3.2.1. Coupling of Photoexcited Rhodopsin to Inositol Phospholipid Hydrolysis

TRP is an illuminating example of a novel signaling protein of prime importance whose physiological function and interactions with other signaling proteins have been elucidated due to the powerful genetic tools and functional tests that have been developed in Drosophila. These studies have led to the identification and characterization of TRP as a light-sensitive and Ca²⁺-permeable channel (Minke, 2010; Montell, 2011; Hardie, 2011).

Illumination of fly photoreceptors induces a cascade of enzymatic reactions, which result in activation of the light-sensitive TRP channels (Minke, 2010; Devary et al., 1987). To function as a reliable light monitor, each stage of the phototransduction cascade needs an efficient mechanism of activation as well as an equally efficient mechanism of termination, ensuring that, at the cessation of the light stimulus, the photoreceptor potential will rapidly reach dark baseline. The use of specific nonphysiological conditions and mutated Drosophila strains has assisted in uncovering conditions in which specific stages underlying the light response failed to activate or terminate, and the TRP channels remain inactive or active in the dark. These experiments have revealed important mechanisms that regulate the enzymatic cascade controlling the activation and termination of the Drosophila TRP channel. These mechanisms are likely to also control the activation and regulation of nonvisual mammalian TRP channels.

3.2.2. Rhodopsin and the Photochemical Cycle

The G protein-coupled receptor (GPCR), rhodopsin (R), is composed of a seven-transmembrane (TM) protein, opsin and the chromophore, 11-cis 3 hydroxyretinal (Vogt and Kirschfeld, 1984). Isomerization of the chromophore by photon absorption to all-trans retinal induces a conformational change in the opsin resulting in a dark stable and physiologically active photoproduct, metarhodopsin (M), that is stable in the dark (Figure 3.3). The action spectrum of this reaction depends on the specific R type and spans a wavelength range between ultraviolet (UV) and green light. To ensure high sensitivity, high temporal resolution, and low dark noise of the photoresponse, the active M has to be quickly inactivated and recycled. The latter requirement is achieved, in invertebrates, by two means; the absorption of an additional photon by dark-stable M, which photoconverts M back to R, or by a multistep photochemical cycle (Byk et al., 1993; Selinger et al., 1993). The photochemical cycle begins by M phosphorylation (Mpp) at multiple sites by rhodopsin kinase (Doza et al., 1992) and consequently the binding of fly Arrestin 2 (ARR2) protein (Byk, 1993; Alloway et al., 2000), which is accelerated by Ca²⁺ influx acting via calmodulin (CaM) and the NINAC (see Section 3.3) (Liu et al., 2008). ARR2 is then phosphorylated by Ca²⁺ calmodulin-dependent kinase (CaMK II) (Matsumoto et al., 1994). Absorption of a photon, by the phosphorylated M-ARR2 complex (Mpp-ARRp), results in the release of the phosphorylated ARR2 (ARRp) from Mpp and the conversion into phosphorylated R (Rpp). Upon photoregeneration of Mpp to Rpp, Rpp is exposed to phosphatase activity by the rhodopsin phosphatase RDGC (Steele et al., 1992; Steele and O'Tousa, 1990), which converts it back to R (Byk et al., 1993; Selinger et al., 1993). Alternatively, in the dark, Mpp-ARR is internalized and degraded by a clathrin-mediated endocytosis (Byk et al., 1993; Kiselev et al., 2000).

Figure 3.3

The photochemical cycle: the “turn-on” and “turn-off” of the photopigment. Upon photoconversion of rhodopsin (R) to metarhodopsin (M), by illuminating with blue light, M is phosphorylated at multiple sites by rhodopsin (more...)

3.2.3. Photoreceptor Potential and Prolonged Depolarizing Afterpotential: A Tool for Discovery of Phototransduction Components

Failure of response termination at the stage of R activation was designated the prolonged depolarizing after (PDA) potential by Hillman, Hochstein, and Minke (Hillman et al., 1983; Minke, 2012). The PDA, like the light coincident receptor potential, arises from light-induced opening of the TRP channels in the plasma membrane. However, in contrast to the light coincident receptor potential, which quickly declines to baseline after the cessation of the light stimulus, the PDA is a depolarization that continues long after light offset (Figure 3.4) (see Hillman et al., 1983; Minke, 2012 for reviews). The PDA has been a major tool to screen for visual mutants of Drosophila and a powerful experimental tool to unravel phototransduction (Pak, 1991; Pak et al., 2012), because it drives the phototransduction cascade to the upper limit of its activation. The PDA is obtained due to (1) the bi-stable property of fly photopigment with separated peak absorption spectra of R and M states, in Drosophila this is obtained by a blue absorbing R spectrum and orange absorbing M spectrum, and (2) the low expression level of ARR2 relative to M, while ARR2 is required to terminate M activity (Byk et al., 1993; Selinger et al., 1993; Dolph et al., 1993). Thus, massive R to M photoconversion by intense blue light induces a PDA, while M to R photoconversion by intense orange light suppresses the PDA. Therefore, the major cause of termination for the active photopigment is either the absorption of an additional photon by M that converts it to R, or the binding of arrestin to M. In summary, the PDA is observed only when a considerable amount of photopigment (>20%) is converted from R to M. The larger the net amount of R to M conversion, the longer the PDA. The PDA can be depressed at any time by photopigment conversion from M to R. The degree of PDA depression depends also on the net amount of M to R conversion (Hillman et al., 1983). After the depression of maximal PDA, additional PDA can be induced immediately by R to M conversion (Figure 3.4). Indeed, the PDA screen yielded a plethora of novel and interesting visual mutants. One group of mutants exhibited a loss in several features of the PDA, and they were termed nina mutants, which stands for “neither inactivation nor afterpotential.” Most nina mutants were found to be caused by reduced levels of R (Figure 3.4). The second group of PDA mutants lost the ability to produce the voltage response associated with the PDA but were still inactivated by strong blue light, and the inactivation could be relieved by orange light. These mutants, consisting of seven allelic groups, were termed inaA–G, which stands for “inactivation but no afterpotential” (Figure 3.4). The ina mutants were found to have normal R levels but deficiencies in proteins associated with the function of the TRP channel. The nina and ina mutants have led to the identification of many of the crucial components of Drosophila phototransduction, some of which are novel proteins of general importance for many cells and tissues (Pak, 1991; Pak et al., 2012).

Figure 3.4

The prolonged depolarizing afterpotential (PDA) response of wild-type Drosophila and its modifications in the nina and ina mutants. Upper trace: ERG recordings from a wild-type white-eyed fly in response to a series of intense blue (B) and orange (O) (more...)

3.2.4. Light-Activated G Protein

Heterotrimeric G proteins mediate a variety of signaling processes by coupling heptahelical receptors to their downstream effectors. It has been well established in photoreceptors of several invertebrate species that photoexcited rhodopsin activates a heterotrimeric G protein (Blumenfeld et al., 1985; Fein, 1986). Direct observations of G protein–mediated chemically induced noise have suggested that there is a three- to fivefold gain at the G protein stage (Minke and Stephenson, 1985; Fein and Corson, 1979), and this was confirmed and extended by mutation analysis (Hardie et al., 2002). Direct demonstration of light-dependent G protein activity in fly eye was first demonstrated in membrane preparations using an α-³²P-labeled azidoanilido analog of GTP (Devary et al., 1987; Minke and Selinger, 1991). Polyacrylamide gel electrophoresis and autoradiography revealed in the blue-illuminated membranes a labeled 41-kDa protein band that was not observed when illuminated with red light (Figure 3.5a). Binding assays, which show strict dependence of GTPγS binding to the eye membrane preparation upon production of metarhodopsin with blue light (Figure 3.5b), also revealed the involvement of a G protein in fly phototransduction (Devary et al., 1987). Later studies in Drosophila, using genetic screens, isolated three genes encoding eye-specific G protein subunits; DG_qα (Lee et al., 1990), G_qβ_e (Dolph et al., 1994), and Gγ_e (Schulz et al., 1999). The isolated eye-specific DG_qα, showed ~75% identity to mouse G_qα, known to activate phospholipase C (PLC) (Lee et al., 1990). The most direct demonstration that DG_qα participates in the phototransduction cascade came from studies of mutants defective in DG_qα which showed highly reduced sensitivity to light. In the isolated Gα_q¹ mutant, DG_qα protein levels are reduced to ~1%, while rhodopsin, G_qβ, and PLC protein levels are virtually normal. The Gα_q¹ mutant exhibits ~1000-fold reduced sensitivity to light and slow response termination (Scott et al., 1995), thus strongly suggesting that there is no parallel pathway mediated by G proteins, as proposed for the Limulus eye (Dorlochter et al., 1997). Manipulations of the DG_qα protein level by an inducible heat-shock promoter made it possible to show a strong correlation between DG_qα protein levels and sensitivity to light, further establishing its major role in Drosophila phototransduction (Scott et al., 1995). Thus, the reason for the reduced sensitivity in G_qα mutants is its role as mediator of the downstream effector PLC. In addition to the reduced sensitivity to light, the Gα_q¹ mutant also shows slow response termination. This phenotype was first explained as a result of reduced Ca²⁺-dependent receptor inactivation. Accordingly, it was suggested that in the Gα_q¹ mutant fly, the imbalance between the number of activated receptors and the light-activated currents and Ca²⁺ influx result in attenuation of receptor inactivation (Liu et al., 2008; Scott et al., 1995). A more recent study has suggested an explanation in which M/G_q interaction affects M/ARR2 binding. According to this hypothesis, the reduced level of G_qα in the Gα_q¹ mutant fly attenuates the binding of ARR2 to the active M state, resulting in slow receptor termination and hence the slow termination of the light response (Hu et al., 2012).

