Effect of Rhodiola rosea L. extract on Alzheimer's Disease-Like Symptoms in Caenorhabditis elegans

1. Introduction

The genus Rhodiola (family Crassulaceae) comprises over 200 species and more than 20 of which are used as medicine, mainly in Asia, including R. rosea L, R. heterodonta (Hook.f. & Thomson) Boriss, R. quadrifida (Pall.) Fisch & Mey, R. kirilowii (Regel) Maxim. and R. crenulata (Hook.f. & Thomson) H.Ohba among others [1,2]. Among them, R. rosea (Sedum rosea (L.) Scop., Golden Root, Arctic Root, Roseroot) was the primary subject of phytochemical studies and used in human medicine and therapies [3].

In modern medicine, R. rosea is regarded as an adaptogen with multiple documented health-promoting effects, including anti-depressive, anti-tumor, anti-inflammatory, neuroprotective, cardioprotective, hepatoprotective, and immunomodulating activities [4,5,6] and in clinical practice it was associated with improved stress resilience and psychophysical health [7,8]. Its main phytochemical markers, salidroside and rosavin, are largely responsible for these pharmacological activities [5].

Among these properties, the neuroprotective and antioxidant potential of R. rosea is particularly relevant to age-related neurodegenerative diseases, which remain one of the greatest public health challenges today.

Alzheimer’s disease (AD) is the most common cause of dementia and remains one of the major public health challenges worldwide [9]. Its pathology involves progressive neuronal dysfunction, synaptic loss, memory decline, and accumulation of amyloid-beta (Aβ) and Tau protein aggregates [10,11,12]. Beyond amyloid pathology, oxidative stress and impaired cellular stress responses are now recognized as central, rather than merely secondary contributors to AD progression [13]. Excessive production of reactive oxygen species (ROS) can damage lipids, proteins, and nucleic acids, thereby amplifying neurodegenerative processes and worsening neuronal vulnerability [14,15].

Because of the complexity of the disease, in vivo model systems that are genetically tractable and experimentally robust are essential for mechanistic and preclinical studies. Caenorhabditis elegans (C. elegans) is a widely used model in neurodegeneration research because of its short lifespan, conserved stress-response pathways, fully mapped cell lineage, and availability of transgenic strains expressing human disease-associated proteins [16]. In particular, C. elegans models expressing human Aβ are useful for studying behavioral defects, aggregation-related toxicity, and protective effects of candidate compounds. These models enable the rapid assessment of phenotypes such as locomotion, chemotaxis, reproduction, and oxidative stress responses, making them particularly well-suited for screening natural products with potential neuroprotective activity [17,18].

The involvement of plant adaptogens in preventing age-associated diseases and decelerating senescence was intensively studied. Previous studies have suggested that R. rosea and its constituents may exert neuroprotective effects, increase lifespan, and ameliorate memory impairment by modulatin stress signaling, redox balance, and cellular resilience pathways [19,20,21]. These observations suggest that R. rosea and its constituents might have a feasible effect on age-related disease, including AD. However, their specific impact on AD-like phenotypes in vivo remains incompletely characterized, and to date no study has systematically and comparatively evaluated the whole R. rosea extract against its individual phytochemical markers (salidroside and rosavin) in a transgenic AD model. This gap is relevant since botanical extracts typically act through multiple compounds with overlapping or complementary mechanisms and determining whether the neuroprotective effect resides in the whole phytocomplex or in isolated individual markers has direct implications for the development of standardized supplements based on this plant.

This study aimed to determine if R. rosea extract and its main components can alleviate AD-related traits in C. elegans. We hypothesized that these substances could enhance behavioral performance, decrease oxidative stress, influence critical stress-response genes, and reduce Aβ-related pathology. Additionally, the research sought to explore the mechanisms behind these effects, focusing on the conserved SKN-1/Nrf2 pathway. The study also examined whether the complete extract has mechanisms or benefits that surpass those of its individual marker compounds, again with particular attention to the SKN-1/Nrf2 pathway as a potential common mediator.

