Impairment of Drosophila Orthologs of the Human Orphan Protein C19orf12 Induces Bang Sensitivity and Neurodegeneration

Mutations in the orphan gene C19orf12 were identified as a genetic cause in a subgroup of patients with NBIA, a neurodegenerative disorder characterized by deposits of iron in the basal ganglia. C19orf12 was shown to be localized in mitochondria, however, nothing is known about its activity and no functional link exists to the clinical phenotype of the patients. This situation led us to investigate the effects of C19orf12 down-regulation in the model organism Drosophila melanogaster. Two genes are present in D. melanogaster, which are orthologs of C19orf12, CG3740 and CG11671. Here we provide evidence that transgenic flies with impaired C19orf12 homologs reflect the neurodegenerative phenotype and represent a valid tool to further analyze the pathomechanism in C19orf12-associated NBIA.


Introduction
NBIA refers to a heterogeneous group of rare (1-3 cases/1 million population) inherited neurological disorders whose hallmark is iron deposition in the basal ganglia. Currently, nine different subtypes of NBIA have been recognized and associated to their genetic counterpart. PANK2, PLA2G6 and C19orf12 are hot spot genes, with C19orf12 accounting alone for up to 30% of cases.
Most of the knowledge about C19orf12-associated NBIA refers to genetic, clinical and radiological features. So far, 67 patients have been reported in literature [1][2][3][4][5][6][7][8][9] but still no hypothesis has been made concerning the pathomechanism underlying this complex disorder. Two postmortem brain samples were examined and showed iron accumulation, Lewy bodies, tau pathology and spheroids with a strong immunoreactivity for ubiquitin. However, the scarcity of cases makes it difficult to state whether the overlap between C19orf12-associated NBIA and Parkinson's disease is a rule or a private trait.
C19orf12 expression occurs mainly in brain, blood cells and adipocytes. During in vitro adipocyte differentiation C19orf12 expression increases constantly and is co-regulated with genes involved in the fatty acid metabolism suggesting a role for C19orf12 in lipid metabolism [1].
C19orf12 harbors two trans-membrane domains. Immunofluorescence experiments have shown that C19orf12 is localized in mitochondria. Based on these evidences, MPAN (Mitochondrial membrane Protein-Associated Neurodegeneration) has been proposed as acronym to refer to C19orf12-associated NBIA. Although C19orf12 is a mitochondrial protein, mutations in C19orf12 do not appear to affect the mitochondrial bioenergetics in fibroblasts under basal conditions [1].
The scarcity of data available from neuropathology, the limitations of investigating the biochemical basis of the disease directly in the brain, the absence of a phenotype in fibroblasts and the lack of homology to proteins with known function, led us to work on the creation of a model that would allow us to study the consequences of impaired C19orf12 in tissues which are primarily affected. We decided to set up a D. melanogaster model. Drosophila was successfully employed to study pantothenate kinase-associated neurodegeneration (PKAN) resulting from mutations in PANK2 [10,11] and tested in rescue experiments with pantethine [12]. We identified two D. melanogaster orthologs of C19orf12, CG3740 and CG11671. We used double heterozygous deletion flies and an RNAi approach to down-regulate the expression of both genes providing evidence that the orthologs are required for the maintenance of the brain tissue and the normal locomotor function. Our results suggest that Drosophila can be used to gain further insight in MPAN disorder.

Results and Discussion
Drosophila has Two Orthologs of the Human C19orf12 Comparative genomics revealed that the human C19orf12 gene has two orthologs in D. melanogaster, CG3740 and CG11671, which share 63% and 55%, respectively, of similarity at protein level with C19orf12 (NP_001026896.2) (Figure 1). CG3740 is localized on the X-chromosome in position 2B5-2B5 while CG11671 maps to chromosome 3R in position 84F8-84F8. The two Drosophila genes are proposed to originate by a duplication event at the speciation node explaining their 72% similarity. The transmembrane domain is the only feature of C19orf12 that is predicted by the primary amino acidic sequence. PolyPhobius program [13] predicted two transmembrane domains for CG3740, similar to the human protein, and one long one for CG11671. Aside the prediction, it is also possible that CG11671 has two transmembrane domains which would also fit with the length of the predicted domain ( Figure 1). CG3740 and CG11671 are differentially expressed ( Figure 2). The analysis of the mRNA expression pattern in head, thorax and abdomen of adult w 1118 flies revealed that expression levels of CG3740 are comparable among the tissues tested, while CG11671 expression is low in head and high in abdomen.

