Printer Friendly

Opsins and Their Expression Patterns in the Xiphosuran Limulus polyphemus.

Introduction

The American horseshoe crab Limulus polyphemus (Linnaeus, 1758) has a long history in vision science beginning in the 1930s (Hartline and Grahm, 1932). Limulus polyphemus, hereafter referred to as Limulus, is particularly attractive to electrophysiologists studying vision because the photoreceptors in its eyes are large and can be recorded relatively easily with sharp electrodes; its tissues are hardy and easily maintained for long periods in vitro; and it has three different types of eyes, all of which are amenable for experimentation: lateral eyes (LEs), median eyes (MEs), and larval eyes. Electrophysiological studies of Limulus eyes and photoreceptors have been critical for furthering our understanding of basic mechanisms of vision, including the role of phospholipids in phototransduction (Brown et ai, 1984), C[a.sup.++] in light and dark adaptation (Payne et al., 1986a, b), lateral inhibition in visual information processing (Hartline et al., 1956), and circadian rhythms in regulating retinal and photoreceptor responsiveness and sensitivity to light (Barlow etal., 1977, 1980; Barlow, 1983). From the perspective of visual system evolution, work on Limulus is particularly important because of the animal's placement in the arthropod family tree. Limulus is a xiphosuran chelicerate, the sister group to arachnids. Limulus also is thought to have retained many plesiomorphic characteristics and therefore has been considered a proxy for the euchelicerate ancestor (Regier et al., 2010; Edgecombe and Legg, 2014). This review brings together extensive knowledge of the organization and function of the Limulus visual system with newer information about Limulus opsins and their expression patterns. The aim is to provide a comprehensive examination of photosensitivity in this well-studied xiphosuran that may provide key insights into photoreception in euchelicerates and the evolution of photosensitivity among chelicerates and arthropods in general.

Organization of the Limulus Visual System

The organization of the Limulus visual system and its eyes is illustrated in Figure 1. The LEs are located dorsolaterally on the cephalothorax (Fig. 1 A). They are compound eyes and the animal's only image-forming eyes. Horseshoe crabs are the only extant chelicerates with compound eyes. In an adult animal, each LE contains more than a thousand ommatidia consisting of a conical lens below which are located 5-12 photoreceptors, also called retinular cells, arranged like the sections of an orange. Pigment cells surround the photoreceptors, the base of the lens, and the aperture at the base of the lens. Directly below the lens aperture, the photosensitive microvilli of neighboring retinular cells interdigitate, creating a rhabdom that in cross section appears like the spokes of a wheel (Fig. 1B). At the very center of each ommatidium are the dendrites of one or more neurons called eccentric cells that are electrically coupled to the retinular cells (Fahrenbach, 1969, 1975). Both the retinular cells and the eccentric cells project to the brain; however, since retinular cells produce graded depolarizing potentials when stimulated by light, it is unclear whether these depolarizations reach the brain in an adult animal because the lateral optic nerve is long (Fig. 1A). However, when eccentric cells depolarize in response to the light-stimulated depolarization of retinular cells, they produce trains of action potentials that clearly propagate to the brain. Therefore, eccentric cells are thought principally responsible for transmitting visual information from the LE to the brain (Waterman and Wiersma, 1954). Eccentric cells are also responsible for lateral inhibitory interactions within the eye (Hartline et ai, 1956; Fahrenbach, 1985).

MEs are located anteriorly on the cephalothorax (Fig. 1 A). They are ocelli consisting of a single lens below which lie elongated photoreceptors with rhabdoms located close to the base of the lens (Fig. 1C). The photoreceptors are arranged into loosely organized clusters containing both ultraviolet (UV)-sensitive and visible light-sensitive photoreceptors, an arhab-domeric neuron, and guanophores--cells containing reflective guanine crystals. Partitions between the clusters contain glial cells, guanophores, and pigment cells. Pigment cells also surround the ME cup (Jones et ai, 1971; Fahrenbach, 1975; Fahrenbach and Griffin, 1975). The shape of the ME lens and the organization of its retina suggest the eye does not form images (Fahrenbach, 1975), but the results of behavioral experiments showed that illuminating the ME provides the animal with directional information (Lall and Chapman, 1973), and its large lens aperture suggests it is adapted to operate in low light conditions (Fahrenbach, 1975).

The number of UV-sensitive and visible light-sensitive photoreceptors in each ME photoreceptor cluster varies considerably, but estimates from electrophysiological studies suggest that about 70% are sensitive to UV light while the remaining 30% are sensitive to visible light. No photoreceptors with dual sensitivity to UV and visible light have been detected (Nolte and Brown, 1969, 1970, 1972; Lall, 1970). Both UV-and visible light-sensitive ME photoreceptors project to the brain and produce graded potentials when stimulated with light; however, the arhabdomeric cells are thought to provide the major output from the eye to the brain. Like LE eccentric cells, arhabdomeric cells are electrically coupled to the photoreceptors, and they generate action potentials in response to the graded depolarization of photoreceptors. Interestingly, arhabdomeric cells are electrically coupled only to UV-sensitive photoreceptors (Nolte and Brown, 1972). Consistent with this finding, adult Limulus exhibit a positive phototactic response when their MEs are illuminated with UV light but no response when their MEs are illuminated with visible light. The behavioral response to UV light may be relevant to the animal's ability to find shallow water and beaches for spawning (Lall and Chapman, 1973). The functions of visible light photosensitivity in MEs remain largely obscure, but visible light illumination of MEs can phase shift the animal's circadian clock (Home and Renninger, 1988; Renninger et ai, 1997), and it may modulate the UV-light-driven output of the eye (Nolte and Brown, 1972).

The larval eyes--lateral, median, and ventral--develop in the embryo before the LEs and MEs, and they are thought to provide the major photic input to embryos and newly hatched larvae. Each larval eye contains clusters of 2 types of photoreceptors: giant photoreceptors that are about 150 [micro]m long and 70 [micro]m wide in adult animals, and smaller photoreceptors that are about half the size of the giants (Fig. 1E). Larval eyes do not contain pigment cells, but their photoreceptors are tightly ensheathed by glia. All three pairs of larval eyes persist in adults; but in adults, the photoreceptors in the lateral larval eyes cluster at the posterior edge of each lateral compound eye, and those in the fused median larval eyes are located below the carapace between the median ocelli (Fig. 1 A). Therefore, it is not clear whether lateral and median larval eyes receive sufficient light in adults to respond and influence behavior. On the other hand, photoreceptors in the ventral larval eyes (VEs) clearly do respond to light in intact adult animals (Kass and Renninger, 1988); and because they are relatively accessible in adults, they are the most extensively studied of the larval eyes. In adults, the VEs consist of a pair of optic nerves that extend anteriorly from the brain and terminate in an end organ that is attached to the ventral organ (Fig. 1A, D)--a specialized, pigment-free region of the ventral cuticle that has a small lenslike structure positioned over each end organ. Giant and smaller photoreceptors cluster near the brain and especially in the end organs. They are also scattered along the length of each optic nerve (Fig. 1D).

In addition to the photoreceptors in its eyes, Limulus has extraocular photoreceptors in each of its segmental ganglia (Mori and Kuramoto, 2004; Mori et al., 2004), which may modulate motor neuron output, and in its tail, which can phase shift the animal's central circadian clock (Hanna et al., 1988; Renninger et al., 1997).

A striking and well-characterized feature of the Limulus visual system is the circadian-clock-driven efferent innervation to each of its eyes (reviewed in Barlow, 2001; Battelle, 2002, 2006, 2013). Briefly, the cell bodies of these efferent neurons are located at the base of the brain; their axons project to the eyes through each of the optic nerves; they synapse on to all cell types in the eyes, including photoreceptors; and their activity is controlled by a circadian clock located in the brain. Their axons begin firing action potentials around dusk, continue firing throughout the night, stop firing before sunrise, and are silent during the day. This activity pattern is circadian because it continues in constant darkness and can be phase shifted. In response to the activation of this efferent input, all three types of Limulus eyes become more sensitive and responsive to light, a response thought to be important for the animal's reproductive success (Barlow et al., 1982). A similar system of clock-driven, efferent input to eyes, with similar impacts on retinal and photoreceptor functions, has been described in scorpions and some spiders (Heinrichs and Fleissner, 1987; Fleissner and Fleissner, 1988; Yamashita, 2002), suggesting that this feature was present in the last common ancestor of xiphosurans and arachnids.

The Full Opsin Repertoire in Limulus Has Been Identified

Opsin is the protein component of visual pigment, and its amino acid sequence primarily determines the spectral sensitivity of visual pigments. Results of detailed electrophysiological recordings made from all known Limulus photoreceptors suggested Limulus expresses four different spectral types of opsins (Table 1): (1) a long wavelength-sensitive (LWS) opsin in LE retinular and giant VE photoreceptors (Millecchia et al., 1966; Nolte and Brown, 1970), (2) a LWS opsin in some ME photoreceptors that is different from that in LE retinular cells and giant VE photoreceptors (Nolte and Brown, 1970), (3) a UV-sensitive opsin in most ME photoreceptors (Nolte and Brown, 1970, 1972), and (4) a short wavelength-sensitive (SWS) opsin in some cells in each abdominal ganglion (Mori and Kuramoto, 2004; Mori et al., 2004). As was mentioned above, the Limulus tail is also photosensitive, but tail photoreceptors have not been identified, and their spectral tuning has not been characterized (Hanna et al., 1988). Considering the results of these electrophysiological studies, it came as a surprise when molecular studies revealed 18 opsin genes in the Limulus genome, 17 of which are expressed in the eyes, central nervous system (CNS), or tail. Each of these opsins has now been characterized along with its gene structure and the cellular or tissue distributions of its transcripts. Specific anti-bodies generated against some of the opsins have also allowed studies of diurnal changes in their expression patterns in rhabdoms.

