dtorsin, the Drosophila Ortholog of the Early-Onset Dystonia TOR1A (DYT1), Plays a Novel Role in Dopamine Metabolism

Dystonia represents the third most common movement disorder in humans. At least 15 genetic loci (DYT1-15) have been identified and some of these genes have been cloned. TOR1A (formally DYT1), the gene responsible for the most common primary hereditary dystonia, encodes torsinA, an AAA ATPase family protein. However, the function of torsinA has yet to be fully understood. Here, we have generated and characterized a complete loss-of-function mutant for dtorsin, the only Drosophila ortholog of TOR1A. Null mutation of the X-linked dtorsin was semi-lethal with most male flies dying by the pre-pupal stage and the few surviving adults being sterile and slow moving, with reduced cuticle pigmentation and thin, short bristles. Third instar male larvae exhibited locomotion defects that were rescued by feeding dopamine. Moreover, biochemical analysis revealed that the brains of third instar larvae and adults heterozygous for the loss-of-function dtorsin mutation had significantly reduced dopamine levels. The dtorsin mutant showed a very strong genetic interaction with Pu (Punch: GTP cyclohydrolase), the ortholog of the human gene underlying DYT14 dystonia. Biochemical analyses revealed a severe reduction of GTP cyclohydrolase protein and activity, suggesting that dtorsin plays a novel role in dopamine metabolism as a positive-regulator of GTP cyclohydrolase protein. This dtorsin mutant line will be valuable for understanding this relationship and potentially other novel torsin functions that could play a role in human dystonia.


Introduction
Dystonia is a movement disorder characterized by sustained muscle contraction. While 15 different gene loci have been implicated in primary hereditary dystonia (DYT1-DYT15) [1], the most common and severe form, early-onset dystonia (also known as DYT1 dystonia) is due to mutation in TOR1A (formerly DYT1) and displays dominant inheritance with reduced penetrance [2]. TOR1A encodes torsinA, a 332 amino acid protein from the AAA ATPase family. TorsinA is localized in the lumen of the endoplasmic reticulum and the nuclear envelope [2], but its function is not fully understood. Most cases of DYT1 dystonia are caused by a single mutation, a 3 bp (GAG) deletion that results in the loss of a glutamic acid residue in the carboxyl terminal region of torsinA [3].
In mammals, four paralogous torsin genes (TOR1A, TOR1B, TOR2A, and TOR3A) have been described, while Drosophila melanogaster has only a single torsin gene, dtorsin (previously known as torp4a: CG3024) in 4C11 on the X chromosome [4]. The dtorsinencoded protein, Dtorsin, comprises 339 amino acids with 31.9% identity to human torsinA and also displays the characteristic features of the AAA ATPase gene family [5]. Misexpression in Drosophila of the mutant form of human torsinA, pan-neuronally or in monoaminergic cells results in a locomotor defect [6] and aberrant bouton morphologies at synapses [7], though the function of Dtorsin is unclear. Muraro and Moffat [8] have reported that targeted down-regulation of the dtorsin gene in the Drosophila eye caused progressive degeneration of the retina, but no loss-offunction mutants of dtorsin have previously been characterized.
To investigate the function of dtorsin, we generated loss-offunction mutant fly lines by the 'ends-out' gene targeting method in which homologous recombination at the ends of the gene results in the exchange of the entire targeted genomic sequence with a heterologous sequence [9,10]. By replacing the dtorsin gene with unrelated sequence from the white+ gene, we created a complete deletion of the dtorsin gene in the genome.
Mutant flies showed the following phenotypes: 1) the vast majority of mutant males died by the pre-pupal stage; 2) the few surviving adult males were sterile with reduced pigmentation and thin, short bristles; 3) the late third instar male larvae exhibited locomotor deficits. As some of these phenotypes are associated with mutations that disrupt dopamine metabolism [11,12], we explored this pathway in the mutant flies using both biochemical and genetic approaches. Our findings indicate that disrupted dopamine metabolism is at least one consequence of this null mutation and that dtorsin is a novel positive-regulator of GTPCH protein in Drosophila.

Generation of dtorsin loss-of-function mutants by homologous recombination
To investigate the function of dtorsin, we generated a loss-offunction mutation using 'ends-out' gene targeting designed to remove almost the entire coding sequence [9,10] (Figure S1A and Materials and Methods). We screened approximately 40,000 flies and obtained fourteen lines that had a white+ (w + ) insertion on the X chromosome. Seven of the fourteen were semi-lethal lines, in which only a few adult males survived. PCR analysis with dtorsin gene-specific primers demonstrated that males from each of the seven semi-lethal lines lacked the dtorsin gene, while males of the seven viable lines retained the dtorsin gene (data not shown). The targeted homologous recombination event was expected to yield the same genomic structure in all correctly targeted lines. To confirm this, we performed Southern blot analysis on two semilethal lines, each derived from a different donor line (y w dtorsin KO13 /Y males from B2-1 and w dtorsin KO78 /Y males from B7-1; see Materials and Methods). The results were consistent with precise homologous recombination and replacement of the dtorsin gene with the targeting construct ( Figure S1B). The integrity of the junctions was confirmed by sequencing of the PCR products. Both dtorsin KO13 and dtorsin KO78 lack all dtorsin sequence from the translation initiator ATG to 16 bases upstream of the termination codon (data not shown), leaving no potential to produce Dtorsin protein in these lines. Anti-Dtorsin antibody revealed a 38 kDa protein band, the predicted molecular weight of the Dtorsin protein, in wild type adult whole body extracts and head extracts ( Figure S2 lanes 1 and 3), the corresponding dtorsin KO13 /Y male mutant extracts showed no clear band of 38 kDa protein ( Figure  S2 lanes 2 and 4). The semi-lethality of all seven correctly targeted lines was rescued by introducing a 1.9 kb genomic fragment that contained the entire dtorsin gene (Table S1). These results indicate that all seven semi-lethal lines were dtorsin loss-of-function mutants.
Adult phenotype of the dtorsin mutants All seven mutant lines had identical phenotypes. They were recessive semi-lethal lines with only a few males reaching adulthood. As these males were sterile, it was not possible to generate homozygous mutant females. To determine the stage of male lethality, we made a dtorsin KO13 line carrying the FM7i ActGFP X-chromosome balancer. Hemizygous dtorsin mutant male first instar larvae were identified by their GFP-negative phenotype. Most of these animals developed to late third instar larvae, and died at the pre-pupal stage. Some mutants entered pupation but failed to eclose. In isolated cultures of hemizygous dtorsin-null larvae, about 10% of these developed to the adult stage, while 94% of wild type (y w) larvae developed to the adult stage when maintained under the same conditions. In mixed cultures where larvae of both genotypes were present, very few dtorsin null males developed to the adult stage (Table S1). The few surviving male adults showed the following characteristic phenotypes. The body color appeared paler than wild type, which suggested reduced pigmentation of the cuticle ( Figure 1A, B, C, and D). The flies had thin, short bristles ( Figure S3A and B), moved slowly, and, as noted above, were sterile. All of these phenotypes were rescued by introducing the 1.9 kb genomic fragment that contained the Figure 1. dtorsin loss-of-function mutant flies exhibit pigmentation phenotypes. A-D, Pigmentation phenotype of dtorsin mutants. Adult males of the genotype y w/Y (A), y w dtorsin KO13 /Y (B), Oregon R (C), and w dtorsin KO78 /Y (D) flies. The pictures for A-D were taken one day after eclosion to ensure pigmentation reached the maximum level. Since dtorsin KO78 male has y+ genotype, it was compared to the Oregon R, which is also y+. doi:10.1371/journal.pone.0026183.g001 coding region of the dtorsin gene (Table S1 and Figure S3C). UASdtorsin cDNA driven by the Actin5c-or the Tubulin-GAL4 driver also rescued all phenotypes (Table S1).

