KIN‐4/MAST kinase promotes PTEN‐mediated longevity of Caenorhabditis elegans via binding through a PDZ domain

Abstract PDZ domain‐containing proteins (PDZ proteins) act as scaffolds for protein–protein interactions and are crucial for a variety of signal transduction processes. However, the role of PDZ proteins in organismal lifespan and aging remains poorly understood. Here, we demonstrate that KIN‐4, a PDZ domain‐containing microtubule‐associated serine‐threonine (MAST) protein kinase, is a key longevity factor acting through binding PTEN phosphatase in Caenorhabditis elegans. Through a targeted genetic screen for PDZ proteins, we find that kin‐4 is required for the long lifespan of daf‐2/insulin/IGF‐1 receptor mutants. We then show that neurons are crucial tissues for the longevity‐promoting role of kin‐4. We find that the PDZ domain of KIN‐4 binds PTEN, a key factor for the longevity of daf‐2 mutants. Moreover, the interaction between KIN‐4 and PTEN is essential for the extended lifespan of daf‐2 mutants. As many aspects of lifespan regulation in C. elegans are evolutionarily conserved, MAST family kinases may regulate aging and/or age‐related diseases in mammals through their interaction with PTEN.

The PDZ (PSD-95/Dlg-1/ZO-1) domain-containing proteins (hereafter referred to as PDZ proteins) act as scaffolds for protein-protein interactions and mediate various cellular signaling processes (Reviewed in Kim & Sheng, 2004). PDZ domains are composed of six β-sheets and two α-helices and comprise~90 amino acids. PDZ proteins typically bind the PDZ-binding motifs that are located at the C-terminal regions of the partner proteins. Roles of various PDZ proteins in cellular processes, such as signal transduction in neurons, are relatively well established; however, their function in aging and lifespan regulation is underexplored.
In the current study, we investigated the role of KIN-4 in lifespan regulation conferred by reduced IIS in C. elegans. We used RNA interference (RNAi) to screen for genes encoding PDZ proteins that affected the lifespan of C. elegans. Our results showed that  was required for the long lifespan of daf-2 mutants. kin-4 was partly required for dauer formation and oxidative stress resistance in daf-2 mutants. Moreover, kin-4 in neurons was crucial for the extension of lifespan by daf-2 mutations. Through a large-scale yeast two-hybrid screen and subsequent protein-protein binding assays, we found that DAF-18/PTEN bound to the PDZ domain of KIN-4 through its C-terminus. More importantly, the interaction between KIN-4 and DAF-18/PTEN was required to extend the lifespan of daf-2 mutants.
Our findings suggest that MAST family kinases exert physiological effects on lifespan regulation via modulating IIS pathways through direct interaction with PTEN.

| KIN-4 is required for longevity conferred by reduced IIS
We aimed at identifying PDZ proteins that contribute to C. elegans longevity. We first identified 80 PDZ protein-encoding genes by using Pfam and WormBase and by performing literature searches (Supporting information Table S1). We then examined the effect of each of the commercially available RNAi clones, targeting 49 candidate genes, on the lifespan of control worms and daf-2/insulin/IGF-1 receptor mutants (Figure 1a and Supporting information Table S2); we used daf-2 mutants that display a prominent longevity phenotype (Kenyon, 2010), as the effect of RNAi on the lifespan of these mutants would be more pronounced than on the wild-type . Many PDZ proteins play roles in neurons (Reviewed in E. Kim & Sheng, 2004); because C. elegans neurons are refractory to RNAi (Timmons, Court, & Fire, 2001), we used rrf-3 mutant animals that display enhanced sensitivity to RNAi (Asikainen, Vartiainen, Lakso, Nass, & Wong, 2005) (Figure 1a; Supporting information Table S2). We found that RNAi targeting kin-4, which encodes a MAST family kinase (Manning, 2005), or gipc-1 and gipc-2, which encode GAIP-interacting protein C proteins (Vartiainen, Pehkonen, Lakso, Nass, & Wong, 2006), had greater lifespan-shortening effects on daf-2 mutants than on control animals (Figure 1a-c). We then determined the effects of genetic inhibition of kin-4 or gipc-1/-2 on daf-2(-) longevity using available loss of function mutants. Two inde-  Figure S1c). These data suggest that kin-4 contributes to longevity caused by daf-2 mutations.
Next, we determined the effect of kin-4 mutation on longevity conferred by other gene mutations. The kin-4 mutation fully suppressed the longevity of the sensory neuron-defective osm-5 mutants (Figure 2c), in which IIS is decreased (Apfeld & Kenyon, 1999), but did not affect the longevity of dietary restriction-mimetic eat-2 mutants ( Figure 2d). These data are consistent with the possibility that KIN-4 functions in IIS for C. elegans longevity.

