Human melanopsin forms a pigment maximally sensitive to blue light (λmax ≈ 479 nm) supporting activation of G(q/11) and G(i/o) signalling cascades

Abstract

A subset of mammalian retinal ganglion cells expresses an opsin photopigment (melanopsin, Opn4) and is intrinsically photosensitive. The human retina contains melanopsin, but the literature lacks a direct investigation of its spectral sensitivity or G-protein selectivity. Here, we address this deficit by studying physiological responses driven by human melanopsin under heterologous expression in HEK293 cells. Luminescent reporters for common second messenger systems revealed that light induces a high amplitude increase in intracellular calcium and a modest reduction in cAMP in cells expressing human melanopsin, implying that this pigment is able to drive responses via both Gq and Gi/o class G-proteins. Melanopsins from mouse and amphioxus had a similar profile of G-protein coupling in HEK293 cells, but chicken Opn4m and Opn4x pigments exhibited some Gs activity in addition to a strong Gq/11 response. An action spectrum for the calcium response in cells expressing human melanopsin had the predicted form for an opsin : vitamin A1 pigment and peaked at 479 nm. The G-protein selectivity and spectral sensitivity of human melanopsin is similar to that previously described for rodents, supporting the utility of such laboratory animals for developing methods of manipulating this system using light or pharmacological agents.

Figure 1.

Human melanopsin in HEK293 cells. (a) Western blot of opsin protein expressed and extracted from HEK293 cells. Blots were stripped and re-probed with an antibody against total ERK protein to indicate protein amount. Predicted sizes of the opsin proteins as calculated with ExPASy.com are provided at foot of blot. (b) Immunocytochemistry photomicrograph showing detection of human melanopsin expressed in HEK293 cells and labelled with a 1D4 antibody (green). Cells are also stained with DAPI, a blue nuclear stain. Melanopsin can be seen mostly localized at the cell membrane and in cytoplasmic inclusions. Scale bar = 6 µm. (c) A luminescent biosensor was used to monitor changes in cAMP production. The camera flash (yellow arrow) induced a large increase in cAMP-dependent luminescence indicative of G_s activity in cells expressing JellyOp but not those expressing human Opn4. (d) In this assay, there was no difference in the change in luminescence following light exposure in cells expressing human Opn4 and no opsin controls (t-test p > 0.05; n ≥ 3). (e) Reductions in cAMP as a result of G_i/o activity are measured by artificially raising cAMP with forskolin (black arrow) before exposing cells to light (yellow arrow). Human rod opsin (Rh1) induces a prolonged reduction in cAMP production. Luminescence was normalized to values immediately following light exposure (n ≥ 6). (f) Comparison of the reduction of forskolin-induced cAMP production 1 min following light in human melanopsin or Rh1 expressing cells treated or untreated with 100 ng ml⁻¹ pertussis toxin (ptx). A pairwise comparison shows the light-dependent reduction is significantly reduced when treated with ptx, a G_i/o-specific inhibitor (paired t-test; **p < 0.01 both opsins). The inhibition appears greater in Rh1 expressing cells. (g) Aequorin luminescence in HEK293 cells expressing opsins as labelled. A camera flash of white light was applied (yellow arrow) and aequorin responses in a 96 well plate reader were monitored. n ≥ 4 except human rod opsin (Rh1), n = 2. (h) There was a significant difference in peak of luminescence immediately following light exposure between no opsin controls and human melanopsin (Opn4; t-test ***p < 0.001). All plots show mean ± s.e.m.

Figure 2.

G-protein coupling of melanopsins from other species. G_q/11 assays showed robust calcium-dependent increases in aequorin luminescence from mouse and amphioxus Opn4 (a) and chicken Opn4m and Opn4x (b). (c) Amphioxus and mouse Opn4 showed a small inhibition of forskolin (black arrow)-induced luminescence following light exposure (grey arrow) indicative of a minor G_i/o response. (d) Conversely, both chicken Opn4x and Opn4m cause an increase in cAMP reporter luminescence following forskolin and light exposure. (e) This light-dependent increase in cAMP production (indicative of G_s coupling) is also measurable when no forskolin is first added. All plots show mean ± s.e.m.; n = ≥ 3. (f) The maximum difference in cAMP reporter luminescence before and after light exposure is significantly greater in Opn4m but not Opn4x expressing cells compared with when no opsin is expressed (Mann Whitney test *p < 0.05). Mean ± s.e.m., n = 4.

Figure 3.

Spectral sensitivity of melanopsin. Action spectra for (a) mouse and (b) human melanopsins were defined by plotting the relative sensitivity at each wavelength (mean ± s.e.m. of the three replicate irradiance response curves). The action spectra were fit with the predicted absorbance spectrum of opsin : vitamin A-based photopigments with λ_max between 450 and 550 nm, and the best fit assessed by minimizing sums of squares. Mouse melanopsin best matches a template with λ_max 484 nm, and human melanospin λ_max 479 nm. Sum of squares for comparisons between our data and opsin templates with λ_max between 470 and 490 nm are shown as an indication of the confidence of these estimates for (c) mouse and (d) human melanopsin.