Generation of Double RNAi and Double Heterozygous Deletion Flies
In order to analyze the consequences of impaired activity we made use of the bipartite GAL4-UAS system. Transgenic lines carrying the yeast Upstream Activating Sequence (UAS) in front of the RNAi constructs for both CG3740 and CG11671 were available at the Vienna Drosophila RNAi stock collection [14]. Both RNAi lines were separately crossed with specific driver lines expressing GAL4 under the control of elav, tubulin or actin promoters, enabling knockdown of CG3740 and CG11671 gene products at the neuronal level (elav) or in the whole body (actin, tubulin). Quantitative PCR performed in RNA isolated from heads of 10 days-old flies revealed that the strongest knockdown was induced under the actin promotor expression, leaving mRNAs levels of only 25% for CG3740 or 10% for CG11671 ( Figure 3A). However, down-regulation of the product of each single gene did not result in any significant difference in lifespan or climbing activity compared to control flies (data not shown).
Based on the homology between CG3740 and CG11671 and the overlapping expression pattern, we hypothesized a compensative mechanism between the two genes that can justify the absence of any pathological phenotype. Hence we generated a stock carrying both RNAi transgenes and crossed these flies with the driver line act-GAL4 in order to reduce compensatory effects between CG3740 and CG11671 ( Figure S1). Quantitative PCRs performed on heads of 10 days-old flies showed a reduction of at least 40% for CG3740 and 75% for CG11671 in double RNAi transgenic flies ( Figure 3A).
In a second approach, heterozygous deletion strains for CG3740 and CG11671 were generated to validate RNAi fly phenotypes. For this purpose, w 1118 flies were crossed with Df(1)BSC589 and Df(3)BSC579 flies, two lines carrying macro deletions encompassing CG3740 and CG11671 respectively and several other genes. Moreover, Df(1)BSC589 and Df(3)BSC579 lines were crossed between each other in order to generate double heterozygous deletion flies ( Figure S2). All heterozygous deletion flies were morphologically normal. Quantitative PCRs performed showed expression levels comparable to RNAi transgenes ( Figure 3B).

Double RNAi Flies Have Reduced Median Lifespan
The lifespan of control w 1118 and double RNAi flies was determined under standard growth conditions by counting the number of dead flies every second day. Down-regulated flies revealed a significantly reduced median lifespan of about 8 days at 50% survival ( Figure 4).

Double RNAi and Double Heterozygous Deletion Flies Show Defective Climbing
NBIA patients with defective C19orf12 display spasticity and progressive cognitive decline. Onset generally occurs in childhood up to early adulthood with slow progression and survival well into adulthood [1,15]. Although double RNAi and double deletion flies are able to walk, climb and fly, we noticed that they recovered more slowly after the application of a force by tapping them to the lower part of the vial. Those flies able to recover and climb were slower than the controls. We quantified the behavioral defects. Since locomotion can be affected by the age, we tested flies of two different ages using age-matched controls for all tests. In addition, two different control strains were used in the assay, the canonical wild-type w 1118 and the act-GAL4/+ flies obtained by backcrossing act-GAL4 for 6 generations with w 1118 (Text S1). When analyzing young flies no differences between double RNAi, double deletion   and control flies were observed. However, 4-weeks-old double RNAi and double deletion flies performed poorly in the climbing assay compared to controls ( Figure 5). Only few down-regulated flies were able to reach the end point fixed at the 150 ml line (17.5 cm) of a cylinder in fifteen seconds. The majority only initiated climbing and stopped before the endpoint was reached, fell back or did not initiate climbing. This behavior suggested a neuromuscular phenotype in double RNAi and double heterozygous deletion flies.

Double RNAi Flies Present Vacuoles in the Brain
Since the hallmark of NBIA is iron accumulation in the brain, we searched for iron accumulation in fly brains of double RNAi flies. The presence of iron deposits was investigated using the standard Prussian blue staining on paraffin-embedded brains of aged flies using sections of mouse spleen as a positive control for the staining. Under our experimental conditions we could not find iron deposits ( Figure 6).
We morphologically analyzed brains for signs of neurodegeneration in order to investigate whether the defective climbing coincided with abnormalities in the nervous system. In Drosophila, the tissue loss is typically detected as accumulation of degenerative vacuoles or holes in the brain [16,17]. For this we analyzed the morphology of brains from 28 days-old double RNAi and agematched control flies. A high number of vacuoles with an area up to 30 mm 2 was found in the brain and in the optical lobe of all investigated down-regulated flies suggesting a neurodegenerative process. Few small vacuoles below 2 mm 2 were also found in one control fly as a normal result of the aging process [18] (Figure 7A-B).

