Drosophila lines with mutant and wild type human TDP-43 replacing the endogenous gene reveals phosphorylation and ubiquitination in mutant lines in the absence of viability or lifespan defects

Mutations in TDP-43 are associated with proteinaceous inclusions in neurons and are believed to be causative in neurodegenerative diseases such as frontotemporal dementia or amyotrophic lateral sclerosis. Here we describe a Drosophila system where we have engineered the genome to replace the endogenous TDP-43 orthologue with wild type or mutant human TDP-43(hTDP-43). In contrast to other models, these flies express both mutant and wild type hTDP-43 at similar levels to those of the endogenous gene and importantly, no age-related TDP-43 accumulation observed among all the transgenic fly lines. Immunoprecipitation of TDP-43 showed that flies with hTDP-43 mutations had increased levels of ubiquitination and phosphorylation of the hTDP-43 protein. Furthermore, histologically, flies expressing hTDP-43 M337V showed global, robust neuronal staining for phospho-TDP. All three lines: wild type hTDP-43, -G294A and -M337V were homozygous viable, with no defects in development, life span or behaviors observed. The primary behavioral defect was that flies expressing either hTDP-43 G294A or M337V showed a faster decline with age in negative geotaxis. Together, these observations implied that neurons could handle these TDP-43 mutations by phosphorylation- and ubiquitin-dependent proteasome systems, even in a background without the wild type TDP-43. Our findings suggest that these two specific TDP-43 mutations are not inherently toxic, but may require additional environmental or genetic factors to affect longevity or survival.


Introduction
Cytoplasmic protein aggregates are a common pathological feature of many neurodegenerative diseases [1]. The TAR DNA binding protein of 43 kDa (TDP-43) has been identified as a major component of the ubiquitinated aggregates in neurons of patients with either frontotemporal lobar degeneration (FTLD) or amyotrophic lateral sclerosis (ALS) [2,3]. The plasmids by BestGene (Chino Hills, CA, USA). One plasmid contained a donor template coding for the hTDP-43 cDNA sequence and a red fluorescent protein under the control of an eye-specific promoter flanked by portions of the TBPH 3' and 5' UTR was injected at a concentration of 500 ng/μl, and the other contained two guide RNAs and injected at a concentration of 200 ng/μl. The guide RNAs target cas9 to sites on either side of the TBPD coding region, resulting in two double strand breaks which are repaired by homologous recombination using the donor template.
To generate the donor template, the fly codon-optimized human TDP-43 cDNA sequence was synthesized (GeneArt, Lifetechnologies; CA, USA) and cloned into pCMV-3FLAG to produce the FLAG-hTDP-43 DNA fusion. This FLAG-hTDP-43 fragment was ligated to the 3' end of Drosophila TBPH 5'UTR (amplified by primers 5' Homology Arm-S and -R, Table 1), and sub-cloned into the AarI sites of 5' MCS in pHD-DsRED attp vector [36]. The 3' UTR of TBPH (amplified by primers 3' Homology Arm-S and -R, Table 1) was incorporated through the SapI sites of 3' MCS in pHD-DsRED attp vector. To incorporate ALS/FTD associated TDP-43 mutations into the donor template, primers against hTDP-43 G294A or M337V were designed through the NEBaseChanger web tool. The pHD-DsRED TDP-43 mutant forms were generated, followed by the instruction of Q5 site-directed mutagenesis kit (New England Biolabs; MA, USA).
The guide RNA sequences were designed using the Drosophila RNAi Screening Center CRISPR2 web tool (http://www.flyrnai.org/crispr2/). The sequences selected were located in the intron within the TBPH 5' UTR and the 3' end sequence was adjacent to the stop codon of the TBPH coding region. These two crRNA (see Table 1, CRRNA-5UTR and CRRNA-3UTR) were cloned into two sibling vectors, pBFv-U6.2 and pBFv-U6.2B, respectively [37]. Finally, these two independent crRNA cassettes were fused into a single plasmid for increased injection efficiency [37].

