Mutation in E1, the Ubiquitin Activating Enzyme, Reduces Drosophila Lifespan and Results in Motor Impairment

Neurodegenerative diseases cause tremendous suffering for those afflicted and their families. Many of these diseases involve accumulation of mis-folded or aggregated proteins thought to play a causal role in disease pathology. Ubiquitinated proteins are often found in these protein aggregates, and the aggregates themselves have been shown to inhibit the activity of the proteasome. These and other alterations in the Ubiquitin Pathway observed in neurodegenerative diseases have led to the question of whether impairment of the Ubiquitin Pathway on its own can increase mortality or if ongoing neurodegeneration alters Ubiquitin Pathway function as a side-effect. To address the role of the Ubiquitin Pathway in vivo, we studied loss-of-function mutations in the Drosophila Ubiquitin Activating Enzyme, Uba1 or E1, the most upstream enzyme in the Ubiquitin Pathway. Loss of only one functional copy of E1 caused a significant reduction in adult lifespan. Rare homozygous hypomorphic E1 mutants reached adulthood. These mutants exhibited further reduced lifespan and showed inappropriate Ras activation in the brain. Removing just one functional copy of Ras restored the lifespan of heterozygous E1 mutants to that of wild-type flies and increased the survival of homozygous E1 mutants. E1 homozygous mutants also showed severe motor impairment. Our findings suggest that processes that impair the Ubiquitin Pathway are sufficient to cause early mortality. Reduced lifespan and motor impairment are seen in the human disease X-linked Infantile Spinal Muscular Atrophy, which is associated with mutation in human E1 warranting further analysis of these mutants as a potential animal model for study of this disease.


Aggregation Prone Neurodegenerative Diseases
Neurodegenerative diseases are a major cause of mortality and can cause a range of devastating symptoms. While these diseases have a number of symptomatic differences, they also share key features that could reflect a common underlying pathology. For example, aggregated proteins are found in the brains of patients in many of these diseases, and it is currently believed that these aggregates play an important role in pathology of the diseases . Currently, at least 4.5 million people in the United States, roughly 1 in 68, suffer from Alzheimer's Disease (AD), the most common form of dementia, and prevalence of this disease increases exponentially with advancing age and afflicts one third to one half of all people over age 85 [22][23]. In AD, a number of proteins have been shown to adopt abnormal conformations and/ or to aggregate. For example, the microtubule-associated protein tau adopts abnormal conformations forming neurofibrillary tangles (NFT), a typical feature of AD and taupoathies [3][4][5][6]. In addition, inappropriate processing of amyloid-beta (Aß) results in Aß peptides, which form extracellular plaques [6][7][8]. Parkinson's Disease (PD), another condition with increasing incidence upon aging, is the second most common cause of dementia [9][10][24][25][26]. Pathology in PD is thought in part to result from aggregation of the protein alpha-synuclein. Given the large population now entering the relevant ages for typical diagnosis, the number of people afflicted with AD and PD will climb dramatically in coming decades. 1 in 10,000 people suffers from Huntington's Disease (HD), a dominant neurodegenerative condition. HD results from expansion of the polyglutamine (polyQ) repeats of the gene huntingtin (htt); polyQ-expanded forms of the htt protein form protein aggregates [11,27]. Expansion of polyQ stretches are also implicated in other neurodegenerative diseases [20][21].

