An aqueous extract of the brown alga Eisenia bicyclis extends lifespan in a sex-specific manner by interfering with the Tor-FoxO axis

Eisenia bicylis extracts extend lifespan of Drosophila melanogaster in a sex-dependent manner

An aqueous extract of the brown alga Arame (E. bicyclis) was identified during a larger screen of plant and algal extracts for their life extending potential using fruit flies as a model system [22, 26]. To further characterize and to verify this lifespan prolonging effect, we used two different concentrations (0.1% and 0.05%) of this extract and measured lifespan in cohorts of mated females of the Drosophila melanogaster w¹¹¹⁸ strain. Animals subjected to 0.05 % of the E. bicyclis extract (EBE) showed a lifespan prolongation (Figure 1A), with an increase in median lifespan of about 40 % (p<0.0001). Application of the extract at a higher concentration (0.1%) showed the absence of adverse effects and that this higher concentration increased the median lifespan to a very similar extent (39.5 %, p<0.0001). The increases in lifespan were not only seen for the median lifespan, but also for the maximal lifespans (10% of animals with highest lifespans) under both conditions (56 d to 63 d and 51 d to 65 d, respectively, p<0.0001 each, see also Supplementary Table 1). We also tested cohorts of mated males, but we did not see any lifespan-extending effect by application of the Eisenia extract (Figure 1C). To exclude that the observed effects on lifespan are only strain-specific, we tested a second Drosophila strain, namely y¹w¹¹¹⁸. Here, we could also show a robust lifespan prolongation induced by the EBE (Figure 1C; p<0.0001). Again, also the maximum lifespans were increased (45d to 48.5 d, p=0.0006).

Figure 1. Application of Eisenia bicyclis extract (EBE) enhances lifespan in female Drosophila. Lifelong application of 0.05% EBE (red) to adult female Drosophila (w¹¹¹⁸) was compared to control flies of the same genotype (blue). The proportion of surviving animals is displayed against time (A). In (B), 0.1 % EBE was used under otherwise identical conditions. A similar analysis was performed with w¹¹¹⁸ males and 0.05% EBE was applied (red) and compared with controls (blue) (C). Female flies of the yw strain were confronted with 0.05 EBE (red and compared to matching controls (blue) (D). (n > 100 per condition). Statistical analyses were done using a log-rank test. Ns means not significant, *** means p< 0.005, **** means p<0.001.

To exclude indirect effects caused by reduced uptake and therewith leading to an induced caloric restriction with its lifespan prolonging effects, we quantified the nutritional intake using the consumption-excretion assay. Here, no statistically significant differences in food consumption during a 24 h period could be observed (Figure 2A, p=0.3015). Moreover, we treated the flies with 0.05 % EBE and analyzed the body composition. Therefore, we analyzed parameters such as body weight, protein content, triacylglycerol levels, and glucose content. Regarding the body weight, the experimental groups did not show any difference (p=0.1606) (Figure 2B). We examined lipid storage and carbohydrate content by measuring triglyceride (TAG) and glucose levels, respectively and have related that to body weight, since this is the only independent reference variable. TAG levels and glucose content of whole flies fed with the extract also showed no significant changes (p=0.1023 and p =0.6134 respectively) (Figure 2C, 2D). Only with respect to the body protein content of EBE-treated flies, a significant but mild reduction compared to the control group was observed (p=0.0033) (Figure 2E). To demonstrate that the observed lifespan extension does not compromise other measures of health, a fecundity test was performed. It revealed that algal treated flies produced similar amounts of eggs as the control group (p=0.1051) (Figure 2F).

Figure 2. Effects of Eisenia bicyclis extract of physiological parameters of female Drosophila. Application of EBE to female adult flies did not change the intake of nutrients over a 24 h period (A). The body weight was also not changed in response to EBE (B). The effects on body fat (C), on body glucose (D) and body protein (E) were quantified. Fecundity was quantified under both conditions (F). N≥10, mean values ± S.E.M. are given. Statistical analyses were performed with unpaired t-tests. Ns means not significant, ** means p< 0.01.