Generation of DsRED minus flies
To generate the fly lines lacking the eye-specific DsRed cassette, fly lines expressing wild type hTDP-43 were crossed with flies expressing Cre recombinase. The DsRED cassette within the donor template is flanked by two loxP sites, and can be excised in the presence of Cre recombinase. Offspring from this cross were screened for lack of red fluorescent eyes, crossed to a second chromosome balancer line and the resulting progeny selected for loss of the balancer chromosome and then genotyped to confirm the presence of hTDP-43 and absence of TBPH.

Real time reverse transcription (RT)-PCR
Total RNA was extracted using Trizol reagent (Life Technologies, CA, USA), pre-cleaned by DNaseI and cDNA synthesis performed using SuperScript III (Life Technologies, CA, USA).
To determine the relative expression of TBPH and hTDP-43 in the control flies and flies expressing wild-type hTDP-43, hTDP-43 G294A and hTDP-43 M337V, primer pairs against the 5' UTR of TBPH gene, and also three internal controls, EF2b, CG9165 and SDha were designed using Primer3 (See Table 1 Adult fly heads were dissected on an ice-cold plate, washed with PBS followed by 70% ethanol and homogenized in lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% TritonX100, 1mM EDTA, 5% glycerol, 2 mM AEBSF, 0.3 μM aprotinin, 130 μM bestatin, 14 μM E-64, and 1 μM leupeptin) and centrifuged at 1,000 × g for 5 min at 4˚C. The supernatant was kept as the Tris-soluble fraction, and the pellet was extracted with 0.5% SDS, followed by 10,000 × g for 5 min at RT. The supernatant was kept as SDS soluble fraction. These Tris-soluble and SDS-soluble fractions were denatured in 1X Laemmli sample buffer at 85˚C and proteins fractionated by SDS-PAGE. Following transfer to polyvinylidine difluoride membranes (Pall life science, NY, USA), the membranes were blocked with either 0.1% gelatin in PBST (PBS containing 0.05% tween-20) or 3% non-fat milk and incubated with either anti-FLAG or anti-hTDP-43 antisera followed by HRP-conjugated goat-anti mouse antisera. The proteins were then detected with enhanced chemiluminescence DuoLuX (Vector laboratories, Inc., Burlingame, CA, USA). The same blot was stripped with stripping buffer (Thermo, Rockford, USA) at room temperature for 10 mins, followed by blocking, probed with 1:1,000 anti-GAPDH antisera, incubation of 1:10,000 HRP-conjugated goat-anti rabbit antisera (Jackson Immunoresearch) and the detection with enhanced chemiluminescence (Invitrogen). The relative TDP-43 levels were normalized to GAPDH levels by use of Gel plugin of ImageJ.

Immunofluorescence
To visualize the distribution of FLAG-hTDP-43 in the central nervous system (CNS), 7 daysold flies were prefixed in 4% formaldehyde in 1X PBS, 1%TritonX-100 for 20 min at room temp. Dissection of the CNS was performed in ice-cold PBS and the tissue post-fixed in 4% formaldehyde in 1X PBS, 0.3%TritonX-100 (PBSTX) for 20 min. The dissected tissues were then blocked with 5% normal goat serum (NGS) in PBSTX overnight, followed by incubation in mouse anti-FLAG antibody (sigma Aldrich; 1:500) overnight at 4˚C in PBSTX. After washing 5X with PBSTX, the tissue was incubated in 1:500 AlexaFluor 488 goat anti mouse IgG (Jackson Immunoresearch) overnight at 4˚C, and then washe 5Xwith PBSTX. The tissue was then rinsed in a series of concentrated glycerol and mounted in 50%Vectashield (Vector Laboratories, INC., Burlingame, CA). Images were acquired with an Olympus Fluoview FV1000 confocal microscope using the same setting of Olympus Fluoviewer3.1 for experimental and control samples. For the cytoplasmic FLAG-hTDP-43 signal, the fluorescence of the pixels outside the nucleus (delineated by DAPI staining) were quantified by use of the histogram plugin of FIJI, following subtraction of the background signal, and then normalized to the total pixels within the cells.

