Drosophila melanogaster phototransduction proceeds via a phospholipase C (PLC)–triggered cascade of phosphatidylinositol (PI) lipid modifications, many steps of which remain undefined. We describe the involvement of the lipid phosphatidic acid and the enzyme that generates it, phospholipase D (Pld), in this process. Pldnull flies exhibit decreased light sensitivity as well as a heightened susceptibility to retinal degeneration. Pld overexpression rescues flies lacking PLC from light-induced, metarhodopsin-mediated degeneration and restores visual signaling in flies lacking the PI transfer protein, which is a key player in the replenishment of the PI 4,5-bisphosphate (PIP2) substrate used by PLC to transduce light stimuli into neurological signals. Altogether, these findings suggest that Pld facilitates phototransduction by maintaining adequate levels of PIP2 and by protecting the visual system from metarhodopsin-induced, low light degeneration.

Introduction

The phototransduction cycle of Drosophila melanogaster is one of the best understood examples of inositol lipid signaling. As in most invertebrates, phototransduction in D. melanogaster uses a G protein–coupled phosphoinositide pathway (Fig. 1; for reviews see Montell, 1999; Hardie, 2001; Hardie and Raghu, 2001) wherein photoisomerization of rhodopsin to metarhodopsin in the photoreceptor rhabdomeric microvilli activates heterotrimeric Gq. Gq-α then activates a phosphatidylinositol (PI)-specific PLC, which hydrolyzes PI 4,5-bisphosphate (PIP2), generating inositol 1,4,5-triphosphate (IP3) and DAG. Direct action of IP3 on endoplasmic reticulum calcium stores and the stimulation of protein kinase C by DAG are not thought to underlie the subsequent Ca+ surge that triggers depolarization because genetic inactivation of either the single IP3 receptor or the single protein kinase C isoform does not eliminate phototransduction (Raghu et al., 2000a). Instead, through some other mechanism (Raghu et al., 2000b), DAG mediates the opening of several classes of light-sensitive cation channels: transient receptor potential (TRP; Montell and Rubin, 1989; Hardie and Minke, 1992), TRP-like (Phillips et al., 1992; Niemeyer et al., 1996; Xu et al., 2000), and TRP-γ (Xu et al., 2000). It has been proposed that conversion of DAG into polyunsaturated fatty acids mediates TRP activation (Chyb et al., 1999; Raghu et al., 2000b), but this remains controversial as DAG lipase activity has not yet been demonstrated in D. melanogaster (Agam et al., 2000, 2004; O'Tousa, 2002).

At the base of the rhabdomeric microvilli, a network of subrhabdomeral cisternae (SRC) plays a role in signal termination and in PIP2 replenishment. DAG is eliminated through its conversion to phosphatidic acid (PA) via DAG kinase; PA, in turn, is converted to PI through additional steps, including the formation of CDP-DAG by CDP–DAG synthase (Wu et al., 1995). PI is then transported to the rhabdomeric microvillar membrane via one or more actions (Milligan et al., 1997) undertaken by the PI transfer protein retinal degeneration B (RdgB; Hotta and Benzer, 1970; Vihtelic et al., 1991, 1993) because flies lacking RdgB exhibit light-stimulated PIP2 depletion (Hardie et al., 2001). Finally, the PI is converted back to PIP2, presumably by sequential phosphorylation via PI4 and PI 4-phosphate 5-kinase (P14P5K).

PA is presumed to constitute a critical component of the phototransduction pathway as a precursor in the replenishment of PIP2 after PIP2's depletion through the light-mediated activation of PLC. PA could also play a direct role in the regulation of PIP2 synthesis because in mammals it regulates the rate-limiting enzyme PI4P5 kinase (Moritz et al., 1992; Honda et al., 1999). Finally, PA could also be dephosphorylated to generate DAG by PA phosphatase (PAP; Starz-Gaiano et al., 2001) and promote TRP/TRP-like channel activation.

One source of PA comes from the hydrolysis of phosphatidylcholine by the lipid signaling enzyme phospholipase D (Pld; for review see Frohman and Morris, 1999; McDermott et al., 2004). Pld is expressed in mammals in the retina (Lee et al., 2001) and in rod outer segments (ROS; Salvador and Giusto, 1998). In the ROS, PA can be converted to DAG via PAP and could play a role in signaling events there (Pasquare de Garcia and Giusto, 1986; Pasquare et al., 2000); it has also been proposed that it is involved in the membrane fusion processes that occur between disk membranes and the plasma membrane (Boesze-Battaglia et al., 1998) and in transport of rhodopsin-laden membranes to the ROS during development (Deretic et al., 2004).

The single Pld gene found in D. melanogaster has been characterized, and a role for it in cellularization has been uncovered (unpublished data). D. melanogaster Pld exhibits classic Pld activity, increases its activity in signaling contexts, and is inactivated by a point mutation to the conserved catalytic domain (see Fig. S1, available at http://www.jcb.org/cgi/content/full/jcb.200502122/DC1). We describe genetic evidence for the role of Pld-generated PA in the regulation of the phototransduction cycle via regeneration of the PIP2 substrate pool and in the protection from apoptotic degeneration by modulation of the light-transducing signal through PA dephosphorylation to DAG.