Phenotypic analysis of the effects of EBE treatment

We analyzed the effect of EBE intake on additional phenotypic characteristics. In doing so, we subjected the influences on the activity behavior to a special analysis (Figure 3). The general activity pattern hardly differs between control and EBE-treated animals (Figure 3A), the same is true for the cumulative activity over a 24 h period (Figure 3B). We also analyzed sleep amounts and sleep patterns, because sleep duration has been associated with lifespan [27]. A more detailed analysis of sleep behavior revealed slightly, although not significantly, increased sleep episodes after EBE treatment over a 24 h period (Figure 3C). However, further breakdown into daytime sleep activity (Figure 3D) and nighttime sleep activity (Figure 3E) revealed that the amount of nighttime sleep was significantly increased.

Figure 3. Influence of Eisenia bicyclis extracts on activity and sleep of adult female flies. Activity traces derived from a DAM-based analysis (A) of mated female flies in a 24h period. The total mean activity in a 24h period (B), the accumulated total sleep time (C), and the accumulated daytime sleep time (D) and the accumulated nighttime sleep time (E). N≥10, mean values ± S.E.M. are given (B–E). Average sleep amounts are stated as a fraction of 1. Statistical analyses were performed with unpaired t-tests. Ns means not significant, * means p< 0.05.

E. bicyclis extracts improve resistance to some stress conditions

We next analyzed if addition of the EBE changed the lifespan in response to stressful conditions. We first assessed resistance to starvation by placing control and EBE-treated flies on agar only, to prevent energy intake, while sufficient water is supplied. Under these conditions, the median lifespan was 72 h and 70 h for the EBE treated and the control group, respectively. However, the difference was not statistically significant (p=0.1997) (Figure 4A and Supplementary Table 1). In contrast, adding EBE to the flies’ food increased the resistance to desiccation. The median lifespan for the control group was 24 h, whereas this increased to 25 h in EBE-treated flies (p=0.0027) (Figure 4B). Subsequently, we tested the effects of the major nutritional stressors, namely high-fat, and high sugar diets. Flies treated with 0.05 % EBE had a slight, but significant lifespan extension on a high-fat diet (HFD). The median lifespans were 29 and 30 days for the control group and the EBE treated groups, respectively (p=0.0140) (Figure 4C). Moreover, we reared flies on food containing 30 % sugar and asked whether the administration of EBE reduced the deleterious effects of the high sugar diet. Here, we found that the EBE addition led to an increase in median lifespan by about 30% from 35 days under control conditions to 46 days in response to EBE application (Figure 4D; p-value <0.0001). The maximum lifespans also showed a significant increase (from 46 in controls to 50 after EBE, p<0.0001).

Figure 4. Influence of Eisenia bicyclis extracts of lifespan in response to different stressors. Lifespan of w¹¹¹⁸ females in response to starvation (A) and desiccation (B) in control flies and those subjected to 0.05 % EBE. Lifespan of W1118 females on a high-fat diet (C) and a high sugar diet (D) (n > 100 per condition). Statistical analyses were done using a log-rank test. ns means not significant, * means p<0.05, ** means p< 0.01, **** means p<0.001.

Due to the effects of EBE on HSD mentioned earlier, we quantified the body composition of flies reared on a HSD (Figure 5). The body weight, total TAG levels and whole-body protein did not change under these conditions (Supplementary Figure 1). Only the body glucose level in glucose level of EBE treated flies was significantly lower if compared to the control group (p= 0.0443) (Supplementary Figure 1).

Figure 5. Mode of action of the Eisenia bicyclis extract induced lifespan prolongation. The alpha-Amylase activity of control flies and EBE treated ones (A) are displayed. Lifespan analyses of control and EBE-treated female flies of the genotypes: dSir2-deficient (B), TOR-deficient (C), and FoxO-deficient (D). Western blot analysis of samples from control flies (Co) and EBE-treated ones (EBE) of puromycin treated flies (E) and the quantification of bad intensities (F). lifespan did not increase on concentrated medium containing 0.05 % EBE, (n > 100 per condition). Protein synthesis of flies treated with 0.05 % EBE was not affected (E, F). N≥10, mean values ± S.E.M. are given (A, F). Statistical analyses were performed with unpaired t-tests. ns means not significant. (B–D, n > 100 per condition). Statistical analyses were done using a log-rank test. ns means not significant, **** means p<0.001.