Phenotyping
All the hTDP-43 replacement flies were outcrossed to the control flies (described above) at least five times to minimize the effects of differing genetic backgrounds and potential CRISPR off-target effects. To determine if there were any differences in developmental time mated females from each line were allowed to lay eggs for 24 hrs and the time taken for the progeny to eclose was recorded. The longevity of each line was assessed by collecting virgin males and females of each line and placing them in separate vials of 10-15 animals in each vial. Vials were inspected for dead flies every other day and the remaining flies placed in new vials each week. Fecundity was assessed by placing a single virgin male and female in a vial for 7 days and then counting the total number of adult flies produced from each vial.

Behavioral assays
Larval locomotion was determined for each line as previously described [11] and adult activity assessed in an activity recorder as previously described [38].
Negative geotaxis assays were performed by placing flies (ranged from 1~28 days-old of age) in separate vials (10 flies per vial), one day before the assays. Flies were transferred to adapted 25 ml tubes to be used as vertical climbing columns. Flies were placed in a 25˚C incubator for recovery for 30 min before the experiment and then tapped to the bottom of the tubes and allowed to climb for 60 s and a digital image captured of the entire tube. The whole climbing tubes was captured as an image. The number of flies in the top one tenth (n top ) and the bottom one tenth (n bottom ) of the tubes was recorded. A performance index (PI) was calculated as follows: ½[(n total + n top − n bottom )/n total ] as described previously [39]. The mean PI of each genotype was calculated from at least three groups of flies.

Generation of novel fly lines that express hTDP-43 at equivalent levels to the endogenous TDP-43 orthologue
Previous studies using Drosophila to model TDP-43 proteinopathies have shown that loss of function of TBPH, the Drosophila TDP-43 orthologue, is lethal [9][10][11]15]. Similarly, overexpression of hTDP-43 in fly models also highlights the toxicity of over-expression of both wild type and mutant hTDP-43 [40][41][42][43][44][45]. To minimize the effects of over-expression, we have developed an alternative method of expressing hTDP-43 cDNA under the control of the endogenous TBPH regulatory regions in the absence of endogenous TBPH expression, using the CRISPR/Cas9 genome editing technique ( Fig 1A). Immunoblots from the resulting fly line demonstrated the absence of TBPH and presence of FLAG-hTDP-43 ( Fig 1C). The donor template used to replace the TBPH coding region with hTDP-43 also included a sequence for expressing a red fluorescent protein (DsRed) under the control of an eye-specific promoter to identify flies that incorporated hTDP-43 ( Fig 1A). To verify that the presence of the DsRed cassette had no effect on the expression of hTDP-43, we also generated a DsRed free-hTDP fly line by excising the DsRed cassette utilizing the flanking loxP sites ( Fig 1A). We assessed the levels of hTDP-43 and TBPH transcripts by quantitative RT-PCR using a probe against the 5'UTR of TBPH, which is common to all three of these lines. These results showed that there were no differences in expression between flies expressing TBPH (control) and wild type hTDP-43 with (hTDP-43 WT DSR+) or without (hTDP-43 WT DSR-) the DsRed cassette ( Fig  1B). Similarly, the protein levels of hTDP-43 showed no difference in flies that retained or had the DsRed cassette excised (Fig 1D & 1E), suggesting that the DsRed cassette did not interfere with the expression of hTDP-43.

Reciprocal changes in TDP-43 transcript and protein levels in flies expressing mutant hTDP-43
Using the same strategy, we generated two additional fly lines carrying disease causing mutations in hTDP-43 along with the DsRed marker. One of these, hTDP-43 G294A, was identified in a sporadic ALS case, with disease onset at 65 years of age [32,46]. The other, hTDP-43 M337V, was identified in three cases of familial ALS, with disease onset ranging from 37 to 55 years of age [32, 33, 47].