Results

Pld deficiency decreases light sensitivity 50-fold

Potential roles for Pld in phototransduction were explored by examining the effects of Pld deficiency on the generation of a light-induced signal in the retina. Pldnull flies, which exhibit reduced viability during cellularization but, as adults, are overtly normal (unpublished data), were examined using electroretinograms (ERGs). ERGs measure the change over time in membrane potential (depolarization and repolarization) in retinal photoreceptors in response to light and, as such, constitute a summary of the light-triggered signals that are generated within an ommatidium. A typical recording for a control heterozygous (Pldnull/CyO) eye is shown in Fig. 2 A. A relatively high membrane potential is observed under resting (dark) conditions. Upon stimulation of the retina by light, the rhodopsin→metarhodopsin→PLC activation→DAG generation→TRP activation cascade results in an influx of Ca2+ that causes a rapid depolarization (net negative change in current potential). The membrane then remains depolarized until light stimulation ceases, because the TRP channels remain open as long as the light-induced signal continues. With return to dark conditions, the TRP channels close, Ca2+ is pumped out, the membrane repolarizes, and the slow steady wave (receptor wave) returns to its resting potential. Positive on- and negative off-transients are also observed after stimulus onset and offset, respectively. The presence of these postsynaptic responses indicates functional connections from the photoreceptor cells R1–6 to the first optic neuropil, the lamina ganglionaris.

A typical recording for a homozygous Pldnull eye is shown in Fig. 2 B. On- and off-transients were smaller and more variable in the mutant, and the intensity responsivity shifted ∼1.9 log units (Fig. 2, C and D). Calculating sensitivity as the inverse of the stimulus required to elicit a criterion (defined) receptor wave, the comparison of three mutants with three control flies revealed the sensitivity difference to be ∼1.61 orders of magnitude. Altogether, Pldnull mutant eyes exhibit ∼1.9% ± 0.4% of the light sensitivity of wild-type (heterozygous) control eyes. Accordingly, Pld is required for the normal photoresponse.

Pld localizes to the cell body and to the base of the rhabdomere, but is not required for retinal ultrastructure

The findings in Fig. 2 suggested that Pld might function by regulating the phototransduction lipid cycle. As one possibility, Pld could act in the SRC to add PA into the phototransduction lipid cycle in order to generate more PI and ultimately replenish the PLC-depleted PIP2. As a second, nonexclusive possibility, Pld might function in the rhabdomere to directly stimulate the PI4P5K-catalyzed phosphorylation of PIP to generate PIP2. Protein localization was examined using an affinity-purified polyclonal rabbit antiserum that was directed against the Pld NH2 terminus. Pld staining was observed in the photoreceptor cell body, in some cases most strongly at the base of the rhabdomere (Fig. 3, A–C), which would be consistent with partial localization to the SRC. Pld was not observed in the remainder of the rhabdomere where light transduction occurs.

The adult retina in Pldnull mutant flies was examined by using both light and electron microscopy. No apparent changes in morphology were observed (Fig. 3 D), suggesting that Pld is not intrinsically required for photoreceptor and ommatidial structural integrity.

Pld is not activated by light, but does rescue retinal degeneration in the norpA7 mutant

As demonstrated in Fig. 2, the phototransduction response is diminished in the absence of Pld-generated PA. A priori, PA could feed forward as a substrate or could act as a regulator (Fig. 1) to increase the PIP2 available for PLC hydrolysis during light stimulation. Alternately, mammalian Pld is well known as a signaling-regulated enzyme that is activated by a wide variety of agonists through G protein–coupled receptors, including ones in the rhodopsin family (Mitchell et al., 1998). Accordingly, the activation of Pld as a downstream effector of photoactivated rhodopsin could generate PA that, in turn, could be dephosphorylated by PA phosphohydrolase (Starz-Gaiano et al., 2001) to form DAG and to trigger a TRP-mediated Ca2+ influx.

To maximize potential light-stimulated, Pld phototransduction events, we used the UAS–GAL4 system (Brand and Perrimon, 1993) with the ninaE-GAL4 (Rh1) driver to ectopically overexpress wild-type Pld cDNA in photoreceptor cells R1–6. To eliminate PLC-mediated phototransduction events in parallel and, thus, permit the isolated detection of any potential Pld-generated signal, the experiment was performed in flies lacking the eye-enriched isoform of PLC (no receptor potential A [norpA]), which eliminates PIP2 hydrolysis and overt light-stimulated signals (Fig. 4 B). Light-stimulated transduction was not observed in norpA7 flies that overexpressed Pld (Fig. 4 C), indicating that the contribution of Pld to phototransduction responses is not mediated by the rhodopsin-triggered activation of Pld and conversion of the ensuing PA to DAG. This finding only addresses whether light stimulation elicits a Pld-mediated change in membrane potential. It does not indicate whether Pld-generated PA is converted to DAG and subsequently activates TRP at a significant steady-state level, because ERGs only report changes in electrical activity; i.e., they do not provide a means to assess differences in baseline current between different individuals.

In addition to the absence of a phototransduction response, the norpA7 fly is also characterized by progressive, light-dependent retinal degeneration (Meyertholen et al., 1987). The degeneration ensues from a combination of generating metarhodopsin in a normal, light-dependent manner but failing to subsequently hydrolyze PIP2, generate DAG, activate TRP channels, and trigger a light-induced calcium influx (Steele and O'Tousa, 1990). In wild-type flies, metarhodopsin transiently complexes with arrestin2 (Arr2). This interaction ceases once Arr2 is phosphorylated by Ca2+/CaM-dependent kinase II (Fig. 1), which is activated by the TRP-triggered influx of Ca2+ during phototransduction (Kahn and Matsumoto, 1997). In norpA7 flies, Arr2 does not become phosphorylated and, thus, does not dissociate from activated rhodopsin (Alloway and Dolph, 1999; Alloway et al., 2000). The stable metarhodopsin–Arr2 complexes then undergo endocytosis and trigger light-dependent retinal degeneration through an apoptotic (Kiselev et al., 2000) or apoptotic-like (Hsu et al., 2004) pathway.

