The Drosophila phototransduction cascade serves as a paradigm for characterizing the regulation of sensory signaling and TRP channels in vivo. Activation of these channels requires phospholipase C (PLC) and may depend on subsequent production of diacylglycerol (DAG) and downstream metabolites. DAG could potentially be produced through a second pathway involving the combined activities of a phospholipase D (PLD)and a phosphatidic acid (PA) phosphatase (PAP). However, a role for a PAP in the regulation of TRP channels has not been described. This study reports the identification of a PAP, referred to as Lazaro (Laza). Mutations in laza caused a reduction in the light response and faster termination kinetics. Loss of laza suppresses the severity of the phenotype caused by mutation of the DAG kinase, RDGA, indicating that Laza functions in opposition to RDGA. The retinal degeneration resulting from overexpression of the PLD is suppressed by elimination of Laza. These data demonstrate a requirement for a PLD/PAP-dependent pathway for achieving the maximal light response. The genetic interactions with both rdgA and Pld indicate that Laza functions in the convergence of both PLC- and PLD-coupled signaling in vivo (Kwon, 2006).

To identify new proteins that are involved in phosphoinositide (PI) signaling and participate in Drosophila phototransduction, focus was placed on a gene (CG11440), which was identified in a microarray screen for genes expressed predominately in the adult eye. CG11440 mRNA is expressed in the adult eye at a concentration 15-fold higher than in the rest of the head and encodes a 334 amino acid protein that is >35% identical to known lipid phosphate phosphatases. As with other lipid phosphate phosphatases (LPPs), CG11440 is characterized by multiple transmembrane segments and three conserved domains that form the catalytic domain. An LPP would be predicted to catalyze the reverse reaction promoted by RDGA (Kwon, 2006).

To address a potential function for the CG11440 (referred to as lazaro, laza) in phototransduction, a P element insertion line, GE27043 (GenExel), was obtained which was inserted in the 5′ untranslated region (laza^GE). A null mutation in laza was created by genetically mobilizing the P element, because excisions of P elements are frequently associated with deletions of the flanking genomic DNA. Pools of genomic DNA were screened by PCR and two lines (laza¹ and laza²) were obtained that deleted the first ATG and additional portions of the coding region. In addition, one revertant line was retained in which the P element excised precisely (laza^RE) (Kwon, 2006).

Antibodies were raised against the C-terminal region of Laza and a protein of 42 kDa (predicted molecular weight of 37 kDa) was identified that was not detected in any of the laza mutants (laza^GE, laza¹, laza²) or in flies containing laza^GE in trans with a deficiency chromosome, Df(3L)ED230, that removed laza. The 42 kDa protein was restored in laza^RE as well as in laza¹ flies in which a wild-type CG11440 genomic transgene was introduced. These data indicated that expression of the Laza protein was disrupted in laza^GE, laza¹, and laza². Protein levels of other phototransduction proteins were unchanged in each of the lines studied (Kwon, 2006).

To evaluate whether the Laza protein was expressed in the retina, a biochemical approach was used because the antibodies were ineffective for immunostaining. Laza was detected exclusively in extracts from wild-type heads but not bodies. Moreover, the Laza protein was absent in head extracts prepared from a mutant lacking eyes (sine oculis, so) and was greatly reduced in the heads of a mutant (glass), which were missing photoreceptor cells specifically. Thus, Laza was expressed primarily in photoreceptor cells (Kwon, 2006).

The requirement for laza for phototransduction was evaluated by performing electroretinogram (ERG) recordings, which assay the summed responses of all retinal cells to light. Exposure of wild-type flies to light results in a corneal negative receptor potential, which decays to baseline after cessation of the light stimulus. The amplitude of the light response was reduced nearly 40% in laza^GE flies and 3-fold in laza¹ and laza² flies. These reductions in ERG amplitudes were not due to background mutations because they were observed in laza^GE/Df flies and the laza^GE phenotype reverted to wild-type upon precise excision of the P element. Furthermore, the wild-type laza⁺ transgene rescued the reduction in the ERG amplitude. Although it was difficult to formally exclude that the laza mutation had no effects on the overall physiology of the photoreceptor cells, the reduced ERG amplitude in laza flies did not appear to be due to retinal degeneration because 7-day-old adults did not display retinal degeneration and the ERGs were performed on flies <7 days old. As is the consequence of most mutations that affect phototransduction, in older flies (≥15 days old) the laza mutation resulted in retinal degeneration, which was light-dependent. Nevertheless, the decreased ERG responsiveness in the laza flies was not age dependent (Kwon, 2006).