| kin-4 is partly required for enhanced dauer formation and oxidative stress resistance caused by daf-2 mutations
We sought to determine the role of KIN-4 in other daf-2(−)-mediated physiological processes, including development and stress resistance.
kin-4 mutations significantly reduced the formation of daf-2 mutationinduced dauer, a hibernation-like alternative larva (Hu, 2007)  was not significantly affected by the kin-4 mutation (Supporting information Figure S2c). Thus, KIN-4 appears to play roles in various IISmediated physiological processes in a context-dependent manner. F I G U R E 3 Neurons are crucial tissues for lifespan regulation by KIN-4. (a) KIN-4a is predicted to have 4 domains; DUF1908 whose function is not known, pre-PK that is mainly found in MAST family kinases, protein kinase, and PDZ domains. The deleted part by kin-4 (tm1049) mutation is marked with a black solid line. (b, c) kin-4::gfp transgene-encoded protein (KIN-4::GFP) was mainly expressed in head and tail (c) neurons (arrowheads) and dimly in the intestine (arrow). Scale bar indicates 50 μm. (d) KIN-4a::GFP increased the shortened lifespan of daf-2(e1370); kin-4(tm1049) [daf-2(−); kin-4(−)] mutants. (e) kin-4 transgenic worms did not display longevity. We also found that daf-2 RNAi did not alter KIN-4::GFP levels (Supporting information Figure S4d). See Supporting information Table S3 for  and mtl-1 (Supporting information Figure S6c-e). We confirmed these results using sod-3p::gfp transgenic animals (Supporting information Figure S6f). These data suggest that kin-4 influences reduced IIS-mediated longevity by acting through factors other than DAF-16/ FOXO.

| Identification of KIN-4-interacting proteins that contribute to the longevity of daf-2 mutants
Because KIN-4 has a PDZ domain, we sought to identify proteins that bound to the PDZ domain, as these may contribute to longevity together with KIN-4. We conducted an exhaustive yeast two-hybrid screen using the PDZ domain of KIN-4 as bait against C. elegans cDNA library and identified 41 prey proteins (Supporting information Figure S7). Of these, 23 proteins contained putative PDZ-binding motifs at their C-termini (Supporting information Table S6). To determine whether any of these 23 proteins were required for the longevity of daf-2 mutants, we performed an RNAi lifespan screen targeting 21 genes (Figure 4a; Supporting information Table S7).

| Interaction between KIN-4 and DAF-18/PTEN is crucial for the longevity of daf-2 mutants
Next, we examined whether the interaction between KIN-4 and DAF-18/PTEN contributed to reduced IIS-mediated longevity. We generated transgenic animals expressing mutant forms of DAF-18/ PTEN proteins carrying Δ4C and Δ12C fused with mCherry (daf-18 Δ4C and daf-18 Δ12C, respectively) and those expressing mCherrytagged wild-type (WT) DAF-18 (daf-18 WT) as a control. Expression of all these three mCherry-fused DAF-18/PTEN proteins was detected, and their subcellular localization was enriched at the plasma membrane (Figure 6a). We found that daf-18 WT fully restored the longevity of daf-2; daf-18 double mutants (Figure 6b).
In contrast, daf-18 Δ4C or daf-18 Δ12C only partially restored longevity in daf-2; daf-18 double mutants (Figure 6c, d). These data suggest that the C-terminal end of DAF-18/PTEN, which mediates the physical interaction with KIN-4, is required for the full longevity of daf-2 mutants. We also found that kin-4 mutation further shortened the lifespan of daf-18 Δ12C-expressing daf-2; daf-18 mutants (Supporting information Figure S9)  Table S3 for statistics and additional repeats longevity via KIN-4-independent targets. Altogether, our data suggest that KIN-4 contributes to longevity in daf-2 mutants partly through binding to DAF-18/PTEN at the C-terminus.