Mutant third instar larvae show a locomotion deficit
The surviving mutant adult flies moved slowly, compared to wild type. Since most of the mutants died at the pre-pupal stage, we studied the movements of the mutants as larvae prior to this stage. Drosophila larval crawling entails repeated, rhythmic, peristaltic contraction. During each peristaltic stride, muscle contractions are propagated from one end of the body to the other, passing through all eleven segments, one at a time [13]. Late third instar larval locomotion has been established in several recent studies as an excellent system to study the regulation of locomotion and has been used to evaluate various parameters of larval movement [13,14,15,16,17] and to examine the effects of various mutants on locomotion [13,15,16,17,18].
To quantify the difference in locomotion between wild type and mutant larvae, we analyzed peristaltic frequency [13] allowing third instar male larvae to crawl freely on an agarose substrate in the absence of any obvious sensory cues for a period of 60 seconds. The entire period was recorded under high magnification with a camera video program. Whereas wild type (y w) male larvae moved at 53.362.1 strides per minute (n = 33), dtorsin KO13 male larvae exhibited approximately a ,50% decrease in stride frequency, 24.663.0 (n = 29, p,0.0001) ( Figure 2A, and Video S1 and S2). The locomotion deficit in mutant male third instar larvae was fully rescued (61.861.4, n = 21) by introducing the genomic DNA fragment containing the entire dtorsin gene ( Figure 2A).
Next we examined the crawling pattern of the larvae. To visualize the locomotion of larvae, the videos were analyzed with a newly developed program ''Prairie Dog''. Wild type male larvae moved in a relatively straight line, with few changes in direction until they quickly reached the edge of the dish ( Figure 2B, C). However, mutant male larvae frequently stopped and changed direction. As a result, mutant larvae often circled around the same location, gaining little distance from the starting point ( Figure  2D, E).

Neural expression of dtorsin can rescue the larval locomotion deficit
To determine whether the larval mobility deficit was specifically due to lack of Dtorsin in neurons, we used the GAL4/UAS system [19] to express a dtorsin cDNA using neuronal drivers ( Figure 3). When the pan-neuronal driver elavGAL4 was combined with the UAS-dtorsin cDNA transgene, a strong rescue of the larval mobility was observed (50.562.5, n = 14, p,0.0001) compared to the equivalent dtorsin-null male larvae (elavGAL4, dtorsin KO13 ) lacking the UAS-dtorsin transgene (28.361.4 strides/min, n = 32). With the dopaminergic tissue-specific driver TH-GAL4 and the UAS-dtorsin transgene, we observed a significant increase of peristaltic frequency (39.462.4, n = 12, p = 0.0003) compared to the control (y w dtorsin KO13 ), although the rescue was weaker than that achieved with elavGAL4, suggesting that other neurons in addition to dopaminergic neurons may play important roles in regulating larval mobility in dtorsin null mutants. Transgenes of elavGAL4 or TH-GAL4 alone had no effect on the mobility of larvae with or without the dtorsin KO13 null mutation ( Figure 3). Expression of the dtorsin cDNA controlled by a muscle-specific driver, MHC-GAL4 [20] did not rescue the dtorsin null mutant larval mobility defect (data not shown). These results suggested that dtorsin expression in neurons plays an important role in regulating larval mobility.

Rescue of larval locomotion deficit by neuronal expression of human torsinA
Next, we examined if the wild type human torsinA cDNA could rescue the dtorsin null mutant mobility defect. We found that human torsinA cDNA expressed with the Actin5c-or the Tubulin-GAL4 drivers rescued all phenotypes (pigmentation, bristle, and mobility) at very low frequency (Table S1). When this same human torsinA was expressed with the neuron-specific driver elavGAL4, a very significant rescue of the 3rd instar dtorsin null male larval mobility was observed (56.363.8 strides/min, n = 14, p,0.0001) compared to the control dtorsin null male larvae (28.361.4 strides/ min, n = 32) (Figure 3). The peristaltic frequency of the dtorsin null male larvae that expressed the human torsinA cDNA in neurons was statistically similar to the wild type control (y w; 55.262.5 strides/min, n = 15). These data indicate that the neuronal functions of the human torsinA and dtorsin are very well conserved with respect to the regulation of the larval locomotion.

Dopamine restored locomotor deficit phenotype in dtorsin mutant lines
In Drosophila, ingestion of dopamine increases dopamine pools in the fly head, though in mammals peripheral dopamine does not enter into the brain [21]. Since expression of dtorsin in dopaminergic cells produced detectable rescue of the locomotion phenotype, we tested whether addition of dopamine to fly food could also rescue this defect. Previous studies have shown that dopamine, serotonin, and octopamine can all affect larval locomotion as well as adult behaviors in Drosophila [16,22,23]. Consequently, we also tested whether serotonin or octopamine could rescue the locomotion deficit of male dtorsin null larvae.
After dopamine ingestion, the dtorsin KO13 male mutant larvae also exhibited relatively straight crawling patterns, similar to those of the wild type larvae with or without dopamine ( Figure 5). The reduced cuticular pigmentation phenotype of surviving male adults was also partially rescued by dopamine supplementation ( Figure S4).