To determine the expression level of hTDP-43 we used real-time RT-PCR with a probe against the 5'UTR of TBPH, which is preserved in all four lines and tested young (~7 days post eclosion) and aged (~28 days post eclosion) flies. For both young and aged flies, these results clearly show that there was no difference in the levels of TBPH/hTDP-43 transcript between control flies (expressing TBPH) and flies expressing wild type hTDP-43. By contrast, both mutant lines had slightly reduced transcript levels in young flies (Fig 2A), however, these Transcript levels. Real time RT-PCR was carried out on the total RNA from adult heads of young (7 day old) and aged (28 day old) animals. For each sample, primers to the 5' UTR of TBPH and three separate housekeeping genes were used (EF2, CG9165 and SDha) and the relative expression levels averaged across each housekeeping gene and then normalized to the relative expression in control flies. No significant difference between control flies and flies expressing wild type hTDP-43 was seen; whereas flies expressing either mutant hTDP-43 had significantly lower levels of hTDP-43 transcriptional differences were absent in aged flies. In addition, there was no significant difference in the transcript levels between young and aged control or hTDP-43 WT flies.
To assess the levels of hTDP-43 protein expressed in these lines we separated the total CNS proteins of all fly lines into two fractions, Tris-soluble and SDS-soluble fractions, and then analyzed these fractions on immuno-blots using an antiserum against hTDP-43. The Tris-soluble fraction revealed a single immunoreactive band at the expected size for hTDP-43 in the three lines expressing hTDP-43 ( Fig 2B). No immunoreactive band was detected in control flies due to the lack of cross reactivity of the antiserum with TBPH. In contrast to the results observed from the real time RT-PCR, both lines expressing mutant hTDP-43 showed higher levels of hTDP-43 protein compared to the flies expressing wild type hTDP-43 in this fraction in young animals (Fig 2B & 2C). The SDS-soluble fraction from young animals also showed elevated levels of hTDP-43 in the two mutant lines (Fig 2D & 2E) and in addition revealed the presence of increased levels of high molecular weight species of hTDP-43 in the two mutant hTDP-43 expressing lines ( Fig 2D). Interestingly, aged flies expressing either wild type or mutant hTDP-43 showed decreased hTDP-43 protein levels as the flies aged in both the Tris (

Ubiquitination and phosphorylation of mutant hTDP-43
To characterize the molecular phenotypes further, we used antiserum specific to phosphorylated TDP (pTDP) and examined immunoblots of head lysates of the fly lines. In addition, we used an ubiquitination linkage-specific K48 antibody to examine immunoprecipitated protein, pulled-down by a TDP-43 antibody. When the lysates were blotted with the pTDP-43 antibody, a band of~53 kDa was visible in all three of the transgenic fly lines and notable was increased in flies expressing hTDP-43 M337V (Fig 3A and 3B). When the immunoprecipitated samples were immuno-blotted with anti-FLAG antiserum, immuno-reactive bands of 50-kDa were detected in all three hTDP-43 transgenic fly samples, and the levels of protein in the mutant samples were highly enriched (Fig 3D). We also blotted these samples with antisera against the K48-specific ubiquitin linkage (Ub-K48). Two major immuno-reactive bands were detected, one at~80-kDa, which was present in all three of the hTDP-43 transgenic fly lines; while the other, at~110-kDa, was present only in mutant samples (Fig 3E). It was notable transcript. However, these transcript variations between hTDP-43 wild type and mutant flies were diminished in aged samples. Mean and SEM from three independent samples are shown, ** p<0.01, **** p<0.0001(ANOVA). B-E. Protein levels. Adult fly heads from each genotype were homogenized in Tris buffer (see Methods), centrifuged and the supernatant (Tris fraction) analyzed by immunoblot with an anti-hTDP-43 antiserum and anti-GAPDH as a loading control. The pellet was re-homogenized in SDS buffer, centrifuged and the supernatant (SDS fraction) analyzed by immunoblot in parallel. B. Representative immuno-blot for Tris-soluble fractions of young (Y) and aged (A) flies for each genotype. C. Quantification of Tris-soluble fraction showing increased levels of hTDP-43 in both lines expressing mutant hTDP-43 relative to flies expressing wild type hTDP-43 (WT hTDP-43). The intensity of the 50kDa band was digitized and the relative intensity compared to that of GAPDH calculated. For each blot the relative expression was normalized to the expression of hTDP-43 in flies expressing wild type hTDP-43. The levels of hTDP-43 in flies expressing mutant hTDP-43 was significantly elevated relative to flies expressing wild type hTDP-43. Notably, the hTDP-43 protein level was decreased with age in all fly lines. There was no difference in hTDP-43 levels among all the aged genotypes.  that the signal for both Ub-K48 bands was always the strongest for the hTDP-43 M337V samples (Fig 3F).