The Ubiquitin Pathway and Neurodegenerative Diseases
One of the major pathways responsible for clearing mis-folded or aggregated proteins from a cell is the Ubiquitin Pathway. The Ubiquitin Pathway consists of a series of enzymes responsible for attaching the small protein ubiquitin to substrate proteins. In the most upstream step, a Ubiquitin Activating Enzyme, E1, charges ubiquitin and transfers ubiquitin to a Ubiquitin Conjugating Enzyme, E2. The E2 then transfers ubiquitin to a Ubiquitin Protein ligase, E3, or works with an E3 to conjugate ubiquitin to a substrate protein. Ubiquitin can be conjugated to a substrate singly or in a poly-ubiquitin chain. Once ubiquitinated, substrates are then directed to a variety of potential fates including endocytosis and degradation [28][29][30][31]. Normally, mis-folded or aggregated proteins can be poly-ubiquitinated and then degraded by the 26S proteasome [1][2][32][33][34][35].
In AD, PD, and HD, ubiquitinated proteins have been shown to accumulate in inclusions and in protein aggregates. Moreover, isolated Aß 1-42 aggregates, tau aggregates, alpha synuclein aggregates, and polyQ aggregates have been shown to inhibit proteasome function in vitro [13][14][15][16][17][18][19]. Other findings have also implicated the Ubiquitin Pathway in neurodegenerative diseases. For example, a reduced level of E1 has been found in the cytosol of AD patients [32], and one of the familial forms of PD is caused by mutation in a gene called parkin that encodes an E3 enzyme [33][34][35]. Some AD patients also show the presence of, UBB+1, a frameshift mutant of ubiquitin that can inhibit the proteasome once it accumulates in a cell but which cannot be attached to substrate proteins to target them for degradation [36][37][38][39][40][41][42].
These findings together raise the question of whether impairment of the Ubiquitin Pathway on its own can promote increased mortality as a general mechanism underlying a broad spectrum of neurodegenerative diseases or if impairment of the Ubiquitin Pathway occurs largely as a side-effect in neurodegenerative processes.

Drosophila Models of Age-related Diseases
Many crucial signaling pathways and important processes are conserved between Drosophila and humans. In fact, more than 70% of genes associated with human diseases have Drosophlia sequence homologs [43]. Because in vivo assays can address growth, proliferation, apoptosis, and longevity in Drosophila, this system confers the ability to address the functional relevance of genes to disease-associated phenotypes by genetic manipulation. Thus, Drosophila can make substantial contributions to understanding human diseases.
We examined lifespan in different genetic backgrounds in Drosophila. Our findings suggest that impairing the Ubiquitin Pathway is sufficient to promote early mortality. We report here that mutation in one or both copies of Drosophila E1 on its own promoted a dramatic reduction in lifespan. Flies carrying two mutant copies of E1 also demonstrated dramatic motor impairment and aberrant Ras signaling in adult brains. Importantly, the reduced lifespan associated with mutation in one copy of E1 was completely suppressed by reducing the gene dosage of Ras while the reduced lifespan resulting from mutation in both copies of E1 was partially suppressed by reducing the gene dosage of Ras.

Results
Impairment of the Ubiquitin Pathway is implicated in normal aging and in a number of neurodegenerative diseases including AD, HD, and PD. In order to evaluate how loss of ubiquitination could affect lifespan using an in vivo model, we utilized loss-offunction mutations in E1, the Ubiquitin Activating Enzyme that we isolated previously [44][45]. E1 is the most upstream enzyme in the pathway and has no specificity for downstream targets. Therefore, loss of E1 is expected to affect all downstream steps. Moreover, a number of variants have been reported in the human E1 gene, Ube1, including confirmed loss-of-function alleles [46]. To avoid confusion, both human Ube1 and Drosophila Uba1 will be referred to hereafter as simply E1, and Drosophila mutant alleles will be referred to with allele-specific designations (Uba1 B1 , Uba1 B2 , Uba1 A1 , Uba1 A3 , and Uba1 A5 ) as appropriate.
Removing one copy of a gene often has no obvious effect because the remaining wild-type copy can allow for production of sufficient levels of the gene product. In some cases, however, loss of one copy of a gene can result in limiting levels of that gene product and can result in attenuation of downstream processes. Therefore, we examined flies carrying one mutant copy of E1. Flies heterozygous for mutation in E1 show no visible abnormalities when compared to wild-type flies (not shown). These flies emerge from their pupal cases at the expected Mendelian frequencies and are fertile. Despite the lack of a visible phenotype, heterozygous mutations can create sensitized genetic backgrounds or even promote disease symptoms on their own. We examined the lifespan of flies carrying a mutant copy of E1. In parallel assays, flies carrying just one mutant copy of E1 showed a dramatic decrease in lifespan compared to wild-type control flies ( Fig. 1A [47][48]. Therefore we did not examine mated females. Virgin females carrying one mutant copy of E1 also showed a statistically significant decline in lifespan compared to controls (P = 0.0041 by log rank Mantel-Cox, P = 0.0293 by Gehan-Breslow-Wilcoxon), although this decline was less dramatic than in males (shown for Uba1 B1 in Fig. 1C-D).
Surprisingly, male flies carrying the E1 null allele Uba1 A1 (which produces a truncated form of the protein [44]) showed statistically significant increased survival compared to flies heterozygous for the hypomorphic mutations Uba1 B1 and Uba1 B2 . Therefore, we examined additional null mutations in E1, alleles Uba1 A3 and Uba1 A5 each of which produces full-length protein that lacks activity [44]. Similar to Uba1 A1 , male flies heterozygous for either Uba1 A3 or Uba1 A5 showed a statistically significant difference in survival compared to male flies heterozygous for the hypomorphic mutation Uba1 B2 (Fig. 1E-F). A number of possibilities could underlie this phenomenon. Because E1 function is essential at a cellular level, perhaps a feedback mechanism senses overall levels of activity to promote increased E1 expression if levels fall short of a critical threshold. Such a feedback mechanism could be activated in flies heterozygous for a null mutation (where the level falls short) but not in flies heterozygous for hypomorphic mutation (presumably the level does not fall short). Alternatively, the null alleles may fail to interact with other pathway components, but if the hypomorphic proteins interact less productively with binding partners, they could thereby sequester such components from the wild-type protein, resulting in reduced pathway activity.
Motor function in Drosophila can be assessed using standard ''negative geotaxis'' climbing assays [49][50]. When tapped to the bottom of a vial, wild-type flies immediately climb towards the top of the vial. In contrast, flies suffering from motor problems cannot easily climb to the same height as wild-type flies in a similar amount of time. We assessed motor function in E1 mutants by counting the percentage of flies capable of climbing to a height of 4 cm within 5 seconds. Despite the decline in lifespan, heterozygous mutation in E1 did not significantly impair motor function (data not shown) when compared to age-matched control flies.

