Regulation of red fluorescent light emission in a cryptic marine fish

Abstract

Introduction: Animal colouration is a trade-off between being seen by intended, intra- or inter-specific receivers while not being seen by the unintended. Many fishes solve this problem by adaptive colouration. Here, we investigate whether this also holds for fluorescent pigments. In those aquatic environments in which the ambient light is dominated by bluish light, red fluorescence can generate high-contrast signals. The marine, cryptic fish Tripterygion delaisi inhabits such environments and has a bright red-fluorescent iris that can be rapidly up- and down-regulated. Here, we described the physiological and cellular mechanism of this phenomenon using a neurostimulation treatment with KCl and histology.

Results: KCl-treatment revealed that eye fluorescence regulation is achieved through dispersal and aggregation of black-pigmented melanosomes within melanophores. Histology showed that globular, fluorescent iridophores on the anterior side of the iris are grouped and each group is encased by finger-like extensions of a single posterior melanophore. Together they form a so-called chromatophore unit. By dispersal and aggregation of melanosomes into and out of the peripheral membranous extensions of the melanophore, the fluorescent iridophores are covered or revealed on the anterior (outside) of the iris.

Conclusion: T. delaisi possesses a well-developed mechanism to control the fluorescent emission from its eyes, which may be advantageous given its cryptic lifestyle. This is the first time chromatophore units are found to control fluorescent emission in marine teleost fishes. We expect other fluorescent fish species to use similar mechanisms in the iris or elsewhere in the body. In contrast to a previously described mechanism based on dendritic fluorescent chromatophores, chromatophore units control fluorescent emission through the cooperation between two chromatophore types: an emitting and an occluding type. The discovery of a second mechanism for fluorescence modulation strengthens our view that fluorescence is a relevant and adaptive component of fish colouration.

Figure 1

Red fluorescence in the iris of Trypterygion delaisi. Illustration of fluorescence in Tipterygion delaisiA. Individual showing red fluorescent eye-ring (= iris) under natural conditions (at -23 m, upside down under shady overhang, STARESO, Corsica, using a Nikon D700 DSLR, image taken without flash or filter and manually adjusted white balance). B and C are two pairs of images, each of which was taken with a 5 s delay showing rapid changes in the brightness of iris fluorescence in the laboratory using blue illumination and a red filter for viewing.

Figure 2

Maximum-normalised excitation and emission spectra in T. delaisi (Y-axis in arbitrary units, ranging from minimum to maximum). Fluorescence peaks at 600 nm (red line). Excitation is most efficient between 500 and 570 nm (green line). The excitation spectrum overlaps strongly with the ambient light spectrum (taken at noon on a sunny day in July at STARESO, Calvi, Corsica, in -10 m depth (black, thin line). Excitation and emission were measured using a spectrofluorometer (QuantaMaster QM-40, Photomed, Germany).

Figure 3

Change in fluorescence intensity after application of KCl. A. Change in percentage of fluorescing iris area over a 450 s period (one measurement per 5 s) averaged for 25 s time intervals for 8 eyes from 8 different individuals. KCl was added at time point 0 s. B. Change in iris fluorescence brightness over time in 6 eyes from 3 individuals expressed as total photon radiance (photons.s^-1.sr^-1.m^-2) in the emission range 550–700 nm at each time point. Colours indicate individuals.

Figure 4

Iris fluorescence brightness expressed as photon radiance, photons.s^-1.sr^-1.m^-2.nm^-1) plotted as a function of wavelength in time steps (see colour legend) starting with the application of KCl at time = 0 s. Data were averaged for 6 eyes from 3 individuals. A. Absolute spectra. B. Area-normalized spectra (each value of a spectrum divided by the sum of all values of that spectrum).

Figure 5

Change in fluorescence spectrum after application of KCl. Micrographs of a section through the iris under bright field (A), bright field and fluorescence (B), fluorescence (C) and polarized light (D), showing that the fluorescence emanates from the stratum argenteum, a layer of iridophores with guanine crystals (see D compared to B and C) (labelling in accordance with [37], p. 47). This layer is invaded by finger-like extensions of melanophores at the posterior side of the stratum argenteum (dark). The anterior (outward facing) side is at the top.

Figure 6

Guanine crystal platelets from the iris of T. delaisi under light microscopy. A: phase contrast. B: fluorescence. Note the weak blue-green fluorescence of the regular guanine crystal at the bottom right, which lacks red fluorescent pigment present in the others. Shapes vary a lot in these crystals, but do not seem to differ systematically between fluorescent or non-fluorescent forms.

Figure 7

Sections of T. delaisi irides, showing chromatophore location. Micrographs of the iris with melanophores before (A) and after (B) KCl treatment. Before KCl treatment, melanosome projections extend over the top of the fluorescent structures and suppress fluorescent emission. After KCl treatment (B) they are retracted and facilitate fluorescent emission. C: SEM picture of a fractured, untreated iris with similar layering, showing the contents of the chromatophores: Guanine crystal platelets from iridophores, and melanosomes (black pigment vesicles) from melanophores.

Figure 8

3D-Model of a reconstructed fluorescent chromatophore unit with one melanophore (black) embracing four fluorescent globular iridophores (red, two shown, two others omitted). Sample taken at an intermediate state with not yet fully aggregated melanosomes.