Dopamine level is significantly reduced in dtorsin mutant heterozygotes
In Drosophila, dopamine metabolism has a fundamental role in pigmentation of the cuticle [11,12], and in many neurologicallymediated processes, including locomotion [24]. Consequently, as the reduced pigmentation and the locomotion deficits in the dtorsin mutants and their rescue by ingestion of dopamine suggested that dopamine metabolism must be disturbed by the dtorsin mutation, we measured dopamine levels in the third instar larval brain and adult heads. Since large numbers of adults or larvae were required for the analyses of dopamine by HPLC (see Materials and Methods), we used heterozygous mutants for these experiments. Wild type (y w) female larvae contained 0.063160.0025 ng dopamine/brain (n = 3, 75 brains per assay), whereas dtorsin heterozygous female larvae (dtorsin KO78 /+) had 0.034060.0014 ng dopamine/brain (n = 3) corresponding to a 46% reduction (p,0.001) ( Figure 6A). Dopamine pools in the heads of adult female heterozygotes were even more dramatically altered, with dopamine levels reduced from 0.33 ng/head in the y w control flies   . Rescue of dtorsin mutant larvae mobility defects by dopamine feeding. Peristaltic frequencies were counted for the wandering stage third instar larvae of the genotype wild type (y w) (n = 15), wild type (y w) (+20 mM dopamine) (n = 10), wild type (y w) (+20 mM octopamine) (n = 8), wild type (y w) (+10 mM serotonin) (n = 9), y w dtorsin KO13 /Y (n = 28), y w dtorsin KO13 /Y (+20 mM dopamine) (n = 15), y w dtorsin KO13 /Y (+20 mM octopamine) (n = 10), and y w dtorsin KO13 /Y (+10 mM serotonin) (n = 13). Results are mean 6 S.E.M. *** p,0.0001, very significant difference between the dtorsin mutant with no drug supplementation and the dtorsin mutant with 20 mM dopamine supplementation. There was no significant difference between the dtorsin mutant males with no drug supplementation and the dtorsin mutant males with 20 mM octopamine (p = 0.4240) or the dtorsin mutant male with 10 mM serotonin (p = 0.8148). There was no significant difference between the wild type with no drug supplementation and the wild type with drug (+20 mM dopamine, +20 mM octopamine, or +10 mM serotonin). doi:10.1371/journal.pone.0026183.g004 to approximately 0.10 ng/head in two different dtorsin mutants (dtorsin KO13 /+ : 0.1160.02 ng/head, n = 3, p,0.001; dtorsin KO78 /+ : 0.1260.04 ng/head, n = 3, p,0.001) ( Figure 6B).
We next sought to determine whether trafficking of dopamine was affected by the dtorsin mutation. Turnover of dopamine during synaptic release can be inferred by the levels of the dopamine metabolite 3,4-dihydroxyphenylacetic acid (DOPAC) and by the ratio of DOPAC to dopamine in flies, as in mammals [21,25]. Although the levels of DOPAC in the dtorsin heterozygous mutant were decreased, the reduction was clearly not proportional to the reduction in dopamine levels. In consequence, the resulting DOPAC:dopamine ratios were elevated over two-fold ( Figure 6C, D). These neurochemical data suggested that the reduction of dtorsin in dopaminergic cells may lead to depressed synthesis of dopamine as well as trafficking abnormalities. Alternatively, the loss of dtorsin might lead to loss of dopaminergic cells.
Tyrosine hydroxylase (TH) catalyses the first and rate-limiting step in the dopamine biosynthesis pathway and is expressed in all dopaminergic neurons. Therefore, we next examined third instar larval brain using anti-dTH antibody to determine whether Drosophila TH (dTH) is expressed in dopaminergic neurons in the brains of dtorsin null mutants [26]. The dopaminergic cell pattern has been well characterized in the central nervous system of Drosophila larvae and consists of four clusters of 4-10 neurons in each brain hemisphere and a stereotyped pattern of three to five paired or unpaired neurons per segment in the ventral ganglion [27] ( Figure 7A). We found that dTH was expressed in the brains of hemizygous dtorsin KO13 third instar larvae and that the number of dTH-positive cell bodies was approximately normal and clearly not sufficiently reduced to account for the dramatic loss of dopamine observed. However, we did observe some apparent morphological changes in the positioning of cell bodies of dTHpositive neurons and in the patterning of dTH-positive axons in the central brain region of the hemizygous dtorsin null male larvae ( Figure 7B). Further studies are underway to understand the exact nature of these structural differences between the wild type and the dtorsin null mutant brains, but we are able to conclude that loss of Dtorsin expression and the consequent reduction in dopamine levels are not associated with failure of dopaminergic neurons to differentiate or their subsequent degeneration.
Genetic interactions of dtorsin with Pu (Punch) or ple (pale) Dopamine synthesis is directly regulated by the specific activity of TH, which, in turn, is modulated by the amount of its cofactor, tetrahydropterin (BH4). In consequence, TH activity is regulated indirectly by the activity of the rate-limiting enzyme in BH4 synthesis, GTP cyclohydrolase (GTPCH) in flies, as in mammals [28,29]. We addressed the hypothesis that dopamine synthesis is affected by mutations in dtorsin and further pursued a putative role in dopamine trafficking by first conducting gene interaction analysies. We tested for genetic interaction of the null dtorsin KO13 allele with mutant alleles of ple (pale), which encodes TH [30,31], Pu (Punch), the Drosophila ortholog of GTP cyclohydrolase (GTPCH) [32], and DAT, encoding the dopamine transporter [33]. Pu and ple loss-of-function mutants show similar phenotypes to each other [34], as Pu mutants exhibit reductions in dopamine pools due to limiting BH4 levels [35]. Homozygous Pu loss-offunction mutants die during late embryogenesis with completely unpigmented and weak cuticles [32]. The human orthologs of these two genes underlie DYT5 and DYT14 (Dopa-responsive) dystonia [1] which is phenotypically similar to DYT1 dystonia, except that DYT5 and DYT14 patients respond very well to the administration of L-dopa [36,37] whereas DYT1 patients do not. Most DYT14 patients have dominantly inherited loss of function mutations in the GCH1 gene, while mutations in the TH gene are responsible for a rare recessive form [36]. The third gene involved in dopamine metabolism, DAT, was found to be mutated in a hyperactive fly line, fumin. Elimination of DAT activity in the DAT fumin mutant caused a dramatic increase in the length of the waking phase [33].
In order to test the possibility of a genetic interaction between the dtorsin null mutation and loss-of-function mutations of Pu, ple, or DAT fumin , we analyzed the peristaltic stride frequency of the double-heterozygous female mutant larvae. No phenotype was seen in any single heterozygous mutant line. The stride frequencies of Pu Z22 /+ males (48.762.0, n = 20) and females (46.861.9, n = 23) were statistically similar to the stride frequency of the dtorsin KO13 /+ females (48.961.6, n = 22). However, the double heterozygous dtorsin KO13 /+; Pu Z22 /+ females showed very significantly reduced peristaltic frequency (28.861.4, n = 30, p,0.0001), relative to either single heterozygote ( Figure 8A). While dtorsin KO13 /Y male third instar larvae were mostly viable, very few dtorsin KO13 /Y; Pu Z22 /+ male of the third instar larvae were observed.
Thus the dtorsin null mutation showed a very strong genetic interaction with the Punch (GTPCH) mutation, a weaker interaction with the ple (TH) mutation, and no detectable interaction with the fumin (DAT) mutation.
GTPCH, but not TH activity, is severely reduced in dtorsin heterozygous mutant lines The above findings suggested that the Dtorsin function may regulate the Pu gene/GTPCH protein activity. To test this hypothesis, extracts were prepared from wild type (y w) and dtorsin heterozygous mutant (y w dtorsin KO13 /+ and w dtorsin KO78 /+) adult heads and TH and GTPCH activities were compared (Figure 9). No significant difference was detected between the TH activities of the dtorsin null heterozygous flies (y w dtorsin KO13 /+ : 1.5360.06 L-dopa nmoles/min/mg protein, n = 3 independent assays, p = 0.2962; w dtorsin KO78 /+ : 1.4560.12 L-dopa nmoles/min/mg Figure 6. Effect of dtorsin mutation on dopamine in larval brains and adult heads. A. Effect of dtorsin heterozygous mutation on dopamine pools in larval brains (nanograms per brain). Monoamines were extracted from third instar larval brains and dopamine was separated and quantified by HPLC. dtorsin mutation significantly reduced dopamine levels. The genotypes of larvae were y w females (n = 3 independent replications) and w dtorsin KO78 /+ females (n = 3 independent replications). Error bars indicate S.E.M. ** p,0.001, significant difference between the wild type and the dtorsin heterozygous mutant. B.-D. Effect of dtorsin heterozygous mutation on dopamine levels (nanograms per head). (B), DOPAC levels (nanograms per head) (C), and DOPAC/dopamine ratio (D) in adult heads. The genotypes of adults were y w females (n = 3 independent replications), y w dtorsin KO13 /+ females (n = 3 independent replications), and w dtorsin KO78 /+ females (n = 3 independent replications). Error bars indicate S.E.M. ** p,0.001, significant difference in dopamine levels between the wild type and the dtorsin heterozygous mutant females (B). No statistically significant difference was observed in DOPAC levels between the wild type and the dtorsin heterozygous mutant females, although the mutant levels were slightly lower (C). * p,0.01, significant difference in DOPAC/dopamine ratio between the wild type females and the dtorsin heterozygous mutant females (D). doi:10.1371/journal.pone.0026183.g006 protein, n = 3 independent assays, p = 0.2657) and the wild type flies (y w) (1.6260.05 L-dopa nmoles/min/mg protein, n = 3 independent assays). However, a significant reduction in GTPCH activity was observed in the dtorsin null mutant heterozygous flies (y w dtorsin KO13 /+ : 0.063360.009 neopterin nmoles/min/mg protein, n = 3 independent assays, p,0.01; w dtorsin KO78 /+ : 0.059060.015 neopterin nmoles/min/mg protein, n = 3, p,0.01) compared to wild type flies (y w) (0.13860.087 neopterin nmoles/ min/mg protein, n = 3 independent assays). These results suggest that the dtorsin is a novel positive-regulator of GTPCH activity in Drosophila.