| C. elegans MAST family kinase KIN-4 promotes longevity via binding DAF-18/PTEN
PDZ proteins function as scaffolds that organize proteins for transducing various intracellular signals. However, the role of PDZ proteins in IIS, an important evolutionarily conserved lifespan-regulatory signaling, has been poorly understood. In this study, we showed that KIN-4/MAST kinase was a lifespan-regulatory PDZ protein acting through DAF-18/PTEN. KIN-4, mainly in neurons, influenced the longevity conferred by reduced IIS. We also identified DAF-18/PTEN as a key interacting protein of KIN-4. The interaction between KIN-4 and DAF-18/PTEN was important for the lifespan regulation by IIS. Our study suggests that KIN-4/MAST family kinase is a novel longevitypromoting protein that acts in IIS via binding DAF-18/PTEN protein.

| KIN-4 affects lifespan differently from previously reported PDZ proteins, MICS-1 and MPZ-1
Previous studies have reported lifespan-influencing PDZ proteins, including MICS-1/synaptojanin 2 binding protein (SYNJ2BP) and multi-PDZ domain scaffold protein (MPZ-1), in C. elegans (Hoffmann et al., 2012;Palmitessa & Benovic, 2010). The genetic inhibition of mics-1 promotes longevity (Hoffmann et al., 2012), and mpz-1 knock-down increases lifespan, perhaps by activating DAF-18/PTEN (Palmitessa & Benovic, 2010). Both mics-1 and mpz-1 genes were included in our RNAi lifespan screen (Supporting information Table S2); however, RNAi targeting of mics-1 or mpz-1 had small or no effect on lifespan in wild-type or daf-2 mutant animals ( Figure 1a; Supporting information Table S2). In contrast, kin-4 mutation or RNAi substantially suppressed the extended lifespan of daf-2 mutants. This difference may have arisen from different genetic backgrounds and/or experimental conditions; for example, mics-1 RNAi lifespan assay was performed at 25°C (Hoffmann et al., 2012), and mpz-1 RNAi was applied for two generations to wild-type worms (Palmitessa & Benovic, 2010). Here, we performed lifespan assays using RNAi-sensitive rrf-3 mutants treated with each of these RNAi clones only during adulthood at 20°C. Despite the difference, previous reports do not conflict with our data showing that KIN-4 is required for the extended lifespan of C. elegans conferred by reduced IIS. We speculate that KIN-4 contributes to IIS-mediated longevity in a manner distinct from MPZ-1 or MICS-1.

| kin-4 mutation appears to elicit a compensatory activation of DAF-16/FOXO in daf-2 mutants
Although a major protein acting downstream of DAF-18/PTEN lipid phosphatase is DAF-16/FOXO, we did not find evidence supporting the possibility that KIN-4 acts with DAF-18/PTEN for increasing DAF-16/FOXO activity. In contrast, DAF-16/FOXO activity in daf-2 mutant animals appears to be further increased by kin-4(−) mutations (Supporting information Figure S6). We speculate that kin-4(−) mutations decrease lifespan in daf-2 mutants via DAF-18/PTEN while acting through factors other than DAF-16/ FOXO and that this in turn enhances DAF-16/FOXO activity as a compensatory response. Genetic inhibition of several DAF-16/ FOXO-independent factors that contribute to reduced IIS-mediated longevity has been shown to elicit compensatory activation of DAF-16/FOXO. For example, we previously reported that mutations in smg-2/UPF1, a key factor for nonsense-mediated mRNA decay, decrease the longevity of daf-2 mutants, while further increasing DAF-16/FOXO activity (Son et al., 2017). In addition, PQM-1, paraquat (methyl viologen)-responsive transcription factor, is required for longevity in daf-2 mutants, but genetic inhibition of pqm-1 increases the nuclear localization of DAF-16 (Tepper et al., 2013). It will be interesting to determine whether KIN-4 and DAF-18/PTEN act with DAF-16/FOXO-independent longevity factors for contributing to long lifespan in C. elegans with reduced IIS.

