A PTEN variant uncouples longevity from impaired fitness in Caenorhabditis elegans with reduced insulin/IGF-1 signaling

Insulin/IGF-1 signaling (IIS) regulates various physiological aspects in numerous species. In Caenorhabditis elegans, mutations in the daf-2/insulin/IGF-1 receptor dramatically increase lifespan and immunity, but generally impair motility, growth, and reproduction. Whether these pleiotropic effects can be dissociated at a specific step in insulin/IGF-1 signaling pathway remains unknown. Through performing a mutagenesis screen, we identified a missense mutation daf-18(yh1) that alters a cysteine to tyrosine in DAF-18/PTEN phosphatase, which maintained the long lifespan and enhanced immunity, while improving the reduced motility in adult daf-2 mutants. We showed that the daf-18(yh1) mutation decreased the lipid phosphatase activity of DAF-18/PTEN, while retaining a partial protein tyrosine phosphatase activity. We found that daf-18(yh1) maintained the partial activity of DAF-16/FOXO but restricted the detrimental upregulation of SKN-1/NRF2, contributing to beneficial physiological traits in daf-2 mutants. Our work provides important insights into how one evolutionarily conserved component, PTEN, can coordinate animal health and longevity.

A ging is accompanied by a decline in biological functions and by pathophysiology of various age-associated diseases. To devise strategies for promoting long and healthy human lives, molecular mechanisms underlying the aging processes have been extensively investigated for the last several decades. Hundreds of mutations in aging-related genes have been identified in various model organisms 1 . However, genetic and environmental factors that have been shown to increase lifespan tend to also cause fitness defects. In addition, the lifespan of wild nematode strains negatively correlates with growth rates 2 , a key developmental fitness parameter. Overall, biological strategies that promote longevity while simultaneously maintaining fitness are rare.
In addition to daf-2, various aging-related factors in IIS pathways have been identified. These include daf-18, which encodes a worm ortholog of phosphatase and tensin homolog (PTEN) phosphatase, an enzyme that dephosphorylates phosphatidylinositol 3,4,5-trisphosphate (PIP 3 ) to phosphatidylinositol 4,5bisphosphate (PIP 2 ) 15 . daf-18(nr2037), a strong loss-of-function mutation [hereafter referred to as daf-18(−)], fully suppresses constitutive dauer formation and the long lifespan of daf-2 mutants 16,17 . The DAF-16/FOXO transcription factor, which is activated downstream of DAF-18/PTEN, is required for both the constitutive dauer phenotype and the longevity of daf-2 mutants 1,4,5 . Other transcription factors, including SKN-1/ nuclear factor erythroid 2-related factor 2 (NRF2) and heat shock transcription factor 1 (HSF-1), are also crucial for longevity and stress resistance in daf-2 mutants 1,4-7, 18 . These transcription factors regulate the expression of common and distinct subsets of target genes, which appear to elicit various physiological effects. However, it remains unclear whether specific IIS components and/or particular targets of these key transcription factors regulate discrete physiological aspects of IIS, including longevity, health span, and development.
In this study, we aimed to uncouple the increased lifespan and decreased fitness, including developmental defects and decreased adult functionality metrics, exhibited by daf-2 mutant C. elegans. We performed a large-scale mutagenesis screen and found that a specific daf-18 missense mutant allele, designated as yh1, fully suppressed slow development and partially rescued the reduced brood size observed in daf-2 mutants with minimal decrease in the extended lifespan. We demonstrated that daf-18(yh1) mutation completely restored the reduced motility observed in young organisms and extended health span as measured by several physiological aspects in daf-2 mutants. Through the analysis of global gene expression profiles, we showed that daf-18(yh1) allele was a weaker allele than the strong loss-of-function daf-18(−). We found that daf-18(yh1) substantially decreased the lipid phosphatase activity of DAF-18/PTEN, while partly maintaining its protein phosphatase activity. Furthermore, we showed that daf-18(yh1) partially retained the activity of the DAF-16/FOXO in daf-2 mutants while preventing the adverse activation of the SKN-1/NRF2 that appears to underlie the reduction in lifespan and health span. These data indicate that the extent of DAF-18/ PTEN activity differentially affects various IIS-regulated physiological processes by calibrating the activities of these key longevity transcription factors. Our findings provide insights into strategies for healthy aging with less undesirable side effects by optimally modulating IIS.