Transcriptomes and genome assemblies were critical for identifying the full repertoire of Limulus opsins and for providing new information about opsin gene duplication and loss in the xiphosuran lineage. Nine different Limulus opsins were characterized before the Limulus genome became available. Transcripts encoding Limulus polyphemus opsins 1 and 2 (LpOps1 and LpOps2) were identified in cDNA libraries of LE and ME, respectively, using classical homology cloning approaches (Smith et al., 1993); and two additional LpOps1-like genes (LpOps3 and LpOps4) were identified through polymerase chain reaction (PCR) and Sanger sequencing of Limulus genomic DNA (Dalai et al., 2003). Transcriptomes of Limulus eyes and CNS (Katti et al., 2010; Battelle et al., 2014, 2015; Speiser et al., 2014) revealed fragments of six additional opsin-like transcripts, LpOps5, LpOps6, LpOps8, LpUVOps1, and peropsin 1 (LpPerOps1), which were subsequently cloned full-length from Limulus VE or ME cDNA libraries. Nine additional opsin genes were identified through tBLASTn searches of a well-assembled Limulus genome. These are LpOps7, LpOps9, LpOpslO, UVOps2, arthropsin 1 and 2 (LpArthOps1 and LpArthOps2), ciliary opsin1 and opsin2 (LpCOpsl and LpCOps2), and PerOps2 (Battelle et al., 2016). The only Limulus opsin gene for which no transcript has been detected in the eyes, CNS, or tail is LpPerOps2 (Battelle et al., 2016). LpArthOps2 transcripts were not detected in the eyes, CNS, or tail of older juveniles (measuring 2.5-3.5 cm across the prosoma) or adults; however, they were detected in the CNS of young juvenile animals, between their first and second juvenile molts.

In a large maximum likelihood phylogenetic analysis of genomic and expressed opsins, Limulus opsins cluster together with other chelicerate opsins in three of the four major opsin groups: R-type, C-type, and RGR/Go-type or Group 4 opsins (Fig. 2) (Battelle et al., 2016). A number of these opsins appear to be paralogous groups: LpOps1-4, LpOps6 and LpOps7, LpOps9 and LpUVOps2, LpArthopsl and LpArthops2, LpPerOps1 and LpPerOps2, LpCOps1 and LpCOps2 (Fig. 3A). Each group has an identical gene structure (intron location and phase) (Fig. 3B), and the proteins they encode are 50% or more identical (Battelle et al., 2016). LpOps1-4 are probably products of recent tandem duplication events. Their genes are on a single scaffold, and the proteins they encode are 99% identical to one another (Battelle et al., 2016). LpOps1 and LpOps2 transcripts can be distinguished from one another by the sequences of their 3'-untranslated regions (Smith et al., 1993), but LpOpsl, LpOps3, and LpOps4 transcripts cannot be distinguished because their genes differ only in their in-trons (Dalai et al., 2003). Other paralogous pairs may be products of a whole-genome duplication that is proposed to have occurred early in the xiphosuran lineage (Nossa et al., 2014; Kenny et al., 2015). The presence of other genes shared between scaffolds encoding the paralogous LpOps6 and LpOps7, LpArthOps1 and LpArthOps2, and LpPerOps1 and LpPerOps2 is consistent with this idea (Battelle et al., 2016). If a whole-genome duplication did occur, it must have been followed by significant gene loss since LpOps5, LpOps8, LpOps10, and UVOps1 do not have paralogs.

When Limulus opsin genes are compared to those recovered from the genomes of 2 other extant horseshoe crabs (Kenny et al., 2015)--at least 14 opsins from Tachypleus tridentatus and at least 15 from Carcinoscorpius routundicaud--a very similar complement of paralogous opsin pairs and non-paralogous opsins is found (Battelle et al., 2016). This is consistent with the idea that a whole-genome duplication followed by opsin gene loss occurred early in the xiphosuran lineage before the species split from one another an estimated 130-150 million years ago (Obst et al., 2012).

The same pattern of paralogous opsin pairs has not been observed so far in other chelicerates, and other chelicerates have fewer opsins--six in the transcriptome of the spider Cupiennius salei (Eriksson et al., 2013), six recovered from the genome of the velvet spider Stegodyphus mimosaurum, and five recovered from the genome of the scorpion Mesobuthus martensii (Battelle et al., 2016). This suggests an expansion of opsins in Xiphosura occurred after their split from the rest of Euchelicerata. But even if only one of each of the paralogous opsin pairs is counted in Limulus, the number of Limulus opsins (10) is still greater than the number detected in other chelicerates. Other chelicerates lack homologs of LpOps5, LpOps6, LpOps7, and LpOps8. LpOps5 homologs were probably lost in other chelicerate lineages (see next section), and LpOps6, LpOps7, and LpOpsS may be unique to the xiphosurans. Interestingly, as is described more fully in a later section, LpOps6, LpOps7, and LpOps8 are all uniquely expressed in Limulus MEs.

The Structures of Limulus Opsin Genes Provide New Insights into Relationships Among Opsin Gene Families

In addition to phylogenetic analyses based on sequence homology, gene structure (intron location and phase) can provide information about relationships among gene families (Rokas and Holland, 2000). Comparisons between the structures of opsin genes from Limulus (Fig. 3) and other species add additional support for the placement of Limulus opsins in phylogenetic trees based on sequence homology. This includes the placement of LpOps5 among medium wavelength-sensitive (MWS) opsins of crustaceans (Fig. 2), a clade originally thought to be crustacean specific (Kashiyama et al., 2009). The structure of the LpOpsS gene is unique among Limulus opsins, but it has an intron in common with Daphnia puelx opsins in the MWS clade most closely related to LpOps5 based on sequence homology (Battelle et al., 2016). The presence of both chelicerate and crustacean opsins in this clade strengthens the idea that an LpOps5 homolog was present in the last common ancestor of arthropods (Henze and Oakley, 2015). This gene has apparently been lost in insects and arachnids, but it is conserved in at least three of the four extant species of horseshoe crabs (Fig. 2).

An unexpected finding was that Limulus opsins with an identical gene structure are distributed in three different opsin clades: (1) LpUVOps1 within the SWS opsin clade; (2) LpUVOps2, LpOps9, and LpOpslO within the pancrustacean UV7 clade; and (3) LpArthOpsl and LpArthOps2 in the more distantly related arthropsin clade (Figs. 2, 3). Introns that match the position and phase of the two introns in these Limulus opsins are also present in R-type opsins from other arthropods and from annelids, echinoderms, and mammals (Battelle et al., 2016). This suggests that the gene structure shared among Limulus SWS opsins, UV7-type opsins, and arthropsins is ancient among R-type opsins.

The structure of Limulus LWS opsin genes may also be ancient at least among LWS opsins in arthropods. Intron 1 in LWS LpOps1-4 and LpOps8 is strictly conserved among insect LWS opsins, and intron 2 of LpOps 1-4 matches in position and phase an intron in some LWS opsins of insects (Battelle et al., 2016).

Opsin gene structure suggests a closer relationship between Limulus LWS opsins and peropsins, a Group 4 opsin, than is expected based on the phylogeny shown in Figure 2. The introns in LWS LpOps1-4 match the position and phase of introns 4 and 5 in LpPerOpsl (Fig. 3), and the position and phase of intron 5 in LpPerOpsl is conserved in all peropsins characterized to date (Albalat, 2012; Battelle et al., 2016). These observations can be interpreted to mean that peropsins and LWS R-type opsins arose from a common ancestral gene. It should be noted that previous phylogenetic analyses have suggested a closer relationship between R-type opsins and Group 4 opsins than is shown in Figure 2 (Porter et al., 2012). Clearly, more data are required to resolve the evolutionary history of peropsins.

The relationship between LpOps8 and the paralogs LpOps6 and LpOps7 presented a particular puzzle that is informed by gene structure. The paralogs LpOps6 and LpOps7 are encoded on different scaffolds. Because both lack introns, they are probably products of an early retrotransposition event. LpOps6 and LpOps8 genes are located close to one another on the same scaffold (within 5.5 kb), which might suggest they are products of a tandem duplication event. Taken together, these findings lead to the hypothesis that LpOps6, LpOps7, and LpOps8 constitute a monophyletic clade; but this hypothesis was rejected in a Swofford-Olsen-Waddell-Hills (SOWH) test (Battelle et al., 2016), and phylogeny (Fig. 2) indicates that LpOps8 is more closely related to LpOps1-4 than it is to LpOps6 and LpOps7. A relationship between LpOpsS and LpOps1-4 is supported further by gene structure. Intron 2 in LpOpsS matches the location and phase of intron 1 in LpOps1-4 (Fig. 3).