Figure 3.5

The first direct biochemical evidence of blue light–activated G protein in fly photoreceptors. (a) Photoaffinity labeling of light-activated G protein in Musca eye membranes. Eye membranes were preilluminated with either red (r) or blue (b) lights (more...)

The eye-specific G_qβ (G_qβ_e) shares 50% amino acid identity with other Gβ homologue proteins. Two defective Drosophila G_qβ_e (Gβ_e¹ and Gβ_e² ) mutants were isolated, exhibiting ~100-fold reduced sensitivity to light and slow response termination (Dolph et al., 1994). The reduced light sensitivity and large reduction in light-stimulated GTPγ³⁵S binding in the Gβe mutant was first interpreted as indicating a major role for G_qβ in the coupling of G_qα with metarhodopsin (Dolph et al., 1994). However, later studies conducted on these mutants revealed that G_qα is dependent on G_qβ for both membrane attachment and targeting to the rhabdomere, suggesting that the decreased light sensitivity of these mutants may result from the mislocalization of G_qα (Elia et al., 2005). Many studies have shown that Gβγ binds to Gα-GDP switch regions, suggesting that this interaction stabilizes the binding of GDP and suppresses spontaneous receptor-independent activation by preventing spontaneous GDP-GTP exchange (Itoh and Gilman, 1991). However, a physiologically relevant in vivo demonstration of this mechanism was lacking. Analysis of the stoichiometry between G_qα and G_qβ in Drosophila photoreceptors, revealed a twofold excess of G_qβ over G_qα. Genetic elimination of the G_qβ excess by using the G_qβ heterozygote mutant (;;Gβ_e¹/+) led to an increase in the spontaneous activation of the visual cascade in the dark, while reestablishing the excess of G_qβ over G_qα in a double G_qα, G_qβ heterozygote mutant fly (;Gα_q¹/+;Gβ_e¹/+) reduced the dark spontaneous activity back to its WT (wild-type) levels. The increase of dark spontaneous activity in the G_qβ heterozygote was found to be rhodopsin independent. These results suggested that G_qβ excess is essential for the suppression of dark electrical activity produced by spontaneous GDP-GTP exchange on G_qα (Elia et al., 2005).

Although Gγ_e has been genetically isolated and both Gγ_e hypomorph mutants and RNAi targeted against Gγ_e exist, the effect of reduced Gγ_e level has never been examined in Drosophila photoreceptor cells (Schulz et al., 1999). However, reduced sensitivity to light has been observed in flies overexpressing a mutated Gγ_e in which the prenylation site was modified, further establishing the participation of heterotrimeric G proteins in Drosophila phototransduction (Schillo et al., 2004). Lipidation, the covalent lipid modifications of proteins, play a major role in targeting heterotrimeric (αβγ) G proteins to cellular membranes. Indeed, the three main types of lipid modifications of proteins (myristoylation, prenylation, and palmitoylation-thioacylation) are all found among different G proteins and their functional subunits (Chen and Manning, 2001). G_qα has two palmitoylation sites at the N-terminal end, while the Gβγ complex is modified by prenylation at the C-terminal of Gγ. No lipid modification has been found on G_qβ. However, overexpression of Gγ_e in which the prenylation site CAAX motif at the C-terminal was modified by replacing a cysteine with a glycine (C69G) resulted in interference with attachment of the β-subunit to the membrane (Schillo et al., 2004). These results support the notion that the peripheral membrane localization of G_qβ depends on Gγ_e.

Little is known about the regulation of G_qα localization within specialized signaling cells. Studies using Drosophila show that prolonged illumination causes massive and reversible translocation of G_qα from the signaling compartment, the rhabdomere, to the cytosol, associated with marked architectural changes in the rhabdomere (Kosloff et al., 2003). Epistatic analysis showed that G_qα is necessary but not sufficient to bring about the morphological changes in the signaling organelle (Kosloff et al., 2003). Long exposure to light (light raised) followed by minutes of darkness resulted in a large reduction of the efficiency with which each absorbed photon elicited single-photon responses, while the size and shape of each single-photon response were unchanged (Frechter et al., 2007). To dissect the physiological significance of G_qα translocation by light, a series of Drosophila mutants were used. Genetic dissection showed that light-induced translocation of G_qα between the signaling membrane and the cytosol induced long-term adaptation. Physiological and biochemical studies revealed that the sensitivity to light depends on membrane G_qα levels, which were modulated either by light or by mutations that impaired membrane targeting. Thus, long-term adaptation is mediated by the movement of G_qα from the signaling membrane to the cytosol, thereby reducing the probability of each photon to elicit a single-photon response (bump) (Frechter et al., 2007). However, the molecular mechanism of light-dependent G_qα translocation is still unknown.

3.2.5. Role of PLC in Light Excitation and Adaptation

Evidence for a light-dependent G_qα-mediated PLC activity in fly photoreceptors came from combined biochemical and electrophysiological experiments. These experiments conducted on membrane preparations and intact Musca and Drosophila eyes showed coupling of photoexcited rhodopsin to phosphatidylinositol 4,5-biphosphate (PIP₂) hydrolysis (Devary et al., 1987). They furthermore showed that light-activated PIP₂ hydrolysis was inhibited in the norpA mutant allele, which suppressed the response to light (Selinger and Minke, 1988). In this study illumination and G_qα-dependent activation result in accumulation of InsP₃ and inositol-bisphosphate (InsP₂), derived from PIP₂ hydrolysis by PLC (Devary et al., 1987).

The key evidence for the participation of PLC in visual excitation of the fly was achieved by the isolation and analysis of Drosophila PLC gene, designated no receptor potential A (norpA). The norpA mutant has long been a strong candidate for a transduction defective mutant because of its drastically reduced receptor potential. The norpA gene encodes a β-class PLC, predominately expressed in the rhabdomeres, which has high amino acid homology to a PLC extracted from bovine brain (Bloomquist et al., 1988). Transgenic Drosophila, carrying the norpA gene on null norpA background, rescued the transformant flies from all the physiological, biochemical, and morphological defects, which are associated with the norpA mutants (Shortridge et al., 1991). The norpA mutant thus provides essential evidence for the critical role of inositol-lipid signaling in phototransduction, by showing that no excitation takes place in the absence of functional PLC. However, the events required for light excitation downstream of PLC activation remain unresolved.

In general, the cytoplasmic GTP concentration in cells is much higher than that of GDP, making the inactivation process of G_qα by hydrolysis of G_qα -GTP to G_qα -GDP unfavorable. In order to accelerate the GTPase reaction and terminate G_qα and PLC activities, specific GTPase-activating proteins (GAPs) exist. In vitro studies of mammalian PLCβ1 reconstituted into phospholipid vesicles with recombinant M1 muscarinic receptor and G_q/11 (Berstein et al., 1992) had shown that upon receptor stimulation the addition of PLCβ1 increases the rate at which G_q hydrolyses GTP by three orders of magnitude, suggesting its action as GAP. A reduction in the levels of PLC in mutant flies affects the amplitude and activation kinetics of the light response (Pearn et al., 1996), but also mysteriously slows response termination. Biochemical and physiological studies conducted in Drosophila have revealed the requirement for PLC in the induction of GAP activity in vivo. Using several Drosophila norpA mutant flies, a high correlation between PLC protein level, GAP activity, and response termination was observed (Cook et al., 2000). The virtually complete dependence of GAP activity on PLC (Cook et al., 2000) provides an efficient mechanism for ensuring the one photon, one bump relationship (Yeandle and Spiegler, 1973), which is critical for the fidelity of phototransduction at dim light. The apparent inability to hydrolyze GTP without PLC ensures that every activated G protein eventually encounters a PLC molecule and thereby produces a response by the downstream mechanisms. The instantaneous inactivation of the G protein by its target, the PLC, guarantees that every G protein produces no more than one bump (Cook et al., 2000). This apparently complete dependence of GTPase activity on its activator PLC, in flies, differs from the partial dependence of GTPase activity on additional GAP factors in vertebrate phototransduction (Chen et al., 2000). Vertebrate phototransduction depends on specific GAP proteins named regulators of G protein signaling (RGSs) (Arshavsky and Pugh, 1998). Accordingly, genetic elimination of RGS proteins reduces and slows down GAP activity and leads to slow termination of the light response (Chen et al., 2000).

