Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
Of the four classes of vertebrate cone visual pigments, the shortwave-sensitive SWS1 class shows some of the largest shifts in λmax, with values ranging in different species from 390–435 nm in the violet region of the spectrum to <360 nm in the ultraviolet. Phylogenetic evidence indicates that the ancestral pigment most probably had a λmax in the UV and that shifts between violet and UV have occurred many times during evolution. In violet-sensitive (VS) pigments, the Schiff base is protonated whereas in UV-sensitive (UVS) pigments, it is almost certainly unprotonated. The generation of VS pigments in amphibia, birds and mammals from ancestral UVS pigments must involve therefore the stabilization of protonation. Similarly, stabilization must be lost in the evolution of avian UVS pigments from a VS ancestral pigment. The key residues in the opsin protein for these shifts are at sites 86 and 90, both adjacent to the Schiff base and the counterion at Glu113. In this review, the various molecular mechanisms for the UV and violet shifts in the different vertebrate groups are presented and the changes in the opsin protein that are responsible for the spectral shifts are discussed in the context of the structural model of bovine rhodopsin.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Visual pigments belong to the large superfamily of G protein coupled receptors that function through the activation of a heterotrimeric guanine nucleotide binding protein, the G protein. Visual pigments have a common basic structure of an opsin protein attached to a chromophore via a Schiff base linkage and a conserved lysine residue. The opsin protein in vertebrates consists of a single polypeptide chain of 340–370 amino acids that forms seven α-helical transmembrane (TM) regions connected by cytoplasmic and luminal loops (1). In the tertiary structure, the seven TM regions form a bundle within the membrane creating a cavity towards the luminal side for the chromophore (2).
Each pigment shows a characteristic peak of maximal absorbance (λmax), the precise location of this peak depending on the interactions between the chromophore and the opsin protein. In vertebrates, the chromophore is either 11-cis-retinal or 11-cis-3,4-dehydroretinal, the derivatives of vitamins A1 and A2, respectively, to give either rhodopsin or porphyropsin pigments (3). Both pigment types may be present in certain fish, amphibians, and reptiles whereas the pigments in birds and mammals are entirely rhodopsins. The effect of 11-cis-3,4-dehydroretinal is to longwave (LW) shift the λmax of the pigment, and this is more pronounced at longer wavelengths (4).
In vertebrates, a single rod class of visual pigment is generally present with up to four cone classes as follows: a longwave or LWS class with a rhodopsin λmax of 505–570 nm, a middlewave or MWS class with a rhodopsin λmax of 480–520 nm, and two shortwave or SWS classes, with rhodopsin λmax values of 355–435 nm for SWS1 and 410–470 nm for SWS2. The SWS1 class shows some of the largest naturally occurring variations in λmax, ranging for example in mammals from around 420 nm (violet-sensitive, VS) in primates (5) to less than 370 nm (UV-sensitive, UVS) in the mouse (6), and in birds from around 420 nm in chicken (7) and mallard duck (8) to around 360 nm in canary (9) and budgerigar (10). UVS pigments are also present in amphibia and reptiles and in many freshwater and marine fish species, with examples distributed across many of the major teleost Orders such as the cyprinids (11–13), the beloniforms (14), the perciforms (cichlids) (15), the salmonids (16), and many species of marine reef fish (17).