To assess whether the decreased amplitude in laza was more likely a reflection of a lower sensitivity to light or a decreased maximal obtainable amplitude, the ERG responses was measured at three different intensities. Over the 130-fold intensity range tested, both wild-type and laza flies showed a linear intensity-responsive relationship when the amplitudes were plotted against the logs of the light intensities. However, in the laza mutant, the plot is shifted down. If there was a reduction strictly in the sensitivity of the light response, the lines generated from the wild-type and laza intensity-responsive relationships should have been parallel. This was not the case, because the amplitude of the laza response increased less sharply than in wild-type. These data are most consistent with a defect in the maximum light response in the mutant; however, the data do not allow assessment of whether or not there is a defect in the sensitivity to light. If laza operates in a pathway downstream of PLD, then there would be expected to be overlaps in the pld and laza mutant phenotypes. Interestingly, in the pld null mutant, there is also a defect in the maximum achievable ERG amplitude (Kwon, 2006).

Mutations in rdgA, which encodes a diacylglycerol (DAG) kinase, prolong termination kinetics. To test whether elimination of laza might accelerate deactivation kinetics, consistent with the predicted biochemical function, the time required for an 80% return to the baseline after cessation of the photoresponse was measured. The termination of the photoresponse was approximately 4-fold faster in the laza¹ and laza² mutant flies than in wild-type. Introduction of the genomic rescue transgene in the laza¹ mutant background restored the deactivation kinetics to that of wild-type (Kwon, 2006).

The apparent increase in the speed of termination of the photoresponse in the laza mutants may have been due to the smaller ERG amplitude. Therefore, wild-type flies were exposed to a lower light intensity so that the amplitudes of the wild-type and laza ERGs were comparable. Under these conditions, the time required for 80% termination of the photoresponse was 2-fold shorter in the laza¹ and laza² mutant flies than in wild-type. Thus, the more rapid termination in laza flies was the reverse of that observed in rdgA flies (Kwon, 2006).

If laza functions in the regulation of the photoresponse by catalyzing the PI-cycle reaction that is the reciprocal of that catalyzed by RDGA, then it is plausible that introduction of laza into an rdgA background would suppress the retinal degeneration observed in rdgA flies. To test this hypothesis, the laza¹ mutation was combined with the rdgA^BS12 mutation and tested for suppression of the retinal degeneration. Wild-type flies contain compound eyes consisting of ~800 repetitive units, referred to ommatidia, each of which contains seven photoreceptor cells in any given cross-section. Each photoreceptor cell includes a microvillar structure, the rhabdomere, which is the site of action of phototransduction. Wild-type flies contain seven rhabdomeres in all ommatidial cross-sections regardless of their age. In contrast, the number and size of the rhabdomeres is dramatically reduced even in 1-day-old rdgA flies as a result of constitutive activation of the TRP channel (Kwon, 2006).

Of significance here, the laza¹ mutation suppresses the degeneration-associated rdgA^BS12. At 1 day posteclosion, the average number and size of the rhabdomeres was significantly increased. By 3 days posteclosion, the double-mutant flies retained at least as many rhabdomeres as 1-day-old rdgA^BS12 flies. However, by 7 days after eclosion, the rdgA^BS12;;laza¹ flies also displayed severe retinal degeneration. Consistent with the partial suppression of retinal degeneration by the laza mutation, the ERG response in the double-mutant flies was also partially restored. Whereas there was virtually no light response in 1-day-old rdgA^BS12 flies, a small ERG was restored in the rdgA^BS12;;laza¹ flies. The suppression of rdgA may not have been complete because of the presence of seven other PAPs encoded in the Drosophila genome, including wunen, wunen 2, CG11425, CG11426, CG11437, CG11438, and CG12746, which could potentially be expressed in the photoreceptor cells (Kwon, 2006).

