Introduction

In dementia disorders that exhibit pathological tau protein, termed tauopathies, the microtubule associated protein tau becomes hyperphosphorylated, misfolded, and aggregates to form neurofibrillary and glial tangles [1, 2]. Tauopathy disorders include Down syndrome, corticobasal degeneration, progressive supranuclear palsy, Pick’s disease, Guam amyotrophic lateral sclerosis/Parkinson’s dementia complex, frontotemporal dementia with parkinsonism chromosome 17 type (FTDP-17) and Alzheimer’s disease[1, 2]. Autosomal dominant mutations in the MAPT gene encoding tau cause FTDP-17, indicating that abnormal tau accumulation can drive neuronal dysfunction and neurodegeneration independent of pathologies that may co-occur in other tauopathy disorders, and suggesting that abnormal tau may possess intrinsic toxicity [3–7]. The mechanistic underpinnings by which abnormal tau leads to neuronal dysfunction remain unclear. One hypothesis is that tau undergoes a gain of toxic function whereby the misfolding and oligomerization of tau creates toxic species [8–10]. A second hypothesis posits that as tau becomes hyperphosphorylated or misfolded, the normal microtubule binding activity is inhibited and results in loss of microtubule regulatory function [11, 12]. While mechanistically distinct, these hypotheses are not mutually exclusive, and both may contribute to a complex environment driving neuron dysfunction and decline.

Microtubules (MTs) are polymeric tubes composed of α- and β-tubulin subunit dimers that provide structural support to maintain the elongated structure of neurons as well as providing the tracks on which motor proteins move cargo between the cell body and the ends of axons and dendrites. MTs also undergo numerous post translational modifications and interactions with microtubule associated proteins (MAPs) to maintain balance between tightly controlled dynamic and stable phases [11, 13]. Multiple lines of evidence implicate MT dysfunction, manifesting as dendritic arbor simplification and disruption of axonal transport, as a component in neurodegenerative diseases [11, 13]. Similar to mammals, the C. elegans tubulin super-family of genes consist of multiple α and β isoforms. C. elegans have 9 α-tubulin genes and 6 β-tubulin genes expressed in tissue specific combinatorial ratios, providing precise regulatory control of a cell’s tubulin pool and resulting microtubule properties [14]. This differential expression is particularly notable in specialized cells like neurons which are highly polarized in their structure and function [15].

Tau plays a critical role in the regulation of the neuronal microtubule cytoskeleton. Tau binds directly to the microtubule lattice, conferring stability and at the plus tip, where it mediates the addition of new tubulin dimers particularly during microtubule rescue [16–19]. Tau’s affinity for MTs is regulated in part by the expression of 6 alternatively spliced isoforms varying in the number of N-terminal and microtubule binding domain repeats [20, 21]. Additionally, tau contains numerous serine, threonine and tyrosine residues which can be reversibly phosphorylated. The addition of phospho-groups at these sites disrupts tau-MT electrostatic interactions causing Tau to dissociate from the microtubule surface [12, 22–25]. Treatment of mice with Epothilone D (EpoD), a MT stabilizing drug, have prevented the development of tau pathology in young mice, cleared existing tau pathology in aged mice and improved cognitive performance [26, 27]. Additionally, EpoD treatment increased MT stability, density, increased fast axonal transport and decreased neuronal dystrophy in PS19 mice [26]. Recently another microtubule stabilizing compound, CNDR51657 was shown to reduce AT8 staining that detects disease-relevant tau phosphorylated at ser202 and ser205 and accumulation of TAR DNA-binding protein 43 (TDP-43) to ameliorate microtubule disruption in the optic tracts of repeated traumatic brain injury (rTBI) treated mice. Additionally, CNDR51657 improved memory and cognitive function in rTBI-treated mice [28]. Taken together, this evidence suggests increasing microtubule stability shows promise as a therapeutic strategy. However, these studies have not directly addressed the molecular underpinnings driving the neuronal protection seen with these compounds. Caenorhabditis elegans offers a number of key advantages as an experimental model for human disease namely, genetic homology to humans, rapid generation time, short lifespans, and tractable genetics. Expression of human tau in C. elegans neurons results in toxic phenotypes similar to those seen in human tauopathies including behavioral abnormalities, axonal degeneration, neuron loss, premature death and accumulation of phosphorylated and aggregated tau protein, which worsen with age [29]. Previously, we have employed these models in forward genetic screens to uncover genetic suppressors of tau pathology [30–36]. Using similar methodology in this study has revealed genetic mutations in 3 of the 9 C. elegans α-tubulin genes, tba-1, tba-2, and mec-12, that suppress pathological tau toxicity in our human-tau transgenic C. elegans models. We used C. elegans models to characterize mutant tubulin rescue of tauopathy phenotypes and probe the underlying molecular mechanisms. Tau phenotype suppressing mutations do not alter overall accumulation of pathological tau or alter overall tubulin expression. Rather, our results suggest mutations act by altering the properties of microtubules. In vitro experiments using semi-isolated and isolated tubulin protein indicate changes to microtubule properties without changing tau-tubulin binding affinity. The studies presented here demonstrate that a single amino acid change in tubulin can have profound effects on the pathological progression of tauopathy related neurodegeneration.

Results

Tau-phenotype suppressing mutations cluster in the tau binding region of α-tubulin

We previously generated a C. elegans model of tauopathy where a genomically integrated multicopy transgene array encoding human tau isoform with all 4 possible microtubule binding repeats, and 1 of 2 possible N-terminal repeats (4R1N) under the pan-neuronal aex-3 promotor was integrated into the C. elegans genome [29]. In this model, human 4R1N tau protein accumulates in a highly phosphorylated state within neurons leading to tau aggregation, neuronal dysfunction manifesting as behavioral deficits, neurodegeneration, and decreased lifespan. To dissect molecular genetic pathways contributing to tauopathy phenotypes, we have employed a classical genetic mutagenesis and screening approach in a tau transgenic C. elegans searching for mutants resistant to tau mediated neurodegenerative phenotypes. Using this approach, we have previously isolated strong suppressor of tauopathy (sut) mutations in the dib-1, sut-1, sut-2, sut-6, and spop-1 genes [30–36].