Results
A genetic screen identified mutations that differentially affected pathogen resistance and developmental defects in daf-2/ insulin/IGF-1 receptor mutants. We performed an ethyl methanesulfonate (EMS) mutagenesis screen to identify the suppressors of the constitutive dauer formation phenotype of daf-2(e1370) [daf-2(−)] mutants with minimal effects on resistance to the pathogenic bacteria, Pseudomonas aeruginosa (PA14) (Fig. 1a). We identified three such mutant alleles (yh1, yh2, and yh3), the penetrance of which in suppressing dauer phenotypes was complete in daf-2 mutants at 25°C, but also conferred resistance against PA14 above that in wild-type animals (Supplementary Fig. 1a-d). Upon sequencing, we found that the yh1 allele bore a mutation that altered the evolutionarily conserved cysteine 150 residue in DAF-18/PTEN to tyrosine (Supplementary Fig. 1e-g and Supplementary Table 1). The yh2 and yh3 alleles resulted in premature termination and defective splicing, respectively, in the daf-16/FOXO (Supplementary Fig. 1h and Supplementary Table 1). These data are consistent with previous reports showing that genetic inhibition of daf-18 or daf-16 suppresses various phenotypes in daf-2 mutants 1,4-7, 16,17,19 .
daf-18(yh1) mutation decreases lipid phosphatase activity while retaining protein phosphatase activity of DAF-18/PTEN. We next sought to determine whether yh1 affected the PIP 3 phosphatase activity of DAF-18/PTEN. To this end, we generated transgenic animals that expressed the pleckstrin homology (PH) domain of mouse Akt fused with cyan fluorescent protein (CFP::PH AKT ), which bound plasma membrane-localized PIP 3 ( Supplementary Fig. 4a-e) 26 . We found that both daf-18(−) and daf-18(yh1) increased the levels of plasma membrane-localized CFP::PH AKT in daf-2(−) animals ( Fig. 3a, b). Moreover, we found that the recombinant human PTEN protein harboring C105Y, the orthologous change caused by C. elegans daf-18(yh1), exhibited substantially decreased lipid phosphatase activity ( Fig. 3c-f); the effect was similar to that of the C124S change, which eliminated the lipid phosphatase activity (Fig. 3c-f) 27 . These data suggest that the C to Y change in DAF-18 and PTEN reduces the lipid phosphatase activity.
Because PTEN also acts as a protein tyrosine phosphatase 27,28 , we performed protein phosphatase assays using a synthesized generic peptide harboring phospho-tyrosine. We found that PTEN C105Y retained a substantial protein phosphatase activity (57.3%) compared with wild-type DAF-18/PTEN, whereas the phosphatase-dead PTEN C124S (negative control) dramatically decreased that (Fig. 3g). We confirmed the results by measuring dose-dependent changes of the tyrosine phosphatase activity ( Fig. 3h and Supplementary Fig. 4f). Thus, daf-18(yh1) appears to retain the partial tyrosine phosphatase activity of DAF-18/PTEN, raising the possibility that the protein phosphatase activity of DAF-18 C150Y contributes to longevity and enhanced immunity in daf-2(−) animals.

Discussion
Uncoupling the association between longevity and reduced fitness, including developmental defects and decreased adult functionality metrics, has been a major challenge in the field of aging research. In this report, we identified a specific missense mutation in the daf-18/PTEN that sustained the long lifespan and enhanced immunity conferred in daf-2/insulin/IGF-1 receptor mutant C. elegans, without apparent accompanying defects in development and health span. Notably, our data revealed that a specific mutation in daf-18/PTEN preserved the partial protein phosphatase activity of DAF-18/PTEN and transcriptional activity of DAF-16/FOXO while preventing the harmful activation of transcription factor SKN-1/NRF2, leading to the differential physiological effects. Thus, a proper balance between DAF-16/FOXO and SKN-1/NRF2 activities appears to promote health span in animals with reduced IIS. These data indicate that the modulation of DAF-18/PTEN activity differentially regulates DAF-16/FOXO and SKN-1/NRF2 in reduced IIS and, in turn, uncouples various pleiotropic phenotypes caused by reduced IIS.