E1 Homozygous Mutants Often Show Patterning Abnormalities
We reported previously that flies homozygous for null mutations in E1 are embryonic lethal, while flies homozygous for hypomorphic mutations have dramatically reduced viability [45]. As described, flies carrying only one mutant copy of E1 have no visible phenotypes although they show a significant reduction in lifespan. In contrast, flies carrying two hypomorphic mutant copies of E1 have visible and severe phenotypes that may result from processes dysregulated in these mutants during development. Importantly, these abnormalities could reveal insights into the mechanisms underlying the mutant phenotypes observed for mutation in one or both copies of E1.
Flies homozygous for the hypomorphic mutation Uba1 B2 die during late larval or pupal stages and do not reach adulthood.
Very few Uba1 B1 /Uba1 B1 flies or Uba1 B1 /Uba1 B2 flies reach adulthood. Those mutant flies that do reach adulthood typically display a number of obvious abnormalities including rare outgrowths [45], and they appear to be infertile. We report here additional abnormalities in the wing as well as mis-patterned bristles including disruption of the bristle patterns on the notum and in the sternopleural region. Uba1 B1 /Uba1 B1 flies frequently exhibit extra sternopleural bristles (Fig. 2C) compared to wild-type flies ( Fig. 2A) that resemble the Sternopleural (Sp) mutant phenotype (Fig. 2B). Extra sternopleural bristles sometimes form as a consequence from local increased wingless signaling [51]. This would be consistent with our previous findings of wingless accumulation in E1 null mutant clones in both the wing and the eye in regions of wingless expression [44]. We also frequently observe loss of one or more dorsocentral mechanosensory bristles ( Fig. 2E-G) compared to the wild-type pattern (Fig. 2D).

E1 Mutants Demonstrate Motor Impairment
The wings of Uba1 B1 /Uba1 B1 flies are typically held out at an abnormal angle (Fig. 2K-L) compared to wild-type posture (Fig. 2J). Such wing posture can reflect problems with muscles or motor neurons [54][55], so we investigated the motor function of these Uba1 B1 /Uba1 B1 and Uba1 B1 /Uba1 B2 mutants. 5 days after emerging from their pupal cases, 79 percent of males flies and 69 percent of female flies of the control genotype w; FRT42D are capable climbing 4 cm in 5 seconds. In contrast, only about 20 percent of 5 day-old and 10 day-old Uba1 B1 /Uba1 B1 and Uba1 B1 / Uba1 B2 E1 mutants can do so (Fig. 3A). Flies were tested in groups, not individually; therefore, climbing ability reflects the motor function in the population. While there appears to be an increase in motor function in Uba1 B1 /Uba1 B2 males from 5 days to 10 days of age, because we tested each genotype as a population and did not track individuals, this increase may reflect that those flies surviving to 10 days were healthier overall than their siblings who were tested at 5 days but did not live until 10 days.
Motor function declines with age in flies just as it does in humans. Control FRT42D flies were tested in climbing assays every 5 days. Both male and flies showed a gradual decline in climbing ability over time (Fig. 3B). Interestingly, the motor impairment of the Uba1 B1 /Uba1 B1 and Uba1 B1 /Uba1 B2 mutants resembled that of control flies of extremely advanced age. 50 dayold female control flies and 55 day-old male control flies showed a motor performance similar to that of 5 and 10 day-old Uba1 B1 / Uba1 B1 and Uba1 B1 /Uba1 B2 flies.