Using the same screening paradigm, we uncovered several distinct genetic mutations in α-tubulin genes that rescue tau-induced motility deficits in C. elegans models of tauopathy. Eleven tau suppressing missense mutations have been identified in multiple tubulin genes. The isolated mutations cluster in the same region across three α-tubulin genes (tba-1, tba-2, mec-12) and are detailed in Supplementary Table 1. To date we have measured suppressive capability of several mutant strains: mec-12 (D431N), tba-1 (E437K), tba-2 (D429A), tba-2 (D429N), and tba-2 (V433D). We selected mec-12 (D431N), tba-1 (E437K), and tba-2 (D429N) for more extensive exploration as they encompassed the range of suppression observed, and further focused on the strongest suppressor: tba-2 (D429N). Mutations in mec-12, tba-1, and tba-2 occur at the C-terminal encoding region corresponding to helix 12 in each of these α-tubulin encoding genes (Figure 1). An alignment of the protein sequences (Figure 1A) revealed near identical homology in primary sequences across the various α-tubulin genes in C. elegans as well as near human homologues, indicating that the mutations occur in a highly conserved region of the proteins. Interestingly, helix 12 is exposed on the outside surface of MTs (depicted in Figure 1B and 1C) and is predicted to play a prominent role facilitating tau-MT binding [16], leading us to hypothesize that the mutations may confer suppression of tau-induced phenotypes through a mechanism of altered tau-tubulin and/or tau-MT binding affinity.

Figure 1.

C. elegans α-tubulin mutations rescue tau-mediated neuronal dysfunction

To investigate the impact of mutations in α-tubulin genes on tau induced neuronal dysfunctions, we measured C. elegans behavioral deficits caused by transgenic tau expression in a swimming motility assay in which C. elegans are placed in buffer and motility is measured (body bends over time) as a readout of neuronal dysfunction. Human tau transgenic worms that express human tau pan-neuronally exhibit an uncoordinated locomotion phenotype driven by neuronal dysfunction that worsens with age, whereas tau-transgenic animals containing a tubulin mutation exhibit rescue of this neuronal dysfunction. We tested representatives from each mutant tubulin gene and observed that two mutant alleles of tba-2 conferred the strongest levels of tau-induced behavioral deficit suppression, rescuing tau transgenic animals to wild type levels of motility. We observed that two of the mutations found in tba-2 (D429N, and D429A) conferred the strongest levels of tau-induced motility deficit suppression, rescuing to wildtype or near wildtype levels (94% and 100% of wildtype levels respectively) of motility in a C. elegans model of tauopathy expressing high levels of human tau (WT-TauH, strain CK144) (Figure 2 A-B). Interestingly, mutant mec-12 demonstrated the lowest level of suppression observed for the tubulin genes tested, conferring suppression to approximately 37% of wildtype levels (Figure 2C). We observed similar patterns of α-tubulin mediated suppression in a transgenic tauopathy model expressing more moderate levels of human tau (strain CK1443, WT-TauM, Supplementary Figure 1), with tba-2 (V433D) mutations suppressing motility deficits in WT-TauM expressing worms to approximately 82% of wildtype levels. Additionally, we tested whether mutant tba-2 could suppress the toxic effects conferred by the FTLD-tau (V337M) mutation (strain CK10, Figure 2D) [29]. Here, the same tba-2 mutant provided only partial (∼87% of wildtype levels) rescue of deficits in the mutant tau-transgenic background (Figure 2D). The suppressor mutant found in tba-1 also confers partial rescue (∼79% of wildtype levels) of neuronal dysfunction as measured in the motility assay (Supplementary Figure 2). The failure of an α-tubulin truncation (Q230Stop) mutant to suppress tau toxicity suggests that this suppression is specific to the amino acid substitutions within helix 12 (Supplementary Figure 3). Notably, when testing the level of suppression seen in C. elegans heterozygous for the tba-2 (D429N)/+ mutation, we observed heterozygous tba-2 D429N mutation rescues tauopathy phenotypes to approximately 65% of wildtype levels, demonstrating that mutant tba-2 appears to be a strong semi-dominant suppressor of tauopathy neuronal dysfunction (Supplementary Figure 4).

Figure 2

Abundance of mutant α-tubulin drives the strength of tauopathy suppression

The well supported multi-tubulin hypothesis asserts that α- and β-tubulin isotypes are functionally distinct and differential spatial-temporal expression impart specific microtubule properties across different tissue types [37–40]. According to the C. elegans Neuronal Gene Expression Network (CeNGEN) dataset [41], tba-2 is the more abundantly expressed α-tubulin isotype in neurons as compared with tba-1 and mec-12, with mec-12 being the lowest expressed of these three genes neuronally. Simplistically, the degree of tubulin expression appears to align with the strength of tauopathy behavioral suppression as observed in the swimming assays. Thus, we hypothesized that the relative strength of tau-neurotoxicity suppression may be imparted by the relative levels of mutant α-tubulin isotype expression. To directly test this hypothesis, we generated integrated multicopy D429N mutant tba-2 transgenic strains of C. elegans in which mutant tba-2 transgenes are restricted to pan-neuronal expression. We selected mutant tba-2 transgenic lines (tba-2-D429N Tg-A and tba-2-D429N Tg-B) that exhibited different levels of tau suppression when crossed into a tau-transgenic line expressing high levels of human tau (WT Tau-H, CK144) as assessed by motility deficits observed in the swimming assay described previously. We observed that the tba-2-D429N Tg-B animals exhibited significantly greater levels of tau induced motility deficit suppression (Figure 3A). Transgene expression analysis by qRT-PCR revealed transgenic expression of mutant tba-2 was five-fold greater in tba-2-D429N Tg-B animals demonstrating that greater levels of mutant tubulin gene expression indeed correlated with a greater degree of motility deficit suppression (Figure 3B). As a control, we measured total, both mutant and endogenous, tba-2 mRNA throughout the worm, which was not different between the two strains (Supplementary Figure 5).

Figure 3.

tba-2 confers strong suppression of tau-Aβ and tau-TDP-43 co-pathology models

Tau neurofibrillary tangles and (Amyloid-β) Aβ plaques define Alzheimer’s disease in humans and act synergistically to enhance pathological phenotypes in animal models including tau/amyloid bigenic C. elegans [42]. Recent neuropathological data from human tissue has revealed that a number of AD cases additionally harbor aggregates of TDP-43 [43]. AD cases with TDP-43 co-pathology exhibit heterogeneous arrays of protein deposits and differ in the severity of toxic phenotypes. Tau and TDP-43 bigenic expression has also been shown to drive synergistic neuronal dysfunction and neurodegeneration in C. elegans [43]. Here we used animal co-pathology models to test the effects of tba-2 suppression when tau phenotypes are enhanced by either Aβ peptide or TDP-43 protein expression. In both co-pathology models, mutant tba-2 only strongly suppresses motility deficits to levels nearing the relative contribution of tau toxicity to the overall phenotype, ∼75% of wildtype motility in Aβ co-pathology model and ∼71% of wildtype motility (Fig. 3C and 3D).