The number, arrangement, and structures of Limulus LWS opsin genes--LpOps 1-4, LpOps6, LpOps7, and LpOpsS--suggest that they have a complex evolutionary history. One scenario that is consistent with an ancient whole-genome duplication within the xiphosuran lineage, phylogeny, and parsimony principles is depicted in Figure 4. It suggests that synteny between LpOps6 and LpOps8 has been maintained a very long time and thus may have strong functional significance, possibly relevant to their unique expression in MEs and coexpression in visible light-sensitive photoreceptors in those eyes (Table 1; Fig. 5) (Battelle et al., 2015).

Most Photoreceptors in Limulus Eyes Express Multiple Opsins (Table 1) That Change Their Relative Concentrations in Rhabdoms with a Diurnal Rhythm

LpOps 1-4 and LpOps5 are coexpressed in LE retinular cells and the giant photoreceptors of larval eyes

LpOps 1-4 and LpOps5 coexpression in rhabdoms in LE retinular cells and giant VE photoreceptors was confirmed with immunocytochemical assays. Their relative concentrations in rhabdoms were determined in immunochemical assays of membranes isolated from LEs at night in the dark, when the concentration of opsins in rhabdoms is highest, and almost all opsin-containing membranes are localized to rhabdoms. Under these conditions the molar concentration of LpOps 1-4 in membranes is four times that of LpOps5. LpOps 1-4 is clearly the more abundant opsin expressed in LE retinular cells, but the concentration of LpOps5 is also significant.

The higher concentration of LpOps 1-4 compared to LpOps5 in rhabdoms may be a consequence of the translation of mRNAs transcribed from all four genes in the LpOps1-4 gene cluster compared to the single LpOpsS gene, and indirect evidence suggests that all LpOps1-4 genes are transcribed. As was described above, LpOpsl, LpOps3, and LpOps4 transcripts, called LpOpsl-like transcripts, cannot be distinguished from one another, and therefore they must be assayed together. However, LpOpsl-like and LpOps2 transcripts can be distinguished from one another based on the sequences of their 3' untranslated regions, and they have been quantified separately using ribonuclease protection assays (Dalai etal., 2003). These assays established that both LpOpsl-like and LpOps2 transcripts are present in LEs and VEs. They showed further that in LEs, LpOps 1-like transcripts are between 2 and 3.5 times more abundant than LpOps2 transcripts, a finding quantitatively consistent with the idea that all 3 LpOps 1-like genes are transcribed.

Although the ratio of LpOps 1-4 to LpOps5 in LE rhabdoms during the night in the dark is 4:1, this ratio is not stable; rather, it changes with a diurnal rhythm. In the LEs of animals maintained under natural illumination, the concentration of LpOps1-4 in rhabdoms (Rh-LpOps 1-4) during the day in the light is 50%-60% lower than it is during the night in the dark, whereas the concentration of Rh-LpOps5 does not change significantly from day to night (Figs. 6, 7). A similar change in the relative concentrations of Rh-LpOps 1-4 and Rh-LpOps5 is observed in the giant VE photoreceptors (Fig. 6) (Katti et al., 2010; Battelle, 2013; Battelle et al., 2013). Limulus photoreceptors provide the first clear evidence that the ratio of coexpressed opsins in rhabdoms can change. Mechanisms underlying the differential regulation of Rh-LpOps 1-4 and Rh-LpOps5 concentrations are not known, but they must involve differences in their turnover. Turnover involves the removal of opsin from rhabdoms by processes referred to as shedding and opsin renewal, the addition of opsin-containing membranes to rhabdoms. One major difference between LpOps 1-4 and LpOps5 is that LpOps 1-4 turnover is significantly influenced by the circadian efferent input to photoreceptors described in the Introduction, whereas Rh-LpOps5 turnover is not.

The impact of circadian clock input on the concentrations of opsins in rhabdoms (Rh-LpOps) has been examined in LEs by measuring the intensities of opsin immunoreactivity in rhabdoms per unit of rhabdom area at different times of the day and night in LEs with and without clock input (Battelle, 2013; Battelle et al., 2013). Clock input to LEs was eliminated by cutting the lateral optic nerve (Barlow et al., 1977). All animals used in these studies were maintained in natural illumination; thus, time of day is expressed relative to sunrise and sunset (Fig. 7).

During the day in the light, when clock input to LEs is silent, the concentration of Rh-LpOps 1-4 does not change significantly even though LpOps 1-4-containing membranes are continuously shed from rhabdoms by a process called light-driven shedding (Fig. 6) (Sacunas et al., 2002). Light-driven shedding is mediated by arrestin and clathrin (Sacunas et al., 2002) and is considered similar to the activity-dependent clathrin-mediated endocytosis of other G-protein-coupled receptors. Since the concentration of Rh-LpOps1-4 does not change during most of the day in the light, LpOps1-4 shedding and renewal must be in equilibrium during this time. However, during dusk, the concentration of Rh-LpOps1-4 increases gradually about 20%. This initial increase occurs in LEs whether the lateral optic nerve is intact or not, and therefore it is independent of clock input. It can be attributed to the cessation of light-driven shedding with the onset of darkness while renewal continues. In LEs without clock input, Rh-LpOps1-4 remains at about 20% above its daytime concentration throughout the night. But in LEs with clock input, the concentration of Rh-LpOps1-4 continues to increase until, by 4 h after sunset, it is about 60% above its daytime concentration (Fig. 7).

Mechanisms responsible for the clock-driven increase in the concentration of Rh-LpOps1-4 at night are not known, but they are driven by a 3',5'-cyclic adenosine monophosphate (cAMP) cascade that is activated by octopamine, a biogenic amine neurotransmitter released from the clock-driven neural projection to the eyes (reviewed in Battelle et al., 2002, 2013); and they must involve renewal since shedding is not observed in LEs at night in the dark. A reasonable hypothesis is that clock-enhanced renewal results from a clock-dependent increase of LpOps1-4 transcripts; however, this is not the case. LpOps1-4 transcript levels are lower during the night in the dark than they are during the day in the light, and LpOps1-4 transcript levels are elevated by light (Dalai et al., 2003). This leads to the hypothesis that clock input enhances LpOps1-4 translation or posttranslational processing.

The concentration of Rh-LpOps1-4 remains high throughout the night in LEs with normal clock input. Then, soon after sunrise, after the clock-driven efferent neurons stop firing (Chamberlain et ai, 1987), the concentration of Rh-LpOps1-4 drops precipitously during a burst of what is called transient rhabdom shedding. During the first hour after sunrise, the same amount of LpOps1-4 that was added to rhabdoms in response to clock input the night before is removed (Fig. 7). Transient rhabdom shedding has been studied in detail (Chamberalin and Barlow, 1979; Chamberlain et al., 1984), and it is very different from clathrin-mediated, light-driven shedding. Transient rhabdom shedding must be primed by at least 3 h of circadian clock input, which is about the same duration required for the clock-dependent increase in the concentration of Rh-LpOps4 at dusk (Fig. 7); it is triggered by the dim light of dawn; and it occurs only during a brief period after dawn. Furthermore, transient shedding is characterized by a unique process of membrane endocytosis. During transient shedding, rhabdomeric microvilli, which are highly organized during the night, become disorganized, and large membranous whorls, consisting of many layers of microvillar membrane, form at the base of the rhabdom. These membranes break away from the rhabdom and coalesce to form multivesicular bodies, which, with time, migrate away from the rhabdom. During the burst of transient shedding, the total rhabdom area decreases up to 70%, but it rapidly returns to its preburst area (Chamberlain et ai, 1984). Although rhabdom area recovers, the concentration of LpOps1-4 in the rhabdom does not; it is about 40% lower. Following its precipitous fall during the first hour after sunrise, the concentration of Rh-LpOps1-4 decreases more gradually, an additional 20%, to its daytime level. The gradual decline following transient shedding is identical to that seen in eyes without clock input; therefore, it is attributed to light-driven shedding.

In view of the dynamic changes in the concentration of Rh-LpOps1-4 in LEs with normal clock input, it is surprising that the concentration of Rh-LpOps5 does not change significantly, especially when LpOps5-containing membranous debris is observed between the rays of the rhabdom throughout the day in the light (Fig. 6), indicating that LpOps5 is shed. The average concentration of Rh-LpOps5 does not increase during the night, and although it decreases an average of 35% during the morning--with a time course that correlates with light-driven shedding and not transient shedding--this decrease does not reach the level of significance. Furthermore, by midday the average concentration of Rh-LpOps5 is at least as high as it was the night before. This suggests that Rh-LpOps5 can rapidly renew in the light and compensate for its light-driven loss. Comparisons between the diurnal changes in the concentrations of Rh-LpOps1-4 and Rh-LpOps5 coexpressed in retinular cell rhabdoms provide clear evidence that the turnover of coexpressed opsins can be regulated differently.