Recent studies have drawn controversy over the effects of daf-2 mutations on health span despite the consensus of the effects of these mutations on extreme longevity [9][10][11][12][13] . For example, daf-2 mutations have been reported to increase lifespan by mostly prolonging the unhealthy period of old age 9 , through decreasing gut colonization by dietary bacteria 11 . Contrarily, we previously revealed that daf-2 mutants exhibit extended healthy periods throughout aging by measuring a maximum physical ability 10 and that temporal inhibition of daf-2 enhances immunocompetence in old age 36 . All these studies used genetic inhibition of the daf-2, mutant alleles and RNAi, to measure multiple aspects of health span, without mutations in other genes [9][10][11][12][13] . Here, we aimed to modulate the activity of additional components of IIS to improve the health span in daf-2 mutant worms. Our data demonstrated that the specific change in DAF-18/PTEN caused by yh1 increased the fitness and health span in daf-2 mutants with minimal unfavorable effect on lifespan. Thus, DAF-18/PTEN can be used as a calibrator for achieving healthy longevity.

daf-2(-) daf-18(yh1)
that strong genetic inhibition of DAF-18/PTEN increases SKN-1/ NRF2 activity in C. elegans with reduced IIS, which contributes to shortened lifespan and health span, consistent with the studies on mammals. In addition, recent studies suggest that hyperactivation of SKN-1/NRF2 can impair worm health 46,54 . Thus, we propose that DAF-18/PTEN acts as a negative regulator of SKN-1/NRF2 in C. elegans with reduced IIS and contributes to the modulation of health span.
Mutations in PTEN, a tumor suppressor, underlie the pathology of various human cancers 55,56 . A human mutation that results in PTEN C105Y , which corresponds to DAF-18 C150Y in C. elegans daf-18(yh1), leads to an autosomal dominant syndrome, Bannayan-Riley-Ruvalcaba syndrome 57 . This disorder is characterized by hamartomatous polyps in the intestine and benign subendothelial lipomas 58 . Here, we showed that the daf-18(yh1) mutation reduced the lifespan and health span in wild-type animals ( Supplementary Fig. 2a-c, f, g) but maintained extended lifespan, health span, enhanced stress resistances, and overall fitness in animals with genetically inhibited daf-2 ( Fig. 1 and Fig. 2). These findings raise a possibility that IIS reduction in mammals bearing mutations in PTEN may improve the fitness and/or extend the health span of these animals. It is noteworthy that both activation and repression of NRF2 is implicated in the progression and development of tumors 59 . Our current work also suggests that various alleles in daf-18/PTEN can differentially affect the transcriptional activity of SKN-1/NRF2. In conclusion, it will be necessary to properly modulate the activity of PTEN and NRF to develop therapeutic strategies for treating human cancer patients.
EMS mutagenesis screen. EMS mutagenesis screen was performed as described previously 60 , with modifications. Synchronized L4 larval daf-2(e1370) worms were washed with M9 buffer until residual bacteria were cleaned and then exposed to 47 mM ethyl methanesulfonate (EMS, Sigma, St. Louis, MO, USA) in M9 buffer for 4 hrs at 20°C with rotation. After washing worms three times with M9 buffer, mutagenized P 0 worms were placed on OP50-containing chicken egg plates (see below for details) until the majority of F 1 worms became adults at 20°C. The F 1 adult worms were then bleached for the synchronization of F 2 eggs. With four independent mutagenesis trials, approximately 25,000,000 F 2 eggs were transferred onto the OP50-seeded nematode growth medium (NGM) plates and cultured at 25°C to screen dauer-suppressor mutants. The 269 F 2 dauer-suppressor mutants that were recovered as L4 or young adult animals were picked and directly transferred onto plates completely covered with PA14 (big lawn) for screening PA14-resistant worms. Simultaneously, 100 daf-16(mu86); daf-2(e1370) mutants were used as a control for each of the four mutagenesis screen trials. When all the daf-16(mu86); daf-2(e1370) animals were dead, each of 21 F 2 worms that were alive at that point was singled and moved onto an OP50-seeded NGM plate for obtaining F 3 animals. Among these 21 singled F 2 animals, 18 mutants produced F 3 progeny. Among them, 14 animals contained the same allele, yh1, and the other four alleles were named as yh2 through yh5. The enhanced resistance against PA14 conferred by yh4 and yh5 was not reproduced, and therefore yh1, yh2, and yh3 were used for further characterization.