GTPCH protein levels are severely reduced in dtorsin heterozygous and hemizygous mutant lines
The above findings suggested that dtorsin might be required for maintaining GTPCH protein levels. To test this hypothesis, we prepared extracts from heads of wild type adult males (y w), dtorsin heterozygous adult females (y w dtorsin KO13 /+), and dtorsin hemizygous adult males (y w dtorsin KO13 /Y) and compared TH and GTPCH protein levels by Western blot analysis ( Figure 10, lanes 1-3). No significant difference was detected between the TH protein levels of dtorsin null hemizygous flies (y w dtorsin KO13 ), heterozygous flies (y w dtorsin KO13 /+ females), and of wild type flies (y w) (Figure 10, lanes 4-6). However, a very severe reduction in the level of GTPCH protein was observed in the dtorsin null mutant heterozygous females (y w dtorsin KO13 /+) compared to wild type (y w). Moreover, only very low levels of GTPCH protein level were detected in dtorsin null mutant hemizygous males, suggesting only a very small amount of GTPCH proteins are present in the hemizygotes ( Figure 10, lanes 1-3). These results support the conclusion that dtorsin is required for maintaining the level of GTPCH protein in Drosophila though the molecular mechanism by which it acts remains to be elucidated.

Discussion
The human genome contains four paralogous genes of the torsin family, while Drosophila has a single torsin gene, dtorsin, encoding a protein of the AAA family of adenosine triphosphatases (ATPases), related to the Clp protease/heat shock family [4]. The precise functions of these genes and the degree to which their activities may overlap are not known, though TOR1A is notable for causing early-onset dystonia due to a codon deletion that removes a particular glutamic acid residue from the torsinA protein.
TorsinA-null knock-out mice (Tor1a 2/2 ) have been created and die at birth with abnormal nuclear membrane morphology in neural cells [38]. In comparison to the human genes, dtorsin is most closely related to TOR1A [4,39], suggesting that investigation of the fly protein's function may be informative concerning the pathway(s) disrupted in human early onset dystonia. Our generation and characterization of a null mutant for dtorsin revealed that it is an essential gene in flies, as mutants typically fail to reach adulthood and those that do survive fail to reproduce. Null dtorsin mutant flies display both morphological and behavioral phenotypes that may provide starting points for elucidating activities of the Dtorsin protein, using both biochemical and genetic strategies. Our initial examination of these phenotypes indicates that at least one major disruption in dtorsin mutant flies involves regulation of dopamine, a function also implicated in human dystonia [40]. We expect that the availability of the dtorsin null mutant will facilitate investigation of this activity and other effects of torsin deficiency that are likely to emerge with more detailed study.
We were led to examine dopamine metabolism in the dtorsin mutants because of the phenotypes exhibited by rare adult survivors, which had a pale body color and thin bristles (Figure 1 and S3), moved slowly and were sterile. A series of reports have shown that dopamine not only functions as a neurotransmitter but also plays a role in cuticle pigmentation, fertility, and other vital functions in Drosophila [11,12,41]. Thus the most evident phenotypes exhibited by the dtorsin null mutants are consistent with previously reported phenotypes caused by dopamine deficiency.
Disruption of dopamine metabolism in the dtorsin mutants was supported by HPLC analysis of heterozygous dtorsin mutant larval and adult brains/heads which demonstrated a significantly reduced dopamine level ( Figure 6) and by rescue of the larval locomotion defect and cuticle pigmentation by feeding dopamine (Figure 4, 5, S4). Thus, dopamine synthesis appears to be disrupted in dtorsin null flies and is likely to be largely responsible for both the deficit in larval locomotion and in cuticle pigmentation. While ingestion of serotonin and octopamine had no significant effect on the dtorsin larval locomotion deficit, the results of this particular assay do not exclude the possibility that other monoamine pathways may also be altered in these flies. Indeed the weaker rescue of the locomotor phenotype by expression in dopaminergic neurons than by more widespread neuronal expression ( Figure 3) suggests a role for other neuronal elements/pathways. Further studies, including the measurement of serotonin and octopamine pools, will be required to examine the effect of dtorsin null mutation on these particular pathways.
Our visualization of dopaminergic neurons in mutant flies indicated that the reduced dopamine levels were not due to reduced numbers of dopaminergic neurons (Figure 7), suggesting a relative decrease in dopamine production in the absence of Dtorsin. Dopamine synthesis involves hydroxylation of the amino acid L-tyrosine by TH to produce 3,4-dihydroxy-1-phenylalanine (L-dopa), which is subsequently decarboxylated by aromatic-Lamino acid decarboxylase to form dopamine. TH activity depends on the availability of its cofactor, BH4 [32], which is produced by conversion of GTP to 2,4-dihydroneopterin triphosphate by GTPCH, followed by reduction and dephosphorylation to the A: A projection of twenty-seven confocal z-sections of wild type (y w) third instar male brain is shown. B: A projection of thirty confocal z-sections of y w dtorsin KO13 /Y third instar male larval brain is shown. The projection was made from a series of confocal Z-sections (optical thickness 2 mm) that covered the whole brain regions using the NIH Image-J program with the Max intensity method. vg: ventral ganglion. doi:10.1371/journal.pone.0026183.g007 final product, BH4, by 6-pyruvoyltetrahydropterin synthase and sepiapterin reductase, respectively [29].