Spectral-sensitivity of the ancestral SWS1 pigment
The lampreys and hagfishes are members of the agnathous or jawless vertebrates that separated from the main gnathostomatous or jawed vertebrates in the Cambrian period (18) around 540 MYA (Fig. 1). Visual pigment gene sequences have been reported for a number of lamprey species (19–22) but uniquely, only the Southern hemisphere species, Geotria australis, would appear to possess multiple visual pigment genes (23). Amongst these is an orthologue of the SWS1 opsin gene which generates a UVS pigment with a λmax around 360 nm (Fig. 2) when expressed as a recombinant protein in mammalian cells and regenerated in vitro with 11-cis-retinal (J. A. Cowing and W. L. Davies, unpublished data). Teleost fish are members of the gnathostomatomes that separated from the Chondrichthiomorphi (chimeras, rays and sharks) at the base of the gnathostomatous lineage: in all teleost fish where an SWS1 pigment has been shown to be present, it is invariably UVS (Fig. 2). The presence of UVS pigments in extant members of both the jawed and jawless classes indicates therefore that the ancestral vertebrate pigment was UVS (24,25), and this is supported by the UV-sensitivity of an engineered SWS1 pigment based on the inferred sequence of different vertebrate ancestors (26).
UV-sensitive pigments are also found in all the other major vertebrate groups (6,10,12,14,15,27–32), with VS pigments present in some species of amphibia (33), birds (34) and mammals (35), but not in reptiles. The shift from UVS to VS has occurred therefore many times in vertebrate evolution.
Evolution of VS pigments
Alignment of UVS and VS pigments from a wide range of vertebrate species shows a clear association of the residue at site 86 with spectral-sensitivity (36); in all non-avian UVS pigments, this site is occupied by Phe whereas it is replaced in VS pigments by a number of different residues. In the artiodactyls, the cow (Bos taurus) and pig (Sus scrofa), Tyr86 is present (Table 1) and its role in the violet shift has been confirmed by replacement with Phe by site-directed mutagenesis of the bovine pigment, followed by in vitro expression (36). This SW-shifts the λmax to 360 nm, and the converse substitution of Tyr86 into the UVS pigment of the mouse (Mus musculus) and goldfish (Carassius auratus) LW-shifts the λmax to 424 and 413 nm, respectively, in the violet region of the spectrum (24,36,37).
Table 1. Residues at sites 86, 90 and 93 in mammalian UVS and VS pigments.
Closely related to the Artiodactyla are the Cetacea, the whales and dolphins. Surprisingly, however, the SWS1 opsin gene present in these species has become inactive, with numerous deleterious mutations; whales and dolphins possess only a single LWS cone pigment and therefore lack color vision (38). There is, however, sufficient genomic sequence from the nonfunctional SWS1 pseudogenes of the bottlenose dolphin, Tursiops truncatus (AF055458), Blainsville’s beaked whale, Mesoplodon densirostris (AY228441), and the long-finned pilot whale, Globicephala melas (AY228442), to establish that Tyr would have been encoded at position 86 if the gene was functional, identical therefore to the cow and pig SWS1 opsins (Table 1). Unusually, however, the humpback whale, Megaptera novaeangliae, has Leu at site 90; with the exception of avian UVS pigments (see below), all other functional SWS1 pigments have Ser90, and hence it is possible that the Leu90 substitution occurred after gene inactivation and the relaxation of selection pressure on gene sequence.
Gene sequence data on the SWS1 pigments of carnivores is limited to the aquatic pinnipeds. Sequencing of the gene in the harp seal, Phoca groenlandica, and the harbor seal, Phoca vitulina, revealed a two-nucleotide deletion in exon 4 which generates a premature stop codon in the former species (38). The gene in the latter species was, however, intact but RT-PCR with retinal cDNA failed to identify a gene transcript (38). The genes in both species would appear therefore to be nonfunctional, with obvious parallels to the situation in cetaceans. Alignment of the pseudogene sequences does, however, identify codons 86, 90 and 93 and it would seem significant that in both cases, Tyr86 is present. What is unusual is that Val90 is present. As discussed above for the humpback whale, this substitution may also have occurred after gene inactivation.