Consistent with previous papers (Meyertholen et al., 1987; Zinkl et al., 1990), norpA7 flies displayed light-dependent retinal degeneration by 21 d after eclosion (Fig. 4 E). The degeneration was characterized by irregularly shaped ommatidia with intracellular vacuoles and by the loss of multiple photoreceptor cells (Fig. 4, compare E with the wild-type control in D, where a regular pattern of seven photoreceptor cells is observed in all ommatidia). In norpA7 flies overexpressing Pld, however, there was a partial morphological rescue of the norpA7 phenotype (Fig. 4 F). norpA7 retinal degeneration is also inhibited when rhodopsin protein levels are decreased (to <3% of wild-type levels; Alloway et al., 2000); however, decreased rhodopsin was not observed in Pld-overexpressing flies (Fig. 4 G). Moreover, as PIP2 hydrolysis and PLC activity in norpA7 are only 2–3% of the wild type (Inoue et al., 1985, 1988), the substrate (PIP2) should not be limiting. Accordingly, altogether, the rescue of retinal degeneration in the norpA7 mutant by the targeted expression of Pld suggests that the inability to generate DAG via PLC was compensated for by converting Pld-generated PA to DAG via PAP to activate the TRP ion channels and trigger dissolution of the metarhodopsin–Arr2 complexes in a constitutive, nonlight regulated manner.

Pld overexpression results in TRP channel–mediated retinal degeneration

The maintenance of appropriate levels of DAG in the phototransduction system is critical. As discussed above, the failure to generate DAG and to subsequently activate the TRP channels leads to apoptotic or apoptotic-like death via metarhodopsin–Arr2 complex persistence (Alloway and Dolph, 1999; Alloway et al., 2000; Kiselev et al., 2000). Conversely, the rapid conversion of DAG to PA via DAG kinase is fundamental to response termination, and elevated levels of DAG, such as those that occur in the rdgA (DAG kinase) mutant, cause necrotic cell death that is secondary to the sustained opening of the TRP Ca2+ channels (Raghu et al., 2000b).

To explore further the possibility that Pld-generated PA can be converted to DAG and activate the TRP ion channels, we used several approaches. First, because Pld overexpression rescued norpA7 metarhodopsin-mediated retinal degeneration, the levels of DAG that were generated by the Pld overexpression in the norpA mutant background were clearly not sufficient to trigger TRP-dependent cell death. Accordingly, Pld overexpression was examined in wild-type flies in which the physiologically normal DAG contribution from the light stimulation of PLC and subsequent PIP2 hydrolysis would be present. Overexpression of Pld in wild-type flies resulted in the progressive retinal degeneration with clear loss of photoreceptor cells and increased intracellular vacuoles (Fig. 5 B) relative to controls (Fig. 5 A). This degeneration was light (PLC) dependent, as dark-reared, Pld-overexpressing flies displayed substantially less disruption of ommatidial morphology (Fig. 5 C).

Second, to demonstrate that the degenerative phenotype results from the production of PA as opposed to a nonenzymatic consequence of Pld overexpression, we generated UAS-Pld lines encoding a catalytically inactive allele of Pld, Pld-H1095N. These lines, when overexpressed by the ninaE-GAL4 (Rh1) driver, did not undergo retinal degeneration (Fig. 5 D), confirming that the degenerative mechanism requires active Pld.

Finally, although many mechanisms could be envisioned to account for the degeneration caused by elevated levels of Pld-generated PA, a cell death pathway that is mediated by persistently elevated levels of DAG would necessarily need to proceed through activation of the TRP Ca2+ channels. Thus, we examined whether Pld-mediated degeneration could be rescued by a null mutation in the trp gene, as degeneration in the absence of DAG kinase activity, which also leads to persistently elevated DAG levels, is suppressed in a trp-null background (Raghu et al., 2000b). Retinal tissue sections from trp1 mutant flies that overexpressed Pld in R1–6 were analyzed for degeneration. The Pld overexpression phenotype (Fig. 5 E) was diminished (Fig. 5 G), suggesting that the increased Pld results in an increased activation of the TRP channels as a result of the increased levels of DAG.

Altogether, these findings suggest that Pld-generated PA can be converted to DAG in a light-independent manner. Because Pld is also required for normal light-dependent responses (Fig. 2), this implies that it should have at least two sites of action in the PI phototransduction pathway.

Pld expression rescues light responsiveness in the rdgB9 mutant

RdgB is a PI transfer protein (Vihtelic et al., 1991) that has been proposed to mediate the transport of PI from the SRC back to the rhabdomere in order to regenerate the PIP2 substrate that is required for hydrolysis in PLC-mediated phototransduction (Fig. 1). The rdgB9 mutant is characterized by absent or rapidly extinguished ERGs because once the existing PIP2 substrate is consumed, the photoreceptors are unable to readily regenerate it via PI recycling (Hardie et al., 2001). The ERG defects appear before any obvious morphological indication of the retinal degeneration that ensues through a complex mechanism involving, in part, metarhodopsin–Arr2 complex persistence (Alloway et al., 2000).

1 d after eclosion, rdgB9 mutants did not respond appropriately to a light stimulus (Fig. 6 A), even though retinal degeneration was not obvious (Fig. 6 C). Pld expression in these mutants rescued the light response, as evidenced by the presence of the slow depolarization and repolarization waves with both on- and off-transients, which signified intact connections that were downstream of the photoreceptor (Fig. 6 B). Because the conversion of the Pld-generated PA to DAG showed no evidence of being light regulated (Fig. 4 C), the functional rescue of the rdgB9 mutation by Pld overexpression suggests that Pld activity facilitates phototransduction via the maintenance of adequate PIP2 substrate levels for PLC hydrolysis. These results are consistent with the findings described in Fig. 2, in which the loss of Pld activity resulted in blunted phototransduction.