In flies, TH is encoded by the ple (pale) gene and GTPCH is produced from the Pu (Punch) gene. Our finding that genetic interaction of mutations in dtorsin and Pu produces a strong defect in the larval locomotor assay whereas genetic interaction of dtorsin and ple produces a weaker but notable effect (Figure 8), is consistent with the view that dopamine metabolism is further compromised in these double mutants. This genetic effect parallels our biochemical measurement of a severe reduction of GTPCH activity but not TH activity, in dtorsin heterozygous adult flies ( Figure 9). Furthermore, western blot analyses demonstrated that GTPCH protein levels are severely reduced in dtorsin-null heterozygotes and hemizygotes ( Figure 10). Taken together, our A. Peristaltic frequencies were counted for the wandering stage third instar larvae of the genotype y w female (n = 31), Pu Z22 /+ female (n = 23), dtorsin KO13 /+ female (n = 22), and dtorsin KO13 /+; Pu Z22 /+ female (n = 30). Results are the mean 6 S.E.M. **p,0.0001, significant difference between the wild type females (yw, Pu Z22 /+, or dtorsin KO13 /+) and the double heterozygous females (dtorsin KO13 /+; Pu Z22 /+). No significant difference was observed between the wild type females (y w) and the single heterozygous females (Pu Z22 /+ or dtorsin KO13 /+). B. Peristaltic frequencies for the wandering stage third instar larvae of the genotype y w female (n = 20), ple 2 /+ female (n = 20), dtorsin KO13 /+; female (n = 31), and dtorsin KO13 /+; ple 2 /+ female (n = 30). Results are the mean 6 S.E.M. *p,0.05, significant difference between the wild type females (y w, ple 2 /+, or dtorsin KO13 /+) and the double heterozygous females (dtorsin KO13 /+; ple 2 /+). No significant difference between the wild type females (y w) and the single heterozygous females (ple 2 / + or dtorsin KO13 /+). doi:10.1371/journal.pone.0026183.g008 experiments suggest that presence of functional Dtorsin is required to maintain GTPCH protein levels, indicating that, through an as yet undetermined mechanism, dtorsin is a positive-regulator of GTPCH protein and dopamine synthesis in Drosophila.
Interestingly, GTPCH is also critical for the activity of other aromatic acid hydroxylases, nitric oxide synthatase [29], and alkylglycerol monooxygenase [42]. Thus GTPCH activity is also essential to the serotonin synthesis, suggesting that it too may be disrupted in the dtorsin mutant. Why then did addition of serotonin have no effect on the larval mobility in our experiments? There are two possible explanations. The first is that the larval mobility assay that we used may be sensitive only to a defect in the dopamine pathway. The second is that the TH protein has a much lower affinity for BH 4 compared to tryptophan hydroxylase, the ratelimiting enzyme in the serotonin synthesis [43], so that TH activity is preferentially affected when the concentration of BH4 is decreased to submillimolar levels. Further studies will be needed to distinguish these alternative possibilities. Figure 9. Effect of dtorsin mutation on TH activity and GTPCH activity in adult heads. A. Effect of dtorsin heterozygous mutation on TH activity in adult heads (L-dopa nmoles/min/mg protein). The genotypes of adults were y w (n = 3 independent replications), y w dtorsin KO13 /+ (n = 3 independent replications), and w dtorsin KO78 /+ (n = 3 independent replications). Error bars indicate S.E.M. No significant difference between the wild type and the dtorsin heterozygous mutant females. B. Effect of dtorsin heterozygous mutation on GTPCH activity (neopterin nmoles/min/mg protein). The genotypes of adults were y w (n = 3 independent replications), y w dtorsin KO13 /+ (n = 3 independent replications), and w dtorsin KO78 /+ (n = 3 independent replications). Error bars indicate S.E.M. ** p,0.01, significant difference in dopamine levels between the wild type females and the dtorsin heterozygous mutant females. doi:10.1371/journal.pone.0026183.g009 Notably, human GTPCH and TH are defective in two forms of dopa-responsive dystonia (DYT14 and DYT5, respectively), while torsinA is altered in DYT1 dystonia. Deficits in dopamine neurotransmission have been implicated in all these forms of dystonia that have a similar clinical appearance, manifesting as early onset generalized dystonia [37,40]. DYT5 and DYT14 dystonia results from mutations in genes needed for biosynthesis of dopamine, and symptoms can be alleviated with L-DOPA [36]. In the case of DYT1 dystonia, the abnormality seems to involve insufficient neurotransmission by dopamine, although levels of dopamine are normal and, in this case, L-DOPA is not therapeutic. In DYT1 patients, the number of dopamine receptor binding sites in the striatum are reduced [44,45], and, in addition, neuroleptic medications which block D2 receptors can lead to dystonia reactions and tardive dystonia [46]. These studies support a link between TOR1A mutation, altered dopaminergic neurotransmission and dystonic symptoms. This concept has been supported in a variety of DYT1 mouse models showing that expression of mutant torsinA causes an increase in dopamine turnover [47], a decrease of release and uptake of dopamine in response to amphetamine [48,49] or electrical stimulation [50], and impairment of dopamine D2 receptor function [51]. So, although the mechanism through which dysfunctional torsinA leads to compromise of dopaminergic neurotransmission is not clear, many lines of evidence supports this connection [2,40].
It is important to stress that the mutated abnormal torsinA protein is expressed together with normal torsinA in the DYT1 animal models and DYT1 patients, while the null dtorsin fly lines do not express any functional Dtorsin protein. Thus, the fly mutant is fundamentally different from other DYT1 disease model systems. The interaction between the DYT1-mutant form of the torsinA protein (for example DGAG torsinA) and the wild type torsinA protein could be very important for the DYT1 disease manifestations, presumably leading to a decrease in wild type activity. Consequently, it is essential to understand the normal function of the torsinA protein in order to understand how compromise of its function leads to DYT1 dystonia. The availability of fly lines mutant for dtorsin provides the basis for applying the powerful tools of Drosophila melanogaster genetics to define the function(s) of the sole torsin family member in this species, with the expectation that these activities may provide insight into torsinA function in humans. In this regard, the ability of the human torsinA cDNA to rescue the mobility defect of the dtorsin null mutant demonstrates that at least the torsin function involved in regulation of larval locomotion, which requires expression in neurons, is well conserved between the fly and the human. The dtorsin null fly system will therefore permit analysis of the effects of various DYT1-causing mutations on this function, using the biochemical, pharmacological, and genetic approaches.