2. Results

2.1. Phytochemical Analysis of R. rosea Extract

The identification of the key compounds of R. rosea, such as phenylethanoid glycoside (e.g. salidroside, also known as rhodioloside and rhodosin) and the phenylpropanoid glycosides (rosavin, rosarin and rosin) in the rhizome’s extract was performed by comparing their NMR signals with previously published data, as well as the ¹H NMR signals of their reference standards. The chemical structures of the molecules have been presented in Figure 1.

In Table 1 are presented the compounds identified in R. rosea extract according to their NMR signals.

In the ¹H NMR spectrum of the extract, salidroside (p-hydroxyphenylethyl-O-β-D-glucopyranoside) was identified by the observed signals of four aromatic protons δ_H 7.17ppm (2H, d, J=8.4 Hz, H-2/6) and δ_H 6.81 ppm (2H, d, J=8.5 Hz, H-3/5), indicating its aromatic AA’BB’ spin system. The signals of four methylene protons at δ_H 2.88 ppm (2H, t, J=7.7, H-7), δ_H 4.08 ppm (1H, m, H-8a) and δ_H 3.80 ppm (1H, m, H-8b) revealed the presence of salidroside ethyl fragment (ethyl-side chain), which connects its p-hydroxylated aromatic ring to the glucose moiety. The β-glucopyranosyl unit was identified by the anomeric H-atom at δ_H 4.42 ppm (1H, d, J=8.0 Hz, H-1’). The ¹H NMR signals of salidroside are shown in Figure 2 A and B. According to the HSQC spectrum (Figure 2 E and F) were assigned the following hydrogen-carbon correlations: δ_H 7.17 ppm/δ_C 129.42 ppm (C2/C6); δ_H 6.81 ppm/δ_C 114.66 ppm (C3/C5); δ_H 2.88 ppm/δ_C 35.01 ppm (C7); δ_H 4.08 (3.80) ppm/δ_C 75.85 ppm (C8 a/b) and the anomeric proton at δ_H 4.42 ppm/δ_C 101.14 ppm (C1’). Salidroside identification was further supported by the signals of its segments in the COSY experiment (Figure 2 I): δ_H 7.17 ppm/ δ_H 6.81 ppm (H2/H5-H3/H6), corresponding to the aromatic AA’BB’ spin system and δ_H 2.88 ppm/δ_H 4.08 ppm (H7/H8), as well as the respective TOCSY experiment (Figure 2 J): H2/H5-H3/H6; H7/H8 and δ_H 4.42 ppm/δ_H 3.70 ppm (H1’/H6’). The signals of salidroside reported here were in accordance with [1,22,23,24,25,26,27,28].