Mode of action of the lifespan prolonging effects of EBE

To elucidate potential mechanisms by which the extract extends the lifespan of flies, we tested the alpha-amylase activity in flies in response to EBE treatment as this has been shown to be modulated by Eisenia-derived phlorotannins [28]. Here, no differences in the alpha-amylase activities between control animals and those subjected to EBE were seen (Figure 5A). We then assessed which signaling pathways are responsible for the lifespan-prolonging effects of an EBE treatment. We first evaluated Sir2-dependent processes since activation of the Sir2 pathway was clearly identified as life-prolonging in Drosophila [29]. By analyzing sir2-deficient flies, we could show that Sir2 is obviously not the relevant EBE target, as the corresponding flies nevertheless showed EBE-induced lifespan extension (Figure 5B). Nevertheless, the reduced increase in lifespan could also be interpreted in a way that Sir2 participates in the life-prolonging effect. Furthermore, we analyzed the Tor signaling pathway and tested Tor-deficient animals for this purpose. Here, we could show that EBE application did not show any lifespan-prolonging effect in these animals, proving that the Tor signaling pathway is closely linked to the EBE-induced effect (Figure 5C). Since FoxO signaling is closely linked to lifespan on the one hand and can act downstream of Tor on the other [30], we also checked this factor and analyzed foxo-deficient animals. Again, we could not detect any EBE-induced positive effect on lifespan (Figure 5D). This means that FoxO is also required for EBE-mediated effects on lifespan. Since Tor-signaling is also more important for proteostasis, we checked protein synthesis using the puromycin assay (Figure 5E, 5F). Here, we could not determine any effect of EBE treatment on new protein formation (Figure 5E, 5F).

Qualitative HPLC-MS analyses of aqueous extracts

We performed a comparative HPLC-MS study to characterize the algae material and find the very first clues about the possible substances responsible for the life-prolonging effects. For this purpose, we compared aqueous extracts of three marine plants/algae with different properties to focus especially on those substances that are specifically found in EBE. The extracts in question were those of Eisenia bicyclis, of the brown alga Saccorhiza polyschides, and of the picklegrass (Salicornia spec.). The EBE is in the center of the current study and the Saccorhiza polyschides extract has been characterized in detail to exert life- and/or health-span extension in flies and mice [22, 23]. The Salicornia extract on the other hand is of marine origin, but not an algae and serves as a control. The extracts were separated by chromatography on C18, compounds detected by Time-of-Flight mass spectrometry (ToF-MS) and identified based on library search (cf. methods and Supplementary Material). The MS analysis gave relative concentration differences between the samples. Principal component analysis (PCA) showed that the compositions of the extracts were clearly distinguishable (Figure 6A). By performing a partial least-squares discriminant analysis (PLS-DA) (Supplementary Figure 2), the top 100 features which are higher abundant in Eisenia bicyclis samples could be determined and sorted by their importance (Supplementary Table 5). Moreover, we show differentially present metabolites as a heatmap (Supplementary Figure 3). Of these 100 selected metabolites with differential occurrence in the different samples, we focused on those metabolites that are specifically found in EBE and for which information on life-prolonging potential is already available. In particular, 2 substances stood out, 7-phloroeckol (Figure 6B) and caffeic acid quinone (Figure 6C). In the case of 7-phloroeckol, however, a differentiation to the isoflavonoid C-glyoside UNPD147636 is not possible due to identical mass numbers. It should be noted, however, that 7-phloroeckol is one of the "lead substances" of the brown alga of the genus Eisenia, whereas isoflavonoid C-glyosides are hardly known in algae.

Figure 6. Untargeted HPLC-MS analysis of algae extracts. Principal component analysis (PCA) (A) of algae extracts measured using the ESI+ and ESI- acquisition mode. Underlying peak areas are log10 transformed and pareto-scaled. (B, C) LC-ToF-MS intensities of 7-phloroeckol (B) and caffeic acid quinone (C). Intensities are displayed as the mean of triplicate injections (duplicate injections for Saccorhiza).