Fly stocks
Flies were kept on standard medium, except for larval locomotion analysis (see below). The y w; P{70FLP}11 P{70I-SceI}2B Sco/CyO, S 2 line for the gene targeting, GAL4 driver lines, and the FM7i, P{w[+mC] = ActGFP}JMR3/C(1)DX, y f were obtained from the Bloomington Stock Center. The MHC-GAL4 line was obtained from Dr. Aaron DiAntonio (Washington University School of Medicine, MO). The DAT fumin line was a kind gift from Dr. F. Rob Jackson (Tufts University School of Medicine, MA).

Generation of the dtorsin loss-of-function mutant lines
The ''ends-out'' targeting scheme was a modified version of a method described previously [9,10]. The donor vector was constructed using pw25 [10] (a gift from Dr. Kent Golic, University of Utah, UT) in two steps. In step 1, a 4 kb genomic fragment including the 59 portion of dtorsin was amplified from genomic DNA (y w male) by PCR using the following primers: 59-AGCGGCCGCGATATTATGACCCTGGCCAAGGGCTACT-TGGAGCAGCTACCAGGACCGGGCAAGGCATCCTCCTA-TCGCT -39 (TK4A) and 59-CGGTACCGCAAGGAAAATTC-CGCGGTGGAATGTATAGATGGGCAAAAGCGATAGGCA-GCATAAGACTAGA-39 (TK3A). The product was digested with NotI and Acc65I and cloned into the NotI and Acc65I sites of pw25.
In step 2, a 3.5 kb genomic fragment of the 39 portion of dtorsin was amplified from genomic DNA by PCR using the following primers: 59-AGGCGCGCCGGCCATGGCCATTTACTGAG-AAAGGACCGACCGACTCCACATTTTAATGC -39 (TK2) and 59-CCGTACGCAATGCTGTATATATTAATCGTTTGA-GTATAAACAAATATCACGACTTA-39 (TK1). The product was digested with AscI and BsiWI and cloned into the AscI and BsiWI sites of the plasmid made in step 1. Nine fly lines carrying independent insertions of the donor construct were generated using standard procedures. Two independent lines, B2-1 (2nd chromosome insertion) and B7-1 (3rd chromosome insertion) were used for further experiments. Males heterozygous for each donor insertion were crossed to y w; P[70FLP]11 P[70I-SceI]2B Sco/S 2 CyO females, and resulting larvae were heat-shocked at 38uC for 1 hour at 3 days after egg laying (AEL). The resulting unbalanced females were mated in groups of 10 to y w; P[70FLP]11 P[70I-SceI]2B Sco/S 2 CyO males. The progeny from this cross were heatshocked at 3 days AEL, and 405 vials were screened for red eye female upon eclosion. Non-mosaic, unbalanced flies were saved as FLP-insensitive integrations. In total 14 FLP-insensitive integrations were found that mapped to the X chromosome. Of those, seven lines were hemizygous viable and seven (y w dtorsin KO13 , y w dtorsin KO14 , y w dtorsin KO18 , y w dtorsin KO22 , y w dtorsin KO26 , y w dtorsin KO38 , and w dtorsin KO78 ) were semi-lethal. Few male adults survived from the seven semi-lethal lines, and no homozygous females survived.