PLC activity is known to be regulated by Ca²⁺ (Rhee, 2001). It has been shown that Ca²⁺ is bound at the catalytic site of PLC and is required as a co-factor for the catalytic reaction (Essen et al., 1996). These studies show that the positive charge of Ca²⁺ is used to counterbalance local negative charges formed in the active site during the course of the catalytic reaction. Accordingly, Ca²⁺ performs electrostatic stabilization of both the substrate and the transition state, thus providing a twofold contribution to lower the energy of the enzyme reaction (Essen et al., 1997). In Drosophila, both in vitro (Running Deer et al., 1995) and in vivo (Hardie, 2005) measurements revealed Ca²⁺ dependence of PLC activity. This activity shows a bell-shaped dependency of NORPA catalytic activity as a function of Ca²⁺ concentration ([Ca²⁺]), with maximal basal activity in the range of 10⁻⁷–10⁻⁵ M [Ca²⁺]. This complex dependency affects both excitation and adaptation of the photoresponse. While at physiological conditions cellular Ca²⁺ levels are sufficient for the activation of NORPA, it was demonstrated that highly reduced Ca²⁺ levels eliminated excitation completely (Minke and Agam, 2003). Furthermore, it was shown that during the light response, rhabdomere Ca²⁺ concentration can reach mM levels (Oberwinkler and Stavenga, 2000; Postma et al., 1999). At this high Ca²⁺ level, NORPA activity is attenuated raising the possibility of physiological relevance (Hardie, 2005). Later studies have shown that the inhibition of NORPA activity by high Ca²⁺ concentration participates in the mechanism of fast light adaptation and prevents depletion of membrane PIP₂. Accordingly, the high Ca²⁺ concentration, reached at the peak response, attenuates NORPA activity preventing PIP₂ depletion and adapting the cells by reducing excitation (Gu et al., 2005). However, this hypothesis still needs to be substantiated by specific mutations of the NORPA, decoupling its activity from its Ca²⁺ regulation.

3.2.6. Phosphoinositide (PI) Cycle

In the phototransduction cascade of Drosophila, light triggers the activation of PLCβ. This catalyzes the hydrolysis of the membrane phospholipid PIP₂ into water-soluble InsP₃ and membrane-bound diacylglycerol (DAG) (Berridge, 1993). Genetic elimination of the single Drosophila InsP₃ receptor had no effect on light excitation (Acharya et al., 1997; Raghu et al., 2000b), questioning the role of InsP₃ in phototransduction (but see Kohn et al., 2015). The continuous functionality of the photoreceptors during illumination is maintained by rapid regeneration of PIP₂ in a cyclic enzymatic pathway (Figure 3.6).

Figure 3.6

The phosphoinosite (PI) cycle. In the phototransduction cascade, light triggers the activation of phospholipase Cβ (PLCβ, encoded by norpA). This catalyzes hydrolysis of the membrane phospholipid PI(4,5)P₂ (PIP₂) into InsP₃ and DAG. DAG (more...)

The phospholipid branch of the phosphatidylinositol (PI) cycle following PLC activation, begins by DAG transport through endocytosis to the endoplasmic reticulum (submicrovillar cisternae [SMC]), and subsequently, inactivation by phosphorylation into phosphatidic acid (PA) via DAG kinase (DGK; encoded by the rdgA) (Masai et al., 1993; Masai et al., 1997) and to cytidine diphosphate diacylglycerol (CDP-DAG) via CDP-DAG synthase (encoded by the cds) (Wu et al., 1995). Both retinal degeneration A (RDGA, the gene product of the retinal degeneration A gene) and CDS are located in extension of the smooth endoplasmic reticulum called SMC (Figure 3.6). PA can be reconverted back to DAG by lipid phosphate phosphohydrolase (Kwon and Montell, 2006) (LPP, also designated phosphatidic acid phosphatase, PAP, encoded by laza) (Kwon and Montell, 2006; Garcia-Murillas et al., 2006) or produced from phosphatidylcholine (PC) by phospholipase D (PLD, encoded by Pld) (LaLonde et al., 2005). DAG is also hydrolyzed by DAG lipase (encoded by inaE) (Leung et al., 2008) predominantly localized outside the rhabdomeres, into polyunsaturated fatty acids (Figure 3.6). Subsequently, CDP-DAG is converted into PI, which is transferred back to the microvillar membrane, by the PI transfer protein (PITP; encoded by the rdgB) (Vihtelic et al., 1991) located in the smooth endoplasmic reticulum. PIP and PIP₂ are produced at the microvillar membrane by PI kinase and PIP kinase, respectively (Raghu et al., 2012).

Mutations in most proteins of the PI pathway result in retinal degeneration. The retinal degeneration phenotypes in these mutants are thought to occur due to Ca²⁺ influx through the light-activated TRP and TRPL channels, making the PI pathway crucial for understanding phototransduction and TRP channel activation. Although it is possible to partially rescue the degeneration phenotypes by genetically eliminating the TRP channels (Raghu et al., 2000a), it is unclear whether these mutations promote channel opening directly or through indirect changes in the photoreceptor leading to their activation. Recent detailed studies by Raghu and colleagues (reviewed in Raghu et al., 2012) have shown that the principles of PI signaling seem largely conserved between Drosophila and other metazoan models. They concluded that with a fully sequenced, relatively compact genome and with sophisticated molecular genetic technology, studies in Drosophila will be influential in understanding the details and physiological implications of PI signaling in general (Raghu et al., 2012).

3.2.7. Single-Photon Responses and Spontaneous Dark Bumps

In Drosophila photoreceptors, light activates PLC, which hydrolyzes PIP₂ into DAG and InsP₃ promoting the opening of the TRP and TRPL channels and resulting in membrane depolarization. The absorption of a single photon promotes the opening of ~15 TRP channels generating currents of ~12 pA (at a holding potential of –70 mV) (Henderson et al., 2000). At total darkness, spontaneous unitary events of ~3 pA, called dark bumps, are also observed at a rate of ~2 per second. The phenomenon of discrete current fluctuation as a result of dark bumps and single-photon absorption raises many questions on the underlying transduction mechanisms, channel regulation, and anatomical structures that enable it.

3.2.7.1. Spontaneous Dark Bumps

The dark bumps originate from spontaneous GDP to GTP exchange on the G_qα subunits. This has been demonstrated by showing that spontaneous bump frequency is highly correlated with the level of membrane attached Gα_q, by eliminating dark bumps in the Gα_q¹ mutant fly (Hardie et al., 2002; Katz and Minke, 2012), by the high rate of spontaneous bumps in the Gβ_e¹/+ heterozygote mutant fly (Elia et al., 2005), and by the reduction of spontaneous bumps as a result of cellular GDP elevation (Katz and Minke, 2012). The spontaneous bumps in Drosophila have a mean amplitude of ~3 pA under physiological conditions, which correspond to ~4 open TRP channels at the peak amplitude (Hardie et al., 2002; Katz and Minke, 2012). However, their amplitude can be enhanced to ~9 pA by removing Mg²⁺ from the extracellular solution, which when present, reduces the single-channel conductance of the TRP channel by an open-channel block mechanism (Hardie and Mojet, 1995; Parnas et al., 2007). The waveform and time to peak of the spontaneous bumps are only slightly faster than that of the light-induced bump supporting the notion that the kinetic parameters of both dark and quantum bumps are determined downstream of PLC (Figure 3.7) (Katz and Minke, 2012). A surprising feature of the spontaneous bumps is their sensitivity to cellular ATP level. Accordingly, reduction in cellular ATP results in enhancement of both bump rate of occurrence (frequency) and amplitude (Katz and Minke, 2012). These two phenotypes are also observed in the rdgA¹/+ mutant fly with reduced DAG kinase activity (Raghu et al., 2000a). Hence, accumulation of putative PLC products due to the reduction in cellular ATP or reduced DAG kinase activity by mutations facilitates spontaneous bump production and affects channel opening. Several mutants have been shown to alter dark bump frequency— these are retinophilin and ninaC—however, the underlying mechanism is still unknown (Chu et al., 2013).

Figure 3.7

Initiation and properties of single-photon responses (quantum bumps) and spontaneous dark bumps. (a) A scheme of the initial stages of phototransduction underlying dark bump production (top) and quantum bump production (bottom). (b) A demonstration of (more...)

In general, the dark spontaneous bumps are considered as the noise of the transduction system constituting a limiting factor for reliable single-photon detection that should be reduced to a minimum (Katz and Minke, 2012).

3.2.8. Single-Photon Responses: Quantum Bumps

Dim light stimulation induces discrete voltage (or current) fluctuations in most invertebrate species, which are called quantum bumps (Yeandle and Spiegler, 1973). Each bump is assumed to be evoked by the absorption of a single photon and obeys a stochastic process described by Poisson statistics (Yeandle and Spiegler, 1973). In Drosophila the mean amplitude of the light-induced bumps is ~12 pA under physiological conditions, which is approximately fourfold larger than that of the dark spontaneous bumps. This has led to the suggestion that light-induced bumps are a result of synchronous activation of ~3–5 G_qα molecules by a single activated rhodopsin, which synchronously activates ~5 PLC molecules and promotes the opening of ~15 TRP channels (Hardie et al., 2002). The light-induced bumps vary in latency, time course, and amplitude for identical stimulation with half-width duration of ~20 ms and a characteristic variable latency of between 20 and 100 ms.

The discrete nature of the bumps is not due to the quantized nature of light. This has been demonstrated by the application of nonquantized stimulus such as GTPγS (nonreversible activation of G_qα), eliciting quantum bump–like events (Fein and Corson, 1981). Further genetic evidence for the discrete nature of the bumps came from two Drosophila mutants, arr2 and ninaC, in which rhodopsin inactivation is attenuated. These mutants respond to the absorption of a single photon with a train of bumps, which do not overlap. This phenotype was shown to be the consequence of failure in the rhodopsin inactivation process and by the assumption that a “refractory period” exists in bump production (Liu et al., 2008; Dolph et al., 1993; Song et al., 2012). Hence, continuous rhodopsin activation elicits trains of discrete events.