Rosavin (cinnamyl-(6’-O-α-L-arabinopyranosyl)-O-β-D-glucopyranoside), rosarin (cinnamyl-(6’-O-α-L-arabinofuranosyl)-O-β-D-glucopyranoside) and rosin (cinnamyl-O-β-D-glucopyranoside) frequently assigned as rosavins were identified based on their relevant NMR signals denoted in the ¹H- and 2D-NMR spectra of the extract. Structurally, rosavin and rosarin represent rosin glycosylated with arabinose. In the ¹H NMR spectrum, the signals of the cinnamyl moiety were shared for rosavin, rosarin and rosin as previously noted by [29]: δ_H 7.48 ppm (2H, d, J=7.6 Hz, H-2/6); δ_H 7.37 ppm (2H, t, J=7.5 Hz, H-3/5); δ_H 7.30 ppm (1H, t, J=7.5 Hz, H-4); δ_H 6.73 ppm (1H, d, J=16.5 Hz, H-7); δ_H 6.41 ppm (1H, dt, J=16.0; 6.0 Hz, H-8) and δ_H 4.56 ppm (1H, ddd, J=13.5; 6.5; 1.5 Hz, H-9a) and δ_H 4.41 ppm (1H, ddd, J=13.8; 6.5; 1.4 Hz, H-9b). The large coupling constant value of olefinic protons H-7 and H-8 (d, J=16.5 Hz) clearly suggested a trans geometry stereochemistry of the double bond. The anomeric proton at δ_H 4.49 ppm (1H, d, J=8.0 Hz, H-1’) revealed the β-glucopyranose common for the three pehenylpropanoids, while the anomeric proton at δ_H 4.37 ppm (1H, d, J=8.0 Hz, H-1’’) and δ_H 5.05 ppm (1H, d, J=1.4 Hz, H-1’’) revealed the α-arabinopyranose and α-arabinofuranose in rosavin and rosarin, respectively. A distinct feature among the three compounds were the chemical shifts at 6’ position: δ_H 4.11 ppm (1H, dd, J=11.5; 2.0 Hz, H-6a’) and δ_H 3.82 ppm (1H, dd, J=11.6; 5.3 Hz, H-6b’) for rosavin; δ_H 4.07 ppm (1H, m, H-6a’) and δ_H 4.04 ppm (1H, m, H-6b’) for rosarin and δ_H 3.90 ppm (1H, dd, J=12.0; 1.4 Hz, H-6a’) and δ_H 3.72 ppm (1H, m, H-6b’) for rosin. A difference in 6’’ was also observed between rosavin and rosarin: δ_H 3.60 ppm (1H, m, H-6a’’) and δ_H 3.47 ppm (1H, m, H-6b’’) for rosavin and δ_H 3.91 ppm (1H, m, H-6a’’) and δ_H 3.78 ppm (1H, dd, J=13.6; 4.0 Hz, H-6b’’) for rosarin. The typical ¹H NMR signals of rosavins can be observed on Figure 2 C and D. The observed correlations according to the HSQC experiment (Figure 2 G and H) were as follow: δ_H 7.48 ppm/δ_C 125.64 ppm (C2/C6); δ_H 7.37 ppm/δ_C 127.20 ppm (C3/C5); δ_H 7.30 ppm/δ_C 127.65 ppm (C4); δ_H 6.73 ppm/δ_C 113.76 ppm (C7); δ_H 6.41 ppm/δ_C 142.16 ppm (C8); δ_H 4.56 (4.41) ppm/δ_C 71.46 ppm (C9 a/b) and the anomeric proton at δ_H 4.49 ppm/δ_C 101.24 ppm (C1’). The anomeric proton of α-arabinopyranose at δ_H 4.37 ppm/δ_C 104.14 ppm (C1’’) in rosavin were also identified in the HSQC spectrum, however the ¹H-¹³C correlation at δ_H 5.05 ppm indicating the anomeric proton of α-arabinofuranose in rosarin was hampered due to the low signal intensity. Rosavins identification was further supported by signals of their aromatic ring in the COSY experiment (Figure 2I): δ_H 7.48 ppm/δ_H 7.37 ppm (H2/H5-H3/H6), related to its aromatic AA’BB’ spin system and δ_H 6.73 ppm/δ_H 6.41 ppm (H7/H8), as well as the respective TOCSY experiment (Figure 2J): H2/H5-H3/H6 and H7/H8. The signals of rosavins reported here were in accordance with [1,22,23,26,27,29,30,31].

In agreement with the NMR profiling, salidroside, rosavin and rosarin have been unambiguously identified according to their MS spectra presented in Table 2 together with several other compounds. In addition, a supplementary Table S1 is presented in the supplementary material.

The HPLC results revealed that the content of the characteristic compounds was as follow: p-tyrosol 0.07±0.01, salidroside 27.23±0.04, rosavin 29.05±0.54, rosarin 9.59±1.73 and rosin 3.76±0.12 mg/g dry extract (mg/g DE). Calculated as mg/g dry rhizomes weight (mg/g DW) were obtained the following values: p-tyrosol 0.03±0.01, salidroside 12.51±0.01, rosavin 13.34±0.03, rosarin 4.40±0.24 and rosin 1.73±0.02 mg/g DW or salidroside 1.25±0.01%, rosavin 1.33±0.02%, rosarin 0.44±0.08% and rosin 0.17±0.01%.

2.2. Locomotor Activity Assay Results

The CL2006 transgenic strain, expressing human Aβ1-42 peptide under the unc-54 promoter, displayed significantly reduced locomotor activity compared to the N2 wild-type control, confirming the established locomotor deficit phenotype characteristic of this AD model in C. elegans.

Following 4 days of liquid culture treatment starting from synchronized L1 larvae, R. rosea extract, salidroside and rosavin, produced a dose-dependent improvement in locomotor activity counts measured by Wmicrotracker^® MINI over 20 minutes. Notably, the 1 µg/mL of R. rosea extract and 1 µM concentration of each compound treatment yielded the most robust recovery, with activity levels approaching (but not fully reaching) those of the N2 control, while higher doses (10 µg/mL and 10 µM) showed more modest effects across all compounds (Figure 3).