Mutant tubulin eliminates tau-mediated neurodegeneration

Transgenic C. elegans neuronally expressing high levels of human tau exhibit neurodegeneration that worsens with age [29]. We investigated whether tba-2 and mec-12 suppressed tau-induced neuronal loss by measuring the loss of GFP-marked GABAergic neurons in the ventral nerve cord [29, 44]. C. elegans expressing the human WT-Tau transgene along with mutant α-tubulin were compared to wildtype, tba-2, and tau-transgenic controls. Wildtype C. elegans develop the same number of neurons in each animal, facilitating the quantification of GFP marked neurons lost to neurodegeneration in mutant and transgenic strains. We observed the expected significant mean loss of ∼1.5 neurons by day 1 adulthood in WT-Tau worms (Figure 4A, 4B), whereas both tba-2;WT-Tau (Figure 4A, 4C) and mec-12;WT-Tau (Figure 4B) animals exhibited no significant neuron loss as compared to wildtype animals. These data suggest that both mutant tba-2 and mec-12 suppress tau-induced neuron loss and the differences observed in motility deficits result from differences in rescue of neuronal dysfunction rather than rescue of neurodegeneration.

Figure 4.

Mutant tba-2 reduces levels of tau protein

To understand the molecular mechanisms underlying mutant-tubulin mediated suppression of tau toxicity, we assessed alterations to tau protein in mutant tau-transgenic strains versus tau-transgenic strains without tubulin mutations. We immunoblotted equivalent amounts of staged day-1 adult age matched tau-transgenic C. elegans using a phosphorylation and pan-isoform antibody to detect total tau levels across strains. We observed that tba-2 moderately decreased overall tau levels by ∼45% while mec-12 showed no significant effect on tau protein (Figure 5 A-B).

Figure 5.

Tau phosphorylation accumulates in neurons of AD affected brains as well as the neurons of C. elegans models of tauopathy. We, therefore, assessed for mutant-tubulin mediated alterations in tau phosphorylation. We immunoblotted staged day-1 adult C. elegans with antibodies specific to p-ser396/p-ser404 (PHF-1), p-ser202 (CP13), and p-thr231 (AT180) (Figure 5A, C-E, antibodies described in Supplementary Table 3). We observed that neither tba-2 or mec-12 altered levels of tau phosphorylation at these examined tau phospho-sites.

Tubulin mutation does not influence tau aggregation

Progressive tau aggregation is a key pathological event in the progression of AD and related tauopathies and is recapitulated in C. elegans models. We hypothesized that mutant tubulin may confer suppression by preventing the aggregation of tau protein. We tested this hypothesis by extracting aggregated tau from whole C. elegans protein lysates using sequentially applied buffers of increasing solubilization strength to separate soluble from detergent insoluble tau species. The final solubilization buffer, formic acid (FA) fraction contains the most insoluble tau aggregates. Upon immunoblot analysis of the aggregated tau fractions, we detected no alteration of tau aggregation in tau transgenic C. elegans (WT-Tau, CK144) animals compared to tba-2 mutants (Figure 6).

Figure 6:

Mutant tba-2 decreases sedimented microtubule mass

Given that the tauopathy suppressing mutations occur in the region of the tubulin protein sequence that when folded and assembled into polymer, exists on the outer surface of the microtubule and within a predicted MAP interaction site, we hypothesized that the mutations may alter tubulin properties or function. To examine this possibility, we employed a microtubule sedimentation assay (based on [45], experimental schematic shown in Supplementary Figure 6A) in which C. elegans cellular extracts were used as the base material to build taxol stabilized microtubules. Taxol stabilized microtubules were isolated by ultracentrifugation allowing us to compare pelleted microtubule mass. (Figure 7). Signal from pelleted MT mass was normalized to corresponding input signal and mean post-incubation mass was calculated. We observed that tba-2/WT-TauH C. elegans extracts yielded less taxol-stable microtubule mass as compared to WT-TauH extracts C. elegans extracts (Figure 7A, 7B). Here we used nocodazole, a drug that prevents MT growth, as a negative, no microtubule, control (Figure 7A). Capillary western data from all 4 biological replicates is presented in Supplementary Figure 6B.

Figure 7.

Mutant tubulin does not alter the binding affinity between tau and soluble tubulin

The suppressor mutations map to helix 12 of α-tubulin on the outer lattice of the microtubule, which is predicted to be involved in tau-MT interaction. As such, we hypothesized that the mutations may confer suppression by altering the binding affinity between tau and tubulin. We tested this hypothesis using biolayer interferometry (BLI) to determine binding affinities (K_d) between 6xHis-Tau(4R1N) (His-Tau) and wildtype and tba-2 (D429N) tubulin association and dissociation measurements. Here, purified recombinant His-Tau was immobilized to NTA BLI sensors and introduced to solutions containing soluble tubulin isolated from wildtype and tba-2 (D429N) C. elegans (tubulin association). After incubation in tubulin containing solutions, sensors were returned to buffer alone (BRB80), for tubulin dissociation. Changes to optical interference as detected through the sensor tips were used to calculate equilibrium association and dissociation constants. No significant difference in binding affinity was observed between the soluble tubulin isolated from wildtype animals as compared to tba-2 (D429N) mutant C. elegans (Figure 8).

Figure 8.

Discussion

Within the work presented here, we have shown mutations clustering in the tau-binding region of α-tubulin genes suppress tau-induced phenotypes in multiple C. elegans models of tauopathy, indicating the importance of this region in modulating microtubule function and protection from neurodegenerative assaults. The strongest suppressors identified completely rescue wildtype tau-induced phenotypes and strongly rescue mutant tau phenotypes and the phenotypes seen with tau/Aβ and Tau/TDP-43 co-pathology induced neuronal dysfunction. We have also identified similar mutations in other C. elegans α-tubulin genes that differentially suppress tauopathy to a level related to the relative level of tubulin expression from that gene; in other words, greater mutant tubulin expression confers a greater level of tauopathy suppression. Thus, titration of mutant tubulin levels may allow a fine-tuning of neuron specific tubulin expression to rescue tauopathy using a gene-based therapy approach.