The diurnal change in the concentration of Rh-LpOps1-4 in both LE retinular cells and giant VE photoreceptors probably contributes to observed diurnal changes in LE and VE sensitivity to light (Barlow et al., 1977; Kass and Renninger, 1988), but the functional relevance of LpOps1-4 and LpOps5 coexpression and diurnal changes in their relative concentrations in rhabdoms is unclear. The spectral sensitivities of LpOps1-4 and LpOps5 were initially hypothesized to be different because in phylogenetic trees LpOps1-4 cluster among LWS opsins, whereas LpOps5 clusters among MWS opsins (Katti et al., 2010). If their spectral sensitivities were different, their coexpression might be expected to broaden photoreceptor spectral sensitivity, as has been documented in other systems (Arikawa et al., 2003; Hu et al., 2011, 2014; Dalton et al., 2014), and a diurnal change in their relative concentrations might produce a diurnal change in the photoreceptor's spectral tuning. However, results of electrophysiological studies showed that LpOps5 is a LWS opsin and that its spectral sensitivity is indistinguishable from LpOps 1-4 (Battelle et al., 2014). Although the spectral sensitivities of LpOps1-4 and LpOps5 are indistinguishable, they may differ in other aspects of their biochemistry that could subtly influence the efficacy or kinetics of the photoresponse.

LpUVOps1 and LpOps5 are coexpressed in small larval eye photoreceptors

LpUVOps1 was first identified in a transcriptome from larval VEs, larval eyes that are readily accessible in adult animals (Battelle et al., 2014). This was a surprise because extensive prior electrophysiological studies of the VEs provided no evidence for UV photosensitivity. The failure to detect UV sensitivity in VEs is probably because previous studies focused on the giant ventral photoreceptors, which, as described above, coexpress LpOps1-4 and LpOps5. LpUVOps1 is expressed in a morphologically distinct class of smaller VE photoreceptors (Herman, 1991) where it is coexpressed with LpOps5 (Battelle et al., 2014). Small VE photoreceptors are most abundant at the distal end of the VE nerve in and near the end organ where they cluster among giant photoreceptors. Small LpUVOps1-LpOps5-containing photoreceptors were also detected in larval MEs and LEs (Battelle et al., 2014). Spectral sensitivity curves recorded from small VE photoreceptors show 2 peaks, one at 360 nm and a second at 520 nm, indicating that both opsins form functional photopigments and that the small photoreceptors are sensitive to both UV and visible light and therefore are classified as UV-VIS cells. Electrophysiological recordings suggested further that the 2 opsins utilize the same downstream transduction machinery and that in cells near the end organ, the ratio of LpUVOps1 to LpOps5 is about 10:1. However, as in the rhabdoms of LE retinular cells and giant VE photoreceptors, the ratio of the opsins coexpressed in small VE photoreceptor rhabdoms changes with time of day. In animals maintained under natural illumination, the concentration of Rh-LpOps5 in small VE photoreceptors in the end organ is significantly lower during the day in the light compared to during the night in the dark (Battelle et al., 2014), whereas the concentration of Rh-LpUVOps1 in these same cells does not change significantly from day to night even in animals exposed to full-spectrum natural daylight. The daily changing ratio in the concentrations of Rh-LpOps5 to Rh-LpUVOps1 predicts that the spectral sensitivity of small VE photoreceptors changes from day to night, with the cells becoming relatively more sensitive to visible light during the night compared to during the day.

Further studies of VE photoreceptors in vivo showed that the concentration of Rh-LpOps5 observed in both giant and small VE photoreceptors during the day depends on the photoreceptor's location along the VE nerve. Specifically, the concentration of Rh-LpOps5 decreases significantly during the day only in VE photoreceptors located in the end organ at the distal end of the VE nerve and not in photoreceptors at more proximal locations along the VE nerve (Battelle et al., 2014). A major difference between photoreceptors at these two locations is probably the amount of light they are exposed to. In vivo, end organ photoreceptors are located directly below clear, lenslike regions within the pigment-free ventral organ. They are probably exposed to much brighter light than more proximal VE photoreceptors, which are surrounded by the hepatopancrease and shadowed by a layer of pigment cells below the cuticle. But light clearly penetrates to proximal VE photoreceptors at least to an intensity sufficient to drive LpOps1-4 shedding and sharply reduce the concentration of Rh-LpOps1-4 in giant VE photoreceptors during the day. This reduction is probably due mostly to Rh-LpOps1-4 transient shedding, which is triggered by dim light. Since Rh-LpOps5 appears to undergo only light-driven shedding (Fig. 7), results with VE photoreceptors indicate that light-driven shedding can produce a significant reduction in the concentration of Rh-LpOps5 but only in photoreceptors exposed to very bright light, like the photoreceptors in the VE end organ.

The paralogs LpOps6 and LpOps7 are coexpressed with LpOps8 in visible light-sensitive photoreceptors in MEs

As was described in the Introduction, electrophysiological studies showed that the MEs contain both UV- and visible light-sensitive photoreceptors. Electrophysiological studies also provided the first hint that the opsin expressed in visible light-sensitive ME photoreceptors might be different from that expressed in LE retinular cells and giant VE photoreceptors. Although the maximum sensitivity of visible light-sensitive photoreceptors in all 3 eye types is about 520 nm, ME photoreceptors showed greater sensitivity at longer wavelengths (Nolte and Brown, 1970).

LpOps6, LpOps7, and LpOps8 are all classified as visible light-sensitive opsins based on their amino acid sequences and their phylogenetic relationships to other opsins (Fig. 2) (Battelle et al., 2015). A series of double-label in situ hybridization studies and double-label studies involving immuno-cytochemistry and in situ hybridization (Battelle et al., 2015) revealed that LpOps6, LpOps7, and LpOps8 are consistently expressed together in ME photoreceptors; that none of them are detected in LpUVOps1-expressing ME photoreceptors; and that LpOps6, LpOps7, and LpOps8 expressing photoreceptors are less abundant in MEs than those expressing LpU VOps1. All of these observations are consistent with findings from previous electrophysiological studies (Nolte and Brown, 1969, 1970, 1972).

Much remains to be learned about LpOps6, LpOps7, and LpOps8; however, an extensive search for their transcripts in each eye type, the CNS, and the tail provided clear evidence that all three are ME specific (Fig. 5) (Battelle et al., 2016). The individual spectral sensitivities of LpOps6, LpOps7, and LpOps8 are not known; however, the spectral sensitivity curve recorded from visible light-sensitive ME photoreceptors has a single peak that is not unusually broad (Nolte and Brown, 1970). This suggests the spectra of LpOps6, LpOps7, and LpOps8 are very similar. Immunocytochemical assays showed that LpOps6 is present in rhabdoms. It is also detected in vesicles distributed throughout the photoreceptor cell body, which suggests that LpOps6 is shed from rhabdoms (Battelle et al., 2015). However, the relative concentrations of LpOps6, LpOps7, and LpOps8 in rhabdoms are not yet known because specific antibodies directed against LpOps7 and LpOps8 are not yet available, nor is it known whether their relative concentrations in rhabdoms change from day to night. Although the visible light-sensitive opsins in ME photoreceptors have now been characterized, the functional relevance of visible light photoreception in MEs remains largely a mystery (see Introduction).

The LpUVOps1-expressing ME photoreceptors and LE eccentric cells are the only retinal neurons in which opsin coexpression has not been detected

The LpUVOps1-expressing ME photoreceptors are the only exceptions to the observation that classical eye photoreceptors in Limulus express multiple opsins. As was described above, none of the visible light-sensitive opsin transcripts are coexpressed with LpUVOps1, and LpUVOps1 is the only UV opsin detected in MEs (Fig. 5). The concentration of LpUVOps1 in ME rhabdoms also does not change from day to night, and no obvious LpUVOps1-containing multivesicular bodies are detected in photoreceptors even in animals exposed to full-spectrum natural sunlight during the day (Battelle et al., 2014). This suggests that LpUVOps1 undergoes minimal light-driven shedding.

LpUVOps1 is also the only opsin detected in LE eccentric cells where it concentrates in dendrites, which form electrical junctions with retinular cells. Finding any opsin in eccentric cells was a surprise because electrophysiological studies indicated that they were second-order neurons and not photosensitive (Waterman and Wiersma, 1954). Finding LpUVOps1 in eccentric cells was even more surprising because LEs were thought not to be UV sensitive. In LEs, the visible light-sensitive opsins LpOps1-4 and LpOps5 are quantitatively far more abundant than LpUVOps1 (Battelle et al., 2014), and visible light is the major driver of the LE's photoresponse. But the presence of LpUVOps1 in the LE's major output neuron indicates that UV light may modulate the eye's visible light-driven output.

Opsin expression in Limulus eye photoreceptors: summary and perspectives

The findings reviewed above, concerning opsins that are confirmed to be expressed in the photoreceptors of Limulus eyes, have expanded significantly our understanding of the biochemical and functional complexity of these photoreceptors.

1. Most Limulus photoreceptors express multiple opsins. Examples of opsin coexpression are increasing as opsin expression is investigated in more detail in a greater variety of species. Opsin coexpression has also been described in crustaceans (Sakamoto et al., 1996; Rajkumar et al., 2010) and in insects (Kitamoto et al., 1998; Gao et al., 2000; Arikawa et al., 2003; Mazzoni et al., 2008; Hu et al., 2011, 2014; Ogawa et al., 2012; Schmeling et al., 2014), but in most species examined so far, opsin coexpression has been detected in only one or a few specific photoreceptor types that are often located in specialized regions of the retina. By contrast, opsin coexpression in Limulus seems the rule. Opsin coexpression in the LWS photoreceptors of the MEs is probably a consequence of opsin gene duplication in the xiphosuran lineage (Fig. 4). But the coexpression of opsins from different clades--LpOps5 with LpOps1-4 in LE retinular cells and giant VE photoreceptors and LpOps5 with LpUVOps1 in smaller VE photoreceptors--could be a plesiomorphic characteristic that Limulus has retained from its euchelicerate ancestor.