Moreover, treatment of CL2006 worms with R. rosea extract, salidroside and rosavin, resulted in varying degrees of improvement in locomotor performance. Salidroside and R. rosea extract showed the most pronounced effects, increasing the travelled distance closer to that observed in N2 worms. Rosavin treatment led to a modest improvement, though the overall locomotion remained substantially lower than in the N2 control (Figure 4).

2.3. Egg-Laying Assay Results

The CL2006 transgenic strain exhibited a severe egg-laying deficiency compared to the N2 wild-type control, laying significantly fewer eggs per worm over 6 hours, consistent with the strain's characteristic adult-onset paralysis and egg-laying defect induced by muscle-specific expression of human Aβ_1-42 peptide. Treatment with R. rosea extract at 10 µg/mL, salidroside and rosavin at 10 µM partially rescued this reproductive phenotype in CL2006 worms, increasing the number of eggs laid relative to the untreated transgenic control. While the two R. rosea-derived compounds showed beneficial effects, R. rosea extract at 10 µg/mL produced the most pronounced recovery, approaching levels closer to the N2 control, whereas salidroside and rosavin displayed more modest improvements (Figure 5).

2.4. Chemotaxis Assay Results

The chemotactic response toward benzaldehyde was significantly reduced in the transgenic CL2355 strain compared to the control CL2122, confirming the presence of a sensory deficit associated with the AD model. The control strain exhibited a high positive chemotaxis index (CI), whereas CL2355 worms showed a markedly lower CI, indicating impaired attraction to the odorant.

Treatment with salidroside resulted in a clear improvement in chemotactic behavior in CL2355 worms, with CI values significantly increased compared to untreated transgenic animals, although still not reaching control levels. A similar, but slightly less pronounced, effect was observed following treatment with R. rosea extract, which also significantly enhanced the chemotaxis response relative to untreated CL2355 worms. In contrast, rosavin treatment produced only a modest and non-significant increase in CI, with values remaining substantially lower than those of the control strain and comparable to untreated CL2355 animals (Figure 6).

RT-qPCR analysis revealed that the transgenic CL2006 strain exhibited a marked reduction in the expression of skn-1/nrf2 and sod-1 compared with the N2 control, whereas daf-2/IGF1R was strongly upregulated. Treatment with salidroside, R. rosea extract, and vitamin C restored skn-1/nrf2 expression to different extents, with salidroside showing the most pronounced effect. A similar trend was observed for sod-1, where salidroside and R. rosea significantly increased gene expression relative to the CL2006 untreated group. In contrast, rosavin produced a more moderate effect on both genes. Regarding daf-2/IGF1R, all treatments reduced its elevated expression in CL2006 worms, with rosavin showing the strongest decrease. The expression of sek-6/MAP2K4 showed no major changes among the experimental groups, suggesting that this pathway was less responsive to the treatments under the conditions tested. Overall, these results indicate that the selected compounds may exert neuroprotective effects in C. elegans through modulation of the SKN-1/Nrf2-associated antioxidant response and related signaling pathways (Figure 7).

2.5. Determination of ROS Level Results

Intracellular reactive oxygen species (ROS) levels increased progressively over time in all experimental groups, as indicated by the rise in DCF fluorescence intensity during the 160-minute measurement period. However, clear differences in the rate and magnitude of ROS accumulation were observed between treated and untreated nematodes.

Treatment with R. rosea extract, as well as its active compounds salidroside and rosavin, markedly attenuated ROS accumulation compared to the vehicle control. While fluorescence intensity increased in all groups, the control worms exhibited a steeper and more pronounced rise, particularly at later time points, indicating higher oxidative stress.

Among the treatments, R. rosea extract produced the greatest reduction in ROS levels, followed by salidroside and rosavin, both of which also significantly lowered fluorescence intensity relative to the control group. In contrast, the untreated group maintained consistently higher ROS levels throughout the assay, with only minimal changes over time (Figure 8).