All three mutant tubulin genes: the abundant mutant tba-2, the moderately abundant tba-1 and the less abundant mec-12 suppress the neurodegeneration seen in tau transgenic C. elegans. Both mutant tba-2 and mec-12 completely rescued neuron loss in tau transgenic animals to that of non-transgenic animals with minimal loss of GABA-ergic ventral nerve cord neurons. While these two mutations varied in their ability to suppress neuronal dysfunction, they are equally able to fully suppress neurodegeneration. Additionally, these data point towards neuron dysfunction as a major contributor to tau-induced motility phenotypes as compared to gross loss of neurons.

Mutant tba-2 but not mec-12 moderately reduces levels of total tau protein. The moderate reduction in total tau levels by tba-2 and the absence of reduction with mec-12 suggest that the mechanism underpinning the disproportionately remarkable rescue in neuronal dysfunction acts independently of tau levels. While we only see reduction in tau levels with the strongest suppressor, the strength of suppression seems outsized compared to the level of tau protein reduction and points to a mechanism independent of total tau levels. Additionally, we observed no alteration to phosphorylation or aggregation of tau with mutant tba-2, suggesting these mutants to confer suppression through a mechanism independent of pathological tau accumulation.

The amino acid substitutions that confer tauopathy suppression occur in or immediately adjacent to helix 12 of α-tubulin. Helix 12 is positioned on the outer surface of the microtubule once tubulin subunits are assembled into the polymer. Since helix 12 is important for MAP interactions we hypothesized that the mutations might affect microtubule properties in assembled microtubules. We tested this by forming MTs from endogenous C. elegans tubulin with the stabilizing drug, taxol, and measured the relative microtubule mass pelleted upon ultracentrifugation. Interestingly, we found that less microtubule mass pelleted from lysates that contained mutant tubulin, indicating that the microtubules containing mutant tubulin are either taxol resistant or have different assembly properties in the presence of taxol. Microtubule stabilizing compounds have previously demonstrated rescue of tau pathology in mammalian models improving axon density, cognitive function and clearance of tau pathology [26, 27]. Taken together, these data support the notion that microtubules containing mutant α-tubulin possess different properties as compared to those made solely of wild type tubulin and altering microtubule properties may contribute to increased functionality in the presence of pathological tau.

The cluster of tau suppressing mutations occur in helix 12, a region playing a prominent role in the charge-charge interactions facilitating tau-MT binding. To test the hypothesis that mutations in this region may alter tau-MT binding we used recombinant human tau immobilized to the biolayer interferometry (BLI) sensor. Using solutions containing tubulin isolated from wildtype tubulin or tba-2(D429) C. elegans we could determine association and dissociation parameters for each. The binding affinities between tau and soluble wild type tubulin compared to tau and mutant tubulin were not observably different. However, this does not rule out a difference in tau binding to polymerized microtubules containing mutant tubulin. Additional testing will be needed in order to explore whether mutations affect tau or other MAP binding with polymerized MTs.

Conclusions

While more experimentation is needed to understand the specific molecular details of tauopathy suppression, it is already evident that the mutations described in this study provide protection from the adverse effects of tau-mediated neuronal dysfunction and neurodegeneration. As the mutations do not reduce the level of phospho-tau or tau aggregates, it seems the α-tubulin mutations do not act by directly reducing pathological tau species but rather protect neurons via a tubulin/microtubule function mediated mechanism. Given recent advances in gene therapy technologies for neurodegenerative and neurodevelopmental disorders [46, 47], the possibility of harnessing mutant helix 12 α-tubulin to suppress tauopathy phenotypes appears attractive. Therefore, exploration of the mechanism of action of these mutations will contribute to a more complete understanding of the conceptual basis for developing effective disease modifying therapeutics for tauopathy.

Materials and Methods

C elegans strains and maintenance

For a list of strains used, refer to Supplementary Table 2. Worms were maintained at 20°C (with the exception of Aβ_1-42 strains and controls, which were maintained at 15°C prior to temperature upshift for Aβ_1-42 peptide expression for 24 hours at 25°C prior to experimentation, on NGM plates seeded with OP50 E. coli as described in [48]. The C. elegans tubulin alleles described in Supplementary Table 1 were generated through chemical mutagenesis using ethyl nitroso urea as the mutagen using established methods previously described in detail [34, 35, 49].

For mRNA and protein extraction, worms were grown on nutrient rich media plates containing 5 times peptone (5xPEP) seeded with OP50 [48]. For microtubule co-sedimentation experiments, worms were grown to dense populations on 5xPEP plates and subsequently washed onto egg plates which were prepared by seeding OP50 mixed with separated chicken egg yolk. Animals were subsequently harvested from 5xPEP plates using sequential washes with M9, or by floating on sucrose cushion if harvested from egg plates [32, 50, 51]. Pellets were snap frozen in liquid nitrogen and stored at −80°C until use.

C. elegans motility assays

C. elegans were synchronized and grown to day 1 of adulthood on NGM growth plates. Plates containing the worms were placed in the assay room and allowed to acclimate for 20 minutes, prior to assay initiation. One strain at a time, worms were washed from NGM plates to food free video plates with 2mL of M9 and allowed to acclimate to the buffer for 10 seconds before a 1 minute video was recorded. Video data was collected using the WormLab platform (MBF BioScience, VT). Worm locomotor behavior was tracked and analyzed using the WormTracker software (MBF BioScience, versions 2019, 2020 and 2022). Body bends from the mid-point body location of each worm were counted. The total number of body bends was divided by the track length(s) to give the frequency of body bends per second, this was then converted to body bends per minute. Mean body bends per minute was compared across strains using the Kruskal-Wallis test with post-hoc Dunn’s test.

Manual thrashing assay

A manual thrashing assay [50] was performed to assess the efficacy of suppression of a single copy of the tba-2(D429N) mutation. tba-2 (D429N)/+ heterozygous strains were generated by mating the homozygous tba-2; WT-Tau strain with WT-Tau males to make obligate heterozygotes for tba-2(D429N) in the F1 generation. F1s were subjected to a manual thrashing assay where worms were placed in a 10 µL drop of M9 buffer on a microscope slide. After worms were allowed to acclimate for 10 seconds, the number of thrashes were counted for 1 minute. One thrash is defined as a bend to the opposite direction across the midline and a return to starting position or two complete bends to the same side where the worm straightens to the midline in between. To confirm heterozygosity worms were removed from the buffer and placed on a fresh NGM plate and a clonal population was allowed to grow from the single worm. The populations were then genotyped for tba-2(D429N) heterozygosity.