The Mesaxonia are related to the Cetacea and Artiodactyla and include elephants (Proboscidea) and manatees (Sirenia). In both the African (Loxodonta africana) and Asian (Elephas maximus) elephant, the SWS1 opsin possesses Ser86; the corresponding pigment, when expressed in vitro, generates a pigment with a λmax at 419 nm (39). Moreover, replacement of Ser86 with Phe SW-shifts the λmax to 367 nm, confirming therefore the role of Ser86 in the violetshift in these species. The SWS1 pigment in the West Indian manatee, Trichechus manatus, is also VS with a λmax at 410 nm but possesses Tyr86 (40), identical therefore to the cow, pig and mutated cetacean pigments. This implies that the acquisition of Ser86 in elephants occurred within the Proboscidea lineage following a prior Phe86Tyr substitution that occurred in the common ancestor to the Artiodactyla and Mesaxonia.
The order Rodentia comprises two suborders, the Sciurognathi and the Hystricognathi. The mouse and rat, members of the Sciurognathi, both have UVS pigments as does the hystricognathous caviomorph rodent, the Chilean degu (Octodon degus) (41). VS pigments are, however, found in the guinea pig (Cavia porcellus), a South American member of the Hystricognathi (42), and in the gray (Sciurus carolinensis) and ground squirrel (Spermophilus spp.), members of the Sciurognathi (43–45). The UVS pigments in the mouse and rat have both retained Phe86 with the LW-shift in the gray squirrel achieved via a Phe86Tyr substitution (46). Tree and ground squirrels are strictly diurnal whereas the closely related flying squirrels are nocturnal. If the evolution of a VS pigment is an adaptation to diurnality, this might not have occurred in flying squirrels. Examination of the SWS1 gene in two flying squirrel species, Pteromys volans and Glaucomys sabrinus, revealed, however, that both have Tyr86 (46) and would therefore be VS. The gene has, however, accumulated mutations in both species such that it is now nonfunctional. Like many of the aquatic mammals, the flying squirrel has disposed of a SWS1 pigment and thereby color vision, retaining only a single LWS cone pigment (47).
The VS pigment in the guinea pig, a species from the other branch of the Rodentia, is unusual in having Val86 (48), and replacement with Phe is sufficient to SW-shift the pigment into the UV. The reverse mutation of Phe86Val into goldfish UVS pigment does not, however, LW-shift the λmax, indicating that Val86 requires other changes to LW-shift a UVS pigment. This does not necessarily mean that the original evolutionary event also required multiple substitutions into the ancestral SWS1 opsin. Rather, it implies that goldfish UVS pigment is not a perfect analogue of the ancestral pigment.
Australian marsupials are divided into two major Orders, the Diprotodontia and the Polyprotodontia. Amongst species from the former Order, the Tamar wallaby (Macropus eugenii) and quokka (Setonix brachyurus) have VS SWS1 pigments with Tyr86 present (49,50), whereas microspectrophotometry indicates that the honey possum has a UVS pigment (51), with Phe86 retained (J.A. Cowing, unpublished data). The presence of a UVS pigment with a λmax at 360 nm has been confirmed in the fat-tailed dunnart, a member of the Polyprotodontia, by in vitro expression (J. A. Cowing, unpublished data). Phe86 is retained in this latter species and also in another polyprotodont, the quenda (Isoodon obesulus) (50). The Phe86Tyr substitution seen in the Tamar wallaby and quokka must have arisen therefore within the diprotodont marsupial lineage. This represents convergent evolution with the identical substitutions in the Rodentia, Artiodactyla, Cetacea and Mesaxonia.