E1 Mutants Exhibit Dramatically Reduced Survival
In addition to their reduced survival to adulthood, E1 mutants that reached adulthood were extremely short-lived. In parallel assays, wild-type w; FRT42D control male flies lived 47.2260.70 days on average, whereas Uba1 B1 /Uba1 B1 male mutants lived only an average of 10.2261.12 days (Fig. 3C-D). In a separate trial, wild-type w; FRT42D control female virgin flies lived 51.8861.20 days on average, and Uba1 B1 /Uba1 B1 female virgins lived only an  (Fig. 3E-F). In both males and females, the reduction in lifespan was extremely statistically significant (P,0.0001).

Inappropriate Ras Upregulation in the Brains of E1 Mutants
We showed previously that Uba1 B1 /Uba1 B1 and Uba1 B2 /Uba1 B2 mutant larvae exhibit an increase in Ras signaling through ERK [45]. In Drosophila, Ras activation results in expression of Ras target genes including the high-threshold target argos. Using an argos-lacZ reporter, we investigated if Ras signaling is altered in Uba1 B1 /Uba1 B1 adult brains. At 1 day, 5 days, and 10 days of age, there is no obvious argos expression in the brains of Uba1 B1 /+ flies (Fig. 4A, shown for a 10 day old brain). In contrast, brains dissected from 1 day-old, 5 day-old, and 10-day old Uba1 B1 / Uba1 B1 mutant flies show a number of cells with clear argos expression (Fig. 4B-D) indicating inappropriate Ras activation.
We attempted to examine these brains for caspase activation by staining with antibodies to the activated form of caspase 3 (anti-C3). We did not observe an obvious increase in anti-C3 staining in E1 homozygous mutant brains compared to age-matched control brains (data not shown). If cell death occurred gradually over time, it would be difficult to visualize by analysis of individual time points, and a massive wave of cell death could be missed by examining the wrong time points. Alternatively, it is possible that cell death occurred in a caspase-independent fashion or that the reduction in lifespan did not involve an increase in cell death in these brains.

The Reduced Survival of E1 Mutants is Sensitive to the Gene Dosage of Ras
Increased Ras activation can promote AD-like changes in neuronal cells in culture [56], and we previously found that the reduced survival to adulthood of E1 homozygous mutants was sensitive to the gene dosage of Ras [45]. We tested if mutation in Ras could rescue the mortality of Uba1 B1 /+ and Uba1 B1 /Uba1 B1 mutant adult flies. Reducing the gene dosage of Ras by introducing one copy of the mutant allele Ras e1b significantly increased the lifespan of flies carrying one or two mutant copies of E1 (Fig. 5).
Uba1 B1 /+; Ras e1b /+ male flies had an average lifespan of 46.82 days, a statistically significant improvement (P,0.0001) from 31.91 days of Uba1 B1 /+ male flies in a parallel assay (Fig. 5A-B). In a separate trial, Uba1 B1 /+; Ras e1b /+ virgin females had an average lifespan of 34.69 days, statistically significantly improved (P = 0.0003 by Log-Rank Mantel-Cox, and P = 0.0379 by Behan-Breslow-Wilson) from 29.16 days of Uba1 B1 /+ virgin females (Fig. 5C-D). Under parallel conditions in each of these experiments, control male flies lived an average of 48 days and control female flies an average of 33.04 days. Importantly, the rescued survival of Uba1 B1 /+; Ras e1b /+ flies was not statistically significantly different from wild-type controls for both males and females. Uba1 B1 /Uba1 B1 ; Ras e1b /+ males flies also showed a statistically significant (P,0.0001) increase in lifespan to 13.72 days from 10.22 days of Uba1 B1 /Uba1 B1 male flies (Fig. 5E-F); Uba1 B1 /Uba1 B1 ; Ras e1b /+ virgin females showed a statistically significant (P,0.0001) increase in lifespan to 11.92 days from 8.98 days of Uba1 B1 /Uba1 B1 virgin females (Fig. 5G-H). The suppression of mortality by mutation in Ras in flies mutant in one or both