Quantification of mRNA levels of tubulin sut genes in C. elegans

C. elegans were grown from eggs to day 1 of adulthood at 20°C on 5xPEP plates. Worms were washed off plates using M9 buffer and pelleted by centrifugation to a final pellet size of ∼100mg. Worm pellets were snap frozen in liquid nitrogen and stored at −80°C. RNA was isolated from ∼100mg pellets using TRIzol Reagent (Life Technologies) according to the manufacturer’s instructions. RNA was quantified using a NanoPhotometer NP80 (Implen GmbH, Munich, Germany) and integrity was assessed by 1% Tris/Borate/EDTA (TBE) agarose gel electrophoresis. RNA samples were treated with DNA-free DNAse treatment and removal kit (Invitrogen) following the manufacturer’s instructions. cDNA was made using the iScript Reverse Transcription Supermix (Bio-Rad) following manufacturer’s instructions and qPCR was performed iTaq Universal SYBR Green Supermix (Bio-Rad) in a CFX Connect Real-Time PCR System (Bio-Rad). Data was analyzed using CFX Manager 3.0 (Bio-Rad). Experimental samples were run in duplicate and three independent biological replicates were performed. Total tba-2 and transgenic tba-2 were normalized to rpl-32 internal control gene. Means were compared using a Student’s T Test. Primers used were as follows: Transgenic-tba-2 forward: GAC TCC AAC GAG GGA GGA G, Transgenic-tba-2 Reverse: CGA GAT GGC GAT CTG ATG, Total-tba-2 Forward: TCT CGT CTC GAC TAC AAG TTC G, Total-tba-2 Reverse: CTT CTC GAG AGC AGC CAA GT, rpl-32 Forward: GGTCGTCAAGAAGAAGCTCACCAA [50], rpl-32 Reverse: TCTGCGGACACGGTTATCAATTCC [50].

GABAergic Neuron Loss Assay

GABAergic neuron loss was assessed as a measure of gross neurodegeneration [32, 44, 50]. Tubulin mutant;Tau-transgenic worms were crossed into CZ1200 (juIs76 [unc-25p::GFP + lin-15(+)]) to generate tba-2;WT-Tau;unc-25p::GFP, WT-Tau;unc-25p::GFP, tba-2;unc-25p::GFP, mec-12;WT-Tau;unc-25p::GFP and mec-12;WT-tau;unc-25p::GFP. Worms were synchronized to day 1 of adulthood (approximately 4 days post-hatch) and mounted onto 2% agarose pads and immobilized with sodium azide. The number of visible GABAergic neurons along the ventral nerve cord were counted using a DeltaVision microscope (Applied Precision) at 60x magnification. A minimum of 10 animals were counted per strain per replicate, and a minimum of 3 biological replicates were completed. Statistics were calculated using the Kruskal-Wallis test with a post-hoc Dunn’s comparison, ***p<0.001, error bars reflect SEM.

Protein extraction for total tau levels

Protein extraction and immunoblotting procedures were conducted as previously described in [29, 32]. Worms were grown from eggs to staged day 1 adult populations on 5xPEP plates. Worms were harvested by washing with M9 buffer, and pelleted by centrifugation (3,000xg, 45 sec). Pellets were resuspended and washed three times with 8mL of M9 buffer and aliquoted to approximately ∼100mg final pellet size into pre-weighed Eppendorf tubes. Buffer was aspirated from centrifuged worms and pellets snap frozen in liquid nitrogen prior to storage at −80°C. Whole worm protein lysates were generated as follows. Worm pellets were weighed to determine pellet mass and thawed on ice. SDS proteins ample buffer (0.046 M Tris, 0.005 M ethylenediamine tetraacetate, 0.2 M dithiothreitol, 50% sucrose, 5% sodium dodecyl sulfate, 0.05% bromophenol blue, 1× concentration) was added to the pellets at a volume (µL) four times the pellet weight (mg). Pellets were sonicated three times for 20 seconds each at 70% amplitude, and cooled ice between sonication sessions. Samples were boiled for 10 min at 95°C and centrifuged at 13,000xg for 2 minutes prior to gel loading. Four biological replicates were performed.

Sequential protein extraction

Insoluble tau aggregates were isolated from soluble tau by the collection of sequential fractions after worm lysate treatment with buffers of increasing protein solubilization strength, as previously described in [29]. Briefly, staged day 1 adult worm pellets were resuspended in 2µL per milligram of wet pellet weight of RAB high salt buffer (0.1 M MES, 1 mm EGTA, 0.5 mm MgSO4, 0.75 M NaCl, 0.02 M NaF, pH 7.0, plus Complete Protease Inhibitor Cocktail (Roche Diagnostics) and 0.5 mm PMSF). Samples were sonicated three times at 70% amplitude for 8 seconds. An appropriate amount of worm lysate was reserved as the total fraction used to determine the relative level of total tau (both soluble and insoluble), while the remaining RAB sample (minimum volume of 100 µL) was subjected to centrifugation at 40,000xg for 40 min. The supernatants (RAB soluble tau) were collected and pellets resuspended in RIPA buffer (150 mm NaCl/1% Nonidet P-40/0.5% deoxycholate/0.1% SDS/50 mm Tris, pH 8.0, plus Complete Protease Inhibitor Cocktail (Roche Diagnostics) and 0.5 mm PMSF), using a volume of RIPA buffer equal to half that of the RAB sample starting volume. Samples were sonicated 1-2 times at 70% amplitude for 8 seconds and subsequently centrifuged at 40,000xg for 20 min. The supernatants were collected as the RIPA-soluble fractions (detergent-extractable insoluble tau). The remaining pellets were resuspended in a volume of 70% formic acid (FA, in water), equal to the volume of the starting RAB sample. FA resuspended pellets were sonicated 1-2 times at 70% amplitude for 8 s and centrifuged under vacuum in a SpeedVac with no additional heat until FA had evaporated. The remaining pellets (FA insoluble tau) were resuspended in a volume of 2x SDS protein sample buffer (0.046 M Tris, 0.005 M ethylenediamine tetraacetate, 0.2 M dithiothreitol, 50% sucrose, 5% sodium dodecyl sulfate, 0.05% bromophenol blue, 2x concentration). FA samples were then sonicated 1-2 times at 70% amplitude for 8 s. Samples were then boiled for 5 minutes (95°C) and centrifuged at 13,000xg at 4°C for 5 minutes. RAB samples were boiled for 5 min (95°C), chilled on ice for 5 min then centrifuged at 13,000xg at 4°C for 15 minutes and supernatants were collected as RAB-soluble fractions. SDS protein sample buffer was added to a final concentration of 1x to total, RAB-soluble and RIPA-soluble fractions, samples were boiled for 5 min at 95°C, chilled on ice for 5 min, then centrifuged at 13,000xg for 5 min. All samples were stored at −20°C prior to SDS-PAGE electrophoresis. Four biological replicates were performed.