The platyrrhine (New World) and catarrhine (Old World) primates, the simians, possess VS SWS1 pigments (5,52) and the tuning of the human pigment has been examined in detail (53). A chimeric opsin that combined TM 1-3 from human VS and TM 4-7 from mouse UVS generated a pigment with a λmax very close to the native human pigment. TM1-3 which includes site 86 is important therefore for violet spectral shifts in primates as in other vertebrates. Moreover, the simultaneous replacement of the residues present at seven sites (Phe46Thr, Phe49Leu, Thr52Phe, Phe86Leu, Thr93Pro, Ala114Gly and Ser118Thr) in mouse UVS with those in human VS pigment (53) produced a shift from 359 nm to 411 nm, and the reverse substitutions shifted the λmax of human VS to 360 nm. Overall therefore, these substitutions achieve the spectral shift between mouse UVS and human VS pigments, although their effects must be synergistic since single substitutions had no effect on λmax. Subsequently, it has been shown (54) that although Leu86 and Pro93 substitutions into mouse UVS do not generate a spectral shift into the violet, the addition of substitutions at sites 114 and 118 do result in a pigment with a λmax at 399 nm. Since site 114 is not conserved across other primate species, the key substitutions in the generation of primate VS pigments from an ancestral UVS pigment are most likely Phe86Leu, Thr93Pro and Ser118Thr, although for the full shift to around 415 nm (55), substitutions at sites 46, 49, and 52 are also required (54).
The prosimians are the other major primate group. They are divided into two strepsirhine groups, the Lorisiformes and the Lemuriformes, and a haplorhine group, the Tarsiiformes (56). Among the nocturnal lorisiforms, a functional SWS1 gene appears to be absent, as demonstrated in the greater (Otolemur crassicaudatus) and lesser (Galago senegalensis) galago and in the slow loris (Nycticebus coucang) (57,58). However, this is not the case in the lemuriforms and tarsiers where a functional SWS1 gene is present, as demonstrated for the western tarsier (Tarsius bancanus) and the brown lemur (Eulemur coucang) (58). Moreover, retention of sensitivity to SW light is found in the ringtail (Lemur catta) and brown (Lemur macaco) lemur, and the presence of S cones in the retina has been demonstrated in the gray mouse lemur (Microcebus murinus) and the eastern tarsier (Tarsius spectrum) (59,60). The SWS1 genes in prosimians consistently have Pro93 and, with the single exception of Phe90 in the pygmy slow loris, all also have Ser90. However, the tarsiers have Leu86, making them the group with pigments most like those of Old and New World primates (Table 1). Amongst the aye-aye, lemurs and sifakas, the most common residue at site 86 is Ser (gray mouse lemur, Coquerel’s dwarf lemur, greater dwarf lemur, fat-tailed dwarf lemur), followed by Cys in two species (brown lemur, ring-tailed lemur), and Val (woolly lemur), Leu (Verreaux's sifaka) and Phe (aye-aye) in one species each. Cys, Ser, Leu and Val would all be expected to generate VS pigments whereas Phe86 is generally associated with a UVS pigment (24). Amongst the species with mutated pseudogenes, the lesser galago and slow loris have Tyr86, the greater and small eared galago have Asn86, and the pygmy slow loris has Cys86. The variation at site 86 is therefore extensive and although this may be associated with loss of function, it is not the case for the lemuriforms. Because the effect of Cys86, Ser86 and Val86 are most likely the same, namely the production of a VS pigment, this would imply that the acquisition of a VS pigment occurred separately in the different lineages. In contrast, the aye-aye may have a UVS pigment with Phe86 retained from the ancestral sequence.
The pattern of substitution at site 86, 90 and 93 in the evolution of VS pigments in mammals is summarized in Fig. 3.
Met86 is present in the VS pigment of the clawed frog (61), combined with Pro93. However, neither a Met86Gln nor a Pro93Thr substitution into the VS pigment had any effect on the λmax (62), and a Phe86Met substitution into goldfish UVS pigment also failed to generate a LW-shift (36). Neither substitution by itself is capable therefore of shifting the spectral-sensitivity between VS to UVS. The double mutation of Phe86Met and Thr93Pro does, however, shift the λmax of an ancestral genetically engineered recombinant UVS pigment (26) from 359 to 394 nm, although the full shift to 423 nm requires, in addition, substitutions at sites 91, 109, 113, 116 and 118 (63).