2. The relative concentrations of coexpressed opsins in Limulus LE and VE rhabdoms change with a diurnal rhythm. Studies of LE retinular cells showed that this is, in part, because the coexpressed opsins differ in how their turnover is influenced by signals from an internal circadian clock. The literature contains many examples of light-driven rhabdom shedding (White and Lord, 1975; Blest, 1980; Arikawa et al., 1987; Eakin, 1988), but studies with Limulus eyes are the first to demonstrate quantitatively that shedding can produce a change in the concentration of opsins in rhabdoms and that the concentrations of coexpressed opsins in rhabdoms can be regulated differently. More recent studies have shown that diurnal changes in the concentration of opsins in rhabdoms are not unique to Limulus. A dramatic light-driven decrease in the concentration of opsins occurs in mosquito rhabdoms (Hu et al., 2012). However, to date, Limulus photoreceptors provide the only clear example for the differential regulation of coexpressed opsins in rhabdoms. Examples in other species may be discovered as opsin expression is examined in more species at the protein level.

3. The visible light-sensitive opsins expressed in ME photoreceptors are unique to that eye. Thus, chelicerates join crustaceans and insects in the list of arthropod groups in which visible light-sensitive opsins expressed in MEs are unique to that eye (Henze and Oakley, 2015). This finding also explains why the spectral sensitivities of visible light-sensitive ME and LE photoreceptors are different (Nolte and Brown, 1970).

4. LpOps5 is clearly a functional visual pigment that is present at a sufficiently high concentration in rhabdoms to impact the photoresponse. Intracellular electrophysiological recordings of small VE photoreceptors show that LpOps5 is a LWS visual pigment (Battelle et al., 2014) even though phylogenetic analyses show it clusters with crustacean MWS opsins (Fig. 2). This finding emphasizes that the spectral sensitivity of opsins in this group cannot be assumed based on phylogeny.

5. Each of the three types of Limulus eyes contains LpUVOps1 -expressing cells, and therefore each eye can detect UV light. This finding expands our understanding of photosensitivity in Limulus eyes. Previous studies suggested that a sensitivity to UV light was restricted to MEs. A sensitivity to UV light in the larval eyes may be particularly important for the survival of newly hatched larvae that lack LEs and MEs. The functional consequences of LpUVOps1 expression in the dendrites of LE eccentric cells are not yet known, but the presence of a UV opsin in the major LE output neuron indicates that LE responses to visible light may be modulated by UV light.

LpPerOpsl Is Expressed in Pigment Cells or Glia in Each Eye Type, CNS, and Tail

Peropsins were first characterized in vertebrate retinas (Sun et al., 1997; Baily and Cassone, 2004), subsequently in spider (Nagata et al., 2010, Eriksson et al., 2013) and Limulus (Battelle et al., 2015) eyes; and, most recently, peropsinlike transcripts have been identified in myriapods, crustaceans, and insects (Henze and Oakley, 2015). Vertebrate and spider peropsins are bistable photopigments that, in the dark, bind all-trans retinal, which is converted to 11-cis retinal by light (Nagata et al., 2010). In eyes, peropsins are most often, but not exclusively, found in glia and pigment cells surrounding photoreceptors (Sun et al., 1997; Baily and Cassone, 2004; Nagata et al., 2010; Eriksson et al., 2013; Battelle et al., 2015). Taken together, these observations have led to the hypothesis that peropsins are retinal photoisomerases (Sun et al., 1997; Chen et al., 2001; Nagata et al., 2010). In rhabdomeric photoreceptors, photoisomerases may recycle chromophore released during the degradation of opsins internalized during rhabdom shedding (Wang et al., 2010, 2012). However, peropsins are clearly not required for recycling chromophore from shed rhabdomeric membranes. For example, rhabdoms in both the principal and secondary eyes of the spider Cupiennius sali undergo dramatic light-driven shedding, but peropsin is expressed only in the secondary eyes (Eriksson et al., 2013).

The cellular distribution of LpPerOps1 transcripts and protein was confirmed in Limulus eyes and CNS. Its cellular distribution in the Limulus tail could not be determined because of the difficulty of performing anatomy on this structure.

In Limulus LEs, in addition to being present in pigment cells surrounding photoreceptors, LpPerOps1 is found in distal pigment cells surrounding the base of the lens and the aperture at the base of the lens (Fig. 1B). The presence of LpPerOps1 in distal pigment cells leads to the additional hypothesis that LpPerOps1 functions in intracellular signaling. Distal pigment cells surrounding the aperture undergo dramatic structural changes from day to night that result in changes in the diameter and length of the aperture. These changes are driven by signals from the clock-activated efferent projection to the eyes and are amplified by light (Chamberlain and Barlow, 1987). The light receptor responsible for this response is not known. Could it be LpPerOps1?

Limulus provides the first example for peropsin expression outside of eyes, which indicates that this opsin may have functions beyond vision. In the brain, LpPerOps1 transcripts are associated with ventral photoreceptor cell bodies that cluster at the brain. This is expected because LpPerOpsl is highly expressed in glia surrounding ventral photoreceptors (Battelle et al., 2015). No LpPerOpsl transcripts are associated with cells within the corpora pedunculata, which makes up the bulk of the brain (Fig. 8A), but they are detected in glia surrounding the lateral optic nerve and in a fiber track possibly within the central body. In the synganglion, transcripts are associated with neuronal clusters located between nerve bundles that project to the periphery (Fig. 8B). Transcripts are also detected in two or three bilateral cell clusters in each segmental ganglion (Fig. 8C). Immunocytochemistry confirmed that LpPerOpsl is not expressed in neurons within the synganglion or segmental ganglia but rather in glia surrounding neurons (Battelle et al., 2016). LpPerOpsl's distribution in the CNS is clearly not uniform. This has led to the speculation LpPerOpsl-containing glia are uniquely associated with opsin-expressing neurons (Battelle et al., 2015). As is described in the next section, a number of opsin transcripts are detected throughout the CNS. Clearly, to fully understand the functions of peropsins, many more studies are required to clarify relationships between peropsin-expressing glia and opsin-expressing neurons.

Many Opsin Transcripts Detected in Limulus Eyes and CNS by the Reverse-Transcriptase Polymerase Chain Reaction Are Not Detected by in situ Hybridization, and Their Proteins Are Not Detected by Immunocytochemistry: A Cautionary Tale

The discussion so far has focused on opsins with expression patterns that have been confirmed by immunocytochemistry or in situ hybridization. However, several opsin transcripts identified in eyes and all those identified in the CNS, except LpPerOps 1, are detected by reverse-transcriptase polymerase chain reaction (RT-PCR) only (Table 1; Fig. 5). These include LpOps9, LpOps10, and UVOps2, which cluster with pancrustacean UV7 opsins, as well as LpArthOps 1, LpCOps1, and LpCOps2. In addition, transcripts encoding LpOps1-4 and LpOps5, the major opsins expressed in LEs and larval eyes, are detected in the MEs, throughout the CNS, and in the tail, where their expression could not be confirmed by in situ hybridization or immunocytochemistry. A reasonable assumption is that where an opsin transcript is detected by RT-PCR only, its copy number in cells is low, and therefore its functional significance must be viewed cautiously. This is especially true in tissues where other opsins are highly expressed. For example, LpOps9 transcripts are detected by RT-PCR in LEs where LpOps1-4 and LpOps5 are the major opsins in retinular cells and LpUVOps1 is the major opsin in eccentric cells. LpOps1-4, LpOps5, and LpOps9 transcripts are routinely detected in MEs by RT-PCR where LpOps6, LpOps7, LpOps8, and LpUVOps1 are clearly expressed in photoreceptors. Finally, LpOps9, LpOps10, and LpUVOps2 are detected in VEs where LpOps1-4, LpOps5, and LpUVOps1 are clearly expressed in photoreceptors. In Limulus, opsin transcripts detected in eye photoreceptor so-mata are also detected in photoreceptor axons and terminals in the brain where their proteins are not detected (Battelle et al., 2016). Therefore, the functional relevance of opsin transcripts in regions of the brain innervated by eyes must also be interpreted with caution.

On the other hand, the nonuniform tissue distributions of opsin transcripts with low copy numbers may point to functional specificity. For example, the C-type opsin transcripts are expressed in the CNS and the tail but not the eyes; LpUVOps2 and LpArthOpsl transcripts are expressed in the VEs, throughout the CNS, and in the tail but not in LEs or MEs; and LpOps 10 transcripts are detected only in the tail. Furthermore, the opsins responsible for extraocular photosensitivity in Limulus abdominal ganglia and the tail have not been identified.

The spectral sensitivity of the tail photoreceptors is not known, so it is difficult to predict what opsin(s) might be responsible; but LpOps10, a pancrustacean UV7-type opsin, is particularly interesting because it is expressed exclusively in the tail.