Immunoblotting

Lysed total worm protein or sequentially extracted samples were subjected to SDS-PAGE using precast Tris-HCl gels (catalog number 3450028, BioRad, Hercules, CA), according to manufacturer’s instructions. Gels were run at 200V for approximately 60 minutes and proteins were cold transferred to PVDF membranes at 80V for 30 minutes. Membranes were blocked with 5% milk in PBST (1x PBS with 0.1% Tween-20) for 2 hours and incubated in primary antibody solutions diluted in 5% milk PBST block overnight at 4°C. Membranes were washed with PBST three times for 10 minutes each wash. Membranes were then incubated in secondary antibody solution for 2 hours at room temperature (see Supplementary Table 3 for a list of antibodies and dilutions used), after which they were washed three times in PBST as above. Detection was completed using the Clarity Western ECL substrate kit (Biorad; cat. number 1705060) and LiCOR Odyssey and Image Studio software. Blots were quantitated using ImageJ [52].

Large Scale growth of C. elegans using chicken egg yolk growth media

Worms were grown at large scale on 5X Peptone (PEP) plates seeded with an egg yolk-OP50 mixture based on methods previously described [51, 53]. Briefly, approximately 1 egg per 5X PEP plate to be seeded was sterilized in a 95% ethanol bath for approximately 5 minutes. The eggs were removed and air dried on a sterilized bench for several minutes. Egg yolks were separated using standard culinary methods maintaining as much sterility as possible (cracked on the side of an autoclaved beaker, and a sterile egg culinary egg separator was used). Separated egg yolks were placed in an autoclaved beaker. Using a sterile kitchen fork, as much non-yolk material and debris was removed as possible. Alternatively, equivalent volume of store-bought pasteurized separated egg yolks were used. The remaining yolk was mixed with 1.5-2mL stationary phase OP50 bacteria grown overnight at 37°C in standard LB. The egg-yolk/OP50 mixture was sterile pipetted onto 25cm 5X PEP plates, using 12mL per plate. Each plate was swirled to spread the egg-yolk/OP50 mixture and additional debris was pipetted off. Plates were left for approximately 3 days at room temperature to solidify to custard like consistency at which point they were seeded with C. elegans. C. elegans were grown to a dense lawn and harvested by sucrose floatation.

Sucrose Flotation to clean C. elegans cultures

As described previously in [51], worms from egg plates were harvested using room temperature M9 and collected in 50mL conical vials. Worms were pelleted by centrifugation 2500xg at 4°C in an Eppendorf tabletop centrifuge and resuspended in cold fresh M9 to rinse, and re-pelleted. Supernatant was discarded and pellets were resuspended and brought to 25mL with cold M9 and incubated on ice for ∼10 minutes. 25mL of cold 60% sucrose was added. Tubes containing the worm and sucrose mixtures were inverted 10 times to mix. They were then centrifuged for 5 min at 4°C at 2500xg. Worms float and form a supernatant layer on sucrose whereas bacteria and debris pellet at the bottom of the tube. Worms were removed with a sterile metal spatula and wide bore 1mL micropipette into fresh cold M9 on ice. Total volume was brought up to 50mL with cold M9. Tubes were inverted to mix thoroughly and centrifuged for 2 min at 1000xg at 4°C. The supernatant was discarded and fresh cold M9 was added to 30mL volume to resuspend the pellet and rinse any residual sucrose. Worms were centrifuged a final time at 1000xg for 2 min at 4°C. The supernatant was discarded and the pellet snap frozen in liquid nitrogen then stored at −80°C until use.

Microtubule Sedimentation assay

MT sedimentation assays were based on the protocol published in [45]. Briefly, large scale populations were grown on plates with chicken egg yolk media and sucrose floated to clean. Approximately one gram worm pellets were weighed and thawed on ice. Worm Lysis Buffer (50mM HEPES-KOH, pH 7.6, 1mM MgCl2, 1mM EGTA, 75mM KCl, 0.5mM DTT plus ProteincOmplete™, Mini, EDTA-free Protease Inhibitor Cocktail) volume (mL) was added to 1.5 times the weight (g) of the pellet. The pellets were resuspended and sonicated at 70% amplitude for two rounds of 30 seconds on 60 seconds off, for a 1 minute sonication time per round that was repeated for two rounds to complete 2 minutes of active sonication time. Samples were kept on ice throughout the sonication process and the sonicator tip was cooled with ice bath in between each sonication duration. Crude lysates were ultra centrifuged 20,000xg for 10 minutes (4°C). The Low Speed Spin (LSS) supernatant was saved and subjected to centrifugation at 50,000xg (4°C). The supernatant from this spin was reserved as the High Speed Spin (HSS) sample, this material was snap frozen with liquid nitrogen and stored at −80 until use. HSS material was thawed on ice prior to conducting microtubule sedimentation assays. The total protein concentration for the extracts was determined by Bradford assay and equalized prior to proceeding with the sedimentation reactions. DTT and GTP was spiked into 1mL of the equalized HSS sample (final concentrations of 0.5mM and 1mM respectively) to make the HSS++ input material used in the subsequent reactions. Material was kept on ice until reactions were assembled. Three reaction conditions were assembled using 205µL HSS++ in each reaction: DMSO control, Nocodazole control and Taxol stabilized. DMSO was added to the equivalent volume as the additions of nocodazole (10µg/mL final concentration) and taxol (20µM final concentration). Microtubule reactions were incubated at 25°C for 10 min to allow for microtubule assembly. 200µL of the reactions was placed on a warmed (25C) glycerol cushion (40% glycerol, 50mM HEPES-KOH, 1mM MgCl2, 1mM EGTA, 75mM KCl, 0.5 mM DTT and ProteincOmplete™, Mini, EDTA-free Protease Inhibitor Cocktail). Microtubules were spun over the glycerol cushions at 48,000xg for 30 minutes (25°C). 80 µL of the supernatant was collected and combined with 20 µL 5x SDS sample buffer. The supernatant-cushion interface was washed 3 times with 200 µL worm lysis buffer. The cushion was carefully removed by pipetting and the residual pellet was resuspended in 200 µL 1x SDS sample buffer. Samples were then subjected to capillary electrophoretic analysis using the Simple Western Peggy Sue^TM Capillary Electrophoresis System for analysis of tau and tubulin.