Algal extraction

Eisenia bicyclis extract was prepared according the method described by Onur et al. (Onur et al., 2013). In short, the dried algal material was grinded with an analytical mill (IKA Type A11 basic). 3 grams of grinded alga were transferred into a test tube containing 30 ml of boiling double distilled water. After slight stirring, the suspension was sonicated for 1 min and centrifuged for 2 min at 2000 g. The supernatant was filtered, and the soluble extract was freeze-dried and stored frozen until use.

Fly husbandry

Wild type adult flies were kept as previously described [65]. In brief, they were cultivated on a diet containing 5 % (w/v) yeast extract, 8.6 % (w/v) corn meal, 5 % (w/v) sucrose, and 1 % (w/v) agar-agar supplemented with 1 % (v/v) propionic acid and 3 % (v/v) nipagin named here Drosophila medium. Adults, 3–5-days after hatching were used in the experiments. After this initial period of 3-5 days where the flies had the chance to mate, they were separated according to their sex. All experiments were performed at 25° C, a light-dark cycle of 12h:12h, and 60 % humidity. For most experiments, the wild type strain w¹¹¹⁸ was used. The following fly strains were used for the study: w¹¹¹⁸ (Bloomington stock ID #5905), Sir2-deficient (Bloomington stock ID #8838), Tor-deficient (Bloomington stock ID #11218), dfoxo-deficient (foxo^21/21) (Marc Tatar lab), y¹w¹¹¹⁸ (Bloomington stock ID #111281).

Lifespan assay and analysis

The mated female flies were separated into experimental groups and kept on the food described above. Vials were regularly changed every 2 days and dead flies were counted every day. For the high sugar diet, the sucrose concentration was increased to 30 % and for the high-fat diet, the concentrated medium was supplemented with 20 % palm fat. To supplement the food with the algal extract, the aqueous extract was added on top of the food in the experimental vials at the final concentration of interest. For this reason, the concentration is calculated to correspond to the concentration in the entire medium. Then, the water is allowed to evaporate to dry the food. A total of at least 100 animals were used at minimum per experiment, divided into 5 aliquots with at least 20 animals each, which is well in the range used for lifespan assays [66].

Body fat quantification

The whole-body triacylglycerol (TAG) content was measured using coupled colorimetric assay [67] as previously described [68]. In short, samples were collected (5 females per sample) and weighted. 1 ml PBS/Tween-20 (0.05 %) was added to the samples, and they were homogenized in a bead ruptor apparatus (BioLab Products, Bebensee, Germany) for 2 min at 3.25 m/s. Next, samples were centrifuged for 3 min at 3000 g and the supernatant was transferred to new tubes. The supernatant was heat-inactivated at 70° C for 10 min and centrifuged for 3 min at 2500 g. 50 μl of each sample was added to a 96-well plate and the absorbance was measured at 500 nm (T0). 200 μl prewarmed triglyceride reagent was added to each well and incubated for 30 min at 37° C with mild shaking. The absorbance was again measured at 500 nm (T1). The TAG concentration was determined by subtracting T0 from T1. The TAG content was quantified using a triolein-equivalent standard curve.

Glucose measurement and protein content analysis

The body glucose levels were determined using a Glucose (HK) Assay Kit (GAHK-20, Sigma-Aldrich, Taufkirchen, Germany) according to the manufacturer’s instruction with minor modifications. Samples were collected (5 female flies per sample), weighted, and homogenized using a bead ruptor apparatus (BioLab Products, Bebensee, Germany) for 2 min at 3.25 m/s. For the glucose measurement, the supernatants were heated for 10 min at 70° C and then centrifuged for 3 min at maximum speed at 4° C. 30 μl of the supernatant was added to a well of a 96-well plate. 100 μl of the HK reagent was added to each well and the plate was incubated at room temperature for 15 min. Then, the absorbance was measured at 340 nm. The glucose content was calculated using a glucose standard curve. For the determination of the protein content, the samples were centrifuged for 1 min at 1000 rpm at 4C. The supernatant was transferred to a new 1.5 ml tube and centrifuged again at 6000 rpm, at 4° C for 10 min. The supernatant was again transferred to new tubes and centrifuged at maximum speed, at 4° C for 10 min. The protein content was measured using the Pierce BCA Protein Assay Kit according to the manufacturer’s instructions.