Southern blot analysis
Southern blots were performed using standard techniques. The DNA was prepared from adult male survivors of y w dtorsin KO13 /Y and w dtorsin KO78 /Y. Two DNA fragments were generated to the 59 and 39 adjacent region of dtorsin coding sequence by PCR for radio labeled probes (Probe 1 and Probe 2) ( Figure S1A) and gel purified. A 514 bp DNA fragment (Probe 1) was amplified by using the following primers: 59-GGATCCACCCGTGGAGCCGGCGAAACTTT-TCCGGGTGCAG -39 and 59-GGCCATGGCCATTTACT-GAGAAAGGACCGACCGACTCCAC -39 and detected a 1.5 kb BamHI fragment in wild type (y w and w) in genomic Southern blots ( Figure S1B). A 381 bp DNA fragment (Probe 2) was amplified by using the following primers: 59-GATTGCAG-CACTTGCCGGCAACCGTTTCTATCGATAGTTG -39 and 59-GGATCCTGGACAGGATACGCTTGACGGATGGGCC-TTGTTG -39 and detected a 0.5 kb BamHI fragment in wild type (y w and w) ( Figure S1B).
Genomic rescue line A 1.9 kb genomic fragment containing dtorsin was amplified from genomic DNA by PCR using the following primers: 59-GCGGCCGCTTGGACCTGGACTTTTCGGGATCCTGGA-CAGGATACG -39 and 59-GAATTCCCTCAAAGATCTGG-GACTAGACGGAAAGCAGAGCGATTT-39. The fragment contains 0.4 kb 59 non-coding region, the whole coding region of the dtorsin gene, and 0.37 kb 39 non-coding region. The product was digested with EcoRI and NotI and cloned into the EcoRI and NotI sites of pCaSpeR3. Four fly lines (GDT101-1,4) carrying independent insertions of the donor construct were generated using standard techniques. All four lines were crossed to individual dtorsin mutant lines.
The transgenic line B5 with UAS-dtorsin transgene on the second chromosome was used for the rescue experiments. A 1.0 kb human torsinA cDNA was amplified from pcDNA3-htorA [52] by PCR using the following primers: 59-GCGGGATC-CATTCATGAAGCTGGGCCGGGCCGTGCTGGGCCTGC-39 and 59-CTCGAGCGGCCGCTCAATCATCGTAGTAATA-ATCTAACTTGGTG-39. The product was digested with Acc65I and NotI and cloned into the Acc65I and NotI sites of pUAST.
The transgenic line #8 with UAS-htorA transgene on the second chromosome was used for the rescue experiments.

HPLC analysis
Brains of third instar larvae or heads of 48-72 hr post-eclosion adult flies were homogenized in 0.1 M perchloric acid. Monoamines were separated by HPLC analysis using a CoulArray HPLC system (model 5600A; ESA, Chelmsford, MA) and a Synergi 4 mm Hydro-RP column (4.66150 mm; Phenomenex, Torrance, CA) as described [21]. One hundred third instar larval brains or 75 to 200 adult heads were extracted in 100-200 ml of 0.1 M perchloric acid. Ten microliters of each extract were injected for each sample. Pool sizes were determined relative to freshly prepared standards (Sigma-Aldrich). Analysis was performed using ESA CoulArray software.

Tyrosine hydroxylase assay
The specific activity of tyrosine hydroxylase (TH) in tissue extracts was determined as described previously [53] with some modifications. Briefly, 50 fly heads were homogenized in 100 ml of cold homogenization buffer. Homogenates were centrifuged at 9,300 x g for 10 minutes at 4uC. The protein concentration of the supernatants was determined using BioRad protein assay reagent (BioRad, Hercules, CA). Extracts were adjusted to equal protein concentrations (20-30 mg/ml), and 50 ml of the extract were mixed with 1 mM m-hydroxybenzylhydrazine dihydrochloride (NSD 1015; Sigma-Aldrich), 2 mM ferrous ammonium sulfate and 0.1% catalase. After incubation for 5 minutes at 37uC, 200 mM Ltyrosine and 1 mM (6R)-5,6,7,8-tetrahydro-L-biopterin (6R-BH4) were added to the mixture and incubated for another 20 minutes at 37uC. The reaction was terminated by adding 100 ml of stop solution (0.1 M perchloric acid, 0.4 mM sodium metabisulphite, and 0.1 mM EDTA) and the mixture was incubated on ice for 10 minutes. The assay mixture was centrifuged for 10 minutes at 1,000 x g at 4uC, and the supernatant was filtered through 0.2 mm PVDF microspin filters (Analytical, Pompton Plains, NJ) and analyzed by HPLC, as described above.

GTP cyclohydrolase activity assay
Approximately 50 heads from 48-72 h post-eclosion adults were homogenized in 100 ml of cold 50 mM Tris, pH 8.0 solution containing 2.5 mM EDTA and 5% sucrose. After centrifugation of the homogenates at 9,300 x g for 10 minutes at 4uC, 5 mg of activated charcoal was added to 100 ml supernatant with gentle mixing to absorb eye pigment. Samples were centrifuged to remove the charcoal, and protein concentrations were determined using BioRad protein assay reagent. Forty-five mg of protein from tissue extracts was mixed with GTP to a final concentration of 0.25 mM in a final volume of 70 ml. After incubation at 37uC for 1 hour, the product of the reaction, 7,8-dihydroneopterin triphosphate, was oxidized to its fluorescent form in 30 ml of 1% iodine and 2% potassium iodide in 1 mM HCl, and incubated for 1 hour at room temperature. The samples were then centrifuged at 14,500 x g for 5 minutes at room temperature. The oxidation reaction was terminated and the samples decolorized with the addition of 15 l of 3% ascorbic acid. The samples were adjusted to pH 8.0, and 50 ml of each sample was mixed with 2 ml of calf intestine alkaline phosphatase, 7 ml of alkaline phosphatase buffer (Roche Applied Science, Indianapolis, IN) and 11 ml of distilled water. The mixtures were incubated at 37uC for 30 minutes to dephosphorylate the neopterin triphosphate product. The supernatant was filtered and then separated by HPLC as described above, except that neopterin was detected by fluorescence at excitation 360 nm/emission 465 nm, and its content was determined by comparison to commercial neopterin (Sigma-Aldrich) as standard using ESA CoulArray software.

