Determination of the architecture of an invertebrate photoreceptor protein, squid rhodopsin, is a notable event. It illuminates the mechanism of invertebrate vision and a ubiquitous intracellular signalling system.
Many invertebrates have excellent visual systems1. Squid, for example, are formidable hunters that rely on their acute visual abilities to catch their prey. As in vertebrates, the properties of the photoreceptor protein rhodopsin contribute significantly to those abilities. So one reason for the attention that will be paid to the paper on page 363 of this issue2, in which Murakami and Kouyama present a high-resolution crystal structure of rhodopsin from the squid retina, is that it will help in understanding invertebrate vision. But the paper's significance extends far beyond that.
Rhodopsin is located in the cell membrane of photoreceptor cells. When activated by light, it undergoes a conformational change that triggers the action of a heterotrimeric GTP-binding protein (G protein) lying just beneath the cell membrane. Rhodopsin, therefore, is a member of the superfamily of G-protein-coupled receptors (GPCRs), all of which contain seven structural domains that each span the membrane. However, invertebrate rhodopsin signals through the α-subunit of a Gq type of G protein (rather than transducin, the α-subunit for vertebrate rhodopsin), leading to activation of phospholipase C and eventually the opening of a calcium channel (rather than activation of cyclic GMP phosphodiesterase, leading eventually to closing of a cation channel, which occurs in vertebrates3).
In other words, Murakami and Kouyama2 are the first to determine the structure of a Gq-coupled GPCR. The wider significance of the paper is that many hormone and neurotransmitter receptors signal through Gq, including vasopressin and oxytocin receptors as well as serotonin and acetylcholine receptors in the brain. And many drugs, such as antihistamines and angiotensin antagonists, target Gq-coupled receptors.
The functional unit of squid rhodopsin consists of a light-sensitive chromophore, retinal, covalently bound to an evolutionarily conserved lysine amino-acid residue in helix VII of the opsin protein. This covalent bond between a carbon atom from the aldehyde group of retinal and a nitrogen atom from the amino group of the side chain of lysine is called a Schiff base. The protonated nitrogen atom of this bond is thought to require a negatively charged 'counterion' for it to be neutralized.
Activation of rhodopsin works in the following way. Retinal has a double bond in its carbon chain that is isomerized upon absorption of light. This light-induced isomerization of the carbon chain from a kinked to a straightened form leads to a conformational change of the entire protein, triggering exchange of GDP for GTP in the Gq α-subunit. The GTP-bound form then activates phospholipase C, leading to the formation of inositol phosphates, membrane-bound diacylglycerol and polyunsaturated fatty acids3. Further events, which are not completely understood, lead to the opening of a calcium channel and the generation of nerve impulses that carry the relevant visual information to the brain.
Invertebrate rhodopsins are larger than vertebrate rhodopsins as they have an extension in the carboxy-terminal part of the protein that lies on the cytoplasmic side of the membrane and may contribute to photoreceptor organization in the retina4. This extension does not affect G-protein activation, so Murakami and Kouyama2 removed part of it to obtain three-dimensional crystals of the transmembrane domains. This transmembrane core was extracted from isolated photoreceptor membranes and crystallized5, the high-resolution structure being determined by X-ray crystallography. From the arrangement of the light-absorbing chromophore within the crystal packing, the authors propose a detection mechanism for linear polarized light on the principle that it is similar to the in vivo arrangement of the chromophore in the squid retina.
The core of squid rhodopsin consists of two cytoplasmic helices as well as seven transmembrane domains (Fig. 1). The crystal structure2 shows that the arrangement of helices I to VIII is similar to that of the rhodopsins isolated from frogs and cattle, so the overall three-dimensional structure of the protein is very similar in vertebrates and invertebrates. But helices V and VI of squid rhodopsin protrude 25 Å farther into the cytoplasm from the membrane surface, with an additional helix IX contributing to the cytoplasmic part of rhodopsin. The extended helices V and VI, together with helix IX, explain the additional electron density on the cytoplasmic surface that was found in an electron-density map of two-dimensional crystals of squid rhodopsin4. However, the architecture of this helical dome-like structure is entirely new and could not have been predicted.