Preparation of chicken egg plates. Chicken egg plates were prepared for a largescale worm culture as follows. Chicken eggs were rinsed with 100% ethanol (DAE-JUNG, Siheung, South Korea), and separated egg yolks were transferred into a sterile beaker. Sterilized double distilled water (25 ml/egg) was mixed with the yolk by using stirrer. To inactivate lysozymes, the egg yolk mixture was incubated at 60°C for one hr and was subsequently cooled to room temperature. Concentrated E. coli OP50 (50X) was mixed with the egg yolk mixture (1:3 ratio) and diluted with M9 buffer. The OP50-yolk mixture was then seeded on 100 mm NGM plates (2 ml/plate).
Generation of transgenic animals. mCherry::daf-18 transgenic animals generated in a previous report 20 was used to rescue the phenotypes of daf-18(yh1) animals. rpl-28p::CFP::PH AKT -expressing animals were generated in this study as follows. For generating CFP::PH AKT construct, the CFP::PH AKT -expression vectors (CFP-PH AKT ) 63,64 were linearized by NheI restriction enzyme and were treated with Klenow (F. Hoffmann-La Roche, Basel, Switzerland) for blunt end generation. The linearized vectors were then digested with XbaI to obtain CFP::PH AKT DNA fragments. pPD129.57 vectors (L4455, Fire lab C. elegans vector kit) containing a rpl-28 promoter were digested with SmaI and NheI restriction enzymes, and the CFP::PH AKT fragments were inserted into the linearized pPD129.57 using T4 DNA ligase (New England Biolabs, Ipswich, MA, USA). The transgenic strain was generated by injecting the plasmid (25 ng/µl) and a co-injection marker (odr-1p::RFP, 75 ng/µl) into the gonads of day one adults. The extrachromosomal array transgenes were integrated with UV irradiation 65 .
Pathogen resistance assays. Pathogen resistance assays were performed as described previously 66 , with modifications. For small-lawn assays, Pseudomonas aeruginosa PA14 was cultured in LB media overnight at 37°C, and 5 μl of the liquid culture was subsequently seeded on each high-peptone NGM plate (0.35% bactopeptone). For big-lawn assays, 15 μl of overnight-cultured PA14 was seeded onto each high-peptone NGM plate with a glass spreader. The PA14-seeded plates were cultured at 37°C for 24 hrs and moved into a 25°C incubator for 24 hrs before assays. L4-stage worms that were grown on OP50-seeded NGM plates were transferred to PA14 plates containing 50 μM FUDR (5-fluoro-2'-deoxyuridine, Sigma, St. Louis, MO, USA) that prevents progeny from hatching. The assays were performed at 25°C, and the survival of worms was scored at least once a day. The worms were counted as dead if the worms did not respond to prodding. All the assays were conducted at least twice independently. OASIS (https:// sbi.postech.ac.kr/oasis/) and OASIS2 (https://sbi.postech.ac.kr/oasis2/) were used for statistical analysis 67,68 , and p values were calculated using a log-rank (Mantel-Cox method) test.
Stress resistance assays. Stress resistance assays were performed as described previously 20 , with modifications. Gravid adults were allowed to lay eggs for 12 hrs on NGM plates seeded with OP50. For the oxidative stress assay, L4-stage worms were transferred onto 5 μM FUDR-treated NGM plates with E. coli bacteria and 7.5 mM tert-butyl hydroperoxide (t-BOOH, Sigma, St. Louis, MO, USA) solution. For the thermotolerance assay, L4-stage worms were placed in a 35°C incubator. The number of live worms was counted every 2 or 3 hr and recorded as dead when the worms did not respond to tactile stimuli with a platinum wire. All the assays were conducted at least twice independently. OASIS (https://sbi.postech.ac.kr/ oasis/) and OASIS2 (https://sbi.postech.ac.kr/oasis2/) were used for statistical analysis 67,68 , and p values were calculated using a log-rank (Mantel-Cox method) test.