Photoreceptors in each segmental ganglion are maximally sensitive to visible light at or below 420 nm (Mori et al., 2004). Of the opsin transcripts detected in segmental ganglia, LpOps1-4 and LpOps5 probably do not contribute to the photosensitivity of these cells because they are LWS opsins. LpUVOps2 also probably does not contribute because it is a predicted UV-sensitive opsin based on its amino acid sequence. The remaining candidates are LpOps9, which is most closely related to the pancrustacean UV7 opsins, the C-type opsins, and LpArthOpsl (Fig. 2).

The cellular functions of pancrustacean UV7 opsins and arthropsins are unknown. A member of the UV7 opsin clade that is expressed in photoreceptors of the Aedes mosquito, AaRh10, was able to elicit a photoresponse when expressed in transgenic Drosophila (Hu et al., 2014); however, these responses were smaller and slower than expected for a visual rhodopsin. This result might only reflect poor compatibility between AaRh10 and the Drosophila photoreceptor environment. But it is also possible that opsins in the pancrustacean UV7 clade have a function different from that of classical rhabdomeric opsins. The arthropsins have been proposed to function in olfaction (Schumann and Mayer, 2016); however, recent evidence suggests they can also form a photopigment. The opsin responsible for producing diapause in Daphnia magna was recently identified (Roulin et al., 2016), and a phylogenetic analysis of this opsin places it among the arthropsins (B.-A. Battelle, unpubl. data).

Looking at an adult Limulus, with its deeply pigmented dorsal carapace, it is easy to conclude that if there are photosensitive cells in its CNS, they are of no functional relevance to the animal because light will not reach them. In this regard, it is important to recall that illuminating the tail of an adult animal can phase shift its circadian clock (Hanna et al., 1988; Renninger et al., 1997). Thus, light clearly penetrates the carapace surrounding the adult tail. Light may also penetrate other regions of the dorsal carapace. Furthermore, since light-driven shedding of LpOps1-4-containing membranes occurs in vivo in the giant VE photoreceptors located along the proximal VE nerve (Katti et al., 2010; Battelle et al., 2014), light must also penetrate the ventral cuticle of adults. The VEs are exposed to light anytime the animal walks, and adult Limulus swim ventral-side up. It is also important to keep in mind that for at least several years after hatching, juvenile Limulus are nearly transparent; thus, CNS photoreception may be particularly important for the survival of juveniles.

Conclusion

Much is known about the Limulus visual system. The anatomy of Limulus eyes and photoreceptors have been described in detail. The central projections of Limulus photoreceptors, eccentric cells, and arhabdomeric cells are known. The physiological responses of Limulus photoreceptors to light have been characterized in detail. Much is known about the circadian efferent projections to Limulus eyes and their impact on eye and photoreceptor anatomy and physiology. The full repertoire of Limulus opsins is now known, as well as the distribution of their transcripts in the visual system, CNS, and tail. Furthermore, dynamic changes in the rhabdomeric expression of several Limulus opsins have been documented. The Limulus visual system is now arguably the best understood visual system among chelicerates and one of the best characterized among arthropods. Much remains to be learned about the Limulus visual system, but because of the wealth of knowledge already in hand, the availability of a well-assembled genome, and the key position of Limulus in the arthropod family tree, Limulus is a critical preparation for future studies of the evolution and diversification of the structure and function of arthropod visual systems.

Acknowledgments

Work described here that was performed in the Battelle laboratory was supported by the National Science Foundation and the Whitney Laboratory for Marine Bioscience. The author thanks her collaborators and students who contributed to this work, especially Karen E. Kempler, Dr. Megan L. Porter, and Dr. Joseph F. Ryan. She also thanks Dr. Joseph E. O'Tousa for helpful discussions, the reviewers for their thoughtful comments and editing, and the organizers of this Virtual Symposium for the invitation to participate.

Literature Cited

Albalat, R. 2012. Evolution of the genetic machinery of the visual cycle: a novelty of the vertebrate eye? Mol. Biol. Evol. 29: 1461-1469. A Aikawa, K., K. Kawamata, T. Suzuki, and E. Eguchi. 1987. Daily changes of structure, function and rhodopsin content in the compound eye of the crab Hemigrapsus sanguineus. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 161: 161-174.

Arikawa, K., S. Mizuno, M. Kinoshita, and D. Stavenga. 2003. Coexpression of two visual pigments in a photoreceptor causes an abnormally broad spectral sensitivity in the eye of the butterfly Papilio xuthus. J. Neurosci. 23: 4527-4532.

Bailey, M. J., and V. M. Cassone. 2004. Opsin photoisomerases in the chick retina and pineal gland: characterization, localization, and circadian regulation. Investig. Ophthalmol. Vis. Sci. 45: 769-775.

Barlow, R. B. 1983. Orcadian rhythms in the Limulus visual system. J. Neurosci. 3: 856-870.

Barlow, R. B. 2001. Circadian and efferent modulation of visual sensitivity. Prog. Brain Res. 131: 487-503.

Barlow, R. B., and E. Kaplan. 1977. Properties of visual cells in lateral eye of Limulus: in situ intracellular recordings. J. Gen. Physiol. 69: 203-220.

Barlow, R. B., S. C. Chamberlain, and J. Z. Levinson. 1980. Limulus brain modulates the structure and function of the lateral eyes. Science 210: 1037-1039.

Barlow, R. B. J., L. Ireland, and L. Kass. 1982. Vision has a role in Limulus mating behavior. Nature 296: 65-66.

Battelle, B. A. 2002. Circadian efferent input to Limulus eyes: anatomy, circuitry, and impact. Microsc. Res. Tech. 58: 345-355.

Battelle, B. A. 2006. The eyes of Limulus polyphemus (Xiphosura, Chelicerata) and the afferent and efferent projections. Arthropod Struct. Dev. 35: 261-274.

Battelle, B. A. 2013. What the clock tells the eye: lessons from an ancient arthropod. Integr. Comp. Biol. 53: 144-153.

Battelle, B. A. 2016. Simple eyes, extraocular photoreceptors and opsins in the American horseshoe crab. Integr. Comp. Biol. 56: 809-819.

Battelle, B. A., A. Dabdoub, M. A. Malone, A. W. Andrews, C. Cacciatore, B. G. Caiman, W. C. Smith, and R. Payne. 2001. Immunocy-tochemical localization of opsin, visual arrestin, myosin III, and calmodulin in Limulus lateral eye retinular cells and ventral photoreceptors. J. Comp. Neurol. 435: 211-225.

Battelle, B. A., K. E. Kempler, A. K. Parker, and C. D. Caddie. 2013. Opsin 1 -2, [G.sub.q]a and arrestin levels at Limulus rhabdoms are controlled by diurnal light and a circadian clock. J. Exp. Biol. 216: 1837-1849.

Battelle, B.-A., K. E. Kempler, A. Harrison, D. R. Dugger, and R. Payne. 2014. Opsin expression in Limulus eyes: A UV opsin is expressed in each eye type and co-expressed with a visible light-sensitive opsin in ventral larval eyes. J. Exp. Biol. 217: 3133-3145.

Battelle, B.-A., K. E. Kempler, S. R. Saraf, C. E. Marten, D. R. Dugger, Jr., D. I. Speiser, and T. H. Oakley. 2015. Opsins in Limulus eyes: characterization of three visible light-sensitive opsins unique to and co-expressed in median eye photoreceptors and a peropsin/RGR that is expressed in all eyes. J. Exp. Biol. 218: 466-479.

Battelle, B. A., J. F. Ryan, K. E. Kempler, S. R. Saraf, C. E. Marten, W. C. Warren, P. J. Minx, M. J. Montague, P. J. Green, S. A. Schmidt etal. 2016. Opsin repertoire and expression patterns in horseshoe crabs: evidence from the genome of Limulus polyphemus (Arthropoda: Chelicerata). Genome Biol. Evol. 8: 1571-1589.

Blest, A. D. 1980. Photoreceptor membrane turnover in arthropods: comparative studies of breakdown processes and their implications. Pp. 217-245 in The Effects of Constant Light on Visual Processes, T. P. Williams and B. N. Baker, eds. Plenum, New York.

Brown, J. E., L. J. Rubin, A. J. Ghalayini, A. P. Tarver, R. F. Irvine, M. J. Berridge, and R. E. Anderson. 1984. Myoinositol polyphosphate may be a messenger for visual excitation in Limulus photoreceptors. Nature 311: 160-163.

Caiman, B., and S. Chamberlain. 1982. Distinct lobes of Limulus ventral photoreceptors. II. Structure and ultrastructure. J. Gen. Physiol. 80: 839-862.

Chamberlain, S. C, and R. B. Barlow, Jr. 1979. Light and efferent activity control rhabdom turnover in Limulus photoreceptors. Science 206: 361-363.

Chamberlain, S. C., and R. B. Barlow, Jr. 1984. Transient membrane shedding in Limulus photoreceptors--control mechanisms under natural lighting. J. Neurosci. 4: 2792-2810.

Chamberlain, S. C, and R. B. Barlow, Jr. 1987. Control of structural rhythms in the lateral eye of Limulus: interactions of natural lighting and circadian efferent activity. J. Neurosci. 7: 2135-2144.