A detailed study in Limulus and Locusta photoreceptors has indicated that the latency of the bump is not correlated with the bump waveform and amplitude, thus strongly suggesting that the triggering mechanism of the bump arises from different molecular processes than that determining the bump waveform (Dorloechter and Stieve, 1997; Howard, 1983). This finding was further confirmed by experiments conducted in Drosophila, demonstrating that in norpA^H43 mutant fly, in which PLC catalytic activity is highly reduced, the bump triggering process is attenuated, while little effect is observed on bump waveform (Katz and Minke, 2012). However, both the triggering process and bump waveform are highly modulated by Ca²⁺. At low external Ca²⁺ (100 μM), bump kinetics becomes slow with two distinct “on” phases (an initial slow phase) accompanied by a faster phase (Henderson et al., 2000), while little effect is observed on bump frequency. Further reduction in Ca²⁺ (10 μM) reduces bump kinetics to a single slow phase, and the excitation efficiency of bump production is reduced (Henderson et al., 2000). These results have led to the hypothesis that Ca²⁺ participates in both positive and negative feedback of bump production (Henderson et al., 2000) and bump triggering (Katz and Minke, 2012), but the molecular targets of Ca²⁺-dependent positive and negative feedback in bump production are still unknown.

The single photon–single bump relationship, typical for Drosophila photoreceptors, requires that each step in the visual cascade must have both “turn on” and “turn-off” mechanisms (Figure 3.7). This principle has been demonstrated for rhodopsin which is inactivated by the binding of arrestin and for G_qα and PLC, which are inactivated by the GAP activity of PLC and promote dissociation of G_qα from PLC (Byk et al., 1993; Dolph et al., 1993; Cook et al., 2000). However, the still unknown mechanism by which TRP channels are inactivated, although highly dependent on PLC activity, PKC (protein kinase C; encoded by InaC), and Ca²⁺ concentration, may require an additional molecular component (Gu et al., 2005; Parnas et al., 2007; Parnas et al., 2009). The functional advantage of such a transduction mechanism is obvious; it produces a sensitive photon counter, very well suited for both the sensitivity and the temporal resolution required by the visual system.

A bump is thought to represent the activation of most of the light-activated channels within a single microvillus, thus representing a functional anatomical element. This has been proposed by comparing the estimated channel number underlying each bump (calculated from the division of the single-channel conductance with the mean bump amplitude) (Henderson et al., 2000) and the estimated number of channels in each microvillus (Huber et al., 1996). Moreover, it has been shown that reduced cellular ATP increases low-amplitude bumps of several mutant flies but does not significantly increase the amplitude of WT bumps (Hardie et al., 2002), demonstrating an upper limit for bump size. Further demonstration has been given by showing that conditions that highly reduce bump production by PLC mutation cannot be overcome by increasing the light intensity to a level that activates all the microvilli at once (Katz and Minke, 2012). Hence, the biochemical reactions that promote channel openings do not sum up across different microvilli.

A recent study has presented a quantitative model explaining how bumps emerge from stochastic nonlinear dynamics of the signaling cascade. Three essential “modules” govern the production of bumps in this model: (1) an “activation module” downstream of PLC but upstream of the channels, (2) a “bump-generation module” including channels and Ca²⁺-mediated positive feedback, and (3) a Ca²⁺-dependent “negative-feedback module.” The model shows that the cascade acts as an “integrate and fire” device conjectured formerly by Henderson and Hardie (Henderson et al., 2000), much like the generation of the action potential. The model explains both the reliability of bump formation, low background noise in the dark, were able to capture mutant bump behavior and explains the dependence on external Ca²⁺, which controls feedback regulation (Pumir et al., 2008). This view was extended by showing experimentally that a critical level of PLC activity (threshold) is necessary to trigger bump production (Katz and Minke, 2012). Hence, the synchronization of ~5 PLC by rhodopsin results in sufficient PLC product to surpass the threshold in bump production, while single spontaneous G_qα molecules activating single PLC molecules have low probability of surpassing this threshold and producing a bump. This mechanism of spontaneous dark bump suppression maintains the fidelity of single-photon detection. Moreover, it was demonstrated that the Ca²⁺-dependent bump kinetics are not due to effects on the catalytic activity of PLC but rather to actions downstream, probably at the channel level (Katz and Minke, 2012).

3.3.1. Complex Phylogeny of TRPC Channels in Invertebrate Species

Identification of genes orthologous to Drosophila trp and trpl may be helpful in identifying candidate genes in other species that mediate phototransduction. This is because orthologous genes are more likely to perform similar functions across species (Conant and Wolfe, 2008; Altenhoff et al., 2012; Rogozin et al., 2014; but see Nehrt et al., 2011). In order to identify orthologs, TRP family phylogenetic trees containing representative sequences from protostome genomes have been constructed (Matsuura et al., 2009; Rivera et al., 2010; Peng et al., 2014; Speiser et al., 2014). The tree shown in Figure 3.8a is an extension of those efforts, displaying 174 TRPC protein sequences mined from 43 animal genomes and transcriptomes. The tree predicts four major clades of protostome TRPC channels, three of which have deuterostome sister clades.

Figure 3.8

Evolution of TRPC channels: The complex phylogeny of TRPC channels. (a) Phylogram of TRPC, TRPV, and TRPM channel protein species mined from 43 animal genomes or transcriptomes released to GenBank. (b) Expanded phylogram of the clade, trpγ-like (more...)

3.3.2. TRPC Channels from Two Major Clades Shown to Be Expressed in Microvillar Photoreceptors

Drosophila TRP, TRPL, and TRPγ fall within just one major clade, trpγ-like (Figure 3.8b), which also contains c-elegans TRP-2 and is a sister to that containing mammalian TRPC1, TRPC4, and TRPC5 (Venkatachalam and Montell, 2007; Denis and Cyert, 2002). The trpγ-like clade contains sequences from all of the protostome genomes sampled for the tree and comprises a set of orthologs of TRP and TRPL. Although Drosophila melanogaster is the only species for which there is functional evidence for members of the trpγ-like clade having a role in phototransduction, the expression of orthologs of trp or trpl has been found to be enriched in microvillar photoreceptors in the eyespots of the platyhelminth, Schmidtea mediterranea (trpC-2) (Lapan and Reddien, 2012), the simple eyes of the mollusc, Loligo forbesii (strp) (Monk et al., 1996), and the ventral eye of the chelicerate, Limulus polyphemus (Lptrpl; R. Payne, personal communication). The localization of a trpγ-like channel, trpC-2, to the eyespot of a platyhelminth is illustrated in Figure 3.9a. Orthologs of trp and trpl therefore constitute a set of candidate genes for future functional studies of light-sensitive channels with a role in vision in a very broad range of protostome species.

Figure 3.9

In situ hybridization patterns of probes directed against trpC-2 (a), trpC-1 (b), and rhabdomeral r-opsin mRNA sequences in the head of the platyhelminth, S. mediterranea. Densest areas of r-opsin localization indicate the photoreceptor cell bodies. (From (more...)

Trp-1-like (Figure 3.8c) is a second clade of TRPC channels whose expression pattern provides evidence for a possible role in phototransduction, although physiological evidence is lacking. Trp-1-like clade members are limited to protostome species basal to the neoptera and to deuterostome species basal to vertebrates. Members of the trp-1-like clade are expressed in microvillar photoreceptors in the platyhelminth S. mediterranea (trpC-1) (Lapan and Reddien, 2012) (Figure 3.9b) and in Limulus ventral photoreceptors (Bandyopadhyay and Payne, 2004), although whether they have a function in phototransduction is unknown. Localizing members of this clade to the microvillar photoreceptors of more species as well as demonstrating a functional role in at least one of those species may open up new avenues of research. Deuterostome members of this clade may also be of interest as potential mediators of the response of microvillar photoreceptors in the cephalochordate, amphioxus (Gomez et al., 2009; Angueyra et al., 2012; Ferrer et al., 2012; Pulido et al., 2012).

The gene trees of Figure 3.8 provide evidence of an evolving palette of TRPC channel subtypes available to be coupled to phototransduction in different species. For example, trp, trpl, and trpγ evolved following a double duplication of an ancestral trpγ-like gene at the base of the arthropods, accompanying the evolution of the compound eye. Following these duplications, trp, trpl, and trpγ may have further evolved to mediate phototransduction and other sensory processes to differing extents across species (French et al., 2015). In addition, trp-1-like channels appear to have been lost in neopterous insects, accompanying the evolution of fast, complex flight behaviors to track mates and capture prey.

3.3.3. A Variety of Putative Pore Selectivity Filter Sequences Have Evolved

One possible reason for the duplication and loss of channel clades within phyla may be the adaptive significance of channel ionic selectivity. In microvillar photoreceptors of several species, light-induced elevation of intracellular calcium (Ca²⁺_i) regulates the amplification and speed of phototransduction. The relative contributions of light-induced Ca²⁺ release from intracellular stores versus Ca²⁺ influx from the extracellular space appear to vary across species (Peretz et al., 1994a; Fein and Cavar, 2000; Gomez and Nasi, 1996; Ziegler and Walz, 1990). The expression of TRPC channel clade members with different ionic selectivities could, in principle, provide a mechanism for this variation. An aspartate, D621, in the channel selectivity filter of Drosophila TRP is critical for that channel's high Ca²⁺ selectivity (Liu et al., 2007). Thus, it is interesting that the acidity of the amino acid residue that aligns with Drosophila TRP D621 varies across channel clades (Figure 3.10), the consensus being acidic (aspartate) in insect TRP, insect TRPGAMMA, and in the trpγ-like channels of species basal to the arthropods but neutral in most insect TRPL channels (glycine) (Liu et al., 2007) and neutral or basic in members of the trp-1-like clade (proline or lysine). By expressing a different mix of clade members, these substitutions may allow an adjustment of relative light-induced Ca²⁺ fluxes in the photoreceptors of different species.