Lifespan assays. Lifespan assays were performed at 20°C or 25°C on NGM plates seeded with OP50 for experiments with mutants or HT115 for experiments using RNAi as described previously 69 , with minor modifications. For lifespan assays with FUDR that prevents progeny from hatching, synchronized young (day 1) adult worms were transferred onto 5 μM FUDR-treated NGM plates with E. coli and moved onto NGM plates freshly treated with FUDR after 24 hrs. For the experiments without FUDR, young (day 1) adults were placed on new plates every 1-2 days until the worms stopped laying eggs. For RNAi experiments, 1 mM isopropyl-β-D-thiogalactoside (IPTG, Gold Biotechnology, St. Louis, MO, USA) was supplemented onto RNAi bacteria-seeded plates, containing 50-100 μg/ml ampicillin (USB, Santa Clara, CA, USA), and incubated at room temperature for 24 hrs before the assays. The number of live or dead worms was scored every 2 or 3 days until all the animals were dead. Worms that ruptured, displayed internal hatching, burrowed, or crawled off the plates were censored but included for statistical analysis. All the assays were conducted at least twice independently. OASIS (https://sbi.postech.ac.kr/oasis/) and OASIS2 (https://sbi.postech.ac.kr/ oasis2/) were used for statistical analysis of lifespan assays 67,68 , and p values were calculated using a log-rank (Mantel-Cox method) test.
Dauer formation assays. Dauer assays were performed as previously described 20 , with minor modifications. Gravid adult worms were allowed to lay eggs on NGM plates at 25°C or 27°C depending on assay conditions and removed after 3-6 hrs for synchronization of eggs. The F 1 progeny were examined for dauer formation after 3 or 4 days at 25°C or 27°C. Dauer formation was visually determined 14 under a dissecting stereomicroscope (SMZ645, Nikon, Tokyo, Japan).
PA14-GFP accumulation assays. Intestinal accumulation of PA14 expressing GFP (PA14-GFP) was measured as previously described 66 , with minor modifications. Big-lawn PA14 plates, for which the whole surface was covered by the bacteria, were prepared by spreading 15 μl of overnight culture of PA14-GFP in LB media containing 50 μg/ml kanamycin (Sigma, St. Louis, MO, USA) onto NGM plates that contained 0.35% peptone. The NGM plates were incubated at 37°C for 24 hrs, and subsequently stored at 25°C for additional 24 hrs before use. L4-stage larvae were infected with PA14-GFP for 36 to 48 hrs. p values were calculated by using chi-squared test.
Microscopy. Fluorescence images of worms were captured by using Axiocam (Zeiss Corporation, Jena, Germany) mounted on a HRc Zeiss Axioscope A.1 (Zeiss Corporation, Jena, Germany) equipped with EC Plan-Neofluar (Zeiss Corporation, Jena, Germany) objective lens. Green fluorescence was detected by using Zeiss filter set 38 Endow GFP shift free emission filter (Zeiss Corporation, Jena, Germany). Animals used for PA14-GFP accumulation assays or CFP::PH AKT localization experiments were placed on 2% agarose pads and were anesthetized with 100 mM sodium azide (DAEJUNG, Siheung, South Korea) before imaging.
Measurement of developmental time. Developmental time was measured as previously described 2 , with minor modifications. Adult worms were washed off NGM plates using M9 buffer, and the remaining eggs were incubated at 20°C for 1-2 hrs. Newly hatched L1-stage worms were transferred onto OP50-seeded plates and were cultured at 20°C. After 40 hrs of synchronization, the numbers of adult and non-adult worms were counted at 20°C. Worms that contained at least one egg in their bodies were considered as adults, and the number of adult worms was counted every 2 hrs. The assay was conducted at least twice independently. Twotailed Student's t test was used for statistical analysis.
Measurement of total brood size. Total brood size was measured as previously described 2 , with minor modifications. A single L4 hermaphrodite was transferred onto an NGM plate seeded with OP50 and maintained at 20°C. Each of the individual worms was transferred onto a freshly OP50-seeded plate every day until the worms stopped laying eggs for 2 days in a row. The number of viable larvae that reached L4 stage descended from a single hermaphrodite was set as the brood size. The brood size measurement was performed five times independently at 20°C, and two-tailed Student's t test was used for statistical analysis.
Measurement of swimming. Swimming rate (body bend in liquid per min) of worms was measured as described previously 9,70 , with minor modifications. Ten worms at indicated ages were transferred into a well in 24-well plates containing 1 ml of M9 buffer. After 1 min for stabilization in a new environment, the body bending of worms in liquid was recorded by using a digital microscope (DIMIS-M, Siwon Optical Technology, Anyang, South Korea). The body bends of the individual worms were counted for 30 sec and converted to the number of bending per min. Dead worms were excluded from the assays, and two-tailed Student's t test was used for statistical analysis for the measurement of swimming at day 0 (L4stage worms) and day 7.
Measurement of moving worms in population. Percentage of moving worms in population was measured as described previously 12,71 , with minor modifications. The movements of age-synchronized worms were categorized as class A, class B, and class C as described previously 71 ; class A animals are healthy and mobile with typical sinusoidal movement, class B animals display mobile but irregular movement and require a prodding stimulus for the movement, and class C worms do not move. The movements of worms were scored every 2-3 days starting from L4 stage until all the worms on the plates were dead. Dead worms were excluded from the assays.
Feeding assays. Feeding (pharyngeal pumping) rate of worms was measured as described previously 70 , with minor modifications. Ten worms at indicated ages grown on OP50 were transferred onto experimental plates containing PA14-GFP or OP50 as indicated. The number of pumping was counted for 30 sec by observing the pharyngeal pumping of a worm under a dissecting microscope, and the measurements were re-scaled to the number of pumping per min. Dead worms were excluded from the assays, and two-tailed Student's t test was used for statistical analysis for the measurement of feeding at day 0 (L4-stage worms).
Quantitative RT-PCR analysis. Quantitative RT-PCR was performed as described previously 69 , with modifications. Synchronized pre-fertile or day 1 adult worms at 20°C were harvested with M9 buffer, and total RNA was extracted using RNAiso Plus (Takara, Shiga, Japan). cDNA templates were synthesized by using ImProm-II Reverse Transcriptase (Promega, Madison, WI, USA) with random primers. cDNA samples were used for quantitative RT-PCR with SYBR green dye (Applied Biosystems, Foster City, CA, USA) by using StepOne Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Data were analyzed by using comparative C T method following the manufacturer's protocol. The average values of ama-1 or pmp-3 mRNA levels were used as a control for normalization, and the average of at least two technical repeats was applied for each biological data point. See Supplementary Dataset 5 for primer details.
Analysis of RNA-sequencing data. mRNA-sequencing analysis was performed as described previously 72 , with modifications. Sequencing pairs were aligned to the C. elegans genome WBcel235 (ce11) and Ensembl transcriptome (release 95) by using STAR (v.2.7.0e). Aligned pairs on genes were quantified by using RSEM (v.1.3.1). Alignment and quantification of RNA-seq data in this study were conducted based on the parameters described in the guidelines of ENCODE long RNA-Seq processing pipeline (https://www.encodeproject.org/pipelines/ENCPL002LPE/). The batch effects of samples were removed by upper-quartile normalization followed by RUVSeq (v1.16.1) with internal control genes. Global expression changes of previously published gene sets in a comparison daf-2(e1370) vs. daf-2(e1370); daf-18(nr2037) or daf-2(e1370) vs. daf-2(e1370); daf-18(yh1) were represented as normalized enrichment scores (NES) by using gene set enrichment analysis (GSEA) (v.3.0) or calculating cumulative fractions with read counts of all expressed genes. Gene sets whose false discovery rate q value < 0.05 in any comparison were regarded as significant in GSEA. The significance of difference in calculating cumulative fractions was computed by using two-tailed paired permutation test using asymptotic approximation. Differentially expressed genes (fold change > 2 and adjusted p value < 0.05) were identified by using DESeq2 (v. 1.22.2). Wald test p values were adjusted for multiple testing using the procedure of Benjamini and Hochberg. Gene ontology terms enriched in genes whose expression changes were greater by daf-18(−) (fold change > 2) than by daf-18(yh1) in daf-2(e1370) mutants were identified by using GOstats (v.2.48.0), and summarized by using Revigo. Subsequently, these genes were compared to the genes expressed in different tissues based on Worm tissue 73 . R (v.3.6.1, http://www.r-project.org) was used for plotting data.
Preparation of recombinant PTEN protein. Recombinant PTEN expression was performed as described previously 74 . Wild-type and mutant PTEN constructs (WT, C105Y, and C124S) were subcloned into pFastBac containing an N-terminal six histidine tag. pFastBac-HTA was digested with EcoRI and HindIII. Using pCMV-FLAG-PTEN (Addgene, Watertown, MA, USA) as a template, a 1.2 kb fragment containing human PTEN was amplified with primers including digestion sites (forward 5′-GCGCCATGGATCCGGAATTCATGACAGCCATCATCAAAGA-3′ and reverse 5′-GTACTTCTCGACAAGCTTTCAGACTTTTGTAATTTGTG-3′). The PCR product was then subcloned into pFastBac-HTA. Recombinant PTEN proteins were expressed in Sf9 cells with the Bac-to-Bac expression system (Invitrogen). Sf9 cells were transfected with recombinant bacmid for 72 hrs, and the cell culture media containing baculoviruses were then harvested. Following the baculovirus infection, the cell pellets were resuspended in buffer A [20 mM Tris-HCl (pH 7.5), 50 mM NaCl, and 1 mM DTT], and underwent the Ni-NTA affinity chromatography for purifying his-tagged PTEN proteins. Protein fractions were further dialyzed in buffer B [25 mM Tris-HCl (pH 7.5), 100 mM NaCl, and 1 mM DTT]. Purified proteins were then separated on a 10% SDS-PAGE gel with Coomassie blue staining, and the proteins that displayed appropriate sizes were confirmed by using western blotting with an anti-His antibody (1:1,000, #2365, Cell signaling technology, Danvers, MA, USA).
In vitro phosphatase assays. In vitro lipid and protein phosphatase assays were performed with a malachite green phosphatase assay kit (K-1500, Echelon Biosciences, Salt Lake City, UT, USA) following the manufacturer's instruction. For lipid phosphatase assays, soluble phosphatidylinositol 3,4,5-trisphosphate diC8 (PIP 3 diC8, Echelon Biosciences, Salt Lake City, UT, USA) was diluted to 1 mM in distilled water. Indicated amounts of purified PTEN recombinant proteins were incubated with 3 μl of the 1 mM PIP 3 solution in 25 μl Tris-buffer [25 mM Tris-HCl (pH 7.4), 140 mM NaCl, 2.7 mM KCl and 10 mM DTT] for 40 min at 37°C. For protein tyrosine phosphatase assays, a synthetic phospho-tyrosine peptide (YEEEEpYEEEE) was used as a substrate 75 . Indicated amounts of recombinant His-PTEN proteins (WT, C105Y, and C124S) and the protein tyrosine phosphatase 1B (PTP1B: positive control, R&D system, Minneapolis, MN, USA) were used as enzymes. The peptide substrates (100 μM) were incubated with each enzyme in 25 μl reaction buffer [50 mM Tris-HCl (pH 7.4), 10 mM DTT] for 60 min at 37°C. One hundred microliters of malachite green solution was then added to terminate the enzyme reaction and incubated for 20 min at room temperature. The released phosphate was quantified by measuring absorption spectrum at 620 nm using a microplate reader. A standard curve derived from 0.1 mM phosphate provided by the assay kit was used to convert the absorbance value at 620 nm to amount of free phosphate. To obtain dose-response curves of PTEN enzymes, the protein concentration was increased stepwise by 2-fold to reach a final 10 nM concentration for the protein phosphatase assay. The dose-dependent activity was not observed for the phosphatase activity-dead variant, PTEN C124S .