Mutation in E1 as a Factor in Normal Age-related Decline and Age-related Neurodegenerative Diseases?
We have presented studies in Drosophila showing that loss-offunction mutations in only one copy of E1 have a dramatic effect on lifespan even in the absence of other mutations. In humans, E1 is encoded by the gene Ube1 on the X chromosome. Given the high conservation of genes in the Ubiquitin Pathway, this could mean that women carrying one mutant copy of E1 might be at risk for reduced lifespan. How does loss of only one copy of E1 cause such a change in lifespan? The Ubiquitin Pathway controls a number of crucial cellular activities including signal transduction, apoptosis, and proteasome-mediated protein degradation. Proteasome activity and assembly decline with increased age [57][58][59]. Therefore, it is possible that at a young age, the threshold of E1 is easily met by only one functional genomic copy, but that as age advances and the proteasome becomes more limiting, that one copy of E1 is no longer sufficient to allow for clearance of misfolded or aggregating proteins. Thus, one possible explanation is that increased protein aggregation in flies with only one functional copy of E1 could cause increased mortality.
Disease-associated mutations in specific genes have been identified in familial forms of a number of neurodegenerative diseases including HD, AD, and PD as reviewed earlier. In HD, the length of the expanded polyQ region in part determines the age of onset of the disease; longer repeats often result in onset of symptoms at an earlier age. Intriguingly, however, patients with the same polyQ length do not always exhibit the same time of onset and course of the disease [60][61][62]. Therefore, polyQ length alone cannot explain all differences in disease presentation. Environmental factors and genetic background likely also contribute to variations in disease progression [63][64]. It will be exciting to explore if human E1 variants could create sensitive genetic backgrounds with adverse effects on the course of disease progression in patients suffering from HD. Moreover, there are familial cases of other neurodegenerative diseases in which causal mutations have not been identified. In addition, for some diseases, there are sporadic cases with no family history. In fact, sporadic AD is far more prevalent than familial AD, and the causes of sporadic AD also remain unclear [65][66]. Thus, it is highly likely that there are a number of genes whose mutation or dysregulation serve as risk factors or even causes of sporadic AD cases. We speculate that human E1 variants may serve as risk factors for the age-related decline in AD and other diseases. In the future, it will be important to address how loss of E1 affects lifespan in Drosophila neurodegeration models including models of HD and AD.
The Ubiquitin Pathway also regulates a number of signaling pathways including (but not limited to) Ras signaling. Upstream RTKs are down-regulated by ubiquitination [67][68][69], as is Ras itself [45,70]. Therefore, another possibility is that upon aging, specific signaling pathways are dysregulated and contribute to reduced lifespan. In fact, examination of the brains of AD patients found evidence of increased Ras signaling [71][72][73][74][75]. Also, expressing activated Ras in neurons causes AD-type phenotypes in neurons in culture [56]. Importantly, we have shown here that reducing the gene dosage of Ras in flies carrying only one mutant copy of E1 restores lifespan to that of wild-type controls.
A Drosophila Model for XL-SMA?
There are a number of variants reported for human E1 including loss-of-function alleles. In humans, the E1 gene Ube1 is located on the X chromosome and has been lost from the Y chromosome [76], so a male inheriting a loss of function variant in E1 would have no wild-type copy. Some human E1 variants are associated with X-linked Infantile Spinal Muscular Atrophy (XL-SMA), a rare and severe form of Spinal Muscular Atrophy [46]. XL-SMA is a tragic condition in which males who inherit a mutant copy of E1 typically live less than two years and during which time they suffer terribly [46,[77][78]. Mothers who are carriers for a mutant copy of E1 often have a history of miscarriages presumably because many of their affected male children do not make it to term. XL-SMA has a similar presentation to the severe Type 1 SMA caused by mutation in the SMN1 gene, but also presents with congenital contractures [46,[77][78].
As we report here, flies homozygous for null mutations in E1 do not survive, but flies homozygous for hypomorphic E1 mutations can survive to adulthood at a very reduced rate, and these flies show a number of patterning abnormalities and severe motor impairment. Their lifespan is dramatically reduced compared to heterozygous mutants and wild-type controls.
To our knowledge, there is currently no animal model in which to study XL-SMA. We showed here that Drosophila E1 homozygous mutants recapitulate some aspects of human XL-SMA such as motor impairment and reduced lifespan. Thus, these Drosophila mutants warrant further study to determine if they recapitulate other aspects of this disease, such as degeneration of motorneurons reminiscent of the loss of anterior horn cells in XL-SMA, to establish if they could serve as an animal model to increase our understanding of this devastating disease. We previously showed that reducing the gene dosage of Ras in homozygous E1 mutants increases their survival to adulthood [45], and in this investigation we reported that it also extends their adult lifespan. If Ras signaling contributes to XL-SMA pathology in humans as it does to reduced lifespan in Drosophila E1 mutants, targeting Ras may serve as a potential therapeutic strategy for XL-SMA.