Double RNAi and Double Heterozygous Deletion Flies Present Bang Sensitivity
In order to further analyze possible locomotor defects, we subjected young and adult double RNAi and double heterozygous flies to mechanical stress by vortexing them shortly and banging

Concluding Remarks
Drosophila models have been successfully applied as a tool to unravel basic mechanisms underlying a large number of human neurodegenerative disorders such as Parkinson's disease [20], Alzheimer [21,22], Non-coding Trinucleotide Repeat diseases [23] and PKAN [10,11]. Here we provide compelling evidence that an in vivo model for MPAN has been generated by downregulating the two homologues of C19orf12. The uncoordinated behavior or lack of orientation in the bang test, the impaired climbing activity as well as the shorter lifespan and the morphological alterations in brain suggest that double RNAi flies can be used as a model for MPAN. The presence of locomotor impairment and bang sensitivity also in double heterozygous deletion flies provides additional genetic evidence that the phenotypes detected in double RNAi flies are due to the knockdown of CG3740 and CG11671 rather than due to off-target effects. The availability of CG3470; CG11671 down-regulated flies provide affected tissues, like brain or muscles which are the main targets of NBIA disorders but seldom available for functional studies, and may represent a valid tool for the screening of therapeutics.

Drosophila Genetics
Fly stocks were maintained routinely in our lab at 22uC according to standard protocols. As wild type stock w 1118 and act-

RNA Extraction and qPCR
Flies were cut in three parts: head, thorax and abdomen. Four specimens from each part were collected, resuspended in 500 ml of Trizol (Invitrogen) and homogenized using a teflon pestle. Total RNA was isolated using the Trizol method. One microgram of RNA was reversely transcribed using the M-MLV Reverse Transcriptase (Promega) and oligo dT. CG3740 and CG11671 expression were evaluated in comparison to the ribosomal protein 49.

Behavioral Assays
Two groups of flies, corresponding to ''young'' (i.e. 24 h after eclosion) and ''aged'' (i.e. 4 weeks after eclosion) populations of wild-type, double RNAi and double heterozygous deletion flies, were tested in the behavioral assays. Assays were carried out with male and female down-regulated flies and with female double heterozygous deletion flies. Flies were CO 2 -treated and separated into groups with the same number of individuals per vial one day before the test. Hence flies were subjected to the climbing assay and to the bang test.
The climbing assay was performed according to a modified version of Benzer [24]. 50 flies were placed in a 250-ml glass graduated cylinder that was sealed at the top with parafilm to prevent escape. The flies were gently knocked to the bottom of the cylinder and the number of flies able to cross the 150-ml line (17.5 cm) in 15 seconds was recorded.
The test for bang sensitivity was performed as described by Zhang [25]. Flies were vortexed at maximum setting for 10 sec (Bang I) and the time until all flies upright themselves was recorded. To test for refractory period, the flies were vortexed again 10 min later to see if the flies were still bang sensitive (Bang II). For control as well as mutants, 50 flies were tested for each age.
Statistical analyses were performed using Student's t-test. Data are presented as mean 6 Standard Deviation.

Longevity Test
100 males for control and mutant were collected immediately after eclosion. Every 2 or 3 days flies were passed on fresh food and the number of survived flies plotted in an excel graph until their death.

Brain Histology
Analysis of heads from 28-day-old double RNAi flies and aged matched controls were fixed in 0.1M sodium cacodylate, 2.5% paraformaldehyde, 2.5% gluteraldehyde and embedded in epon. One micrometer horizontal sections were cut for each head. Section from different brain depths were checked for the presence of vacuoles. Sections were stained with 1% toluidine-blue and 1% borax and investigated with an Olympus BX50 light microscope. The number and size of vacuoles was quantified with Image J.
For detection of ferric iron, paraffin embedded sections were incubated for 45 minutes with Prussian blue stain (10% K 4 Fe(CN) 6 : 20% HCl, 1:1) prior to de-paraffinization. Three washes with water followed and staining was enhanced using DAB following standard procedures. Afterwards, the sections were deparaffinized using ascending gradients of ethanol to xylene and preparation were mounted in Permount mounting medium (Fisher Scientific) and imaged on a Olympus BX50 light microscope.