Starvation, desiccation resistance, and fecundity assay

To assess flies’ resistance to starvation, flies transferred to vials containing 1 % agar after feeding with food of interest. The survivorship of flies was recorded every two h at 25° C and 60 % humidity and 12-h light-dark cycle. To determine the resistance to desiccation, flies were transferred to empty vials after feeding them with the food of interest. The flies were kept at 25° C and 60 % humidity and 12 h light-dark cycle. The survival of flies was monitored every 1 to 2 h. The mated female flies were kept in vials containing the food of interest and the laid eggs were counted every day for 7 days.

Activity and sleep analysis

Activity and sleep were monitored using the Drosophila Activity Monitor (DAM2) system and analyzed by ShinyR [69]. The mated female flies were allowed to feed on the food of interest for 2 weeks and then were placed into tubes containing Drosophila medium. The tubes were loaded into the DAM system and maintained at constant conditions (25° C, 60 % humidity, and 12/12- h light-dark cycle). The first two days were spared for acclimatization and the data were collected on the third day. Sleep measurements were done according as described [69] with a period of 5 min without activity interpreted as a sleep phase. Data were recorded every 60 s.

Food intake

The food consumption of flies was measured using the consumption-excretion method by Shell et al. [70]. First the 0.5 % (w/v) blue dyed (Brilliant Blue FCF food dye; E133) was prepared. The Drosophila medium or blue dyed Drosophila medium (0.5% (w/v) Brilliant Blue FCF food dye; E133) was dispensed into caps of 2 ml screw cap vials. For adaptation, individual flies were transferred to 2 ml screw cap vials with Drosophila medium. After a few h feeding on the concentrated medium the flies were transferred to 2 ml vials with blue dyed food. After 24 h feeding on blue dyed food three ceramic beads and 500 μl H₂O were added to the vials. The samples were essentially treated and measured as described earlier [71].

Puromycin assay and Western blotting

Young adult flies were starved overnight and transferred to a vial containing concentrated medium and puromycin (10 μg/ml). After 1 h treatment with the food, the flies were homogenized in 200 μl RIPA buffer using a bead ruptor apparatus (BioLab Products, Bebensee, Germany) for 2 min at 3.25 m/s. The protein concentration was quantified using BCA Protein Assay Kit (Thermofisher, Karlsruhe, Germany). 40 μg of protein was loaded on SDS-PAGE gel and the protein were resolved through electrophoresis. The proteins were transferred to a PVDF membrane using semi-dry transfer for 1 h. To block unspecific binding, the membrane was treated with 5 % non-fat dry milk powder in TBST for 1.30 h. The membrane was incubated with anti-puromycin (3RH11) antibody (Kerafast, Boston, USA, EQ0001) over night at 4° C. After 3 times washing with TBST, HRP-coupled anti-rabbit secondary antibody was used. The bands were visualized using Clarity™ Western ECL substrate and the ChemiDocImager (BioRad, Feldkirchen, Germany).

Alpha-amylase activity

The alpha-amylase activity was determined using a high sensitivity alpha-amylase kit (K-AMYLSD, Megazyme Ltd., Bray, Ireland) according to the manufacturer’s instruction with minor modifications. 5 female flies per treatment group were homogenized in 200 μl PBS. 10 μl of the supernatants were added to 90 μl of amylase solution (65 μl amylase SD reagent mixed with 25 μl buffer). The assays were incubated for 30 min at 37° C, before adding 100 μl of stop reagent. For each sample, a blank (t = 0 min) was generated, where the stop solution was added prior to the amylase reagent. Absorbance of samples and corresponding blanks was read at 405 nm in a microplate reader. 4-5 samples per treatment group were measured in duplicate in two independent determinations. Amylase activities were calculated with the Mega-Calc™ Data Calculator sheet. One unit of amylase activity is defined as the amount of enzyme required to release one micromole of p-nitrophenol from 4,6-O-ethylidene-α-4-nitrophenyl-maltoheptaoside per min under the defined assay conditions.

