female, are present (Wourms, 1977). This dimorphism strongly supports the involvement of luminescence from this region in reproduction, as already suggested (Claes and Mallefet, 2009b). For example, the differences in L_tot and TL_max found in this study might reveal the presence of a luminescent courtship display between male and female E. spinax specimens. Such luminous sexual signals have already been documented/suggested in other marine animal species including polychaetes (Tsuji and Hill, 1983), crustaceans (Rivers and Morin, 2008), squids (Herring, 1988) and bony fishes (Crane, 1965; Sasaki et al., 2003; Herring, 2007). Since this sexual dimorphism is only observed in response to PRL, it should be more pronounced when luminescence response to PRL are higher, which appears to be in April, at the winter–spring transition (Claes and Mallefet, 2009a). In the Mediterranean, E. spinax is believed to mate in winter (when the proportion of active males is higher), with the female storing the sperm until fertilisation, which occurs in summer (Coehlo and Erzini, 2008). It can therefore be expected that similar specific timing may exist in Norwegian velvet belly lantern sharks with mating occurring at a specific time of year, and that possessing a more pronounced luminescent dimorphism at this period increases sexual signalling and reproductive success.

This work was supported by a grant from the F.N.R.S. to J.M.C. and another grant (FRFC 2.4516.01) to J.M. We would like to acknowledge A. Aadnesen, manager of Espeland Marine Biological station, T. Sørlie for skilful help during field collections, and the two anonymous reviewers for their comments that greatly improved the earlier version of the manuscript. J.M. is research associate of FSR-F.N.R.S. Contribution to the Biodiversity Research Centre (BDIV) and to the Centre Interuniversitaire de Biologie Marine (CIBIM).

 
• L_max

  maximum intensity of light emission

 
• L_tot

  total light emission

 
• MT

  melatonin

 
• P_D

  photophore density

 
• PRL

  prolactin

 
• TL_max

  time between stimulation and maximum intensity of light emission

 
• α-MSH

  α-melanocyte stimulating hormone