Dried CA plant material was purchased from international vendors (Organic India) and supplied for the project by Oregon’s Wild Harvest (Redmond, Oregon, United States). A voucher sample of the material (designated “BEN-CA6”) is maintained at Oregon Health & Science University (OHSU), and a second voucher sample (OSC-V-265416) has been deposited at the Herbarium at Oregon State University (Corvallis, OR). BEN-CA6 was authenticated as CA by DNA barcoding. DNA was extracted from the sample using the DNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Primers to amplify the ITS genomic region were designed using Primer BLAST using already published (Ye et al., 2012) Centella ssp ITS sequences. Designed primers consisted of a 5′ adapter sequence (pJet) facilitating direct sequencing with universal pJet primers of single PCR products. Primers (Table 1) were synthesized by Integrated DNA Technologies (Coralville). Extracted DNA was then diluted 1:10 and the ITS region amplified with the designed primers. Amplification consisted of two rounds of PCR. The first PCR consisted of a 25-μL reaction mixture containing 2 μL of a 1:10 DNA dilution, 1 × PCR reaction buffer, 0.2 mM dNTP mixture, 0.2 μM of each forward and reverse primers CentITSF and CentITSR, 1.5mM MgCl2, and 1 U of Platinum Taq DNA Polymerase (Invitrogen). The first PCR program consisted of an initial denaturation step at 94°C for 3 min, followed by 10 cycles at 94°C for 80 s, 58°C for 25 s and 72°C for 1 min. The second PCR consisted of a 25-μL reaction mixture as mentioned above using 1 μL of the first PCR as a template and pJet forward/reverse primers. The second PCR program consisted of 35 cycles with denaturation at 94°C for 30 s, 65°C for 20 s, and 72°C for 1min, with a final extension at 72°C for 3 min. After amplification, 10 µL were analyzed by electrophoresis on 1.5% borate agarose gel and visualized under UV light. PCR products were compared to the molecular size standard 1 kb plus DNA ladder (Invitrogen). Resulting PCR products were cloned into Vector pCR4-TOPO TA (Invitrogen) and transferred into E. coli according to the manufacturer’s instructions. Eight transformants per cloning event were subjected to colony PCR using M13F and M13R primers. The PCR products of four transformants per DNA were sequenced with M13 forward and reverse primers at Genewiz/Azenta. Derived sequences were trimmed and subjected to homology search against the NCBI nucleotide database using BLAST available in Geneious Prime 2023.2.1 (Biomatters. LTD) [ (Altschul et al., 1990)]. Additionally, a neighbor-joining consensus tree was built with the Tamura-Nei Distance model without an outgroup, bootstrap 500 replicates, available in Geneious Prime 2023.2.1.

The preparation of the CA water extract used in this study (BEN-CAW-7) has been previously described in detail (Holvoet et al., 2023a; Yang et al., 2023a). BEN-CA-6 plant material (4 kg) was extracted by boiling in deionized water (50 L) for 90 min. The extract was filtered, frozen and lyophilized in three batches to yield dried extracts BEN-CAW-7, 8 and 9 with a total extract weight of 820g (20.5% of the starting dried plant material). Voucher samples of BEN-CAW 7, 8 and 9 are stored at OHSU. CAW-7 was used in the present study and will be referred to as “CAW” in the rest of this article.

Four purified triterpenes (TTs) and eight caffeoylquinic acids (CQAs) known to occur in CAW (RdS et al., 2022) were purchased from commercial sources and their identity verified by us as previously described (Holvoet et al., 2023b). The content of these compounds in the CAW batch used in this study was determined by liquid chromatography coupled to multiple reaction monitoring mass spectrometry (LC-MRM-MS) as previously reported by us (Holvoet et al., 2023b; Yang et al., 2023b). Content (% w/w) of the selected TT and CQA compounds in CAW were asiatic acid, 0.042%; madecassic acid, 0.075%; asiaticoside, 1.475%; madecassoside 3.589%; chlorogenic acid (3-CQA), 0.750%, cryptochlorogenic acid (4-CQA), 0.298%; neochlorogenic acid (5 CQA), 0.339%; 1,3-diCQA, 0.258%; 1,5- diCQA, 0.389%; Isochlorogenic acid A (3,5- diCQA), 0.177%; Isochlorogenic acid B (3,4- diCQA), 0.229%, and Isochlorogenic acid C (4,5- diCQA), 0.195%. These values were used to make up solutions of individual compounds or mixtures equivalent to their content in the concentrations of CAW tested in primary neuron and Drosophila experiments.