Capillary electrophoresis and immunodetection

We used the Peggy Sue^TM (ProteinSimple) Simple Western system to immunoblot MT sedimentation assay samples following the manufacturer’s instructions. Briefly, samples were diluted with 0.1X sample buffer and 5× Fluorescent Master Mix, boiled for 5 minutes at 95°C, cooled and loaded into a 384-well assay plate (12-230kDa Size Separation module, Catalog No. SM-S001, ProteinSimple). Samples were immunoblotted using the E7 (1:100) β-tubulin primary antibody, SP70 (1:2000) and DAKO (1:10,000) total tau primary antibodies, and goat anti-mouse-HRP (1:100) and goat anti-rabbit (1:100) secondary antibodies (Supplementary Table 3) that were diluted in Antibody Diluent 2. Samples were run and analyzed using the default assay in Compass for SW Version 4.0.0 (ProteinSimple). Signal for sedimented MT mass was normalized to total pellet signal then compared to input signal (e.g. taxol pellet WT/(taxol-pellet(WT) + taxol-(tba-2))/input tubulin signal) and means were calculated using four biological replicates. A Student’s T Test was used to determine the statistical significance of the difference between the means.

Isolation of C. elegans tubulin

Tubulin isolation methods were based on those previously reported in Widlund et al. 2012, [54]. Large scale populations were grown on plates containing chicken egg yolk media and sucrose floated to clean. Approximately 10g sucrose floated worm pellet material was thawed, mixed with equivalent volume (weight to volume) of BRB80 buffer (80mM PIPES, 1mM EGTA, 1mM MgSO₄, pH 6.8) to form a slurry. Slurry was re-frozen dropwise in liquid nitrogen to make pellets. Pellets were enclosed in a folded foil envelope, ∼15ml pellets per batch, and hammered over a brick of dry ice. Once pellets were ground into fine powder, powder was transferred to pre-chilled 50ml conical on dry ice. BRB80 was added to bring the volume to 50mL, adding more BRB80 as the worm powder thawed to maintain the 50mL volume. The crude lysate was subjected to a low speed spin at 3500 x g at 4°C. The supernatant was decanted into a fresh 50mL conical and pellet discarded. The supernatant was subjected to high-speed ultracentrifugation at 301,311 x g (50,000 rpm in Beckman Ti50.2 rotor) for 35 minutes at 4°C. The centrifuged supernatant was then filtered through a cell strainer (Falcon 70µM, catalog number 352350) 0.22µM filter to further clear lysate. Cleared lysate was loaded, by gravity flow, onto an affinity column with TOG1/2 domains from the Saccharomyces cerevisiae protein Stu2. Column was washed with 10 column volumes of wash buffer comprised of (1x BRB80, 100uM GTP and 100uM MgCl2). Tubulin was eluted off the column using 3 column volumes of elution buffer (1xBRB80, 10uM Mg2 GTP, and 500mM (NH4)2SO4). Protein containing fractions were determined using Bradford dye (10µl fraction sample to 100µl diluted Bradford reagent, Biorad catalog number 5000006) to assess protein presence by eye. Protein containing fractions were pooled, mixed and desalted over BRB80 equilibrated PD10 desalting columns (Sigma, catalog number GE17-0851-01). Desalted protein was concentrated to 250µl at 4°C at 3500 x g with 30 kDa molecular weight cut off Amicon Ultra centrifugation filters (Millipore catalog number UFC9030) previously equilibrated with BRB80. Glycerol was added to a final concentration of 5% v/v and mixed thoroughly. Tubulin was aliquoted and snap frozen in liquid nitrogen and stored at −80°C until use.

TOG1/2-column for affinity purification of C. elegans tubulin

Purification of GST tagged Stu2 TOG1/2 domains and affinity column creation was based on methods in Widlund et al [54]. A 90 L growth and induction of pGEX-6P-1 Stu2 1-590 (Addgene plasmid #38314) was expressed in Rosetta (DE3) cells and bacterial cell paste was frozen into a brick and stored at −80°C until use. At the time of TOG1/2 protein isolation, chunks of the frozen E coli cell paste (∼20g) were thawed on ice. Bacteria was resuspended in 140mL of lysis buffer (2xPBS, 12.5mM DTT, and cOmplete Mini-EDTA free protease inhibitor pellets (Millipore, catalog number 11836170001). 0.1g lysozyme, 100uL DNAse, 25uL RNAse A, 80uL 1M MgCl2 was added to bacterial resuspension to initiate lysis. Resuspension was inverted to mixed every ∼5 min for 20 minutes incubating on ice. Resuspension was aliquoted into 4 conicals and sonicated for one round at 70% amplitude for 10 seconds. Lysate was then transferred to oakridge tubes for centrifugation in a Sorvall SA-600 rotor at 16,000 RPM at 4°C for 30 minutes. Cleared lysate was filtered through a 70µm cell strainer (Falcon, Fisher Scientific catalog number 352350) then passed through a 0.22µM filter to further clear. Lysate was applied to a 5mL GSTrap 4B Sepharose column (Cytiva 28401748) connected to a peristaltic pump and equilibrated with equilibration buffer (2xPBS, 1mM DTT) at a flow rate of 2mL per minute. The column was washed with 5 column volumes (CV) of wash buffer 1 (W1, 2x PBS, 0.1% Tween20, 1mM DTT) followed by 2.5 CV of wash buffer 2 (W2, 6X PBS, 1mM DTT), 2.5 CV of wash buffer 3 (W3, 2x PBS, 1mM DTT). Protein was eluted with Elution buffer (2x PBS, 50mM glutathione, 10mM DTT, pH 9.0) collecting 1 ml fractions. Protein presence was detected using a small-scale Bradford assay as described above. Protein containing fractions were desalted using PD10 desalting columns equilibrated with pH 10 coupling buffer (0.1M sodium citrate, 0.05M Sodium carbonate, pH brought to 10 with NaOH). Protein was snap frozen in liquid nitrogen and stored at −80°C until used for coupling to resin.

Resin coupling was performed using the AminoLink™ Plus Immobilization Kit, 2 mL format (Thermo Scientific, 44894). Approximately 40mL volume and 50g of Stu2 TOG1/2 protein was thawed and mixed together, resin from two columns (4mL) was washed using pH10 coupling buffer and using coupling buffer, resuspended and transferred to with the thawed Stu2 TOG1/2. Resin was incubated in the protein solution with end over end rotation for 1 hour at room temperature. Coupled resin was divided and packed into two columns. Resin was then washed twice with 2mL of pH 10 coupling buffer. Each column was then washed with 10mL of 1xPBS followed by 3 washes with 2mL of storage buffer (1xPBS, 50% v/v glycerol). After the final wash in storage buffer, additional storage buffer was used to top off the column head space, the top and bottom caps were securely replaced, columns parafilmed to seal and placed upright at −20°C for storage. Columns were dedicated either to wildtype or tba-2(D429N) tubulin isolations and re-used for at least 5 isolations.