Increased cytoplasmic puncta in neurons from flies expressing mutant hTDP-43
We also examined the expression of hTDP-43 in neurons in the adult CNS using immunofluorescence to localize the FLAG-tag, fused at the N-terminus of hTDP-43 (Fig 4). These results showed that most, if not all, neurons expressed hTDP-43, similar to results reported for TBPH expression in the adult CNS [10,48]. In flies expressing wild type hTDP-43, the majority of the FLAG-TDP-43 signal was located in the nucleus (Fig 4B). By contrast, in flies expressing mutant hTDP-43, although the majority of the signal was detected in nuclei, we also observed many neurons that contained small fluorescent puncta located in the cytoplasm (Fig 4C & 4D). To quantify this, we determined the fraction of cytoplasmic versus total FLAG immuno-fluorescence. As expected, in flies that expressed wild type hTDP-43 there was a low level (about 10%) of FLAG-immuno-fluorescence localized in the cytoplasm, as judged by the lack of overlap with the DAPI signal. This is consistent with previous studies showing that most of the TDP-43 is located in the nucleus with some protein shuttling between the nucleus and the cytoplasm [49]. By contrast, in flies expressing mutant hTDP-43 G294A or its M337V mutant there were significant increase in the amount of TDP-43 localized to the cytoplasm with approximately 30% in the cytoplasm (Fig 4E). The majority of this cytoplasmic fluorescent signal was present as small (approximately 0.2μm) puncta in the two mutant lines (Fig 4C  & 4D arrows).
The punctate nature of the cytoplasmic TDP-43 in the flies expressing mutant hTDP-43 was reminiscent of stress granules. To determine whether the observed cytoplasmic puncta shared characteristics of stress granules, we co-labeled the CNS using anti-FLAG antiserum as well as an antiserum against Drosophila fragile X mental retardation 1 (FMR1), a component of stress granules in Drosophila [50][51][52]. Although the majority of the FMR1 staining was present as cytoplasmic puncta as expected ( Figure  Neurodegenerative diseases that include TDP-43 proteinopathies are usually characterized with phosphorylated TDP-43 and ubiquitin positive inclusions in the cytoplasm or the nuclei of neurons [3]. To determine whether the introduction of mutant hTDP-43 caused the phosphorylated TDP and ubiquitin-positive inclusions, adult brains were labeled with anti-phospho-TDP-43 (409/410) and antisera specific to the K48 ubiquitin linkage. Immuno-fluorescence for the ubiquitin staining of hTDP-43 WT and G294A were comparable, with no significant phospho-TDP-43 or ubiquitin puncta detected (Fig 5A & 5B). For the flies expressing hTDP-43 M337V, kDa), and also a truncated fragment (~40 kDa). Samples from both mutants also revealed significantly higher levels of hTDP-43 compared to flies expressing wild type hTDP-43. E. Representative immuno-blot probed for the K48 specific ubiquitin linkage (Ub K48). An approximately 80-kDa UB K48 signal was present in all of the transgenic hTDP-43 fly samples (*), whereas a larger, approximately 110-kDa Ub K48 band was only detectible in samples from the mutant hTDP-43 lines. F. Quantification of the 80-kDa Ub K48 signal shows that the sample of the M337V mutant has the significant increase of Ub K48 intensity. ** p<0.01,***p<0.001 (ANOVA) (mean and SEM, n = 3). https://doi.org/10.1371/journal.pone.0180828.g003 Phosphorylation and ubiquitination in Drosophila lines with mutant TDP-43 however, there was a global enrichment of phospho-TDP-43 signal and a higher incidence of neurons containing ubiquitin-positive puncta within the nuclei and cytoplasm (Fig 5C, 5D & 5E), similar to pathological reports in FTLD and ALS cases [3,53]. Occasionally, we observed small ubiquitin positive puncta scattered around the nucleus, however, only rarely did they overlap with hTDP-43 puncta in the mutant lines ( Figures C&D in S3 Fig). Only mild phenotypes are associated with expression of mutant hTDP-43 All three of the fly lines were viable as homozygotes and to determine whether there were any defects in development, we measured the length of time between hatching and adult emergence (eclosion). No significant differences were observed between control flies and those expressing wild type hTDP-43 and no delays in