Retinal degeneration was apparent in rdgB9 mutant flies by 3 d after eclosion (Fig. 6 F), and ommatidial architecture was completely destroyed by 7 d after eclosion (not depicted). Given that Pld overexpression ameliorated the retinal degeneration in norpA7 mutant flies (Fig. 4) and that Pld overexpression restored the photoresponse in the rdgB9 mutant, it might be anticipated that morphological improvement would be observed in rdgB9 mutant flies overexpressing Pld. Surprisingly, however, Rh1-driven Pld expression in this mutant did not prevent or slow the rate of retinal degeneration (Fig. 6, C–H). However, the absence of RdgB did slow the degeneration caused by Pld overexpression (Fig. 6, compare G with H). This further suggests that Pld overexpression causes retinal degeneration by increasing the DAG-mediated TRP channel activation; this is achieved by increasing the levels of the PIP2 substrate for PLC and by increasing the DAG production from PA. When one arm of the pathway is hindered (by decreasing PLC-mediated hydrolysis of PIP2), Pld overexpression results in less extensive degeneration.

Pld deficiency exacerbates the apoptotic-like degeneration observed in light-stimulated norpA7 flies

Our findings thus far suggested that Pld may function to maintain the PIP2 substrate pool that is required for PLC-mediated light responses, and that the PA generated via Pld overexpression may undergo conversion to DAG in opposition to the DAG kinase step that terminates phototransduction responses. To explore whether the potential conversion of Pld-generated PA to DAG occurs in a physiologically significant context through endogenous Pld action, we examined the consequences of removing the Pld function in the background of the norpA7 mutant, which exhibits metarhodopsin–Arr2-mediated neurodegeneration that is secondary to the inability to activate TRP channels via PLC-generated DAG. If the endogenous Pld pathway functions to elevate DAG levels constitutively, then the Pldnull; norpA7 double mutant would be anticipated to have a more severe phenotype than either mutant alone.

1 d after eclosion, neither the Pldnull (not depicted, but see Fig. 3 D) nor the norpA7 (Fig. 7 A) null mutants individually displayed retinal degeneration, and, in fact, retinal degeneration in norpA7 did not become evident until several days later (Fig. 7 C). However, at 1 d after eclosion, the Pldnull; norpA7 double mutant (Fig. 7 B) already exhibited initial signs of degeneration with small rhabdomeres and decreased cell body size. By 6 d after eclosion, retinal degeneration in the Pldnull; norpA7 double mutant was sufficiently advanced so that only R7/8 remained intact (Fig. 7 D). Together with the previous findings, these results demonstrate that Pld plays physiological roles both in supporting PIP2 levels during phototransduction and in maintaining photoreceptor viability in the absence of ongoing phototransduction events.

Discussion

In this study, we used genetic approaches to explore a novel player that is active in the phototransduction pathway in D. melanogaster. We show that Pld, through its generation of PA, modulates both the photoreceptor responsiveness and the structural integrity of the retina.

Facilitation of phototransduction

The diminished, wild-type light responsiveness and the exacerbation of the norpA phenotype in the absence of Pld reveals a role for Pld in the phototransduction pathway in regulating light sensitivity and photoresponse amplitude. There are four potential steps in which Pld could be involved: provision of additional PA for the purpose of generating PI; movement of the PI from the SRC to the rhabdomere; stimulation of PI4P5K to spur conversion of PI to PIP2; and production of DAG via PAP dephosphorylation of PA.

The exacerbation of the norpA phenotype in the absence of Pld (Fig. 7) indicates that the PA generated by Pld functions physiologically in a context other than by simply increasing the pool of PI in the SRC, moving the PI to the rhabdomeres, or promoting PIP2 synthesis from PIP through the stimulation of PI4P5K, because the norpA flies ultimately lack the ability to use PIP2. Although we do not provide biochemical evidence for the following, the most likely role for PA in this setting would be its dephosphorylation to DAG, which would then trigger transient opening of a limited number of TRP channels, the influx of a small amount of Ca2+, the activation of CaM kinase II, the dephosphorylation of Arr2, and the destabilization of metarhodopsin–Arr2 complexes that would otherwise internalize and initiate apoptotic-like pathways. Increasing DAG levels via the inhibition of DAG kinase has previously been shown to rescue norpA degeneration (Hardie et al., 2003), and, most likely, overexpressed Pld rescues norpA degeneration (Fig. 4) by similarly raising DAG levels via production of PA, followed by the conversion of PA to DAG via PAP.

Could the elevation of DAG via the Pld production of PA affect phototransduction? The absence of a light-stimulated response when Pld is overexpressed in the norpA background indicated that Pld is not light responsive itself (Fig. 4). However, the provision of a low level of DAG could lower the threshold for PLC-mediated, light-stimulated depolarization, and the blunted phototransduction response that is seen in Pldnull flies (Fig. 2) could reflect, in part, a loss of this “priming” level of DAG.

Even if Pld-generated PA does function to prime DAG levels, this cannot be the only step in the cycle at which it exerts an effect; DAG priming would not explain how Pld overexpression rescues the phototransduction response in rdgB (phosphatidylinositol transfer protein) mutant flies in which the transfer of PI from the SRC to the rhabdomeres is presumably diminished (Alloway et al., 2000). Although Pld-generated PA may help to maintain PI levels, this is unlikely to underlie the rdgB rescue because it is not PI production that is limiting in this mutant. Conversely, there would be challenges in proposing that the mechanism underlying the rescue involves the local production of PA to stimulate PI4P5K: Pld resides within the photoreceptor cell body (Fig. 3) rather than in the rhabdomere, where conversion of PI to PIP2 presumably takes place; and rhabdomeric PI is already depleted in this mutant. We favor the hypothesis that Pld plays a role, similar to that undertaken in many mammalian systems, by promoting the exocytic vesicular trafficking outward from the Golgi and perinuclear storage vesicles toward the plasma membrane (Chen et al., 1997; Brown et al., 1998; Vitale et al., 2001; Choi et al., 2002; Du et al., 2003). Pld-promoted trafficking of exocytic vesicles could serve to transfer PI to the rhabdomeres and, if PA were also enriched in the vesicles, could additionally provide a means to stimulate PI4P5K in the absence of locally available Pld. The function of RdgB is complex, with the promotion of vesicular trafficking being one potential mechanism through which it helps to maintain rhabdomeric levels of PIP2 (Lev, 2004). The loss of RdgB and the ensuing consequence for vesicular trafficking may be compensated for by the extensively studied mechanisms through which Pld promotes vesicular trafficking in yeast and in mammalian systems (for review see McDermott et al., 2004).