Mice were housed in an AALAC certified facility and maintained in a climate-controlled environment with a 12-h light/12-h dark cycle and fed a Pico Lab Rodent Diet 5LOD (LabDiets, St. Louis, MO, United States). Diet and water were supplied ad libitum. All procedures were conducted in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals and were approved by the institutional Animal Care and Use Committee of OHSU.

C57BL6 breeding pairs were acquired from Jackson Laboratories. Hippocampal neurons were isolated from C56BL6 embryonic mice, based on the methods of (Kaech and Banker, 2006). Embryos were harvested at 18 days of gestation from anesthetized females. Hippocampi were dissected, gently minced, trypsinized, and triturated to generate suspensions of dispersed neurons. Neurons were plated at a density of 130,000 cells in 60 mm dishes in MEM medium (GIBCO/Life Technologies, Waltham, MA, United States), 5% FBS (Atlanta Biologicals, Flowery Branch, GA, United States), 1× Anti-Anti (GIBCO/Life Technologies) and 0.6% glucose (Sigma-Aldrich, St. Louis, MO, United States), each dish containing 3 poly-l-lysine-coated nitric acid treated glass coverslips with paraffin wax spacers. After 3 h, the coverslips were flipped into 60 mm dishes containing mouse neural stem cell-derived glial cells (provided by Dr. Gary Banker, Jungers Center, OHSU) and maintained in 6 mL Neurobasal Medium supplemented with 1× GlutaMAX (GIBCO/Life Technologies), 1× Anti-Anti (GIBCO/Life Technologies) and 1× GS21 neural supplement (ThermoFisher, Waltham, MA, United States). Dishes were fed every week by removing 1 mL of the culture medium and adding 1 mL fresh Neurobasal medium that included GlutaMAX, Anti-Anti and Neuronal Culture Medium Supplement, with the first feed at 5 days in vitro (DIV) containing 6 μM (1 μM final) cytosine β-D-arabinofuranoside hydrochloride (AraC; Sigma-Aldrich).

After 14 days in vitro (DIV) cells were treated with either CAW (50 μg/mL), TT mix (containing asiatic acid, madecassic acid, asiaticoside and madecassoside), CQA mix (containing 3-CQA, 4-CQA, 5-CQA, 1,3-diCQA, 1,5-diCQA, isochlorogenic acid A, isochlorogenic acid B and isochlorogenic acid C), TT + CQA mix, individual TTs or individual CQAs. The vehicle used to resuspend the compounds was 25% methanol and treatment stocks were prepared at 100x so the final concentration of methanol in the cell culture media was 0.25%. The concentrations of TTs and CQAs tested individually or in mixtures were equivalent to their relative abundance in CAW (50 μg/mL). Results were compared to vehicle-treated neurons.

At 19 DIV coverslips were fixed in 4% paraformaldehyde and stained with Anti-MAP2B (Sigma-Aldrich #M4403; 3.3 μg/mL) and Goat anti-mouse IgG1-Cy3 (Jackson ImmunoResearch #115-165-205; 1.5 μg/mL). Immunostained neurons were imaged with a Zeiss ApoTome2 microscope and blinded Sholl analyses were performed to assess dendritic complexity using the Fiji platform. Analysis consisted of quantifying the number of intersections of dendritic processes with concentric circles drawn at 10 μm intervals from the cell body out to 300 μm. Plotting the number those intersections as a function of distance from the cell body generates a curve and results are represented as area under the curve. Thirty isolated, non-overlapping cells were analyzed per coverslip. Arborization data was pooled across 3 independent experiments (3–5 coverslips per treatment condition in each experiment) providing between 180 and 270 cells per condition over 3 experiments (60-90 cells per condition in each experiment).

Statistical analyses were done using GraphPad (v.5 for windows, San Diego CA, United States). ANOVAs were used to compare multiple groups with Tukey’s post hoc pairwise testing. Significance levels are indicated by asterisk with *p < 0.05, **p < 0.01, ***p < 0.001.

The D. melanogaster wild-type Canton S (CS) stock was originally provided by M. Heisenberg, University of Würzburg, Germany. Flies were maintained on standard Drosophila food in a 12:12 h light and dark cycle at 24°C. For treatment, newly eclosed flies were collected daily and aged on standard food for 4 weeks with fresh food provided every week. Flies were then divided into a control and a treatment group. The first treatment paradigm involved mixing CAW into standard fly food at a final concentration of 10 mg/mL (from a stock solution of 100 mg/mL CAW in water) whereas the food for the control group was mixed with the same amount of water. A second treatment paradigm was to provide the supplement (CAW 10 mg/mL) in 10% yeast paste (Fleischmann’s Dry Yeast), whereby the yeast paste was applied to a tissue on the surface of standard food. Yeast paste with water added instead of the CAW supplement was used for the controls.