Figure 3.10

Aligned consensus sequences of portions of the pore loop and selectivity filter of selected TRPC channel groups included in the trees of Figure 3.8. Inclusion in the consensus requires more than 20% presence of an amino acid at that position. Amino acid (more...)

In summary, the gene family of protostome TRPC channels is as complex as that of the deuterostomes, with the potential to couple clade members with different ionic selectivity and gating properties to phototransduction. While little is currently known and even the channel sequence and phylogeny are preliminary, the abundance of predicted TRPC channel sequences now available and improved techniques for heterologous expression, editing, and targeted knockdown of genes in nonmodel organisms may lead to a future understanding of the evolution of phototransduction mechanisms with possible insight into the still-unknown gating mechanism of TRP channels in all physiologically investigated species.

3.4.1. General Structural Features of TRP Channels

The amino acid sequence of TRP channels in general, distantly relates them to the superfamily of the voltage-gated channel proteins, particularly to the voltage-gated K⁺ channels and the bacterial Na⁺ channel (Kelley and Sternberg, 2009). On the basis of their homology to voltage-gated K⁺, bacterial Na⁺ and cyclic nucleotide–gated (CNG) channels, crystal and cryo-EM solved structures a tetrameric assembly of TRP channels has been suggested (Liao et al., 2013; Saotome et al., 2016). The tetrameric channels may be composed of four identical subunits (homomultimers) or of different subunits (heteromultimers). Each subunit contains cytosolic N- and C-terminals and six transmembrane segments (S1–S6). The ion-conducting pore is formed by S5, the S5–S6 linker, and S6 (Figure 3.11). The voltage-sensing domain of voltage-gated channels, which is located at the S4 transmembrane domain and contains positive charged residues, is missing in the TRP channels (reviewed in Minke and Cook, 2002). The N- and C-terminals of TRP channels interact with various proteins, lipids, and signaling molecules and vary greatly between the different members of the TRP superfamily. There are, however, several common domains to most TRP channels (Figure 3.11):

1. Ankyrin repeats: The ankyrin repeat is a 33-residue motif that mediates specific protein-protein interactions with a diverse repertoire of macromolecular targets (Sedgwick and Smerdon, 1999). In TRP channels they are located at the N-terminal. The number of ankyrin repeats varies among the different TRP channels ranging from four repeats in the case of the Drosophila TRP, TRPL, and TRPC1(Minke and Cook, 2002); six repeats in the TRPV subfamily (Jin et al., 2006; Schindl and Romanin, 2007); 14–18 repeats for TRPA1 (Story et al., 2003); and 28 repeats for NompC (Walker et al., 2000; Sidi et al., 2003). The ankyrin repeat domain was proposed to participate in channel assembly; however, an interesting hypothesis for a functional role of long ankyrin chains (e.g., for the TRPA1) is that this repeated structure forms the gating spring of mechanoreceptors (Corey et al., 2004; Howard and Bechstedt, 2004). It was shown that the ankyrin repeats of TRP channels can bind ATP, CaM, and PIP₂. The role of the ankyrin repeats in the TRP channels is not clear. They have been suggested to function as dimerization domains (Lepage et al., 2009), but isolated ankyrin repeats from TRP channels behave as monomers in every biochemical assay (Jin et al., 2006; Phelps et al., 2010). For TRPV1 the ankyrin repeat domain was shown to bind ATP and induces channel desensitization in the presence of the agonist capsaicin (Lishko et al., 2007) and channel activation in the presence of allicin (Salazar et al., 2008). In the case of TRPA1, which has 14 ankyrin repeats, the use of point mutations (Jabba et al., 2014) and swapping of ankyrin repeats from different species (Cordero-Morales et al., 2011) has established a role of ankyrin repeats in thermosensitivity. Whether this holds true for other TRP channels is currently unknown. Mutations in the ankyrin repeats of TRPV or TRPC channels are associated with genetic diseases, for example, TRPV4 with Charcot-Marie-Tooth disease type 2C (Landoure et al., 2010) and TRPC6 with familial focal segmental glomerulosclerosis (Winn et al., 2005). All known disease-causing TRP-channel mutations induce higher basal channel activity by an unknown mechanism. Although the structure of several mammalian TRPV ankyrin repeat domains is known, there is no knowledge about the structural and biochemical properties of TRPC ankyrin repeat domains. In the case of TRPM, TRPP, and TRPML subfamilies, there are no known ankyrin repeats in either the N- or the C-termini (Venkatachalam and Montell, 2007). Hence, the function of the ankyrin repeat domain in TRP channels is still an open question.
2. Coiled-coil motif: The hallmark structural feature is a heptad repeat, denoted (abcdefg)_n, with hydrophobic residues at the “a” and “d” positions forming a nonpolar stripe along the helical surface that is used for multimerization (Crick, 1952). The identities of the “a” and “d” amino acids provide the dominant feature that determines whether a given coiled-coil (CC) helix will associate into a two-, three-, four-, or five-stranded bundle (Harbury et al., 1993; Malashkevich et al., 1996). The CC domain has been suggested to participate in channel assembly and channel activation (Gaudet, 2008; Lepage et al., 2006). The function and structures of the CC domain of TRP channels is still not clear.
3. TRP domain: The TRP domain is a highly conserved 23–25 amino acid region, located at the C-terminal adjacent to the S6. The TRP box is highly conserved among the TRPV, TRPC, TRPN, and TRPM subfamilies. In TRPM8 the TRP box was suggested to mediate PIP₂ sensitivity (Rohacs et al., 2005), while in TRPV1 it was proposed to participate in channel assembly and determine the activation energy needed for channel opening (Garcia-Sanz et al., 2007). The recently solved atomic structure of rat TRPV1 suggests that the TRP domain assumes an α-helical structure that runs parallel to the inner leaflet of the membrane and functions as an integrator of various channel domains facilitating allosteric coupling (Cao et al., 2013b). However, it is unclear if the role of the TRP domains of the TRPV subfamily could be generalized for all members of the TRP family.
4. The calmodulin (CaM) binding site (CBS domain): The number of CBS sites varies among the different TRP channels, ranging from one for the Drosophila TRP, to three for the TRPC4. There are CBS domains at the N- and C-termini. All TRPCs and some of the TRPVs have CBS domains (Lepage et al., 2006). These CBS domains are suggested to mediate Ca²⁺-dependent desensitization of TRP channels and tachyphylaxis (Lepage et al., 2006).

Figure 3.11

Structural features of the Drosophila TRP and TRPL channels. The trp and trpl genes encode a 1275– and 1124–amino acid long protein, respectively. The TRPL channel contains one to three putative CC domains and two CaM-binding sites (CBS) (more...)

3.4.2. Structural Features of Drosophila TRP and TRPL Channels

The Drosophila trp and trpl genes, which constitutes the founding member of the TRPC subfamily (Minke and Cook, 2002), encode a 1275– and 1124–amino acid long protein, respectively (Figure 3.11). Several studies have helped understand the structural domains and amino acids participating in specific channel functions and properties. For example, the replacement of Asp621 (D621) with glycine or asparagine reduced Ca²⁺ permeability of the TRP channel and is critical for the selectivity filter in the pore domain of the channel (Liu et al., 2007). Another is the F550I substitution of TRP, which forms a constitutively active channel leading to extremely fast light-independent retinal degeneration (Yoon et al., 2000). This site located at the beginning of S5 (Figure 3.11) was also found to increase channel activity of other TRP members, pointing to the importance of this amino acid in channel opening. The N-terminal regions of the TRP and TRPL proteins contain four ankyrin repeats and a CC domain. Both domains are believed to mediate protein-protein interactions. The N-terminal region also contains a TRP_2 domain with unknown function, predicted recently as involved in lipid binding and trafficking (van Rossum et al., 2008). At the C-terminal, the TRPL channel contains two CaM-binding sites (CBSs) (Phillips et al., 1992). One of these is unconventional in the sense that it can bind CaM in the absence of Ca²⁺. The C-terminus fragment of the TRP sequence also has been reported to bind CaM in a Ca²⁺-dependent manner. Both channels have a designated TRP domain located adjacent to the S6 with a EWKFAR motif found in many members of the TRP family. At the C-terminal region of the TRP there is a proline-rich sequence with 27 KP repeats, which overlap with a multiple repeat sequence, DKDKKP(G/A)D termed 8 × 9 (Montell and Rubin, 1989). Such proline-rich motifs occur widely and are predicted to form a structure involved in binding interactions with other proteins, including cytoskeletal elements such as actin. This region is unique to TRP and has not been found in any other member of the TRP family. The last 14 amino acids in the C-terminal of TRP are essential for binding to the INAD scaffold protein and form a PDZ binding domain. This has been demonstrated by truncation experiments (Shieh et al., 1997; Tsunoda et al., 1997; Chevesich et al., 1997). Using Web-based servers for the identification of putative coiled-coil (CC) domains in TRP and TRPL (Predictor of CC-domains in proteins [Fariselli et al., 2007], PairCoil2 [McDonnell et al., 2006], and COILS [Lupas et al., 1991]), it was suggested that in the TRP channel a CC domain in the N-terminal exists with high confidence, while at the C-terminal, one CC domain was identified with low confidence. For the TRPL, one CC domain at the N-terminal were identified with high confidence. At the C-terminal, three CC domains, one with high confidence and two with low confidence, were identified (Katz et al., 2013).