Chen, P., W. S. Hao, L. Rife, X. P. Wang, D. W. Shen, J. Chen, T. Ogden, G. B. Van Boemel, L. Y. Wu, M. Yang et al. 2001. A photic visual cycle of rhodopsin regeneration is dependent on Rgr. Nat. Genet. 28: 256-260.

Clark, A. W., R. Millecchia, and A. Mauro. 1969. Ventral photoreceptor cells of Limulus. I. Microanatomy. J. Gen. Physiol. 54: 289-309.

Dalai, J. S., R. N. Jinks, C. Cacciatore, R. M. Greenberg, and B. A. Battelle. 2003. Limulus opsins: diurnal regulation of expression. Vis. Neurosci. 20: 523-534.

Dalton, B. E., E. R. Loew, T. W. Cronin, and K. L. Carleton. 2014. Spectral tuning by opsin coexpression in retinal regions that view different parts of the visual field. Proc. R. Soc. Biol. Sci. B 281: 20141980.

Dereeper, A., V. Guignon, G. Blanc, S. Audic, S. Buffet, F. Chevenet, J. F. Dufayard, S. Guindon, V. Lefort, M. Lescot et al. 2008. Phylogeny.fr: robust phylogenetic analysis for the non-specialist. Nucleic Acids Res. 36: W465-W469.

Dereeper, A., S. Audic, J. M. Claverie, and G. Blanc. 2010. BLAST-EXPLORER helps you building datasets for phylogenetic analysis. BMC Evol. Biol. 10: 8.

Eakin, R. M. 1988. Turnover of photoreceptoral membranes in invertebrates--a review. Comp. Biochem. Physiol. C Comp. Pharmacol. 91: 247-250.

Edgecombe, G. D., and D. A. Legg. 2014. Origins and early evolution of arthropods. Palaeontology 57: 457-468.

Eriksson, B. J., D. Fredman, G. Steiner, and A. Schmid. 2013. Characterization and localization of the opsin protein repertoire in the brain and retinas of a spider and an onychophoran. BMC Evol. Biol. 13: 186.

Fahrenbach, W. H. 1969. Morphology of eyes of Limulus. 2. Ommatidia of the compound eye. Z. Zellforsch. Mikrosk. Anat. 93: 451-487.

Fahrenbach, W. H. 1975. Visual system of horseshoe crab Limulus polyphemus. Int. Rev. Cytol. 41: 285-349.

Fahrenbach, W. H. 1985. Anatomical circuitry of lateral inhibition in the eye of the horseshoe crab, Limulus polyphemus. Proc. R. Soc. Biol. Sci. B 225: 219-249.

Fahrenbach, W. H., and A. J. Griffin. 1975. Morphology of Limulus visual system. 6. Connectivity in ocellus. Cell Tiss. Res. 159: 39-47.

Fleissner, G. 1988. Information Processing in Animals, Vol. 5, Efferent Control of Visual Sensitivity in Arthropod Eyes with Emphasis on Circadian Rhythms. Gustav Fischer, New York.

Gao, N., R. G. Foster, and J. Hardie. 2000. Two opsin genes from the vetch aphid, Megoura viciae. Insect Mol. Biol. 9: 197-202.

Hanna, W. J. B., J. A. Home, and G. H. Renninger. 1988. Circadian photoreceptor organs in Limulus. 2. The telson. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 162: 133-140.

Hartline, H. K., and C. H. Graham. 1932. Nerve impulses from single receptors in the eye. J. Cell. Comp. Physiol. 1: 277-295.

Hartline, H. K., H. G. Wagner, and F. Ratliff. 1956. Inhibition in the eye of Limulus. J. Gen. Physiol. 39: 651-673.

Heinrichs, S., and G. Fleissner. 1987. Neuronal components of the circadian clock in the scorpion, Androctonus australis--central origin of the efferent neurosecretory elements projecting to the median eyes. Cell Tiss. Res. 250: 277-285.

Henze, M. J., and T. H. Oakley. 2015. The dynamic evolutionary history of pancrustacean eyes and opsins. Integr. Comp. Biol. 55: 830-842.

Herman, K. G. 1991. Two classes of Limulus ventral photoreceptors. J. Comp. Neurol. 303: 1-10.

Home, J. A., and G. H. Renninger. 1988. Circadian photoreceptor organs in Limulus. I. Ventral, median, and lateral eyes. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 162: 127-132.

Hu, X., M. T. Leming, A. J. Metoxen, M. A. Whaley, and J. E. O'Tousa. 2012. Light-mediated control of rhodopsin movement in mosquito photoreceptors. J. Neurosci. 32: 13661-13667.

Hu, X. B., M. A. Whaley, M. M. Stein, B. E. Mitchell, and J. E. O'Tousa. 2011. Coexpression of spectrally distinct rhodopsins in Aedes aegypti R7 photoreceptors. PLoS One 6: e23121.

Hu, X. B., M. T. Leming, M. A. Whaley, and J. E. O'Tousa. 2014. Rhodopsin coexpression in UV photoreceptors of Aedes aegypti and Anopheles gambiae mosquitoes. J. Exp. Biol. 217: 1003-1008.

Jones, C., J. Nolte, and J. E. Brown. 1971. Anatomy of median ocellus of Limulus. Z. Zellforsch. Mikrosk. Anat. 118: 297-309.

Kashiyama, K., T. Seki, H. Numata, and S. G. Goto. 2009. Molecular characterization of visual pigments in Branchiopoda and the evolution of opsins in Arthropoda. Mol. Biol. Evol. 26: 299-311.

Kass, L., and G. H. Renninger. 1988. Circadian change in function of Limulus ventral photoreceptors. Vis. Neurosci. 1: 3-11.

Katti, C., K. Kempler, M. L. Porter, A. Legg, R. Gonzalez, E. Garcia-Rivera, D. Dugger, and B. A. Battelle. 2010. Opsin co-expression in Limulus photoreceptors: differential regulation by light and a circadian clock. J. Exp. Biol. 213: 2589-2601.

Kenny, N., K. Chan, W. Nong, Z. Qu, I. Maeso, H. Yip, T. Chan, H. Kwan, P. Holland, K. Chu, and J. Hui. 2015. Ancestral whole-genome duplication in the marine chelicerate horseshoe crabs. Heredity 116: 190-199.

Kitamoto, J., K. Sakamoto, K. Ozaki, Y. Mishina, and K. Arikawa. 1998. Two visual pigments in a single photoreceptor cell: identification and histological localization of three mRNAs encoding visual pigment opsins in the retina of the butterfly Papilio xuthus. J. Exp. Biol. 201: 1255-1261.

Lall, A. B. 1970. Spectral sensitivity of intracellular responses from visual cells in median ocellus of Limulus polyphemus. Vis. Res. 10: 905-909.

Lall, A. B., and R. Chapman. 1973. Phototaxis in Limulus under natural conditions: evidence for reception of near-ultraviolet light in median dorsal ocellus. J. Exp. Biol. 58: 213-234.

Mazzoni, E. O., A. Celik, F. W. C. Mathias, D. Vasiliauskas, R. J. Johnston, T. A. Cook, F. Pichaud, and C. Desplan. 2008. Iroquois complex genes induce co-expression of rhodopsins in Drosophila. PLoS Biol. 6: e97.

Millecchia, R., J. Bradbury, and A. Mauro. 1966. Simple photoreceptors in Limulus polyphemus. Science 154: 1199-1201.

Mori, K., and T. Kuramoto. 2004. Photosensitivity of the central nervous system of Limulus polyphemus. Zool. Sci. 21: 731-737.

Mori, K., T. Saito, and T. Kuramoto. 2004. Physiological and morphological identification of photosensitive neurons in the opisthosomal ganglia of Limulus polyphemus. Biol. Bull. 207: 209-216.

Nagata, T., M. Koyanagi, H. Tsukamoto, and A. Terakita. 2010. Identification and characterization of a protostome homologue of peropsin from a jumping spider. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 196: 51-59.

Nolte, J., and J. E. Brown. 1969. Spectral sensitivities of single cells in median ocellus of Limulus. J. Gen. Physiol. 54: 636-649.

Nolte, J., and J. E. Brown. 1970. Spectral sensitivities of single receptor cells in lateral, median, and ventral eyes of normal and white-eyed Limulus. J. Gen. Physiol. 55: 787-801.

Nolte, J., and J. E. Brown. 1972. Electrophysiological properties of cells in median ocellus of Limulus. J. Gen. Physiol. 59: 167-185.

Nossa, C, P. Havlak, J.-X. Yue, J. Lv, K. Vincent, H. Brockmann, and N. Putnam. 2014. Joint assembly and genetic mapping of the Atlantic horseshoe crab genome reveals ancient whole-genome duplication. GigaScience 3: 9.

Obst, M., S. Faurby, S. Bussarawit, and P. Funch. 2012. Molecular phylogeny of extant horseshoe crabs (Xiphosura, Limulidae) indicates Paleogene diversification of Asian species. Mol. Phylogenet. Evol. 62: 21-26.

Ogawa, Y., H. Awata, M. Wakakuwa, M. Kinoshita, D. G. Stavenga, and K. Arikawa. 2012. Coexpression of three middle wavelength-absorbing visual pigments in sexually dimorphic photoreceptors of the butterfly Colias erate. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 198: 857-867.