Western blot analysis
For analysis of whole flies, thirty adult male flies (y w or y w dtorsin KO13 ) were homogenized in 300 ml of RIPA buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP40, 0.5% deoxycholate, 0.1% SDS (sodium dodecyl sulfate)), and the homogenate was sonicated and then clarified by a brief centrifugation (3,000 rpm for 2 min in a microcentrifuge). For detection of protein in adult heads, heads were removed manually from 50-60 flies of each genotype and were homogenized either in 100 ml phosphatebuffered saline (PBS) with Protein Inhibitor Cocktail (Roche Applied Science, Indianapolis, IN) or in 100 ml RIPA buffer, as above. The protein concentrations were determined with the Dc Protein Assay Reagent (Bio-Rad, Hercules, CA). The proteins (20-30 mg) were separated by electrophoresis in 10% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were blocked with 5% non-fat dry milk in PBS with 0.1% Tween 20 before incubation with primary antibodies. Rabbit anti-Dtorsin antibody raised against amino acids 51-339 of Dtorsin was a gift of Dr. Vijaya Ramesh (Massachusetts General Hospital, Boston, MA) and used at 1:2000 dilution. GTPCH was detected using affinity-purified polyclonal anti-GTPCH isoform B/C antibody [54] and used at 1:1000 dilution. TH was detected using rabbit anti-Drosophila TH [26] and was used at 1:3000 dilution. Rabbit anti-Actin (Sigma-Aldrich) and mouse anti-Tubulin (Sigma-Aldrich) antibodies were used at 1:200 and 1:2000 dilutions. The secondary antibodies used were peroxidase-conjugated anti-rabbit IgG (GE Healthcare, Piscataway, NJ or Jackson ImmunoResearch, West Grove, PA) and peroxidaseconjugated anti-mouse (Abcam, Cambridge, MA). Signals were detected using Pierce ECL Western Blotting Substrate or Supersignal West Pico Chemiluminescent Substrate (Thermo Scientific, Rockford, IL).
Larval locomotion assay and analysis y w dtorsin KO13 was balanced with FM7i Act-GFP. dp cu Pu Z22 px sp and ple 2 were balanced with the CyO arm-GFP and TM3 Act-GFP, respectively. GFP-negative first instar larvae were picked using a fluorescence microscope and transferred into the vials containing 1.5 g of FORMULA4-24 instant Drosophila media (Carolina Biological Supply Company, Burlington, NC) in 7 mls of water.
The larval locomotion assay was carried out at room temperature. Wandering third instar larvae of a particular genotype were picked from the vials. Male larvae were distinguished from females by the presence of larger genital discs as described previously [55]. A single larva was placed at the center of 0.7% agarose in a 100 mm petri dish plate for 30 seconds to recover from handling. The Petri dish was then placed on an evenly illuminated fluorescent light box (TW-26; VWR International, Radnor, PA).
Larval locomotion was recorded for 60 seconds using a Canon Powershot G7 attached to a Nikon SMZ800 microscope. Peristalsis frequency was counted manually from the QuickTime movie.
Larval movement was also recorded at 10 frames per second for 60 seconds using a digital video camera recorder (Handycam DCR-DVD108, Sony). Video tracking was performed using Prairie Dog, an application to trace the movement of an object in a series of TIFF files. Prairie Dog is a new application specifically developed by us to analyze the larval movement and is available upon request from H. M. Some locomotion analysis employed ICARUS software (Tracking/New User Feature) [56]. Larval head and tail locations were extracted and analyzed.

Drug feeding assay
First instar larvae of the wild type (y w) and y w dtorsin KO13 were transferred individually to 1.0 g FORMULA4-24 instant Drosophila media in 5 ml water or 5 ml of 20 mM dopamine-hydrochloride (Sigma-Aldrich), 20 mM octopamine hydrochloride (Sigma-Aldrich), or 10 mM serotonin hydrochloride (Sigma-Aldrich). Larval locomotion assays were performed, as described above.

Antibody staining of larval brains and confocal analysis
Antibody staining of larval brains was carried out as described [57]. Anti-dTH antibody [26] was a gift from Dr. Wendi Neckameyer (St. Louis University, St. Louis, MO) and was used at a 1:5000 dilution. The images were acquired by a Zeiss confocal microscope (LSM5 Pascal) and processed with Photoshop CS4. Multiple z-section confocal images were projected onto a single plane with NIH Image-J software (version 1.40) by Max intensity method. Figure S1 Isolation of dtorsin loss-of-function flies by homologous recombination. A. Ends-out knockout targeting scheme, illustrating how a white mini-gene was inserted into the dtorsin genomic region, resulting in a complete deletion of the dtorsin open reading frame while leaving the surrounding region intact. Other genes in the dtorsin genomic region are indicated by open boxes. Direction and location of primers that were used to make the knockout construct are shown by arrows. Locations of probes 1 and 2 (black boxes) and BamHI sites (B) are indicated. B. Southern blot of BamHI digested genomic DNA, confirming proper integration of the targeting construct. Probe 1 detected 1.5 kb BamHI fragment in the BamHI digested genomic DNA of the wild type (y w/Y and w/Y) males and 5.7 kb BanHI fragment in y w dtorsin KO13 /Y and w dtorsin KO78 /Y males. Probe 2 detected 0.5 kb BamHI fragment in the BamHIdigested genomic DNA of the wild type (y w/Y and w/Y) males and 5.7 kb BanHI fragment in y w dtorsin KO13 /Y and w dtorsin KO78 /Y males. . dtorsin null male dtorsin KO13 (B) had shorter, thinner bristles with reduced pigmentation in the thorax compared to wild type (A). Both larger bristles (macrochaetae) and smaller bristles (microchaetae) were affected. These bristle phenotypes were completely rescued by the introduction of 1.9 kb genomic fragment containing the entire dtorsin gene (GDT101) (C). (TIF) Figure S4 Effect of dopamine feeding on the adult pigmentation. A-D, Effect of dopamine on the pigmentation phenotype of dtorsin mutants. Adult males of the genotype y w (A), y w+50 mM dopamine (B), y w dtorsin KO13 (C), and y w dtorsin KO13 +50 mM dopamine (D) flies. The pictures for A-D were taken one day after eclosion to ensure pigmentation reached the maximum level. Dopamine was added to the fly food during the larval period, as described in Materials and Methods. (TIF) Figure S5 The double heterozygous combinations for dtorsin and DAT mutants have similar mobility compared to the wild type larvae. Peristaltic frequencies for the wandering stage third instar larvae of the genotype dtorsin KO13 /+; female (n = 7), and dtorsin KO13 /+; DAT fumin /+ female (n = 14) are shown. Results are the mean 6 S.E.M. No significant difference between the control females (dtorsin KO13 /+) and the single heterozygous females (dtorsin KO13 /+; DAT fumin /+).

(TIF)
Video S1 Shown is the crawling behavior of wild type (y w) third instar larva for 60 seconds. Wild type larvae showed smooth peristaltic movement of 40 to 60 strides per minute.

(MP4)
Video S2 Shown is the crawling behavior of dtorsin mutant male (y w dtorsin KO13 /Y) third instar larva. dtorsin mutant larvae stopped more frequently and changed directions of movement more often than the wild type. (MP4) Table S1 (DOCX)