3.4.3. Organization in Supramolecular Signaling Complex via Scaffold Protein INAD

An important step toward understanding Drosophila phototransduction has been achieved by the finding that some of the key proteins of the phototransduction cascade are incorporated into supramolecular signaling complexes via a scaffold protein, INAD. The INAD protein was discovered by a PDA screen of defective Drosophila mutant (inaD). The first discovered inaD mutant, the InaD^P215, was isolated by Pak and colleagues (Pak, 1995) and subsequently was cloned and sequenced by Shieh and Zhu (Shieh and Niemeyer, 1995). Studies in Calliphora have shown that INAD binds not only TRP but also PLC (NORPA) and ePKC (INAC) (Huber et al., 1996). The interaction of INAD with TRP, NORPA, and INAC was later confirmed in Drosophila. It was further found that inaD is a scaffold protein, which consists of five ~90 amino acid (aa) protein interaction motifs called PDZ (PSD95, DLG, ZO1) domains (Tsunoda et al., 1997). These domains are recognized as protein modules that bind to a diversity of signaling, cell adhesion, and cytoskeletal proteins by specific binding to target sequences typically, though not always, in the final three residues of the carboxy-terminal. The PDZ domains of INAD bind to the signaling molecules as follows: PDZ1 and PDZ5 bind PLC, PDZ2 or PDZ4 bind ePKC, and PDZ3 binds TRP (Figure 3.12). This binding pattern is still in debate due to several contradictory reports (Shieh et al., 1997; Chevesich et al., 1997; Adamski et al., 1998; Tsunodaet al., 2001; Mishra et al., 2007; Xu et al., 1998). TRPL appears not to be a member of the complex, since unlike INAC, NORPA, and TRP, it remains strictly localized to the microvilli in the inaD¹ null (Tsunoda et al., 1997). Several studies have suggested that in addition to PLC, PKC, and TRP, other signaling molecules such as CaM, rhodopsin, TRPL, and NINAC bind to the INAD signaling complex (Chevesich et al., 1997). Such binding, however, must be dynamic. Biochemical studies conducted in Calliphora and later in Drosophila have revealed that both INAD and TRP are targets for phosphorylation by the nearby PKC (Huber et al., 1998). Accordingly, the association of TRP in a signaling complex together with its activator, PLC, and possible regulator, PKC, could be related to increasing speed and efficiency of transduction. However, the experimental support for this notion is rather limited.

Figure 3.12

The INAD protein complex. The INAD sequence contained five consensus PDZ domains (indicated by numbers 1–5) and identified specific interactions between PKC and PDZ2 (or PDZ4), TRP and PDZ3, and PLC with PDZ1 and PDZ5. This binding pattern is (more...)

TRP plays a major role in localizing the entire INAD multimolecular complex. Association between TRP and INAD is essential for correct localization of the complex in the rhabdomeres as found in other signaling systems. This conclusion was derived using Drosophila mutants in which the signaling proteins, which constitute the INAD complex, were removed genetically, and by deletions of the specific binding domains, which bind TRP to INAD (Tsunoda et al., 2001; Xu et al., 1998). These experiments show that INAD is correctly localized to the rhabdomeres in inaC mutants (where ePKC is missing) and in norpA mutants (where PLC is missing), but severely mislocalized in null trp mutants, thus indicating that TRP but not PLC or PKC is essential for localization of the signaling complex to the rhabdomere. To demonstrate that specific interaction of INAD with TRP is required for the rhabdomeric localization of the complex, the binding site at the carboxyl terminal of TRP was removed or three conserved residues in PDZ3, which are expected to disrupt the interaction between PDZ domains and their targets, were modified. As predicted, both TRP and INAD were mislocalized in these mutants. The study of the above mutants was also used to show that TRP and INAD do not depend on each other to be targeted to the rhabdomeres; thus, INAD-TRP interaction is not required for targeting but for anchoring and retention of the signaling complex. Additional experiments on TRP and INAD further showed that INAD has other functions in addition to anchoring the signaling complex. One important function is to preassemble the proteins of the signaling complex. Another important function, at least for the case of PLC, is to prevent degradation of the unbound signaling protein (Xu et al., 1998).

A structural study of INAD has suggested that the binding of signaling proteins to INAD may be a dynamic process that constitutes an additional level of phototransduction regulation (Mishra et al., 2007). Their study showed two crystal structural states of isolated the INAD PDZ5 domain, differing mainly by the presence of a disulfide bond. This conformational change has light-dependent dynamics that was demonstrated by the use of transgenic Drosophila flies expressing an INAD having a point mutation that disrupts the formation of the disulfide bond. In this study a model was proposed in which ePKC phosphorylation at a still unknown site promotes the light-dependent conformational change of PDZ5, distorting its ligand binding groove to PLC and thus regulating phototransduction. Further studies showed that the redox potential of PDZ5 is allosterically regulated by its interaction with PDZ4 (Liu et al., 2011). Whereas isolated PDZ5 is stable in the oxidized state, formation of a PDZ4-5 “supramodule” locks PDZ5 in the reduced state by raising the redox potential of a disulfide bond. Acidification, potentially mediated via light-dependent PLCβ hydrolysis of PIP₂, disrupts the interaction between PDZ4 and PDZ5, leading to PDZ5 oxidation and dissociation from the TRP channel (Liu et al., 2011). However, demonstration of the physiological significance of these light-dependent changes in INAD is still lacking.

3.4.4. Assembly of TRP Channels

Several mechanisms regulate TRP channel activity and determine its biophysical properties. Examples of such mechanisms include channel assembly and translocation, which participate in the modulation of cellular currents mediated by TRP channels. Channel translocation results in a change in the number of channels present at the signaling membrane and affects the total conductance, partly utilizing the cellular transport and membrane fusion machinery. Channel assembly diversifies the functional properties of channels by assemblage of different subunits into heteromultimeric channels. The later mechanism can affect the gating, activity, and biophysical properties of the channels.

By analogy to other channels with a similar transmembrane structure that have been more extensively studied (e.g., voltage-gated K⁺ channels and CNG channels), TRP channels are most likely tetramers (e.g., TRPV1 [Liao et al., 2013] and TRPV6 [Saotome et al., 2016]). Given that the mammalian TRPC subfamily consists of seven members and the mammalian TRP superfamily consists of a total of 28 isoforms, an important question arises as to the ability of TRP channels to form heteromultimeres and which members of a specific subfamily can combine to form heteromeric channels.

The first suggestion that TRP channels are composed of heteromultimers came from studies on the Drosophila channels TRP, TRPL, and TRPγ (Xu et al., 2000; Xu et al., 1997). According to one study, TRP and TRPL channels co-immunoprecipitated (co-IP) both in the native tissue and in a heterologous expression system (HEK293 cells). This was also shown for TRPL and TRPγ channels (Xu et al., 2000). In studies on the effects of heteromultimerization on channel activity, whole-cell current measurements were carried out in HEK293 cells expressing the Drosophila TRP, TRPL, or both. Expression of TRPL channels resulted in a robust outwardly rectifying current, which resembled the native TRPL current, consistent with other studies (Parnas et al., 2007; Hardie et al., 1997; Kunze et al., 1997; Zimmer et al., 2000). However, TRP channel expression resulted in a nearly linear current-voltage relationship (Xu et al., 1997), different from the strongly rectifying in vivo current carried by TRP channels (Hardie and Minke, 1994). Moreover, coexpression of TRP and TRPL channels resulted in currents almost indistinguishable from the TRPL currents (Xu et al., 1997). Accordingly, while functional TRPL channels can be easily expressed, functional expression of the Drosophila TRP channels could not be reproduced (Minke and Parnas, 2006). Consistent with the notion that TRP and TRPL channels do not form heteromultimers, it was shown that both TRP and TRPL can form functional, light-activated ion channels in photoreceptor cells of Drosophila mutants in isolation, clearly showing that each channel can function independently of the other channel (Niemeyer et al., 1996; Reuss et al., 1997). Moreover, it was shown that the light-dependent conductance of WT flies could be attained by a weighted sum of the individual TRP and TRPL conductance, further supporting a solely homomultimeric TRP and TRPL channel assembly (Reuss et al., 1997). Further support for homomultimerization of TRP and TRPL channels includes the interaction of TRP but not TRPL with the scaffold protein, INAD (Tsunoda et al., 1997; Huber et al., 1996), and the light-dependent translocation of TRPL but not TRP (Bahner et al., 2002). Thus, although many types of TRP channels can assemble as heteromultimers, it was unclear for the Drosophila TRP and TRPL channels. This has led to further investigation of the issue of Drosophila TRP and TRPL channels assembly and its structural determinants. To this end a series of transgenic Drosophila flies expressing GFP-tagged TRP-TRPL chimeric channels were generated. The use of these chimeras indicated that TRP and TRPL assemble exclusively as homomultimeric channels in their native environment. The above chimeras revealed that the transmembrane regions of TRP and TRPL do not determine assembly specificity of these channels. However, the C-terminal regions of both TRP and TRPL predominantly specify the assembly of homomeric TRP and TRPL channels (Figure 3.13).

Figure 3.13

A model for the assembly of Drosophila TRP and TRPL channels. We adapted a previously proposed model in which the N-terminal interaction through the N-terminal coiled-coil (CC) domains (Engelke et al., 2002) and the ankyrin repeat domain, assembles the (more...)