Full sequence and spectral data exist for only three avian VS pigments, the Humboldt penguin (Spheniscus humboldti) (64), the domestic pigeon (Columba livia) and the chicken (Gallus gallus) (8,65); in all cases, Ser86 is present (Table 2). Significantly, a Ser86Phe substitution in the VS pigments of the pigeon and chicken shifts the λmax to 357 and 372 nm, respectively, whereas the converse Phe86Ser substitution into the UVS pigment of the green anole lizard Anolis produces a small but significant LW-shift from 360 to 374 nm (L. Carvalho, unpublished data). These shifts in spectral-sensitivity are consistent therefore with a central role for Ser86 in the evolution of avian VS pigments. In addition, a Val116Leu substitution also SW-shifts the pigeon pigment to 366 nm. It has previously been shown (26) that the triple substitution of Phe49Val, Phe86Ser, and Ser118Ala LW-shifts a genetically engineered UVS pigment from 360 to 374 nm, and that the addition of Leu116Val produces a further shift to 393 nm. Combining these sets of data, the key substitutions in the evolution of avian VS pigments would appear to be therefore Ser86 and Val116.
Table 2. Candidate spectral tuning sites for avian VS/UVS pigments.
Amino acid sites
Evolution of UVS pigments in birds
In contrast to the UVS pigments of all other vertebrates, Phe86 is generally substituted in the UVS pigments of birds; the mechanism for achieving a UVS pigment differs therefore in birds from that in other vertebrates (25). Comparison of the amino acid sequences of avian VS and UVS pigments enabled the identification of Cys90 as the key substitution in the evolution of avian UVS pigments (64) and this was confirmed by replacement of Cys90 with Ser90 in the budgerigar (Melopsittacus undulatus), which LW-shifted the mutant pigment into the violet region of the spectrum. A similar conclusion was reached in another avian species, the zebrafinch (Taeniopygia guttata); in this latter study (65), the reverse substitution of Cys inserted into site 90 of the VS opsins of pigeon and chicken were also made and these generated SW-shifts from 393 to 358 nm in pigeon and from 415 to 369 nm in chicken.
The residues present at site 86 and 90 have been identified in a total of 45 bird species distributed across 14 avian Orders (66). This study utilized genomic DNA and focused on the gene region encoding sites 86 and 90. Although Ser90 is present in most species, a subset possess Cys90 as follows: three of 21 species from the Order Ciconiformes, 4 of the 8 species from the Order Passeriformes, both of two species from the Order Psittaciformes, and one of the two species from the Order Struthioniformes. As discussed above, it would appear that the ancestral avian VS pigment had Ser86 and Ser90, and that UVS pigments, rather then being ancestral as in other vertebrate groups, have been “re-invented” at least four times in avian evolution, each time by a Ser90Cys substitution into a VS pigment. No spectral data are available for many of the species included in this study, hence it remains uncertain whether all these species possess a UVS pigment. One such query arises over the presence of Cys86 combined with Ser90 in a subset of avian pigments. As a Ser86Cys substitution has been shown to generate a SW-shift (26), it is possible that avian species with Cys86 also have UVS pigments. To answer this, we have cloned the full coding sequence of the SWS1 gene from the Common cormorant, Phalacrocorax carbo. The presence of Cys86 was confirmed but spectral characteristics of the regenerated pigment is VS with a λmax at 405 nm (L. Carvalho, unpublished data). Cys86 does not therefore generate a UVS pigment in birds. Phe86 is present in two species, the Common rhea (Rhea americana) and the Blue-crowned trogon (Trogon curucui). The former species also possesses Cys90, so it is highly probable that the pigment is UVS, and this is most likely the case for the Blue-crowned trogon as Phe86 is found in all non-avian UVS pigments and, when substituted into the pigeon VS pigment, SW-shifts the λmax to 357 nm (L. Carvalho, unpublished data). Other substitutions identified in avian pigments (e.g. Ala86 in a number of species from the Order Ciconiiformes) may also result in UVS pigments, but yet, no spectral data are available. In summary, therefore, it would appear that avian UVS pigments have evolved from avian VS pigments via a Ser90Cys substitution on at least five separate occasions (Fig. 4), and by a back mutation to Phe86 on one occasion.