Application by yeast paste was also used when testing individual compounds and compound mixes. If not stated otherwise, the concentration of compounds was equivalent to their levels found in CAW as previously determined (Yang et al., 2023a; Holvoet et al., 2023b). For treatment of Drosophila a 10x concentrated stock solution of the compound or compounds was prepared in ethanol and diluted 1:10 in yeast paste. Control yeast paste for these experiments was prepared by adding ethanol instead of supplement (final concentration 10% ethanol). The flies were given supplemented or control food for two more weeks, changing the food once, before being tested when 6 weeks old.

Fast phototaxis assays were conducted as previously described (Bolkan et al., 2012; Dutta et al., 2016). Briefly, flies were transferred to the apparatus in groups of 10–15 flies, shaken to the bottom of the tubes and allowed to transition towards the light in 5 consecutive runs, each lasting 6s. Flies were then scored based on the tube they were contained in at the end of the final run. A detailed description of the experimental conditions can be found in (Strauss and Heisenberg, 1993). Statistical analyses were done using GraphPad (v.5 for windows, San Diego CA, United States). Normal distribution was addressed with the D’Agostino and Pearson omnibus test and due to the nonparametric distribution, Mann-Whitney tests were used to determine significance between two samples and Kruskal–Wallis ANOVA (with built-in Bonferroni post hoc) were used when comparing multiple groups. Significance levels are indicated by asterisks with *p < 0.05, **p < 0.01, ***p < 0.001.

The genus Centella contains 53 species and we first wanted to confirm that the provided plant material was C. asiatica. We therefore sequenced two samples derived from the provided plant material; BEN-CA6-A and BEN-CA6-B, and found that BEN-CA6-A was 92.4% identical to BEN-CA6-B. In the NCBI database ITS sequences from only three Centella species were available which were used for comparison. The BLAST homology search of BEN-CA6-A resulted in a sequence identity to C. asiatica of 98.7%–99.3%, to Centella capensis 91.4%, and to Centella montana 76.3%. The sequence identity of BEN-CA6-B to C. asiatica was 91.1%–91.8%, to C. capensis 86.8%, and to C. montana 71.1% (Supplementary Figure S1). We therefore concluded that the plant material was C. asiatica.

To evaluate the effects of CAW and its constituent compounds on neuronal complexity, we measured dendritic arborization in an in vitro system. Primary, embryonic, hippocampal neurons isolated from C57BL/6 mice were grown in culture on a glial feeder layer for a total of 19 days with the neurons exposed to treatments during the last 5 days.

We previously showed that treating hippocampal neurons with CAW (50 μg/mL) increased neuronal complexity (Gray et al., 2017). The current batch of CAW at 50 μg/mL also increased the dendritic complexity (Figure 1B) compared to vehicle treated neurons (Figure 1A). To quantify this, we performed a Scholl analysis and measured the area under the curve generated by these values (Figure 1C). Treatment with CQA mix and the TT mix, at concentrations equivalent to their relative abundance in CAW (50 μg/mL), each also increased arborization although the magnitude of increase was less with CQA treatment than was seen with CAW treatment. Surprisingly, the combination of TT and CQA mixes resulted in reduced arborization relative to vehicle treated neurons (Figure 1C). Representative images of the neurons treated with the compound mixes are shown in Supplementary Figure S2A.

We went on to test the effects of the individual CQA and TT compounds present in the mixes on dendritic arborization in the same primary hippocampal neuron system, again, at concentrations equivalent to their abundance in CAW (50 mg/mL). We found that while six out of the 8 individual CQAs tested increased arborization above what was seen with vehicle treated neurons, 5-CQA (neochlorogenic acid) and isochlorogenic acid A (IsoA) did not affect arborization (Figure 2A). In contrast, treatment with each of the 4 TT compounds did result in significant increases in arborization, with MS eliciting a greater magnitude of change than the other TT compounds (Figure 2B). Representative images of the neurons treated with the single compounds are shown in Supplementary Figures S1A, C.

