Exogenous ethanol induces a metabolic switch that prolongs the survival of Caenorhabditis elegans dauer larva and enhances its resistance to desiccation

Abstract The dauer larva of Caenorhabditis elegans, destined to survive long periods of food scarcity and harsh environment, does not feed and has a very limited exchange of matter with the exterior. It was assumed that the survival time is determined by internal energy stores. Here, we show that ethanol can provide a potentially unlimited energy source for dauers by inducing a controlled metabolic shift that allows it to be metabolized into carbohydrates, amino acids, and lipids. Dauer larvae provided with ethanol survive much longer and have greater desiccation tolerance. On the cellular level, ethanol prevents the deterioration of mitochondria caused by energy depletion. By modeling the metabolism of dauers of wild‐type and mutant strains with and without ethanol, we suggest that the mitochondrial health and survival of an organism provided with an unlimited source of carbon depends on the balance between energy production and toxic product(s) of lipid metabolism.


20 21
Worm strains and cultivation 22 Wild-type strain used was C. elegans variant Bristol, strain N2. The following single 23 mutant strains were used in this study: DA2579 sodh-1(ok2799) V, CB1370 daf-24 2(e1370) III, CB1372 daf-7(e1372) III, TJ1052 age-1(hx546) II. All C. elegans strains 25 were obtained from the Caenorhabditis Genetics Center (CGC). daf-2(e1370);aak-26 2(gt33) double mutant was generated as described previously (Penkov et al., 2018). 27 During this study, the following strains have been generated: 28 daf-2;aak-2;mitoGFP: Males of daf-2(e1370); zcIs14 ], 29 (dubbed daf-2;mitoGFP) were crossed to hermaphrodites of daf-2(e1370);aak-2(gt33). 30 Males from the progeny were crossed to the mother strain daf-2(e1370);aak-2(gt33). 31 Eggs from the resulting hermaphrodites were grown at 25˚C and developed into dauers. 32 The dauers were left to recover at 15˚C and eGFP-positive worms were selected and 33 singled. The progeny of these worms was selected based on the fluorescent signal. The 34 presence of aak-2(gt33) was tested by PCR. 35 sodh-1;mitoGFP: Males of zcIs14 ] (dubbed mitoGFP) were 36 crossed to hermaphrodites of sodh-1(ok2799). Resulting hermaphrodites produced 37 eggs that were singled and selected for sodh-1(ok2799) via PCR and mitoGFP via 38 fluorescent signal. 39 Worms were routinely cultured on Nematode Growth Medium (NGM) plates 40 seeded with E. Coli NA22 strain (Brenner, 1974). Worms were placed on the plates 41 either as mixed stage populations or as embryos obtained by hypochlorite treatment. 42 Dauers were obtained by 1% SDS treatment of mixed stage populations from 6 Coomassie blue and the spots of interest were cut out. The proteins in these gel slices 116 were extracted and characterized with geLC-MS/MS (Vasilj, Gentzel, Ueberham, 117 Gebhardt, & Shevchenko, 2012). 118 119 MS Western quantification of the metabolic enzymes 120 All reagents were of the analytical grade. LC-MS grade solvents were purchased from 121 Fisher Scientific (Waltham, MA); formic acid (FA) from Merck (Darmstadt,122 Germany), Complete Ultra Protease Inhibitors from Roche (Mannheim, Germany); 123 Trypsin Gold, mass spectrometry grade, from Promega (Madison). Other common 124 chemicals and buffers were from Sigma-Aldrich (Munich, Germany). Protein 125 quantification was performed using Pierce BCA protein assay kit from Thermo 126 Scientific (Rockford, USA). Ampoules of Pierce BSA standard and isotopically 127 labeled 13 C 6 15 N 4 -L-arginine and 13 C 6 -L-lysine were purchased from Thermo Scientific 128 (Rockford, USA) and Silantes (Munich, Germany) respectively. Worms were washed 129 twice with M9 buffer, counted, collected and snap-frozen in liquid nitrogen for later 130 analysis. The frozen worms were thawed on ice and crushed using a micro hand mixer 131 (Carl Roth, Germany). The crude extract was centrifuged for 15 min. at 13,000 rpm at 132 4 o C to remove any tissue debris. The clear supernatant was transferred to a fresh 133 Protein Lo-Bind tube (Eppendorf, Hamburg, Germany). The total protein content of 134 the samples was estimated using BCA assay and 15 µg of total protein content was 135 loaded to a precast 4 to 20% gradient 1-mm thick polyacrylamide mini-gels from 136 Anamed Elektrophorese (Rodau, Germany). Separate gels were run for the BSA and 137 isotopically labeled chimeric protein standard. Undetectable proteins or proteins 138 without detectable unique peptides (e.g., GPD-1, GPD-3, HXK-1, SODH-2 SUCL-1, and SDHD-1) were not included. From the aldehyde dehydrogenase family, only the 140 peptides from ALH-1, ALH-9 and ALH-12 were included for absolute quantification 141 as the contribution of other family members to the total pool of ALHs was very small 142 ( Fig. S3). The gels were processed according to the protocol described in (Kumar et 143 al., 2018) Ethanol labeling was performed by directly incubating wild-type, sodh-1(ok2977), and 158 age-1(hx546) dauers with 10 μCi of [1-14 C-EtOH] (Biotrend, Germany), in 10 ml 159 complete S-medium. After the incubation was completed, worms were washed three 160 times with ddH 2 O and homogenized by five rounds of freezing in liquid nitrogen and 161 thawing in an ultrasound water bath. Organic compounds were extracted from 162 homogenized samples according to a standard method (Bligh & Dyer, 1957). Aqueous 8 fractions were dissolved in 50% CH 3 OH, while organic fractions were dissolved in 164 CHCl 3 :CH 3 OH (1:2, v/v). Total radioactivity in each sample was measured using a 165 scintillation counter. Samples were normalized according to the number of worms, and 166 loaded on glass HPTLC plates (Merck, Darmstadt, Germany) covered with silicate. 167 2D-TLC of aqueous fractions for the visualization of hydrophilic metabolites was done 168 using 1-propanol-methanol-ammonia (32%)-water (28:8:7:7, v/v/v/v) Here a and l represented the concentrations of the acetate and lipid respectively, where 254 and are the forward and backward reaction rates for acetate to lipid conversion 255 (see below). Toxic compound concentration was denoted by c and it is produced from 256 the lipolysis with the rate . The consumption of acetate was also unidirectional with the rate (see below). If there was exogenous ethanol, its presence was included via 258 a constant influx "& of acetate. The wellbeing of mitochondria was represented by a 259 number denoted by , which took values from 1 (completely functional mitochondria) 260 to 0 (mitochondria completely damaged). Mitochondria could be damaged with a rate 261 if the carbohydrate production, , fell below a threshold %"& , or with a rate 262 when the toxic compound c accumulated above a certain threshold ! . in Eq. (4)  263 denotes the Heaviside step function as defined in Eq.(5). 264 While most of the reaction rates for simplicity were taken as constants, there 265 were some rates that depended on chemical variables. One example was the dependence 266 of on m, where we assumed that energy production requires functioning 267 mitochondria and thus used a simple linear relation: 268 where was a constant. Another rate that depended on concentrations was 270 denoting the conversion rate from acetate to lipids. It reflected the fact that there was a 271 certain maximum amount of lipid ' that could be possibly stored in a single worm and 272 thus: 273 Here was a constant determining at which lipid concentrations the conversion started 275 to slow down and was also a constant. Finally, we chose rate to be of the 276