development were observed with flies expressing either mutant form of hTDP-43 ( Fig 6A & 6B), indicating these mutations had no effect on larval or pupal development. To assess long-term survival we measured the life span of the lines and again no significant reductions were detected in any line expressing wild type or mutant hTDP-43 compared to the control lines (Fig 6C & 6D). We next set up pairwise crosses to test the fecundity of the different lines. These data showed no differences in the number of offspring produced between control flies and those expressing wild type hTDP-43, but a reduced number of progeny in the hTDP-43 G294A and M337V flies (Fig 6E). To determine whether this defect was due to defects in the male or female flies, systematic pairwise crosses between each of the lines were performed. Interestingly, no significant defect were found associated with male flies among all the lines, but a significant reduction in the number of progeny numbers produced from female hTDP-43 G294A and M337V fly lines was detected, suggesting females were more affected by the hTDP-43 mutations ( Fig 6F).
To determine if there were any behavioral effects of mutant hTDP-43, we analyzed larval crawling behavior by measuring the distance 3 rd instar larva crawled during a 5 minute observation period. No differences were detected between any of the lines (Fig 7A). To assess adult locomotion, we placed individual flies in an activity recorder that continuously monitored movement. Flies show diurnal patterns of activity and these patterns were similar in all four lines (Fig 7B). The total level of daily activity was also quantified with this approach and these results showed similar levels of activity between control flies and flies expressing wild type hTDP-43 (Fig 7C). The two mutant lines exhibited significantly different levels of daily activity with flies expressing hTDP-43 G294A showing elevated levels of activity and flies expressing hTDP-43 M337V reduced levels of activity compared to wild type hTDP-43 expressing flies (Fig 7C). Adult flies exhibit negative geotaxis behavior, rapidly climbing to the top of vials. When we measured this behavior in 1 day old flies, we detected no differences between control flies and flies expressing wild type hTDP-43, G294A and M337V. There was a gradual reduction in performance in this assay as the flies aged, with a faster decline in the fly lines expressing each of the mutant forms of hTDP-43 (Fig 7D). By 4 weeks of age, both of the flies with hTDP-43 G294A and -M337V showed a significantly reduced performance, compared to the control flies and those expressing wild type hTDP-43. D'). E. Quantification of the non-nuclear hTDP-43 in adult brain. Cytoplasmic FLAG fluorescence was normalized to the total FLAG fluorescence, and over 100 neurons were quantified for individual prep. As the result the cytoplasmic FLAG is significantly elevated in flies expressing hTDP-43 G294A and M337V. (ANOVA, *** p<0.0005,****p<0.0001, n = 8). Scale bar is 50 μm in A and 2μm in B. https://doi.org/10.1371/journal.pone.0180828.g004 Phosphorylation and ubiquitination in Drosophila lines with mutant TDP-43 separate experiments with males and females of the same genotype in each vial. The number of progeny from each pair was normalized to the mean number of progeny for the control flies in that experiment. Each point represents the number of progeny from a single pair of flies and the lines represent the mean and SEM for between 40 and 60 pairs of flies. Both lines expressing mutant hTDP-43 generated significantly fewer progeny than either control flies or flies expressing wild type hTDP-43, * p<0.05; *** p<0.001; ANOVA followed by Holm-Sidak multiple comparison test. F. Fecundity of males compared to females. Individual males of each genotype were crossed with individual females of each genotype and the total progeny of each pair were then pooled based on the sex and genotype of the parent. The data represents the mean and SEM for between 30 and 35 pairs of flies. There were no significant differences in the number of progeny between the males of each genotype, whereas female flies expressing either mutant hTDP-43 produced significantly fewer offspring. *** p<0.001; two way ANOVA followed by Sidaks multiple comparison test.