Besides improving memory, CA has also been described to improve physical performance in the elderly (Mato et al., 2011b). Drosophila, like humans, show cognitive decline and behavioral changes when aged, including a reduction in locomotion (Iliadi and Boulianne, 2010). We therefore used this model to address whether CA provides resilience against age-related phenotypes using the Drosophila in vivo model. Using the fast phototaxis assay, which tests locomotion and the ability of the flies to detect and orient towards light, we first showed that the performance in this test is continuously reduced with age in male and female flies (Figures 3A, B). We also performed this experiment to determine an appropriate time point for treatment that would allow us to detect effects of CAW during aging. As shown in Figures 3A, B, both sexes show a continuous decline with age up to 42 days (only the difference between 2days and 14 days old females and between 42days and 60 days old male and female flies did not reach significance). We picked 28 days as an age to start treatment because these are middle-aged flies that show a significant decline with age at mid-age (28 days) that progresses further with continuing aging. To determine whether supplementation with CAW can ameliorate further decline in performance when given at mid-age, we kept the flies for 28 days on regular food. We then mixed the food with CAW and kept the flies on it for two more weeks before testing them in the phototaxis assay. We previously found that a dose of CAW 10 mg/mL in the food has beneficial effects in a Drosophila model of oxidative stress (Cabey et al., 2022) and we therefore used this concentration in the present experiment to explore effects on healthy aging. When comparing the performance of control flies and flies treated from day 28 to day 42 with CAW, we found a significant improvement in both males and females (Figure 3C). This shows that CAW supplementation has a beneficial effect even when starting treatment at middle age. We also included untreated 28 days old flies to compare the decline in performance from 28 days old to 42 days old in untreated and CAW treated flies. While both groups showed a significant decrease from 28 days old to 42 days old, the decline in untreated males was 69% whereas it was only 50% in CAW treated males. A similar protection was seen in females, with a 74% reduction in untreated females between day 28 and day 42, compared, to a 59% reduction in CAW treated females (Figure 3C). We also tested a second treatment paradigm by giving the CAW in addition to the regular food by mixing it at 10 mg/mL in 10% yeast paste applied on top of the food. Yeast is a preferred food source for Drosophila that influences adult body weight and survival (Günther and Goddard, 2019). Correlating with the published positive effect on survival, providing the yeast paste did improved the mean performance in untreated flies; in males from 20% to 31% and in females from 16% to 27%. Providing CAW in yeast paste further improved performance to 68% in males and 66% in females (Figure 3D). Together this shows that both treatment paradigms provide resilience against the age-related decline in performance, even when treatment is started at middle age.

To confirm an effect of CAW treatment in another behavioral test, we performed a negative geotaxis test. This test does not require a stimulus and is based on the innate behavior of Drosophila to run upwards in a vertical tube. Using the same treatment paradigm as for the fast phototaxis, we tested 42 days old flies treated for 2 weeks before being tested. Again, we found that CAW supplementation improved the behavior in males and females (Supplementary Figure S3), confirming a beneficial effect of CAW on the fitness of aged flies.

Next, we tested whether the biological activity to provide resilience against the decline in fast phototaxis is mediated by CQAs or TTs. Due to the stronger effect when CAW was given in yeast paste, and the limited quantities of purified single compounds available, we tested CQAs and TTs only given in yeast paste. First, we tested mixes of CQAs and TTs containing the compounds in concentrations equivalent to CAW (10 mg/mL). We previously showed that the levels of these compounds in CAW are not affected by the preparation of the food and are stable for the 7 days the flies are kept on a food vial (Holvoet et al., 2023a). The compounds and compound mixes were prepared in ethanol as 10x stock solutions and therefore the control flies also obtained 10% ethanol in yeast paste. Both, the CQA mix and the TT mix significantly improved the behavior in males and females when given for 2 weeks to 28 days old flies (Figure 4A). Furthermore, both mixes had a similarly protective effect as CAW with 62% for the TT mix and 67% for CQA mix in males (CAW 68%) and 63% for the TTs and 62% for the CQAs in females (CAW 66%). We also tested whether combining both mixes (TT + CQA) further improved the behavior but this was not the case. We then asked whether the lack of a synergistic effect when the CQAs and TTs were combined was due to a “ceiling effect” due to the single mixes already being similarly effective as CAW. To address this, we reduced the concentration of the mixes by 50%. This resulted in a loss of protection by any of the mixes with the exception of the TT mix in males (Figure 4B). Furthermore, we still did not see a positive effect of combining the mixes. In fact, adding the CQAs prevented the protection seen with the TTs.

The results with the mixes suggested that both groups contained biologically active compounds and we therefore tested single CQAs and TTs at concentrations equivalent to their presence in CAW (10 mg/mL). However, none of them had a significant effect on phototaxis in males (Figure 5A) or females (Figure 5B) when testing all of them individually in one experiment and applying the appropriate Bonferroni correction for multiple groups. Nevertheless, some compounds improved the performance non-significantly. One of them was 1,5-diCQA, which increased the transitions towards light, in both males and females, and 1,3-diCQA which increased them in males. We therefore tested these compounds in a separate experiment. Comparing them to the included controls, 1,5-diCQA resulted in a significant improvement in males and females (Figures 5C, D), confirming that this CQA has biological activity in our assay. However, 1,5-diCQA was less protective than the CQA mix or CAW with increasing the locomotion score to about 35% in contrast to the CQA mix and CAW which increased it to about 60%.