| KIN-4 may increase DAF-18/PTEN activity via phosphorylation
Post-translational modifications of PTEN, including phosphorylation, are important for its regulation (Fragoso & Barata, 2015). In many cases, phosphorylation at the C-terminal region of PTEN leads to its conformational change that inhibits the lipid phosphatase activity of PTEN (Fragoso & Barata, 2015). On the other hand, Rak tyrosine kinase and polo-like kinase 3 (Plk3) activate PTEN by phosphorylating the C2 domain and C-terminal part of PTEN, respectively (Xu, Yao, Jiang, Lu, & Dai, 2010;Yim et al., 2009). In addition, RhoA-associated kinase (ROCK) phosphorylates PTEN and increases its activity (Li et al., 2005). Mammalian MAST2 also phosphorylates PTEN in vitro (Valiente et al., 2005), although its effect on the activity of PTEN remains unexplored. Given the similar lifespan phenotypes caused by mutations in daf-18/PTEN and kin-4, we speculate that C. elegans KIN-4 may increase the activity of DAF-18/PTEN through binding and phosphorylation.

| KIN-4 contributes to longevity possibly through altering the protein phosphatase activity of DAF-18/PTEN
Taking into consideration that DAF-18/PTEN has dual phosphatase activity for both proteins and lipids (Fragoso & Barata, 2015), it seems likely that the binding of KIN-4 may affect the phosphatase activity of DAF-18/PTEN for as yet unidentified substrate proteins. In mammals, substrate proteins of PTEN phosphatase include protein kinase B/Akt and cyclic AMP response element-binding protein (CREB; Gu et al., 2011;Phadngam et al., 2016). Therefore, homologs of these proteins may mediate the effects of the interaction between KIN-4 and DAF-18/PTEN on the longevity of C. elegans. One limitation regarding this scenario is that tyrosine 138, which is essential for the protein phosphatase activity of PTEN in mammals (Davidson et al., 2010), is replaced by leucine in C. elegans DAF-18/PTEN. Nevertheless, C. elegans DAF-18/PTEN can act as a protein phosphatase for VAB-1/ephrin receptor (Brisbin et al., 2009). Therefore, it will be important to identify and functionally characterize protein substrates of DAF-18/PTEN in C. elegans in future research, with respect to the role of KIN-4 in protein-protein interaction and longevity.

daf-2 mutants by acting with other factors as well as with DAF-18/PTEN
In this paper, we showed that physical interaction between  and DAF-18/PTEN is required for full longevity of daf-2 mutants.
However, KIN-4 also seems to contribute to reduced IIS-mediated longevity through additional factors. The mammalian MAST kinases bind various proteins, including microtubules (Walden & Cowan, 1993), β2-syntrophin (Lumeng et al., 1999), TRAF6 (Xiong et al., 2004), and ARPP-16 (Andrade et al., 2017). We also identified various factors that bound the PDZ domain of C. elegans KIN-4 through a yeast two-hybrid screen (Supporting information Table S6). Thus, KIN-4 binding to some of these proteins may contribute to longevity in daf-2 mutants. In addition, MAST2 and MAST3 kinases phosphorylate Na + /H + exchanger 3 (NHE3; Wang et al., 2006) and ARPP-16 (Andrade et al., 2017), respectively, and therefore, phosphorylation of C. elegans homologs of these proteins by KIN-4 may affect reduced IIS-mediated longevity. It will be important to test these possibilities by performing biochemical and molecular genetic experiments in future studies.