https://doi.org/10.1371/journal.pone.0180828.g006  = 21-30). B. Adult hourly activity. The activity pattern of individual flies was tracked using an activity recorder and no significant differences were observed for any genotype (mean ± SEM, n = 18-25 flies). C. Adult daily activity. The total activity for each fly was summed for each 24hr period and monitored for 6 days. There was no significant difference between control flies and flies expressing wild type hTDP-43, whereas flies expressing hTDP-43 G294A showed significantly higher levels of activity and flies expressing hTDP-43 M337V had significantly lower levels of activity compared to flies expressing wild type hTDP-43. Two way ANOVA followed by Dunnett's multiple comparison test. Each point represents mean ± SEM n = 25-35 flies. D. Adult negative geotaxis. The ability of flies to climb to the top of a tube was determined as described in the methods section. There was no significant difference in the performance index of control flies and flies expressing wild type hTDP-43. The climbing indexes of hTDP-43 G294A and M337V flies were gradually reduced by time compared to that of flies expressing wild type hTDP-43. Two way ANOVA followed by Dunnett's multiple comparison test. Each point represents mean ± SEM n = 3-20. https://doi.org/10.1371/journal.pone.0180828.g007 Phosphorylation and ubiquitination in Drosophila lines with mutant TDP-43

Discussion
TDP-43 cytoplasmic aggregates are a pathological characteristic of ALS, FTLD and other neurodegenerative diseases [2,3,5,54]. Notably, around 90% of ALS and 50% of FTLD cases have TDP-43 aggregates, strongly implicating the formation of TDP-43 inclusions in the etiology of these diseases [55]. To mimic this TDP-43 proteinopathy, either wild type hTDP-43 or the mutant versions of this protein have been expressed in rodent models. Neurological findings such as axonal degeneration, motor neuron loss, motor dysfunction, learning/memory deficit, brain atrophy as well as TDP-43 inclusions have been reported in these studies [56][57][58][59][60][61][62][63][64][65][66][67][68][69]. Most of these models have shown that many of the phenotypes of these transgenic animals were dependent on the levels of TDP-43. Low levels of expression resulted in TDP-43 cytoplasmic accumulation and age-associated motor dysfunction [64,70], while high levels of expression (usually 3~5 fold compared to endogenous TDP-43) caused early disease onset, but without significant motor neuron losses and TDP-43 inclusions [20]. However, there are some clinical observations concerning TDP-43 proteinopathies that have not been addressed in these models. For example, in contrast to the fast-developing disease onsets of models overexpressing TDP-43, the disease onsets of the patients with familial TDP-43 mutations range from 31 to over 70 years-old of age [26,29,30,71]. Even with specific a TDP-43 mutation, e.g. M337V across three families, the average disease onsets vary from 38 to 52 years-old of age and the disability of the patients within the same pedigree varies considerably [27,32,33]. Similarly, diverse clinical observations are also available in other familial TDP-43 mutations [26, 28, 30, 72]. These clinical observations, however, not only indicate that the TDP-43 proteinopathies vary but also suggests that although TDP-43 mutations clearly contribute to the disease, other genetic and/or environmental factors likely play a significant role in disease onset and progression.