Various molecular determinants have been suggested for the assembly of TRPC family members. Using a yeast two-hybrid, protein-immobilization, and co-IP assays, it was shown that the Drosophila TRP and TRPL associate through their N-terminal regions (Xu et al., 1997). Later study refined this view by showing that TRPL and TRPγ channels assemble through the coiled-coil (CC) domain and by a 33-residue fragment located adjacent to this CC domain, located at the N-terminal (Xu et al., 2000). The involvement of the CC domain in TRPC channels assembly was supported by findings in other TRPC group members (Lepage et al., 2006). Accordingly, using yeast two-hybrid assays, it was shown that TRPC1 channels homotetramerize through their N-terminal CC domains (Engelke et al., 2002). The ankyrin repeat domain, located at the N-terminal of TRPC channels, was also implicated in channel assembly. Accordingly, it was shown that the first ankyrin repeat of TRPC1 is involved in the interaction with the N-terminus of TRPC3 (Liu et al., 2005). A later study utilizing TRPC channels from two distinct functional groups (i.e., TRPC4 and TRPC6) showed that swapping the N-terminal of TRPC4 with that of TRPC6, which contains the CC domain and ankyrin repeat by constructing a TRPC4–TRPC6 chimeric channel, enabled the interaction with TRPC4. However, the replacement of the CC domain or ankyrin repeats of TRPC6 by that of TRPC4 alone did not result in TRPC4 interaction, demonstrating that both domains are necessary for channel assembly (Lepage et al., 2009; Lepage et al., 2006).

3.4.5. Biophysical Properties of TRP Channels

The Drosophila light-sensitive channels, TRP and TRPL, could be studied separately by utilizing the most useful trpl³⁰² and trp^P343 null mutants (Niemeyer et al., 1996; Pak and Leung, 2003). The channels are permeable to a variety of monovalent and divalent ions including Na⁺, K⁺, Ca²⁺, and Mg²⁺, and even to large organic cations such as TRIS and TEA (Reuss et al., 1997; Ranganathan et al., 1991). The reversal potential of the light-induced current (LIC) shows a marked dependence on extracellular Ca²⁺ indicating a high permeability for this ion. Permeability ratio measurements for a variety of divalent and monovalent ions were determined under bi-ionic conditions and confirmed a high Ca²⁺ permeability of ~57:1 = Ca²⁺:Cs⁺ in the trpl mutant (lacking the TRPL channel) and ~4.3:1 = Ca²⁺:Cs⁺ for the trp mutant (lacking the TRP channel) (Reuss et al., 1997). The TRP and TRPL channels show voltage-dependent conductance during illumination. An early study revealed that the light response could be blocked by physiological concentrations of Mg²⁺ ions (Hardie and Mojet, 1995). The block mainly influenced the TRP channel and affected its voltage dependence (Hardie and Mojet, 1995). Later, detailed analysis described the voltage dependence of heterologously expressed TRPL in S2 cells and of the native TRPL channel, using the Drosophila trp null mutant (Parnas et al., 2007). These studies indicated that the voltage dependence of the TRPL channel is not an intrinsic property, as is thought for other members of the TRP family, but arises from a divalent open-channel block that can be removed by depolarization. The open-channel block by divalent cations is thought to play a role in improving the signal-to-noise ratio of the response to intense light and may function in response termination (Parnas et al., 2007).

3.4.6. Stimulus-Dependent TRP Channel Translocation

The physiological properties of cells are largely determined by the specific set of ion channels at the plasma membrane. Besides regulation at the gene expression level, trafficking of ion channels into and out of the plasma membrane has been established as an important mechanism for manipulating the number of channels at a specific cellular site (for reviews see Lai and Jan, 2006; Sheng and Lee, 2001). Furthermore, stimulus-dependent trafficking (translocation) has emerged as an important regulatory mechanism with high physiological significance. Translocation has also been found for several mammalian TRP channels and for the Drosophila TRPL channels. The first study, which demonstrated in vivo translocation of a TRP channel and its physiological implications, came from studies of the light-activated Drosophila TRPL channel. Direct visualization of intracellular movements of TRPL in photoreceptors upon illumination was observed by immunolabeling of cross sections through Calliphora and Drosophila eyes. Accordingly, in dark-raised flies TRPL specific immunofluorescence was confined to the rhabdomere while, in light-raised flies the TRPL specific immunofluorescence was distributed over the cell body of the photoreceptor cells and was not detected in the rhabdomeres. This study further revealed that unlike TRPL, TRP is confined to the rhabdomeres, regardless of the illumination regime (Bähner et al., 2002). Since the translocation of TRPL depended on the illumination regime, a question arose as to whether or not the response to light through activation of the TRP and TRPL channels is the trigger for TRPL movement to the cell body. Using green fluorescent protein (GFP) tagged TRPL and direct fluorescent visualization of intracellular movements of TRPL in intact photoreceptors upon illumination revealed TRPL translocation from the rhabdomere to the cell body (Figure 3.14) (Meyer et al., 2006). This method was used to test TRPL translocation in the nearly null PLC mutant, norpA^P24 and the null mutant of TRP, trp^P343. Light-induced translocation of TRPL to the cell body did not occur in both norpA^P24 and trp^P343 mutant flies indicating that TRP activation is necessary for TRPL translocation (Meyer et al., 2006). This result has led to the hypothesis that Ca²⁺ influx through the TRP channels is the necessary trigger for TRPL channel translocation. This expectation was directly demonstrated when light-dependent movement of TRPL was blocked by EGTA or by overexpression of the Na⁺/Ca²⁺ exchanger CALX (Meyer et al., 2006; Richter et al., 2011).

Figure 3.14

Light-dependent translocation of TRPL-eGFP between the rhabdomere and cell body in vivo. Upper panel: Deep pseudopupil (Franceschini and Kirschfeld, 1971) of TRPL-eGFP expressing flies (trpl-eGFP) raised in the dark (upper left panel). A fluorescence (more...)

Further research has shown that translocation of TRPL occurs in two distinct stages, first to the neighboring stalk membrane then to the basolateral membrane. In the first stage, light-induced translocation occurs within 5 minutes, whereas the second stage is much slower and takes over 6 hours. These two distinct translocation stages suggest that two distinct transport processes occur. The rapid first stage of translocation suggests that channels are released from the rhabdomere and diffuse laterally through the membrane into the adjoining stalk membrane. The slow second stage suggests an active internalization mechanism of the channels from the plasma membrane into vesicles (Cronin et al., 2006). A recent study has shown that TRPL and rhodopsin co-localize in endocytic particles revealing that TRPL is internalized by a vesicular transport pathway that is also utilized, at least partially, by rhodopsin endocytosis. In line with a canonical vesicular transport pathway, it was found that Rab proteins which are member of the Ras superfamily of monomeric G proteins, Rab5 and RabX4, are required for the internalization of TRPL into the cell body (Oberegelsbacher et al., 2011).

In the initial study it was shown that TRPL, but not the TRP channels undergo light-dependent translocation between the rhabdomere and cell body. Therefore, a question arises as to which of the TRPL channel segments are essential for translocation. Using transgenic flies expressing chimeric TRP and TRPL proteins that formed functional light-activated channels, translocation was induced only in the chimeric channel containing both the N- and C-terminal segments of TRPL. These results indicate that motifs present at both the N- and C-termini are required for proper channel translocation (Richter et al., 2011).

3.4.7. TRP Channel Regulation by Phosphorylation

The activity of many proteins is controlled by phosphorylation and dephosphorylation reactions. Protein kinases and phosphatases that are activated during neuronal activity orchestrate cellular events that ultimately reshape the neuronal events via phosphorylation and dephosphorylation of various ion channels, including many members of the mammalian TRP channel superfamily (Voolstra et al., 2010; Por et al., 2013; Voolstra et al., 2013) (for a comprehensive review see Voolstra and Huber, 2014). However, the roles of phosphorylation and dephosphorylation in controlling TRP channel activity are still unclear (Cao et al., 2013a).

TRP phosphorylation: Since ePKC (INAC) is a member of the INAD signaling complex, it was assumed that this protein kinase might phosphorylate other members of the INAD complex. Indeed, ePKC was shown to phosphorylate INAD as well as TRP in vitro (Huber et al., 1996; Liu et al., 2000). Using quantitative mass spectrometry, 28 TRP differential phosphorylation sites from light- and dark-adapted flies (Voolstra et al., 2010; Voolstra et al., 2013) were identified. Twenty-seven phosphorylation sites resided in the predicted intracellular C-terminal region and a single site resided near the N-terminus. Fifteen of the C-terminal phosphorylation sites exhibited enhanced phosphorylation in the light, whereas a single site, Ser936, exhibited enhanced phosphorylation in the dark. To further investigate TRP phosphorylation at light-dependent phosphorylation sites, phosphor-specific antibodies were generated to specifically detect TRP phosphorylation at Thr849, at Thr864, which become phosphorylated in the light, and at Ser936, which becomes dephosphorylated in the light. To identify the stage of the phototransduction cascade that is necessary to trigger dephosphorylation of Ser936 or phosphorylation of Thr849 and Th864, phototransduction-defective Drosophila mutants and the phosphospecific antibodies were used. Strong phosphorylation of Ser936 in dark-adapted WT flies was observed but no phosphorylation in light-adapted WT flies was detected. Conversely, weak phosphorylation of Thr849 and Thr864 was observed in dark-adapted WT flies and strong phosphorylation was observed in light-adapted wild flies. Additionally, in phototransduction-defective mutants, strong phosphorylation of Ser936, and weak phosphorylation of Thr849 and Thr864 were observed, regardless of the light conditions. Conversely, a mutant expressing a constitutively active TRP channel (trp^P365) (Hong et al., 2002) exhibited weak phosphorylation of Ser936 and strong phosphorylation of Thr849 and Thr864 regardless of illumination. These data indicate that in vivo, TRP dephosphorylation at Ser936 and phosphorylation at Thr849 and Thr864 depend on the phototransduction cascade, but activation of the TRP channel alone is sufficient to trigger this process (Voolstra et al., 2013).