Untargeted data acquisition and HPLC-MS analyses

The untargeted data acquisition was performed using the TripleToF 6600 mass spectrometer (Sciex, Darmstadt, Germany) coupled to a Exion LC system (Sciex) operating in the ESI negative (ESI-) and ESI positive (ESI+) ionization mode as described earlier [72, 73]. Algae samples, QC and blank samples were analyzed using the data-independent acquisition (DIA) mode SWATH, by initially performing a ToF-MS scan with an accumulation time of 200 ms (m/z 100 - 1200), followed by 16 fixed Q1 isolation windows with an accumulation time of 81 ms in the m/z range from 100 to 1200 with 1 Da overlap. Following Q1 isolation windows were applied: m/z 169 – 240, 239 – 310, 309 – 380, 379 – 450, 449 – 520, 519 – 590, 589 – 660, 659 – 730, 729 – 800, 799 – 870, 869 – 940, 939 – 1010, 1009 – 1080, 1079 – 1150 and 1149 – 1200. The chromatographic separation was achieved on a C18 column with guard column of the same type (Kinetex C18, 100×2.1 mm, 1.7 μm, Phenomenex, Aschaffenburg, Germany) at an oven temperature of 40° C and a flow rate of 0.3 ml/min. The injection volume was 5 μl, the sample concentration was 10 mg extract per ml in water. Gradient elution with 0.1% formic acid (mobile phase A) and 0.1% formic acid in acetonitrile (mobile phase B) was as follows: 0 min 5% B, 0.5 min 5% B, 12 min 50% B, 13 min 100% B, 16.5 min 100% B, 17 min 5% B and 20 min 5% B. Samples were injected in triplicates in a randomized order with QC injections every eight to tenth injection. Blanks (solvent) were injected regularly, and the ToF calibration for instrument accuracy was performed after every fifth sample. For data acquisition and instrumental control, AnalystTF software (version 1.7.1, Sciex) was used.

Peak picking and feature alignment of acquired data in the ESI^- and ESI⁺ mode was performed using MS-DIAL (version 4.70) [74] (Supplementary Table 2). The feature alignment files for each ionization mode were then exported and further processed for feature reduction and annotation using MS-CleanR [75] (Supplementary Table 3) and MS-Finder (version 3.52) [76, 77] (Supplementary Table 4). For a first feature annotation process, public MS/MS libraries (Supplementary Table 4) were applied to obtain a broad overview of compounds. In a second annotation process, the seaweed specific Seaweed Metabolite Database (SWMD) was applied [78]. To make this library accessible for the MS-Finder annotation, the files storing the structural information (.mol) were downloaded from the SWMD homepage (https://www.swmd.co.in/) and converted to InChI Key, SMILES, formula, and exact mass using the Excel plugin from ChemOffice (version 20.0, PerkinElmer, Waltham, U.S.A). The structural information of the SWMD were further supplemented by compounds known to occur in brown algae (Supplementary Table 4).

After completion of the feature reduction and annotation process, an ESI⁺/ESI^- merged and annotated feature list was exported, which was further statistically analyzed using R. Therefore, NA values were first removed by inserting a half of the minimum observed intensity per feature for all features, then the dataset was log10 transformed and pareto-scaled. For a first overview a principal component analysis (PCA) was performed using the R package mixomics [79] (Figure 6A). To gather further insight, which specific features are responsible for the separation of the algae samples, a partial least-squares discriminant analysis (PLS-DA) was performed using the mixomics package (Figure 6B). A special focus was set on features higher abundant in the Eisenia samples compared to the Salicornia and Saccorhiza samples. Therefore, Salicornia and Saccorhiza samples were put together in one group and compared to Eisenia samples. The PLS-DA was then calculated using two components, while the background was calculated using the maximum distance function and a resolution of 500. The loadings were subsequently exported and sorted by increasing values of component 1 to select only the first 100 features, which are higher concentrated in Eisenia compared to Salicornia/Saccorhiza samples. Based on the two annotation processes using MS-CleanR/MS-Finder, features are shown with up to two annotations (Supplementary Table 5).

Statistical analysis

Statistical analyses were performed with the Prism 9 work package. For analyses of lifespan data, a log-rank test was used (Mantel-Cox or Gehan-Breslow-Wilcoxon). At least 100 individuals were analyzed per treatment, usually in 5 replicates with at least 20 individuals each. Parametric data were analyzed with unpaired Student’s t-test and non-parametric data with Mann–Whitney test (usually with N≥10).