Overexpression of phospholipase D (PLD) causes light-dependent retinal degeneration that is suppressed by mutation of the TRP channel. Because PLD converts phosphatidylcholine (PC) to PA, the retinal degeneration could result from an elevation in phosphatidylinositol 4,5-bisphosphate (PIP₂) production. Alternatively, an increase in production of PA could lead to higher DAG levels, an increase that causes degeneration as a result of higher levels of TRP channel activation. Indeed, it has been shown in mammalian cell culture systems that DAG can be generated from PC by the sequential reactions promoted by PLD and a PA phosphatase. These observations raise the interesting possibility that Laza may function in a common pathway with the PLD. To test this idea, whether the retinal degeneration caused by overexpression of PLD was suppressed by the laza mutation was examined. One-day-old Pld-overexpression flies (Rh1-Pld) displayed early stages of degeneration and were typically missing one rhabdomere per ommatidium. By two days posteclosion, the Pld-overexpression flies displayed clear retinal degeneration; the numbers and sizes of the rhabdomeres were greatly reduced. The degeneration gradually progressed until nearly all of the rhabdomeres disappeared in 7-day-old Rh1-Pld flies. In contrast, no degeneration was detected in 1-day-old Rh1-Pld;laza¹ flies, and the retinal degeneration was significantly reduced in the 2-day-old Rh1-Pld;laza¹ relative to the Pld-overexpression flies. Suppression was detected in 3-day-old flies; however, by 7 days posteclosion, the differences between Rh1-Pld and Rh1-Pld;laza¹ were not significant. These data indicated that the severity of the retinal degeneration in Rh1-Pld flies is delayed by introduction of the laza¹ mutation (Kwon, 2006).

The observation that mutations in laza suppress the degeneration resulting from loss-of-function mutations in rdgA provides genetic evidence that Laza functions opposite to RDGA in the phosphoinositide cycle, consistent with its predicted biochemical function. Furthermore, the faster termination kinetics in laza is the converse of that reported in rdgA mutant flies. These results suggest that the slower response kinetics in wild-type flies is due to Laza-dependent reduction in PA and production of DAG. The DAG or downstream metabolites, such as polyunsaturated fatty acids (PUFAs), may in turn slow down the termination kinetics as a result of prolonged activation of the channels. Several mammalian TRPC channels are activated by DAG, and Drosophila TRP and TRPL have been reported to be activated by PUFAs. The findings in the current report support the model that DAG or downstream metabolites, such as PUFAs, activate the TRPC channels in fly photoreceptor cells. These include the smaller ERG responses and faster termination kinetics in laza mutant flies. Interestingly, a TRPC-like current in rabbit artery myocytes is suppressed by inhibitors of either PLD or PA phosphatases (Kwon, 2006).

PA is generated in fly photoreceptor cells from two sources. The first involves the well-established PLC-dependent hydrolysis of the PIP₂ to produce DAG and phosphorylation by DAG kinase. The second is cleavage of PC to generate PA through activity of PLD. Overexpression of PLD results in retinal degeneration, and it has been proposed that this degeneration results from the subsequent conversion of excess PA to DAG, which in turn leads to overstimulation of the TRPC channels. An indication that Laza may function in a PLD-mediated pathway leading to production of DAG is that the phenotypes resulting from loss-of-function mutations in either the pld or laza are similar. In both cases, there is reduced responsiveness to light. More direct genetic support that Laza functions in a PC/PA/DAG pathway is that elimination of Laza suppresses the retinal degeneration resulting from overexpression of PLD. However, the possiblity that the PLD-dependent production of PA also couples to an alternative Laza-independent pathway cannot be excluded (Kwon, 2006).

It is proposed that the DAG formed through the PLD/Laza pathway, that is, the product of Laza enzymatic function, primes the phototransduction cascade to increase the responsiveness to light. Given the prior evidence that PUFAs, which are formed from DAG, lead to activation of TRP channels, the additional DAG formed through activity of Laza may be necessary for a maximal light response. It is not known whether or not Laza is expressed in the rhabdomeres or cell bodies; however, it is noteworthy that the PLD and RDGA are both in the cell bodies, indicating that lipid metabolism in the cell bodies contributes to the photoresponse. The activity of the PLD has been suggested to be light independent. Nevertheless, given that no light response is observed in the norpA mutant, it cannot be excluded that PLD activity may be indirectly linked to the activity of norpA. It will be of interest to address the nature of the signals that contribute to PLD activation in vivo (Kwon, 2006).

Several LPPs are expressed in the adult retina, and overexpression of each enhanced rdgA³, raising the question of which endogenous LPP functions during phototransduction. To test this, loss-of-function mutants in wun, wun2, and laza were analyzed. Loss-of-function mutants in laza were generated by transposon mutagenesis. A number of mutant alleles were isolated; two of these, laza¹⁵ and laza²², are described. laza²² is a 703 bp deletion that removes the presumed start codon and the first 28 amino acids of the encoded protein. laza¹⁵ is a larger deletion of 2.996 kb that removes the entire open reading frame. Both laza¹⁵ and laza²² leave adjacent genes physically intact and give homozygous viable adults. RT-PCR analysis revealed that the transcript encoding laza was completely absent in laza¹⁵ and laza²², and transcript levels of the other four LPPs were unaffected (Garcia-Murillas, 2006).