In the current study, we have developed a Drosophila system that expresses wild type and mutant hTDP-43, in the absence of endogenous TBPH, to clarify the role of mutant hTDP-43 in disease phenotypes. Although we did observe a transcriptional change in young flies (1 week-old) expressing mutant hTDP-43, however, there was no difference in transcript levels among all the older flies (4 week-old), suggesting that the mutant TDP-43 does not induce the overexpression of the gene itself. We also detected an increase in the levels of soluble hTDP-43 protein in young flies expressing mutant hTDP-43, but these differences were diminished in aged flies, as did the levels of wild-type hTDP-43. It is interesting that in young flies we detected reciprocal changes in transcript and protein levels of hTDP-43. The levels of TDP-43 protein are known to be autoregulated by binding of the TDP-43 protein to the 3'UTR of the TDP-43 transcript [73]. We hypothesize that hTDP-43 is acting through the endogenous tbph UTR to maintain the transcript-protein homeostasis seen in cell-and mammal models. If the mutant protein has a slower turnover rate compared to wild type protein it would result in slightly decreased transcripts in the mutant fly lines as we have observed. Our observation that hTDP-43 protein levels decrease with age is consistent with previous findings reported from both Drosophila and rodent models [74,75]. Interestingly, we also observe an age-dependent reduction in mutant hTDP-43 protein levels. This reduction can be interpreted as an absence of hTDP-43 protein accumulation in these mutants. These results suggest the possibility that the accumulation of TDP-43 in the presence of familial hTDP-43 mutants [27, 76], might not simply be caused by the mutation itself, but by the involvement of other components, e.g. dysfunction of ubiquitin proteasome system [77]. In addition, interestingly, clinical results from sporadic ALS patients show that TDP-43 is not overexpressed in either its transcript or protein levels [20,64,78,79].
In contrast to the relatively small differences in the proteins or transcripts among all the fly lines, the proteins immunoprecipitated with TDP-43 serum showed more substantial changes.
Although the input levels of TDP-43 were equivalent across the three lines, the levels of TDP-43 immunoprecipitated were highly enriched in samples from the two mutant lines compared to the wild type hTDP-43 line, as assessed by FLAG immunoblots. This result suggests that the epitope of the wild type hTDP-43 is masked, possibly by its partner proteins, since TDP-43 usually functions as a protein complex [80]. This would suggest that these complexes are disrupted in the presence of mutant hTDP-43 allowing increased access by the antibody. In addition, in the two mutant samples, there is also an enrichment of ubiquitinated and phosphorylated hTDP-43, which parallels clinical findings [2,3]. The increased levels of ubiquitinated proteins and phospho-hTDP-43 in neurons from flies expressing hTDP-43 M337V were also apparent from immunostaining of the CNS. Interestingly, although most, if not all neurons showed elevated levels of phospho-TDP-43, only a small fraction (less than 10%) of neurons exhibited the ubiquitin inclusions and further studies are required to determine whether these are a defined subset of neurons or whether they are random.
Unexpectedly, there were only mild behavioral defects associated with expression of mutant hTDP-43 in our lines. The lack of any differences between the flies expressing wild type hTDP-43 and control flies expressing TBPH suggests that hTDP-43 can fully fulfill all of the essential functions of TBPH. However, we expected to observe defects in flies expressing mutant hTDP-43. Development and longevity were not diminished and only minor defects in fecundity and negative geotaxis were observed. We did observe slightly increased longevity in the males expressing hTDP-43 M337V, although we do not know whether this is due to the presence of the mutant protein or as a result of differences in the genetic background of the lines, despite outcrossing all the transgenic lines with the control line. These results, nevertheless, stand in marked contrast to results from over-expression of mutant hTDP-43 in a variety of systems where it has been shown to be highly toxic [40,41,44,45,57,64,65,[67][68][69][70][81][82][83][84][85]. This suggests that mutant forms of hTDP-43 are not inherently toxic but require a specific set of conditions that could include time, environmental or genetic interactions to manifest their deleterious effects.

Conclusions
In summary, we have described the preliminary characterization of fly lines designed to allow the examination of the effects of TDP-43 mutations in the absence of endogenous TBPH. Flies that express mutant forms of hTDP-43 display only mild behavioral and developmental phenotypes yet show a number of cellular and molecular characteristics that appear to parallel the histopathological hallmarks of human neurodegenerative disease. These flies will be invaluable tools to identify interacting genes and the downstream components resulting from TDP-43 dysfunction. Interestingly, a recent report describes using similar CRSIPR/Cas9 technology to introduce point mutations into the TDP-43 orthologue in zebra fish, but at this time no phenotypes were reported [86].