Isolation of 6X-histidine tagged recombinant human 4R1N tau protein

E. coli codon-optimized 6X-histidine-tagged human 4R1N tau (His-tau) cDNA was cloned into the pET-28a vector (Merck KGaA, Darmstadt, Germany) and transformed into BL21 (DE3) bacteria (New England Biolabs, Ipswich, MA). Ten mL starter cultures of Terrific Broth (TB) were grown overnight at 37°C in a shaking incubator. For induction, 1 L cultures of TB were inoculated, grown with shaking at 37°C to log phase, and induced with 1mM IPTG for 3 hours at 37°C. Bacterial pellets were resuspended in lysis buffer (1X PBS supplemented with 150mM NaCl, 5mM β-mercaptoethanol) and boiled to lyse. Lysate was cooled in ice-bath before centrifuging at 7000 x g to clear lysate. Cleared lysate was incubated with Ni-NTA agarose resin (Qiagen, Hilden, Germany) which was washed once with 10X bed volume of lysis buffer and then twice using lysis buffer with 10mM imidazole. His-tau was eluted twice: first, using 1 bed volume of lysis buffer with 250 mM imidazole and then using lysis buffer with 500 mM imidazole. Elution 1 was desalted once with PD10 columns (Cytiva, Marlborough, MA) into 50 mM HEPES, 25 mM NaCl, pH 7.4 and Elution 2 was desalted twice.

Biolayer interferometry assay

Wildtype and tba-2(D429N) tubulin was isolated as described above. Recombinant wildtype human 4R1N 6xHis-tagged tau protein (His-tau) was prepared as previously published in McMillan et al 2023 [55]. Recombinant human His-tagged MPRO protein (His-MPRO) was prepared as previously published in Baker et al 2021 [56].y His-MPRO was used as both a quencher and a control for non-specific binding.

All data was obtained using a Sartorius Octet R8 with Sartorius Octet NTA Biosensors (18-5101 Lot number 2412000511) in a 96-well format. BRB80 buffer (80mM Pipes, 1mM EGTA, 1mM MgSO4, pH 6.8) was used to establish an initial baseline. After a 10 minutes shaking (1000 RPM), experimental steps proceeded at 30°C shaking at 1000 RPM in the following order: loading of His-tau (75 nM), quenching with His-MPRO (10 ug/ml), washing with BRB80, running a secondary baseline with BRB80, association, and lastly dissociation with BRB80. After a dissociation step with BRB80, the sensors were regenerated with 10 mM glycine (pH 1.5 – 2.0), reactivated with 10 mM NiCl, and washed with BRB80. Washes, quenches, activation, and loading steps ran for 300 seconds each. Association and dissociation steps ran for 600 seconds each. Baseline steps ran for 180 seconds each. The regeneration step ran for 30 seconds. Shaking was maintained at a speed of 1000 RPM throughout the experiment. Three to four independent isolations, for each wildtype and tba-2 (D429N) tubulin were tested at multiple concentrations within 133 −384.4 nM range. His-MPRO (75nM) was used in place of His-Tau as a test for non-specific binding by tubulin and BRB80 was used as a buffer only control.

Analysis was conducted in the Octet Analysis Studio Software (v. 13.0.0.32). Background signal was accounted for with the His-MPRO control sensor and the buffer only initial sample within the software. Graphs were aligned on the Y-axis according to the baseline averages, along with an inter-step correction aligned to the dissociation step and filtered via Savitzky-Golay. Data was then exported into Microsoft Excel. The mean of the dissociation coefficients (K_d) for a given tubulin preparation was calculated to determine the K_d for that tubulin isolation. The means of the wildtype or tba-2 (D429N) isolations were calculated to compare the overall wildtype-tau K_d versus mutant-tau K_d. Isolation replicate means were graphed in Graphpad Prism 10. Student’s T test was used to test the difference in mean K_d and the displayed association curves are representative curves for wildtype and mutant worm tubulin obtained in the same replicate.

Data availability

Data generated in this work is available at Dryad data repository https://doi.org/10.5061/dryad.15dv41p9s

Supplementary tables and figures

Supplementary Table 1.

Supplementary Table 2.

Supplementary Table 3.

Supplementary Figure 1.

Supplementary Figure 2.

Supplementary Figure 3

Supplementary Figure 4.

Supplementary Figure 5.

Supplementary Figure 6.

Supplementary Figure 7.

Acknowledgements

We thank Charles L. Asbury for guidance and mentorship in producing this work and thank Asia Beale for efforts in growing large populations of C. elegans for tubulin isolations and sedimentation assays. The work presented here is the responsibility of the authors and does not represent the official views of the NIH or government of the United States of America.

Additional information

Funding

This work was supported by research grants from the Department of Veterans Affairs [IK6BX006467 to BK and I01BX005762 to NL] and National Institutes of Health [T32AG052354 and K99/R00AG073455 to SB, R01AG066729 to NL, R01AG084552 to RK, R01NS064131 and R01AG084680 to BK]. Additionally, we acknowledge startup support from University of Washington School of Medicine and Seattle Institute for Biomedical and Clinical Research.

Author contributions

SB conceptualized, designed and executed experiments, analyzed data, drafted manuscript. AS, MB, RU, JS, KK and CD created reagents (i.e. C. elegans strains, purified protein) and executed experiments, RK performed forward genetic screening. LW, NL and BK aided in conceptualization, experiment design and design of analyses. All authors read and edited manuscript.

Funding

HHS | National Institutes of Health (NIH) (T32AG052354)

• Sarah Benbow

HHS | National Institutes of Health (NIH) (K99/R00AG073455)

• Sarah Benbow

HHS | National Institutes of Health (NIH) (R01AG066729)

• Nicole Liachko

U.S. Department of Veterans Affairs (VA) (I01BX005762)

• Nicole Liachko

HHS | National Institutes of Health (NIH) (R01AG084552)

• Rebecca L Kow

HHS | National Institutes of Health (NIH) (R01NS064131)

• Brian C Kraemer

HHS | National Institutes of Health (NIH) (R01AG084680)

• Brian C Kraemer

U.S. Department of Veterans Affairs (VA) (IK6BX006467)

• Brian C Kraemer