To test the potential role of laza in phototransduction, retinal ultrastructure was examined in laza¹⁵ and laza²² using electron microscopy. Essentially similar results were obtained with both alleles and are henceforth presented for only laza²², the smaller deletion. On eclosion, laza²² photoreceptors show normal ultrastructure. However, when grown in bright light, they undergo progressive vesiculation and degeneration of the rhabdomeres that is quite severe by 5 days post-eclosion. Such degeneration was not seen when laza²² flies were grown in dim light. The observation that laza mutants show light-dependent retinal degeneration strongly suggests that this gene product functions during phototransduction (Garcia-Murillas, 2006).

Levels of retinal phospholipids from laza²², comparing samples from flies grown in dark with those grown in bright light. This analysis revealed an ~2.5-fold increase in PA levels when laza²² were grown in light compared to those grown in dark as well as wild-type flies grown in light. By contrast, there were no significant changes in the levels of PC, PE, PI, and DAG. These results demonstrate that light-induced PA levels increase in flies lacking the LPP activity encoded by laza (Garcia-Murillas, 2006).

To test the role of endogenous wun and wun2 in phototransduction, the phenotype was examined of loss-of-function mutants in these genes. No signs of rhabdomeral vesiculation and degeneration were seen in the case of either wun or wun2, suggesting that the LPP encoded by these genes do not participate directly in phototransduction (Garcia-Murillas, 2006).

If endogenous laza is a biochemical antagonist of DGK, an immediate prediction of the findings on overexpression of LPP in rdgA is that loss of the endogenous enzyme will have the opposite effect, i.e., suppress degeneration. To test this hypothesis, rdgA³; laza²² double mutants were generated and their phenotype studied. When grown in dim light, rdgA³ undergoes degeneration that is enhanced by growing flies in bright light (Georgiev, 2005). In dim light, rdgA³; laza²² degenerate at almost the same rate as rdgA³, suggesting that laza is probably not active in antagonizing DGK during low rates of PI(4,5)P₂ hydrolysis. However, when grown in bright light, rdgA³; laza²² showed a dramatic reduction in the rate of degeneration compared to rdgA³ under the same conditions; laza²² heterozygotes slowed the degeneration to a rate intermediate between rdgA³ and rdgA³; laza²². These results clearly demonstrate that the endogenous LPP encoded by laza functions during phototransduction and works in synergy with the DGK encoded by rdgA (Garcia-Murillas, 2006).

Whether the concerted activity of rdgA and laza impacts on Trp and Trpl function during the light response was also tested. To do this, responses to light regulates response termination during phototransduction were analyzed in wild-type, rdgA³, and rdgA³; laza²² double mutants. The rate of deactivation of the light response was compared following a 1 s stimulus of light. Under these conditions, rdgA³ mutants show a pronounced decrease in the rate of deactivation. This defect in deactivation was substantially rescued in the rdgA³; laza²² double mutant. In this study, no clear electrophysiological phenotype was found for the laza²² single mutant using either ERG or whole-cell recording. The effect of overexpressing laza was also studied in rdgA³ photoreceptors. Since rdgA³ + laza photoreceptors show rhabdomeral degeneration which itself could affect deactivation kinetics, rdgA³/+ photoreceptors were used for this analysis. rdgA³/+ photoreceptors do not show retinal degeneration over the time periods used in this study; however, they do show a small but distinct defect in deactivation. This defect was enhanced by overexpression of laza. Overexpression of laza in wild-type photoreceptors did not have any effect on deactivation kinetics. Together, these findings suggest that the combined activity of DGK (encoded by rdgA) and LPP (encoded by laza) can regulate the deactivation of Trp and Trpl during the light response (Garcia-Murillas, 2006).

To test the role of endogenous wun and wun2 in phototransduction, double mutants of these were generated with rdgA³ and their rates of degeneration were studied under bright light illumination. Neither wun nor wun2 were able to slow the rate of degeneration in rdgA³, suggesting that the endogenous LPP encoded by these genes does not participate directly in phototransduction (Garcia-Murillas, 2006).