+ ) 278
where and were constants. Overall the above system of equations contains six independent parameters and 280 requires the knowledge of the initial conditions for all four participating components. 281 While measuring most of the involved rates and concentrations is potentially possible, 282 it certainly goes beyond the scope of this manuscript. Instead, we chose the following 283 strategy: We assumed that the model could predict a certain lifespan of wild-type or 284 daf-2 dauer (for simplicity, we do not mathematically distinguish between wild-type 285 and daf-2 because they have similar lifespans

292
To model the pathway in daf-2;aak-2 and age-1 mutants, we modified the rates of the 293 reactions regulated by AAK-2 and AGE-1 as illustrated on Fig. 5A. In the case of daf-294 2;aak-2, the only part in the model that is modified by the aak-2 mutation is an 295 14 type/daf-2 dauer. Regarding age-1 mutants, we exploited the possibility that they may 297 have a reduced acetate to lipid transformation rate, which reduces both lipid 298 accumulation and toxic production. We tested this assumption by choosing of age-299 1 mutant to be one-third of that in the wild-type/daf-2 dauer. To solve the system of 300 differential equations, we used standard numerical integration tools in MATLAB. 301

303
Bligh, E. G., & Dyer, W. J. (1957). A rapid method of total lipid extraction and 304 purification. Can J Biochem Physiol, 37, 911-917. 305 Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics, 77 (1) The plot combines the mean survival rates of wild-type, daf-2, daf-2;aak-2, and age-1 369 dauers treated or untreated with 1 mM ethanol that are displayed in Fig. 3D and F Survival rates of wild-type dauers treated with ethanol, amino acids, and vitamins. 385 Means ± SD of two-three biological replicates. 386 387 Figure S7. Mathematical modeling of survival rates in daf-2;aak-2 and age-1 388 mutants. Co-plotted are several other trends: "Acetate" is the combined entity 389 representing the free acetic acid and the acetyl-CoA produced in the pathway. "Lipid" represents the bulk complex lipids, mainly TAGs, derived from the fatty acid 391 component. "Toxic" comprises the putative fatty acid-derived toxic compound(s). 392 "Mito" is defined by the degree of activity of the mitochondria and serves as a proxy to 393 the survival rate.  Peak area (a.u.)

Figure S3
Normalized protein abundance