Pattengale, N. D., M. Alipour, O. R. Bininda-Emonds, B. M. Moret, and A. Stamatakis. 2010. How many bootstrap replicates are necessary? J. Comput. Biol. 17: 337-354.

Payne, R., D. W. Corson, and A. Fein. 1986a. Pressure injection of calcium both excites and adapts Limulus ventral photoreceptors. J. Gen. Physiol. 88: 107-126.

Payne, R., D. W. Corson, A. Fein, and M. J. Berridge. 1986b. Excitation and adaptation of Limulus ventral photoreceptors by inositol 1,4,5 trisphosphate result from a rise in intracellular calcium. J. Gen. Physiol. 88: 127-142.

Porter, M. L., J. R. Blasic, M. J. Bok, E. G. Cameron, T. Pringle, T. W. Cronin, and P. R. Robinson. 2012. Shedding new light on opsin evolution. Proc. R. Soc. Biol. Sci. B 279: 3-14.

Rajkumar, P., S. M. Rollmann, T. A. Cook, and J. E. Layne. 2010. Molecular evidence for color discrimination in the Atlantic sand fiddler crab, Uca pugilator. J. Exp. Biol. 213: 4240-4248.

Regier, J. C, J. W. Shultz, A. Zwick, A. Hussey, B. Ball, R. Wetzer, J. W. Martin, and C. W. Cunningham. 2010. Arthropod relationships revealed by phylogenomic analysis of nuclear protein-coding sequences. Nature 463: 1079-1098.

Renninger, G., C. Lajoie, W. J. B. Hanna, D. Fong, C. House, and J. Zelin. 1997. Phase-shifting and entrainment of a circadian rhythm in Limulus polyphemus by ocular and extraocular photoreceptors. Biol. Rhythm Res. 28: 50-68.

Rokas, A., and P. W. H. Holland. 2000. Rare genomic changes as a tool for phylogenetics. Trends Ecol. Evol. 15: 454-459.

Roulin, A. C, Y. Bourgeois, U. Stiefel, J. C. Walser, and D. Ebert. 2016. A photoreceptor contributes to the natural variation of diapause induction in Daphnia magna. Mol. Biol. Evol. 33: 3194-3204.

Sacunas, R. B., M. O. Papuga, M. A. Malone, A. C. Pearson, M. Marjanovic, D. G. Stroope, W. W. Weiner, S. C. Chamberlain, and B. A. Battelle. 2002. Multiple mechanisms of rhabdom shedding in the lateral eye of Limulus polyphemus. J. Comp. Neurol. 449: 26-42.

Sakamoto, K., O. Hisatomi, F. Tokunaga, and E. Eguchi. 1996. Two opsins from the compound eye of the crab Hemigrapsus sanguineus. J. Exp. Biol. 199: 441-450.

Schmeling, F., M. Wakakuwa, J. Tegtmeier, M. Kinoshita, T. Bockhorst, K. Arikawa, and U. Homberg. 2014. Opsin expression, physiological characterization and identification of photoreceptor cells in the dorsal rim area and main retina of the desert locust, Schistocerca gregaria. J. Exp. Biol. 217: 3557-3568.

Schumann, I., L. Hering, and G. Mayer. 2016. Immunolocalization of arthropsin in the onychophoran Euperipatoides rowelli (Peripatopsidae). Front. Neuroanat. 10: 80.

Smith, W. C, D. A. Price, R. M. Greenberg, and B. A. Battelle. 1993. Opsins from the lateral eyes and ocelli of the horseshoe crab, Limulus polyphemus. Proc. Natl. Acad. Sci. U.S.A. 90: 6150-6154.

Speiser, D. I., M. S. Pankey, A. K. Zaharoff, B. A. Battelle, H. D. Bracken-Grissom, J. W. Breinholt, S. M. Bybee, T. W. Cronin, A. Garm, A. R. Lindgren et al. 2014. Using phylogenetically-informed annotation (PIA) to search for light-interacting genes in transcriptomes from non-model organisms. BMC Bioinformatics 15: 350.

Stamatakis, A. 2014. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinfonnatics 30: 1312-1313.

Stamatakis, A., P. Hoover, and J. Rougemont. 2008. A rapid bootstrap algorithm for the RAxML web servers. Syst. Biol. 57: 758-771.

Sun, H., D. J. Gilbert, N. G. Copeland, N. A. Jenkins, and J. Nathans. 1997. Peropsin, a novel visual pigment-like protein located in the apical microvilli of the retinal pigment epithelium. Proc. Natl. Acad. Sci. U.S.A. 94: 9893-9898.

Wang, X. Y., T. Wang, Y. C. Jiao, J. von Lintig, and C. Montell. 2010. Requirement for an enzymatic visual cycle in Drosophila. Curr. Biol. 20:93-102.

Wang, X. Y., T. Wang, J. F. D. Ni, J. von Lintig, and C. Montell. 2012. The Drosophila visual cycle and de novo chromophore synthesis depends on rdhB. J. Neurosci. 32: 3485-3491.

Waterman, T. H., and C. A. G. Wiersma. 1954. The functional relation between retinal cells and optic nerve in Limulus. J. Exp. Zool. 126: 59-85.

White, R. H., and E. Lord. 1975. Diminution and enlargement of mosquito rhabdom in light and darkness. J. Gen. Physiol. 65: 583-598.

Yamashita, S. 2002. Efferent innervation of photoreceptors in spiders. Microsc. Res. Tech. 58: 356-364.

BARBARA-ANNE BATTELLE (*)

Whitney Laboratory for Marine Bioscience, University of Florida, 9505 Ocean Shore Boulevard, St. Augustine, Florida 32080

Received 23 January 2017; Accepted 12 June 2017; Published online 31 October 2017.

(*) E-mail: battelle@whitney.ufl.edu.

Abbreviations: CNS, central nervous system; LE, lateral eye; LpArthOps, Limulus arthropsin; LpCOps, Limulus C-type opsin; LpOps, Limulus opsin; LpPerOps, Limulus peropsin; LWS, long wavelength-sensitive; ME, median eye; MWS, medium wavelength-sensitive; Rh-LpOps, Limulus opsin in rhabdoms; SWS, short wavelength-sensitive; VE, ventral eye.
Table 1
Maximum spectral sensitivities of Limulus polyphemus photoreceptors
identified with electrophysiological recordings, the opsins they are
known to express, and other opsin transcripts in the same tissue


Photoreceptive tissue  Photoreceptor type

Lateral eye            Retinular cell

Larval eyes            Giant

                       Smaller
Median ocelli          Visible light-sensitive

                       Ultraviolet light-sensitive
Segmental ganglia      100 [micro]m neurons



Tail                   Unknown





Photoreceptive tissue  Maximum sensitivity (nm)

Lateral eye            520-525

Larval eyes            520-525

                       520-525 and 350
Median ocelli          520-525

                       350
Segmental ganglia      425 or below



Tail                   Visible light-sensitive,
                       maximum sensitivity
                       unknown


                       Opsins confirmed in
Photoreceptive tissue  identified photoreceptors

Lateral eye            LpOps 1-4, LpOps5

Larval eyes            LpOps 1-4, LpOps5

                       LpOps5, UVOps 1
Median ocelli          LpOps6, LpOps7, LpOps8

                       LpUVOps1
Segmental ganglia      None



Tail                   None



                       Other opsin transcripts
Photoreceptive tissue  expressed in the tissue  References

Lateral eye            LpOps9, PerOps1,         a, b, c, d, f, j, k
                       UVOpsl
Larval eyes            LpOps9, UVOps2,          a, b, c, d, f, i
                       ArthOps1, PerOps1

Median ocelli          LpOps 1-4, LpOps5,       b, c, d, f, j, k
                       LpOps9, PerOps1

Segmental ganglia      LpOps 1-4, LpOps5,       d,g,h
                       LpOps9, UVOps2,
                       ArthOps1, PerOps1,
                       COps1, COps2
Tail                   LpOps 1-4, LpOps9,       d, e
                       LpOps10, UVOps2,
                       ArthOps1, PerOps1,
                       COps1, COps2

Opsin expression was confirmed by in situ hybridization or
immunocytochemical assays.
References: a, Battelle et al., 2001; b, Battelle et al., 2014; c,
Battelle et al., 2015; d, Battelle et al., 2016; e, Hanna et al.,
1988; f, Katti et al., 2010; g, Mori and Kuramoto, 2004; h, Mori et
al., 2004; i, Millecchia et al., 1966; j, Nolte and Brown, 1969; k,
Smith et al., 1993. From B. A. Battelle, "Simple eyes, extraocular
photoreceptors and opsins in the American horseshoe crab," Integrative
and Comparative Biology, 2016, 56, 5, 809-819, by permission of Oxford
University Press.
COPYRIGHT 2017 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Battelle, Barbara-Anne
Publication:The Biological Bulletin
Article Type:Report
Date:Aug 1, 2017
Words:11380
Previous Article:New Insights from Genetic Data Sets on the Function and Evolution of Visual Systems: Introduction to a Virtual Symposium in The Biological Bulletin.
Next Article:Molecular Evolution of Spider Vision: New Opportunities, Familiar Players.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters