Volume 27, Issue 4, Pages 914-925.e5 (April 2018) Impairing L-Threonine Catabolism Promotes Healthspan through Methylglyoxal- Mediated Proteohormesis  Meenakshi Ravichandran, Steffen Priebe, Giovanna Grigolon, Leonid Rozanov, Marco Groth, Beate Laube, Reinhard Guthke, Matthias Platzer, Kim Zarse, Michael Ristow  Cell Metabolism  Volume 27, Issue 4, Pages 914-925.e5 (April 2018) DOI: 10.1016/j.cmet.2018.02.004 Copyright © 2018 Elsevier Inc. Terms and Conditions

Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 1 Impairing gcat/T25B9.1 Expression Extends Lifespan and Promotes Health (A) Threonine is catabolized to glycine and acetyl-CoA by the enzymes threonine dehydrogenase and glycine-C-acetyltransferase (GCAT) via the unstable intermediate 2-amino 3-ketobutyrate. (B–F) (B) Effects of gcat/T25B9.1 RNAi (red) versus control (black) on lifespan (p < 0.0001, log rank test); color coding applies to all subsequent panels and figures. Effect of gcat/T25B9.1 RNAi versus control nematodes regarding (C) aging pigments (∗∗p < 0.005, Student's t test, n = 8 wells × ∼100 worms each) (D), average speed (∗p < 0.01, Student's t test, n = 52), (E) pharyngeal pumping rates (p = 0.51, Student's t test, n = 8 worms × 3 measurements each), and (F) fat content (∗∗p < 0.005, Student's t test, n = 3 worm pellets). Error bars represent the mean ± SD. (G) Lifespan analysis of gcat/T25B9.1 RNAi (p < 0.0001, log rank test) versus control RNAi in daf-2(e1370) mutant nematodes. (H) Lifespan analysis of gcat/T25B9.1 RNAi (p < 0.0001, log rank test) versus control RNAi in daf-16(mu86) mutant nematodes. Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 2 Methylglyoxal Formation Is Necessary for gcat/T25B9.1 RNAi-Mediated Lifespan Extension (A) Lifespan analysis of amx-1/amx-2 double mutant treated with control RNAi and amx-3 (black) RNAi versus amx-3 and gcat/T25B9.1 RNAi (red) (p = 0.87, log rank test), respectively. (B) Relative Amplex Red fluorescence in supernatants of alive nematodes (∗p < 0.05, Student's t test, n = 3 biological replicates from ∼3000 worms each); 100% reflects 1.21 pmol H2O2 in the supernatant per microgram of worm protein. Error bars represent the mean ± SD. (C) Lifespan analysis of N2 Bristol nematodes treated with gcat/T25B9.1 RNAi the presence (blue) or absence of the antioxidant BHA (red) compared with BHA treated worms (gray). (D) Lifespan analysis of N2 Bristol nematodes treated with gcat/T25B9.1 RNAi the presence (blue) or absence of the antioxidant NAC (red) compared with NAC treated worms (gray). (E) HPLC-based measurement of methylglyoxal levels in gcat/T25B9.1 RNAi versus control (∗p < 0.05, Student's t test, n = 5 worm pellets). Error bars represent the mean ± SD. (F) Lifespan analysis of gcat/T25B9.1 RNAi (blue) versus control RNAi (gray) on a glod-4 overexpressing strain (VH725). (G) Lifespan analysis of glod-4 RNAis (from the Ahringer [purple] and ORF [orange] libraries) versus control RNAi (black) in N2 Bristol nematodes (both p > 0.05, log rank tests). (H and I) Lifespan analyses of glod-4 RNAis (from the Ahringer [H] and ORF [I] libraries) combined with gcat/T25B9.1 RNAi (blue) and control RNAi (purple, orange), respectively, in N2 Bristol nematodes (both p < 0.0001, log rank tests). Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 3 MGO Affects Lifespan in a Nonlinear (i.e., Hormetic) Dose-Response Manner (A and B) Effects of MGO supplementation at increasing concentrations (50 μM–10 mM) on N2 Bristol lifespan, depicted by (A) lifespan curves on heat-inactivated E. coli OP50 (log rank tests) and (B) effects on mean lifespan (∗p < 0.05, ∗∗∗p < 0.0005 by log rank tests). Error bars represent the mean ± SEM. (C–F) (C) Average speed (∗p < 0.05, Student's t test, n = 107), (D) aging pigments (p = 0.06, Student's t test, n = 7 wells × ∼100 worms each), (E) pharyngeal pumping rates (p = 0.28, Student's t test, n = 8 worms × 3 measurements) and (F) fat content (p = 0.62, Student's t test, n = 3 worm pellets). (C–F) were performed on heat-inactivated E. coli OP50; error bars represent the mean ± SD. (G) Lifespan analyses of glod-4 RNAi (from the Ahringer library) combined with 100 μM MGO (red) or water (purple), respectively, in N2 Bristol nematodes (log rank test). Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 4 Lifespan Extension of gcat/T25B9.1 RNAi and MGO Treatment Is Mediated by the Transcription Factors SKN-1 and HSF-1 (A) Differentially expressed genes as quantified by deep sequencing analysis upon treatment with gcat/T25B9.1 RNAi; black dots indicate no differential regulation, dark blue and light blue dots indicate regulation according to edgeR and DEseq analyses, respectively. (B) Lifespan analysis of gcat/T25B9.1 RNAi (red) versus control RNAi (black) in skn-1(zu135) mutant nematodes, on alive E. coli HT115 (p = 0.94, log rank test). (C) Lifespan analysis in skn-1(zu135) mutant nematodes without treatment (black) as well as treatment with 50 μM MGO, on heat-inactivated E. coli OP50 (log rank test). (D and E) (D) Expression and nuclear localization of GFP in a gst-4-reporter at control conditions and after exposure to 5 days gcat/T25B9.1 or control RNAi (left set of panels), as well as after exposure to 100 μM MGO (right set of panels) (white size bars reflect 200 μm), and (E) corresponding quantifications (∗∗∗p < 0.0001, χ2 test). (F) Lifespan analysis of gcat/T25B9.1 RNAi versus control RNAi in hsf-1(sy441) mutant nematodes, on alive E. coli HT115 (p = 0.55, log rank test). (G) Lifespan analysis in hsf-1(sy441) mutant nematodes without treatment (black) as well as treatment with 50 μM MGO (red), on heat-inactivated E. coli OP50 (p = 0.09, log rank test). Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 5 Ubiquitin-Proteasome Pathway Is Required for gcat/T25B9.1-Mediated Effects on Healthspan (A–G) (A) qPCR analysis of proteasome subunits upon after exposure to 5 days gcat/T25B9.1 RNAi versus control (∗p < 0.05, xp = 0.10, Student's t test, n = 3 biological replicates each). Error bars represent the mean ± SD. Lifespan analysis in N2 Bristol nematodes the co-presence of gcat/T25B9.1 RNAi versus control RNAi with (B) pbs-3 RNAi (p = 0.35, log rank test), (C) rpn-6.1 RNAi (p = 0.42, log rank test), and (D) pbs-4 RNAi (p = 0.058, log rank test). Lifespan analysis in N2 Bristol nematodes exposed to 100 μM MGO together with (E) pbs-3 RNAi (p = 0.086, log rank test), (F) rpn-6.1 RNAi (p = 0.26, log rank test), and (G) pbs-4 RNAi (p = 0.575, log rank test), or to respective control RNAi, on alive E. coli HT115. (H–J) (H) Chymotrypsin-like, (I) caspase-like, and (J) trypsin-like activities of proteasome in N2 Bristol nematodes treated with gcat/T25B9.1 RNAi versus control (∗p < 0.05, Student's t test, n = 3 worm pellets each). Error bars represent the mean ± SD. (K–M) (K) Chymotrypsin-like (∗p < 0.05), (L) caspase-like (p = 0.63) and (M) trypsin-like (p = 0.09) activities of proteasome in N2 Bristol nematodes treated with 50 μM MGO (for all: Student's t test, n ≥ 4 worm pellets each). Error bars represent the mean ± SD. Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 6 Mechanistic Summary for gcat/T25B9.1- and MGO-Mediated Longevity Impaired expression of gcat/T25B9.1 leads to increased formation of hydrogen peroxide and methylglyoxal in an AMX-dependent manner. Subsequently, and dependent on MGO-induced transcription factors, namely SKN-1 and HSF-1, activation of the ubiquitin-proteasome system mediates extension of lifespan and associated health parameters in nematodes. Cell Metabolism 2018 27, 914-925.e5DOI: (10.1016/j.cmet.2018.02.004) Copyright © 2018 Elsevier Inc. Terms and Conditions