Ultimately, regardless of which multiple mechanism is most important, Pld-mediated generation of PA is a significant factor in phototransduction responses because flies lacking Pld manifest only ∼2% of wild-type light sensitivity.

Rescue of degeneration mutants

As described in Fig. 4, the overexpression of Pld opposed the apoptotic-like degeneration that is associated with the norpA mutant, in which rhodopsin is converted to metarhodopsin but signaling through PLC hydrolysis of PIP2 is disrupted. The effect of Pld overexpression appears to be mediated through conversion of the Pld-generated PA to DAG because the degeneration that is observed with Pld overexpression in wild-type backgrounds is suppressed in the trp1 mutant. This indicates that the DAG-mediated Ca2+ influx through the opening of the TRP channel is a requisite step in cell death progression.

Pld overexpression did not rescue all degenerative phenotypes. The identification of rdgA (Harris and Stark, 1977) as a DAG kinase mutant allele presented two possibilities: that the retinal degeneration observed in this mutant resulted either from excess DAG accumulation or from PA deficiency. Subsequent studies demonstrating that the necrotic degeneration could be rescued with the trp mutation suggested that the key element was excess DAG accumulation (Raghu et al., 2000b; Hardie et al., 2003). We examined this further by crossing the Pld-overexpressing line into the rdgA mutant background. No rescue was observed (unpublished data), indicating that the provision of PA to flies lacking DAG kinase does not address the basis of the degeneration. Accordingly, our findings support published proposals that DAG accumulation underlies the phenotype (Raghu et al., 2000b). Pld also did not rescue the degeneration associated with rdgB; however, the mechanistic pathways underlying degeneration in this setting are complex and are not fully understood because it is rescued in the mutant strains norpA (Harris and Stark, 1977), trp, and inaC (Smith et al., 1991).

Does the production of DAG from PA that is generated by endogenous levels of Pld play a role in retinal protection? Our observation (Fig. 7) that the light-elicited norpA phenotype is exacerbated in the absence of Pld suggests that the levels of DAG that are generated through Pld action are physiologically significant. The most likely scenario in which this would be relevant for wild-type flies is under conditions of low light, in which infrequent photoisomerization might lead to the gradual accumulation of metarhodopsin–Arr2 complexes in the absence of elevating Ca2+ levels sufficient to trigger CaM kinase II dephosphorylation of Arr2. A low, but significant, trickle of DAG production via Pld action could serve to sufficiently stimulate CaM kinase II activity to prevent such accumulation and the ensuing apoptotic-like cell death.

Roles in mammalian phototransduction

Although phototransduction proceeds differently in mammals and flies, mammalian Pld is expressed in the retina (Lee et al., 2001) and in ROS (Salvador and Giusto, 1998). PAPs are also present and active in the retina (Pasquare de Garcia and Giusto, 1986) and appear to be regulated by phototransduction because PAP activity is inhibited by light stimulation (Pasquare et al., 2000). This would be consistent with the proposal that the conversion of DAG to PA, and accordingly, inhibiting conversion of PA to DAG, is important for signal termination. Conversely, the activation of PAP under dark conditions would lead to DAG production, downstream of and consistent with our proposal that Pld mediates protection from low-level, light-induced degeneration. It remains to be determined if Pld alleles associate with retinal dysmorphology or function.

Materials And Methods

Fly strains

D. melanogaster strains were maintained on standard cornmeal, yeast, sugar, and agar media. Transgenic lines carrying P{UAS–Pld}, P{UAS–Pld-H1095N}, or P{>whs, Pld*>} (termed Pldnull) were generated by the embryo transformation of y1w67c23 flies (Rubin and Spradling, 1982). In brief, targeting of the Pld locus was performed using the method of Rong and Golic (2000). The P (ninaE-GAL4) driver (also termed Rh1-GAL4) inserted on the second chromosome was a gift of C. Desplan (New York University, New York, NY). Phototransduction mutants included norpA7 (also termed norpAP24; Pearn et al., 1996), rdgA3 (also termed rdgAPC47; Harris and Stark, 1977), rdgB9 (also termed rdgBks222; Harris and Stark, 1977), and trp1 (also termed trpCM; Montell and Rubin, 1989). All fly lines were in a w+ (red eyed) background.

Histological fixation and sectioning

Unless noted otherwise, light-reared flies for phototransduction experiments were maintained in continuous fluorescent light at 25°C upon eclosion for the desired period of time. Fly heads of the desired genotype were bisected and immersed in 2.5% PFA and 2% glutaraldehyde in 0.1 M sodium phosphate, pH 7.2, for 1 h on ice. A standard ethanol dehydration series was performed on ice, and the tissue was immersed in three 10-min washes of propylene oxide at room temperature. For embedding, Durcapan ACM (Fluka) was used as follows: 50% propylene oxide/50% Durcapan overnight and 100% Durcapan for at least 4 h. The tissue was then transferred to a Durcapan-containing mold and baked at 70°C for 24–48 h. Transverse thick (1–2 μm) sections were stained with toluidine blue (Sigma-Aldrich) for light microscopy. Each experiment was repeated at least twice, and at least two eyes (each from different heads) were sectioned per experiment. The entire set of sections was examined from the proximal to distal eye and from representative sections selected for the figures.

Immunofluorescent staining of pupal retinas and confocal microscopy

Immunofluorescence on whole pupal retinas was performed following standard procedures. In brief, intact pupal heads were fixed in 4% PFA/0.1 M sodium phosphate, pH 7.2, for 15 min at room temperature. Retinas were then dissected in 1× PBS and were incubated with primary antibody overnight at 4°C. After washing, retinas were incubated with secondary antibody and fluorescently labeled phalloidin for 2 h at room temperature. Affinity-purified, rabbit polyclonal anti-Pld antibody was used at 1:1,000. Fluorescent secondary antibody and fluorescently labeled phalloidin were purchased from Molecular Probes. Images were captured using a confocal microscope (model TCS-2; Leica).

Western blot analysis

Total adult head homogenates were prepared as previously described (Kiselev et al., 2000). In brief, five fly heads were manually disrupted in 2× sample buffer (60 mM Tris, pH 6.8, 2% SDS, 0.004% bromophenol blue, 10% β-mercaptoethanol, 20% glycerol, and 1× protease inhibitor cocktail) and were sonicated for 5 s. Western blot analysis was performed according to standard protocols. Primary antibodies were mouse monoclonal anti-rhodopsin1 antibody (1:1,000, 4C5; Developmental Studies Hybridoma Bank) and mouse monoclonal antitubulin antibody (1:20,000; Sigma-Aldrich).

Electron microscopy

Adult heads were fixed and processed as previously described (Colley et al., 1995). The fixed tissue was dehydrated in serial changes of ethanol followed by propylene oxide and was embedded in Spur's medium (Polysciences, Inc.). Ultrathin sections were obtained on an ultramicrotome (model Ultracut E; Reichert), stained with 2% uranyl acetate and lead citrate, and viewed at 80 kV on an electron microscope (model 410; Philips).

Electrophysiology

ERGs were obtained as previously described (Chen and Stark, 1992). A NaCl micropipette was inserted through the cornea of one eye of each immobilized live fly, and an indifferent electrode was grounded on an agar-containing dish underlying the body of the fly. The fly was illuminated with 1-s pulses generated by a standard shutter of either 570-nm or 600-nm light from a 150-W Xenon arc lamp. Electrode output was transmitted through a differential preamplifier to a computer using the PowerLab 410 (AD Instruments) as the interface. Neutral density filters were used to regulate the intensity of the light stimulus.

Online supplemental material

Fig. S1 provides evidence that dPld can be expressed as a full-length protein and can be recognized by an antisera prepared against it. Its activity in mammalian cells is detected under basal conditions and increases upon cellular stimulation by agonist signals.

Acknowledgments

We are grateful to J. Baker for helpful discussions regarding D. melanogaster manipulation, to C. Alonso for assistance with the confocal microscopy and immunostaining techniques, to R. Ward for expert electron microscopy, and to S. Tsirka, M. Kernan, and members of the Center for Developmental Genetics for critical reading of the manuscript.

This work was supported by the National Institutes of Health (NIH) grant GM60452 to M.A. Frohman and NIH grant EY08768 and support from the Retina Research Foundation and Howard Hughes Medical Institute to N.J. Colley. M.M. LaLonde received support from an NIH Medical Scientist Training Grant and NIH grant NS046931.

References

References
Agam, K., M. von Campenhausen, S. Levy, H.C. Ben-Ami, B. Cook, K. Kirschfeld, and B. Minke.
2000
. Metabolic stress reversibly activates the Drosophila light-sensitive channels TRP and TRPL in vivo.
J. Neurosci.
20
:
5748
–5755.
Agam, K., S. Frechter, and B. Minke.
2004
. Activation of the Drosophila TRP and TRPL channels requires both Ca2+ and protein dephosphorylation.
Cell Calcium.
35
:
87
–105.
Alloway, P.G., and P.J. Dolph.
1999
. A role for the light-dependent phosphorylation of visual arrestin.
Proc. Natl. Acad. Sci. USA.
96
:
6072
–6077.
Alloway, P.G., L. Howard, and P.J. Dolph.
2000
. The formation of stable rhodopsin-arrestin complexes induces apoptosis and photoreceptor cell degeneration.
Neuron.
28
:
129
–138.
Boesze-Battaglia, K., O.P. Lamba, A.A. Napoli Jr., S. Sinha, and Y. Guo.
1998
. Fusion between retinal rod outer segment membranes and model membranes: a role for photoreceptor peripherin/rds.
Biochemistry.
37
:
9477
–9487.
Brand, A.H., and N. Perrimon.
1993
. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes.
Development.
118
:
401
–415.
Brown, F.D., N. Thompson, K.M. Saqib, J.M. Clark, D. Powner, N.T. Thompson, R. Solari, and M.J. Wakelam.
1998
. Phospholipase D1 localises to secretory granules and lysosomes and is plasma-membrane translocated on cellular stimulation.
Curr. Biol.
8
:
835
–838.
Chen, D.M., and W.S. Stark.
1992
. Electrophysiological sensitivity of carotenoid deficient and replaced Drosophila.
Vis. Neurosci.
9
:
461
–469.
Chen, Y.G., A. Siddhanta, C.D. Austin, S.M. Hammond, T.C. Sung, M.A. Frohman, A.J. Morris, and D. Shields.
1997
. Phospholipase D stimulates release of nascent secretory vesicles from the trans-Golgi network.
J. Cell Biol.
138
:
495
–504.
Choi, W.S., Y.M. Kim, C. Combs, M.A. Frohman, and M.A. Beaven.
2002
. Phospholipases D1 and D2 regulate different phases of exocytosis in mast cells.
J. Immunol.
168
:
5682
–5689.
Chyb, S., P. Raghu, and R.C. Hardie.
1999
. Polyunsaturated fatty acids activate the Drosophila light-sensitive channels TRP and TRPL.
Nature.
397
:
255
–259.
Colley, N.J., J.A. Cassill, E.K. Baker, and C.S. Zuker.
1995
. Defective intracellular transport is the molecular basis of rhodopsin-dependent dominant retinal degeneration.
Proc. Natl. Acad. Sci. USA.
92
:
3070
–3074.
Deretic, D., V. Traverso, N. Parkins, F. Jackson, E.B. Rodriguez de Turco, and N. Ransom.
2004
. Phosphoinositides, ezrin/moesin, and rac1 regulate fusion of rhodopsin transport carriers in retinal photoreceptors.
Mol. Biol. Cell.
15
:
359
–370.
Du, G., Y.M. Altshuller, N. Vitale, P. Huang, S. Chasserot-Golaz, A.J. Morris, M.F. Bader, and M.A. Frohman.
2003
. Regulation of phospholipase D1 subcellular cycling through coordination of multiple membrane association motifs.
J. Cell Biol.
162
:
305
–315.
Frohman, M.A., and A.J. Morris.
1999
. Phospholipase D structure and regulation.
Chem. Phys. Lipids.
98
:
127
–140.
Hardie, R.C.
2001
. Phototransduction in Drosophila melanogaster.
J. Exp. Biol.
204
:
3403
–3409.
Hardie, R.C., and B. Minke.
1992
. The trp gene is essential for a light-activated Ca2+ channel in Drosophila photoreceptors.
Neuron.
8
:
643
–651.
Hardie, R.C., and P. Raghu.
2001
. Visual transduction in Drosophila.
Nature.
413
:
186
–193.
Hardie, R.C., P. Raghu, S. Moore, M. Juusola, R.A. Baines, and S.T. Sweeney.
2001
. Calcium influx via TRP channels is required to maintain PIP2 levels in Drosophila photoreceptors.
Neuron.
30
:
149
–159.
Hardie, R.C., F. Martin, S. Chyb, and P. Raghu.
2003
. Rescue of light responses in the Drosophila “null” phospholipase C mutant, norpAP24, by the diacylglycerol kinase mutant, rdgA, and by metabolic inhibition.
J. Biol. Chem.
278
:
18851
–18858.
Harris, W.A., and W.S. Stark.
1977
. Hereditary retinal degeneration in Drosophila melanogaster. A mutant defect associated with the phototransduction process.
J. Gen. Physiol.
69
:
261
–291.
Honda, A., M. Nogami, T. Yokozeki, M. Yamazaki, H. Nakamura, H. Watanabe, K. Kawamoto, K. Nakayama, A.J. Morris, M.A. Frohman, and Y. Kanaho.
1999
. Phosphatidylinositol 4-phosphate 5-kinase alpha is a downstream effector of the small G protein ARF6 in membrane ruffle formation.
Cell.
99
:
521
–532.
Hotta, Y., and S. Benzer.
1970
. Genetic dissection of the Drosophila nervous system by means of mosaics.
Proc. Natl. Acad. Sci. USA.
67
:
1156
–1163.
Hsu, C.D., M.A. Whaley, K. Frazer, D.A. Miller, K.A. Mitchell, S.M. Adams, and J.E. O'Tousa.
2004
. Limited role of developmental programmed cell death pathways in Drosophila norpA retinal degeneration.
J. Neurosci.
24
:
500
–507.
Inoue, H., T. Yoshioka, and Y. Hotta.
1985
. A genetic study of inositol trisphosphate involvement in phototransduction using Drosophila mutants.
Biochem. Biophys. Res. Commun.
132
:
513
–519.
Inoue, H., T. Yoshioka, and Y. Hotta.
1988
. Membrane-associated phospholipase C of Drosophila retina.
J. Biochem. (Tokyo).
103
:
91
–94.
Kahn, E.S., and H. Matsumoto.
1997
. Calcium/calmodulin-dependent kinase II phosphorylates Drosophila visual arrestin.
J. Neurochem.
68
:
169
–175.
Kiselev, A., M. Socolich, J. Vinos, R.W. Hardy, C.S. Zuker, and R. Ranganathan.
2000
. A molecular pathway for light-dependent photoreceptor apoptosis in Drosophila.
Neuron.
28
:
139
–152.
Lee, E.J., D.S. Min, W.S. Kang, M.Y. Lee, S.J. Oh, and M.H. Chun.
2001
. The expression and cellular localization of phospholipase D1 in the rodent retina.
Brain Res.
905
:
240
–244.
Lev, S.
2004
. The role of the Nir/rdgB protein family in membrane trafficking and cytoskeleton remodeling.
Exp. Cell Res.
297
:
1
–10.
McDermott, M., M.J. Wakelam, and A.J. Morris.
2004
. Phospholipase D.
Biochem. Cell Biol.
82
:
225
–253.
Meyertholen E.P., P.J. Stein, M.A. Williams, and S.E. Ostroy.
1987
. Studies of the Drosophila norpA phototransduction mutant. II. Photoreceptor degeneration and rhodopsin maintenance.
J. Comp. Physiol. [A].
161
:
793
–798.
Milligan, S.C., J.G. Alb Jr., R.B. Elagina, V.A. Bankaitis, and D.R. Hyde.
1997
. The phosphatidylinositol transfer protein domain of Drosophila retinal degeneration B protein is essential for photoreceptor cell survival and recovery from light stimulation.
J. Cell Biol.
139
:
351
–363.
Mitchell, R., D. McCulloch, E. Lutz, M. Johnson, C. MacKenzie, M. Fennell, G. Fink, W. Zhou, and S.C. Sealfon.
1998
. Rhodopsin-family receptors associate with small G proteins to activate phospholipase D.
Nature
.
392
:
411
–414.
Montell, C.
1999
. Visual transduction in Drosophila.
Annu. Rev. Cell Dev. Biol.
15
:
231
–268.
Montell, C., and G.M. Rubin.
1989
. Molecular characterization of the Drosophila trp locus: a putative integral membrane protein required for phototransduction.
Neuron.
2
:
1313
–1323.
Moritz, A., P.N. De Graan, W.H. Gispen, and K.W. Wirtz.
1992
. Phosphatidic acid is a specific activator of phosphatidylinositol-4-phosphate kinase.
J. Biol. Chem.
267
:
7207
–7210.
Niemeyer, B.A., E. Suzuki, K. Scott, K. Jalink, and C.S. Zuker.
1996
. The Drosophila light-activated conductance is composed of the two channels TRP and TRPL.
Cell.
85
:
651
–659.
O'Tousa, J.E.
2002
. Ca2+ regulation of Drosophila phototransduction.
Adv. Exp. Med. Biol.
514
:
493
–505.
Pasquare de Garcia, S.J., and N.M. Giusto.
1986
. Phosphatidate phosphatase activity in isolated rod outer segment from bovine retina.
Biochim. Biophys. Acta.
875
:
195
–202.
Pasquare, S.J., G.A. Salvador, M.E. Roque, and N.M. Giusto.
2000
. Effect of light on phosphatidate phosphohydrolase activity of retina rod outer segments: the role of transducin.
Arch. Biochem. Biophys.
379
:
299
–306.
Pearn, M.T., L.L. Randall, R.D. Shortridge, M.G. Burg, and W.L. Pak.
1996
. Molecular, biochemical, and electrophysiological characterization of Drosophila norpA mutants.
J. Biol. Chem.
271
:
4937
–4945.
Phillips, A.M., A. Bull, and L.E. Kelly.
1992
. Identification of a Drosophila gene encoding a calmodulin-binding protein with homology to the trp phototransduction gene.
Neuron.
8
:
631
–642.
Raghu, P., N.J. Colley, R. Webel, T. James, G. Hasan, M. Danin, Z. Selinger, and R.C. Hardie.
2000
a. Normal phototransduction in Drosophila photoreceptors lacking an InsP(3) receptor gene.
Mol. Cell. Neurosci.
15
:
429
–445.
Raghu, P., K. Usher, S. Jonas, S. Chyb, A. Polyanovsky, and R.C. Hardie.
2000
b. Constitutive activity of the light-sensitive channels TRP and TRPL in the Drosophila diacylglycerol kinase mutant, rdgA.
Neuron.
26
:
169
–179.
Rong, Y.S., and K.G. Golic.
2000
. Gene targeting by homologous recombination in Drosophila.
Science.
288
:
2013
–2018.
Rubin, G.M., and A.C. Spradling.
1982
. Genetic transformation of Drosophila with transposable element vectors.
Science.
218
:
348
–353.
Salvador, G.A., and N.M. Giusto.
1998
. Characterization of phospholipase D activity in bovine photoreceptor membranes.
Lipids.
33
:
853
–860.
Smith, D.P., R. Ranganathan, R.W. Hardy, J. Marx, T. Tsuchida, and C.S. Zuker.
1991
. Photoreceptor deactivation and retinal degeneration mediated by a photoreceptor-specific protein kinase C.
Science.
254
:
1478
–1484.
Stark, W.S., and G.S. Wasserman.
1972
. Transient and receptor potentials in the electroretinogram of Drosophila.
Vision Res.
12
:
1771
–1775.
Starz-Gaiano, M., N.K. Cho, A. Forbes, and R. Lehmann.
2001
. Spatially restricted activity of a Drosophila lipid phosphatase guides migrating germ cells.
Development.
128
:
983
–991.
Steele, F., and J.E. O'Tousa.
1990
. Rhodopsin activation causes retinal degeneration in Drosophila rdgC mutant.
Neuron.
4
:
883
–890.
Vihtelic, T.S., D.R. Hyde, and J.E. O'Tousa.
1991
. Isolation and characterization of the Drosophila retinal degeneration B (rdgB) gene.
Genetics.
127
:
761
–768.
Vihtelic, T.S., M. Goebl, S. Milligan, J.E. O'Tousa, and D.R. Hyde.
1993
. Localization of Drosophila retinal degeneration B, a membrane-associated phosphatidylinositol transfer protein.
J. Cell Biol.
122
:
1013
–1022.
Vitale, N., A.-S. Caumont-Primus, S. Chasserot-Golaz, G. Du, S. Wu, V.A. Sciorra, A.J. Morris, M.A. Frohman, and M.-F. Bader.
2001
. Phospholipase D1: a key factor for the exocytotic machinery in neuroendocrine cells.
EMBO J.
20
:
2424
–2434.
Wu, L., B. Niemeyer, N. Colley, M. Socolich, and C.S. Zuker.
1995
. Regulation of PLC-mediated signalling in vivo by CDP-diacylglycerol synthase.
Nature.
373
:
216
–222.
Xu, X.Z., F. Chien, A. Butler, L. Salkoff, and C. Montell.
2000
. TRPgamma, a Drosophila TRP-related subunit, forms a regulated cation channel with TRPL.
Neuron.
26
:
647
–657.
Zinkl, G.M., L. Maier, K. Studer, R. Sapp, D.M. Chen, and W.S. Stark.
1990
. Microphotometric, ultrastructural, and electrophysiological analyses of light-dependent processes on visual receptors in white-eyed wild-type and norpA (no receptor potential) mutant Drosophila.
Vis. Neurosci.
5
:
429
–439.

M. LaLonde and H. Janssens contributed equally to this paper.

Abbreviations used in this paper: Arr2, arrestin2; ERG, electroretinogram; IP3, inositol 1,4,5-triphosphate; norpA, no receptor potential A; PA, phosphatidic acid; PAP, PA phosphatase; PI, phosphatidylinositol; PIP2, PI 4,5-bisphosphate; PI4P5K, PI 4-phosphate 5-kinase; Pld, phospholipase D; RdgB, retinal degeneration B; ROS, rod outer segments; SRC, subrhabdomeral cisternae; TRP, transient receptor potential.