Drosophila Genotypes
Adult and larval images were from the following genotypes: w; FRT42D ( W; FRT42D Uba1 B1 /FRT42D Uba1 B1 ; Ras e1b /+ ( Figure 5E, 5G black-filled in squares) * elavgal4 was present in these experimental controls as an additional control for parallel experiments using gal4/UASmediated transgene expression not included in this study. In multiple parallel experiments, elavgal4 did not affect lifespan of these genotypes (data not shown).

Genetic Crosses
Uba1 B1 heterozygous flies with reduced gene dosage of Ras were generated by crossing w; FRT42D Uba1 B1 /SM6-TM6B to flies of the genotype w; Ras e1b /TM6B. w; FRT42D Uba1 B1 /+; Ras e1b /+ flies were identified by the absence of the dominant visible markers Cy, Hu, and Tb found on the SM6-TM6B fused balancer, and the markers Hu, and Tb found on the TM6B chromosome. To generate homozygous Uba1 B1 flies with a reduced gene dosage of Ras, we crossed flies of the genetype w; FRT42D Uba1 B1 /SM6-TM6B to flies of the genotype w; FRT42D Uba1 B1 ; Ras e1b /SM6-TM6B. w; FRT42D Uba1 B1 /FRT42D Uba1 B1 ; Ras e1b /+ flies were identified by the absence of the dominant visible markers Cy, Hu, and Tb found on the SM6-TM6B fused balancer chromosome.
Flies of the genotype w; FRT42D Uba1 B1 /FRT42D were generated by crossing w; FRT42D Uba1 B1 /SM6-TM6B to flies of the genotype w; FRT42D. w; FRT42D Uba1 B1 /FRT42D flies were identified by the absence of the dominant visible markers Cy, Hu, and Tb found on the SM6-TM6B fused balancer chromosome.

Lifespan Assays
Flies of each genotype were collected within 24 hours of eclosion and placed in fresh vials and incubated at 25uC. Surviving flies were counted daily and transferred to fresh food every several days to prevent dessication of the food or growth of mold or bacteria. Data from flies collected on different days was pooled for each genotype.

Climbing Assays
Climbing assays were performed similar to those described previously [55][56]. Age-matched flies of the indicated genotypes were placed into empty vials in small groups. When flies are tapped to the bottom of a vial, they immediately climb back to the top of the vial. To address motor function, flies were tapped to the bottom of the vial, and we counted the number of flies capable of climbing 4 centimeters in 5 seconds. Climbing assays were repeated five times for each group of flies at each time point. Due to the reduced survival of Uba1 B1 /Uba1 B1 and Uba1 B1 / Uba1 B2 flies, flies of these genotypes were collected each day; each small group was tested at age 5 days and 10 days, and the data pooled from the smaller groups.

Statistical Analysis
Analysis of the climbing assays was performed using Microsoft excel spreadsheets. Lifespan averages, standard errors, and medians were calculated using Microsoft excel and Graphpad prism. Averages detailed in the text are lifespan 6 s.e.m. Error bars in the graphs in Fig. 3A-B indicate standard deviation. Standard deviations for motor assays in Fig. 3A-B were calculated based on the deviation from five replicate tests of each population. Kaplan Meier survival analysis/comparison of overall survival curves using both Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests was performed using Graphpad Prism statistical software. P values are indicated in the text for each method, except in cases for which both gave P,0.0001 in which case only one P value is indicated.