Protonation of the Schiff base
The generation of a UVS pigment is most simply achieved by a loss of protonation of the Schiff base and evidence to date indicates that this is the mechanism used. Protonation requires the presence of a charged counterion and this is provided by a negatively charged residue (almost always Glu) at site 113 (67). Replacement of this residue in mouse UVS pigment with uncharged Gln (and consequent loss of protonation in the ground state) does not prevent the generation of a UVS pigment (37), indicating that UVS pigments are unprotonated. Moreover, a VS pigment with a Glu113Gln substitution is SW-shifted to 351 nm at pH 7.0 (37) but at lower pHs, a protonated form of the pigment is produced with a λmax at 424 nm.
In evolution therefore, the generation of vertebrate VS pigments has been achieved by the stabilization of retinylidine Schiff base protonation. The crystal structures of bovine rod opsin (2,68) sets the Schiff base to counterion separation at 3.3 to 3.5 Å and it has proposed that the stabilization of protonation involves a hydrogen bonding network (68) with the hydroxyl group of Thr94 bound to the Glu113 carboxylate oxygen atom. Two water molecules are also involved, one near but not between Glu113 and the Schiff base (69) and a second that hydrogen bonds with Ser186, Cys187 and Glu181. In VS pigments, the residue at site 86 (and site 90 for avian pigments) is critical for Schiff base protonation (Fig. 5a); certain residues (e.g. Tyr, Ser, Val, Leu) may contribute therefore to the hydrogen bonding network, and the distance between the hydroxyl group of Tyr86 and the Schiff base nitrogen of 3.56 Å predicted from modeling the Phe86Tyr mutant goldfish pigment on to the bovine rhodopsin template (Fig. 5b) certainly brings it within hydrogen bonding distance (36). In non-avian UVS pigments, a nonpolar amino acid is invariably found at site 86. This must contribute to the loss of protonation of the Schiff base. Finally, avian VS pigments possess Ser or Cys at site 86 and Ser at site 90; loss of protonation and the generation of a UVS pigment occurs therefore through the replacement of polar Ser by Cys at site 90 (64). Yokoyama et al. (65) have proposed that the hydrophobicity of Cys90 removes a water molecule from the vicinity of the Schiff base and thereby displaces its positive charge. Under these conditions therefore, the Schiff base would be effectively unprotonated.
Phylogenetic analysis indicates that the ancestral vertebrate SWS1 pigment was maximally sensitive in the UV. This shortwave tuning is achieved by an unprotonated Schiff base. The key residues for shifts between UV and violet are at sites 86 and 90; all non-avian UVS pigments studied so far possess Phe86 combined with Ser90 whereas VS pigments have Phe86 substituted with either Tyr, Val, Leu or Ser, depending on species, to generate a protonated Schiff base. The evolution of VS pigments involves therefore the stabilization of protonation, most likely via H-bonding networks in the vicinity of the counterion base residue, Glu113, and sites 86 and 90. VS pigments have evolved on several separate occasions and in many cases, this single amino acid substitution is sufficient by itself to shift the peak sensitivity from UV to violet. Avian UVS pigments differ from those of the other vertebrate classes as they have arisen secondarily from an ancestral avian VS pigment. With only one exception so far identified where Phe86 is present, this has been achieved by a Ser90Cys substitution combined with replacement of Ser86 by either Ala, Cys, Ile or Met.
Acknowledgements— We are grateful to Dr. Rosalie Crouch for the generous gift of 11-cis retinal. This work was supported by grants from the UK Biotechnology and Biological Sciences Research Council and the Leverhulme Trust.