To identify kinases and phosphatases of Thr849 and Thr864, a candidate screen using available mutants of kinases and phosphatases that are expressed in the eye was applied. It was found that Thr849 phosphorylation was compromised in light-adapted inaC null mutants. Diminished phosphorylation in light-adapted PKC^53e mutants was also found, suggesting that these two protein kinases C synergistically phosphorylate TRP at Thr849. AMP-activated protein kinase was not involved in TRP phosphorylation at Thr864. The physiological effects of phosphorylation/dephosphorylation at the above Thr849 and Thr864 sites are still lacking However, the physiological role of Ser936 phosphorylation has been recently elucidated (Voolstra et al., 2017).

Phosphorylation of TRPL: Using mass spectrometry nine phosphorylated serine and threonine residues were identified in the TRPL channel (Cerny et al., 2013). Eight of these phosphorylation sites resided within the predicted cytosolic C-terminal region and a single site, Ser20, was located close to the TRPL N-terminus. Relative quantification revealed that Ser20 and Thr989 exhibited enhanced phosphorylation in the light, whereas Ser927, Ser1000, Ser1114, Thr1115, and Ser1116 exhibited enhanced phosphorylation in the dark. Phosphorylation of Ser730 and Ser931 was not light dependent. To further investigate the function of the eight C-terminal phosphorylation sites, these serine and threonine residues were mutated either to alanine, eliminating phosphorylation (TRPL8x), or to aspartate, mimicking phosphorylation (TRPL8xD). The mutated TRPL channels were transgenically expressed in R1-6 photoreceptor cells of flies as trpl-eGFP fusion constructs. The mutated channels formed multimeres with WT TRPL and produced electrophysiological responses when expressed in trpl;trp double-mutant background, indistinguishable from responses produced by WT TRPL. These findings indicated that TRPL channels devoid of their C-terminal phosphorylation sites form fully functional channels, and they argue against a role of TRPL phosphorylation for channel gating or regulation of its biophysical properties. Since TRPL undergoes light-dependent translocation, subcellular localization of the phosphorylation-deficient, as well as the phosphomimetic TRPL-eGFP fusion proteins by water immersion microscopy were analyzed. After initial dark adaptation, WT TRPL-eGFP was located in the rhabdomeres. After 16 hours of light adaptation, TRPL-eGFP was translocated to the cell body and successively returned to the rhabdomeres within 24 hours of dark adaptation. eGFP fluorescence obtained from the TRPL8x-eGFP displayed marked differences to the WT. After initial dark adaptation, a faint eGFP signal was observed in the cell body, but no eGFP signal was present in the rhabdomeres. After 16 hours of light adaptation, a strong eGFP signal was observed in the cell body akin to that observed in the WT. This indicated that TRPL8x-eGFP fusion construct was newly synthesized during light adaptation. After 4 hours of dark adaptation, TRPL8x-eGFP was present in the rhabdomere, but 20 hours later, only faint eGFP fluorescence was observable in the cell bodies and none in the rhabdomeres (Cerny et al., 2013) This result indicates that the phosphorylation of TRPL participates in TRPL translocation and recycling..

In conclusion, the localization of light and dark phosphorylation sites in both TRP and TRPL is well established. However, the physiological roles of these posttranslational modifications of the channels are unclear.

3.5.1. Lipids Activate Light-Sensitive Channels in THE Dark

The activation of PLC results in the hydrolysis of PIP₂ into DAG and InsP₃ (Figure 3.6). The pathway, which recycles DAG back to PIP₂, the phosphoinositide (PI) pathway, has emerged to be highly relevant for activation of the TRP and TRPL channels, although the detailed mechanism is still unclear. The most familiar action of DAG is to activate the classical protein kinase C (PKC) synergistically with Ca²⁺. However, null mutation of the eye-specific PKC (ePKC, inaC) leads to defects in response termination with no apparent effects on response activation. A second role for DAG in Drosophila photoreceptors is to act as a precursor for regeneration of PIP₂ via conversion to phosphatidic acid (PA) by DGK (Figure 3.6). DGK was identified after the retinal degeneration A (rdgA) mutant was isolated by Benzer and colleagues in a screen searching for defects in eye morphology. The rdgA gene encodes for an eye-specific DAG kinase (Masai et al., 1993) localized to the endoplasmic reticulum (ER) close to the base of the rhabdomere (Masai et al., 1997). The rdgA mutation leads to a severe form of light-independent photoreceptor degeneration. Electrophysiological studies showed that the light-activated TRP and TRPL channels are constitutively active in the rdgA mutant fly and that crossing the rdgA mutant into a TRP null mutant background partially rescued the retinal degeneration (Raghu et al., 2000a). Thus, the suggested mechanism of rdgA mutant degeneration was toxic increase in cellular Ca²⁺, due to the constitutive activity of the TRP channels, via involvement of the phosphoinositide pathway in channel activation (Raghu et al., 2000a).

Another mutation in the PI pathway affects the CDP-diacylglycerol synthase (CDP-DAG synthase) encoded by the cds gene (Wu et al., 1995) (Figure 3.6). The cds mutant was isolated by using a screen based on a collection of P-element enhancer trap Drosophila lines searched by defects in ERG (Wu et al., 1995). An additional mutation in the PI pathway inactivates a PITP encoded by the retinal degeneration B (rdgB) gene (Vihtelic et al., 1991) (Figure 3.6). This mutation, like the cds mutation, led to severe forms of light-dependent photoreceptor degeneration. The rdgB gene product, the PITP protein, is essential for transferring PI from the ER to the rhabdomere (Vihtelic et al., 1993). To account for the effects of the above mutations in the PI pathway and of the rdgA mutation in particular, a hypothesis has been put forward whereby DAG acts as intracellular messenger, leading directly to TRP and TRPL channel activation (Hardie, 2003). However, application of DAG to isolated Drosophila ommatidia either did not activate the channels, or induced single-channel activity after a long delay in an isolated photoreceptor preparation (Delgado and Bacigalupo, 2009). Furthermore, application of DAG to recombinant TRPL expressed in Drosophila S2 cells did not result in channel opening (unpublished observations), thus questioning the DAG hypothesis.

3.5.2. PUFA as Second Messenger of Excitation

DAG is also a precursor for the formation of PUFAs via DAG lipases (Figure 3.6). Studies conducted by Hardie and colleagues showed that application of PUFAs to isolated Drosophila ommatidia, as well as to recombinant TRPL channels expressed in Drosophila S2 cells, reversibly activated the TRP and TRPL channels, respectively (Chyb et al., 1999). These results are consistent with results showing that in the rdgA mutant, TRP and TRPL channels are constitutively active, due to the elimination of DAG kinase, by the rdgA mutation. The suggested accumulation of DAG leads to a still undemonstrated accumulation of PUFAs that constitutively activate the TRP and TRPL channels, leading to a toxic increase in cellular Ca²⁺ and, thereby, degeneration. Recently, Pak and colleagues cloned and sequenced the inaE gene, following the isolation of the inaE mutant by the PDA screen (Leung et al., 2008). The inaE gene encodes a homologue of mammalian sn-1-type DAG lipase that produces the 2-MAG lipid and showed that it is expressed predominantly in the cell body of Drosophila photoreceptors. Mutant flies, expressing low levels of the inaE gene product, have an abnormal light response, while the activation of the light-sensitive channels was not prevented. The 2-MAG lipid is a precursor of PUFA that requires MAG lipase; however, this enzyme was not found in Drosophila (Leung et al., 2008). Thus, the participation of DAG or PUFAs in TRP and TRPL activation in vivo needs to be further explored.

3.5.3. Is Drosophila TRP a Mechanosensitive Channel?

It is important to realize that PLC activation also generates protons that reduce the pH and affects the light response (Huang et al., 2010). In addition, PLC activation converts PIP₂, a charged molecule, containing a large hydrophilic head-group, into DAG, devoid of the hydrophilic head-group, and this conversion causes major changes in membrane lipid packing and lipid-channel interactions. It was therefore hypothesized by Parnas, Katz, and Minke that neither PIP₂ hydrolysis nor DAG or PUFA production affect the TRP and TRPL channel as second messengers but rather act as modifiers of membrane lipid-channel interactions acting as mechanotransducers (Parnas et al., 2009). Indeed, Hardie and Franze have demonstrated by combined patch clamp recordings and atomic force microscopy that by cleaving PIP₂, PLC generates rapid physical changes in the lipid bilayer that lead to contractions of the Drosophila microvilli, and suggest that the resultant mechanical forces that exhibit the kinetics of the light response and are eliminated by the norpA mutation contribute, together with pH changes to the gating of the TRP and TRPL channels (Hardie and Franze, 2012). Clearly, the mechanosensitivity of the Drosophila light-sensitive channels needs further exploration.