| Interaction between MAST kinases and PTEN may play important roles in mammalian physiology
Several lines of evidence indicate that MAST kinases are crucial for the physiology of cancer cells (Eissmann et al., 2013;Robinson et al., 2011;Tomoshige et al., 2015). In some breast cancer cells, MAST1 and MAST2 are fused with other proteins through recurrent gene rearrangements; these fusion proteins act as putative tumorigenic drivers (Robinson et al., 2011). A pedigree study reported that MAST1 variants are associated with lung cancer (Tomoshige et al., 2015). Furthermore, MAST2 influences glioblastoma tumor growth, potentially as an apoptosis suppressor (Eissmann et al., 2013). In this study, we characterized the physiological role of the physical interaction between KIN-4 and DAF-18/PTEN in C. elegans. Considering evolutionarily conserved amino acid sequences of these proteins, our study may provide insights into the role of MAST kinases and PTEN in tumorigenesis and aging, thus contributing toward the development of therapeutic strategies for anti-cancer and anti-aging medicine.

| Strains
All strains used in this study are described in the Supporting Information.

| Identification of PDZ proteins
Identification of potential C. elegans PDZ proteins was performed by using combination of bioinformatic methods similarly to a previous paper . See the supporting information for details.

| Lifespan assays
Lifespan assays were conducted at 20°C as previously described, with some modifications . See the Supporting Information for details.

| Dauer formation
Dauer assays were performed as previously described, with some modifications (Gaglia et al., 2012). Gravid adult worms were placed on OP50-seeded NGM plates and were allowed to lay eggs for 3 hr at 22.5°C and 25°C. The number of dauers and total worms was counted when non-dauer worms reached L4 or young adult stage.
Error bars indicate SEM, and statistics were calculated by using Student's t test.

| Stress resistance assays
Stress resistance assays were performed by following a previous report with some modifications . See the Supporting Information for details.

| Generation of transgenic worms
Transgenic worms were generated as previously described, with some modification (Artan et al., 2016). See the Supporting Information for details.

| Generation of mutant strain using CRISPR
To generate kin-4(nj170), the homologous recombination was performed essentially by following a previous report (Dickinson, Ward, Reiner, & Goldstein, 2013). First, the kin-4 locus was replaced by a repair template (pSN588) that carries the loxP-unc-119(+)-loxP sequence flanked by the left and right arms corresponding to the 1st intron and the 3′ UTR genomic sequence of kin-4, respectively. This repair template was injected into unc-119(ed3) animals with peft-3:: Cas9 and an sgRNA plasmid (pSN598) to generate a double-stranded break at the kin-4 locus. Progeny were screened for animals in which the kin-4 locus was replaced by the repair template. The loxP-unc-AN ET AL. | 9 of 12 119(+)-loxP cassette was subsequently removed by injecting peft-3:: Cre plasmid (pDD104). The animals in which the unc-119(+) gene was excised were isolated and were confirmed by DNA sequence analysis. The resulting strain carries a 17 kb deletion in the kin-4 locus and removes the sequence between 24,629 of the cosmid F22B3 and 11,644 of the cosmid C10C6. The unc-119(ed3) mutation was then crossed out to obtain IK2019.

| Yeast two-hybrid screen
Yeast two-hybrid screen was performed by Panbionet (http://panb ionet.com) following a previous report (Kim et al., 2008). See the Supporting information for details. 4.9 | Prediction of protein domains, structure modeling and sequence alignment, and generation of the phylogenic tree Prediction of protein domains, structure modeling and sequence alignment, and generation of the phylogenic tree were performed similarly to a previous paper , with modifications.
See the Supporting Information for details.

| Co-immunoprecipitation and immunoblotting
Co-immunoprecipitation and immunoblotting were performed following a previous report  with some modifications. See the Supporting Information for details.

| GST pull-down assay
GST pull-down assays were performed following methods described in a previous report (Dev, Nishimune, Henley, & Nakanishi, 1999) with some modifications. See the Supporting Information for details.

| Microscopy
Microscopy experiments were executed as previously described with some modifications (Son et al., 2017). See the Supporting Information for details.

| Quantitative RT-PCR analysis
Quantitative RT-PCR experiments were performed as previously described with some modifications (Son et al., 2017). See the Supporting information for details.

ACKNOWLEDGMENTS
Some C. elegans strains were provided by Dr. Cynthia Kenyon laboratory and Caenorhabditis Genetics Center. We thank all Lee laboratory members for discussion and help. These authors have no competing interests to declare. This study is supported by the