Murakami and Kouyama's structure reveals that the amino-acid residues in contact with retinal, and the orientation of retinal within the protein, are different from those in vertebrate rhodopsins. Retinal is in a less distorted configuration in squid rhodopsin. The binding pocket for it is particularly interesting, because the side chain of a glutamate residue proposed previously to provide the counterion to the positively charged nitrogen atom that provides the covalent bond to retinal is not close enough to the nitrogen atom to act in this way.
This finding has implications for the tuning of the light-absorption maximum and the mechanism of re-isomerization that returns rhodopsin to its original conformation. In this conformation, which rhodopsin assumes in the dark, the side chain of a tyrosine residue that replaces the glutamate counterion of bovine rhodopsin is close to the positively charged nitrogen on rhodopsin's lysine side chain but remains uncharged during photoactivation. This could partly explain why the covalently bound retinal is not released from the lysine side chain after photoisomerization. As a consequence, retinal may be isomerized back to its original form within the photoreceptor to regenerate the ground state and reset the sensory system.
Several other features of the structure2 merit comment. For example, an interhelical cavity containing several ordered water molecules occurs in a similar position in squid and bovine rhodopsin, as well as in vertebrate β-adrenergic receptors6, which also belong to the GPCR family. This water cluster may contribute to the activation mechanism by affecting the transmission of the conformational changes triggered by retinal isomerization to the G-protein-binding cytoplasmic surface7.
The arrangement of the helices, as well as the charged side chains of helices VI and IX on the cytoplasmic side, may provide a binding surface for the carboxy terminus of the Gq α-subunit. In addition, the extended helices V and VI indicate that Gq might bind next to these helices and not on top of the helical dome structure. This is a novel constraint for models of G-protein-receptor complexes in general, given the expected structural conservation of the activation mechanism. The unexpected architecture of this cytoplasmic surface will attract much attention. The arrangement of the helices in this domain could be used to model the receptor interaction with G proteins, as well as for designing drugs that interfere with G-protein binding.
Finally, Murakami and Kouyama's structure2 will be particularly helpful for those investigating the action of human melanopsin, which — on the basis of sequence similarity — is more closely related to invertebrate than to vertebrate rhodopsin. Melanopsin is the photoreceptor that is required for setting our biological clock and for the pupillary reflex8,9, and so the structure of squid rhodopsin, in providing the best template to date for understanding this protein, is of considerable interest.
Hardie, R. C. & Postma, M. in The Senses: A Comprehensive Reference Vol. 1 (ed. Basbaum, A. I.) 77–130 (Academic, San Diego, 2008).
Murakami, M. & Kouyama, T. Nature 453, 363–367 (2008).
Hardie, R. C. & Raghu, P. Nature 413, 186–193 (2001).
Davies, A. et al. J. Mol. Biol. 314, 455–463 (2001).
Murakami, M., Kitahara, R., Gotoh, T. & Kouyama, T. Acta Crystallogr. F 63, 475–479 (2007).
Kobilka, B. & Schertler, G. F. Trends Pharmacol. Sci. 29, 79–83 (2008).
Li, J. et al. J. Mol. Biol. 343, 1409–1438 (2004).
Melyan, Z., Tarttelin, E. E., Bellingham, J., Lucas, R. J. & Hankins, M. W. Nature 433, 741–745 (2005).
Hankins, M. W., Peirson, S. N. & Foster, R. G. Trends Neurosci. 31, 27–36 (2008).
Gebhard F. X. Schertler is a shareholder of, and scientific adviser to, Heptares Therapeutics, 1-3 Burtonhole Lane, Mill Hill, London NW7 1AD, UK (http://home2.btconnect.com/heptares/). This company is involved in research on G-protein-coupled receptors and drug design.
About this article
Biophysical Journal (2019)
Journal of Molecular Endocrinology (2016)
Intermolecularly-induced conformational disorder in ferrocene, 1-bromoferrocene and 1,1′-dibromoferrocene
Journal of Molecular Structure (2014)
Binding Specificity of Retinal Analogs to Photoactivated Visual Pigments Suggest Mechanism for Fine-Tuning GPCR-Ligand Interactions
Chemistry & Biology (2014)
Endocrine Reviews (2013)