2.6. Aβ Aggregate Quantification Results

Thioflavin-S staining revealed prominent Aβ aggregates in the head region of untreated adult CL2006 worms, as expected from muscle-specific expression of human Aβ1-42. Treatments with R. rosea extract at 10 µg/mL, salidroside and rosavin at 10 µM significantly reduced Aβ plaque density (normalized to head area) compared to untreated CL2006 controls, with statistical significance. Salidroside and rosavin produced comparable reductions (~40-50% decrease), while R. rosea extract showed the most potent effect, lowering aggregate levels closest to background fluorescence levels (Figure 9).

3. Discussion

Rhodiola rosea grows naturally on hard-to-reach rocky terrains, sea cliffs and mountains (at 2800-5600 m a.s.l) in the cold regions of the northern hemisphere, ranging from the low Arctic to the high temperate zones of Asia (the subarctic and Siberia, Altai and the mountains of Mongolia), Europe (the cool temperature regions of the Northern, Central and South-eastern parts of the continent), and North America (Iceland, Greenland, Alaska or Canada) [3,32,33]. Golden root is one of the priority species for conservation in many European countries, since its natural habitats are almost completely exhausted because of indiscriminate harvesting of R. rosea plants for their valuable pharmacological sub stances. It is a protected species with vulnerable, endangered and critically endangered status and is also included in the Red Data Book of several countries, including Bulgaria, Ukraine Sweden and others [33]. Although R. rosea is successfully cultivated in Russia, Mongoloia an many European countries, it turns out that it is almost impossible to satisfy the industrial needs for authentic plant material for the production of nutritional supplements based on R. rosea [33]. More than 140 organic compounds have been isolated from R. rosea [4] belonging to different classes (phenylpropanoids, phenylethanoids, flavonoids, phenolic acids, monoterpenes and triterpenes) and still new ones have been reported [32,34]. However, up to now phenylethanoids (salidroside and its aglycone tyrosol), and phenylpropanoids (rosavin, rosarin, and rosin) commonly named “rosavins” found in the rhizomes of the plant, have been recognized as the main bioactive components and pharmaceutical quality markers of this herb [35]. Salidroside was found many Rhodiola species (R. crenulata, R. rosea, R. sachalinensis, and R. kirilowii) [36] and even in other genera (i.e. Salix, Vaccinium and Rhododendron) does not allow for its unequivocal recognition as a marker compound for R. rosea [37]. Consequently, monitoring the presence of rosavin, as a distinctive chemotaxonomic marker, seems to be considered synonymous of recognition of R. rosea from other Rhodiola species, authenticity and high quality of the herbal preparations based on R. rosea, well documented in many studies [6,23,26,36,38,39]). To assure the quality, therapeutic activity and consumer safety are formulated several standards for the content of saidroside and rosavins in R. rosea extracts: the United States Pharmacopeia (0.08% salidroside, 0.3% rosavins), the Russian State Pharmacopoeia (≥0.8% salidroside, ≥1% rosavin), and the Australian standard (≥1.8% phenylpropanoids, 1.2% rosavin, 0.6% salidroside) [40]. The European Medicines Agency classifies the dry extract (DER 1.5–5:1) obtained from R. rosea roots and rhizomes as a traditional herbal medicinal product [41], and most commercial products are standardized to a minimum of 1.0% salidroside and 3.0% rosavins, reflecting their naturally occurring ratio in the plant's rhizomes [8].

P-tyrosol is a natural constituent and a key foundational building block/aglycone found in R. rosea. However, in our study the identification of p-tyrosol according to its characteristic NMR signals and mass-to-charge ration (m/z) in LC/MS data was challenging for several reasons, as stated before from [6]: in one hand, the signals in aromatic regions overlay with those from the salidroside, and on the other, the low amounts of p-tyrosol are difficult to confirm due to the lower sensitivity of NMR compared to chromatographic methods [6]. Similarly, rosin a simple monoside (cinnamyl alcohol-O-β-D-glucopyranoside) was also impossible to confidently identify in a whole-crude and LC/MS profile of R. rosea extract due to is signal overlap (spectral crowding) with structurally similar rosavin and rosarian, combined with its low natural abundance as also reported in previous studies [42]. However, we were able to quantify the small amounts of p-tyrosol and rosarian in the extract.

The content of the phenylethanoids and phenylpropanoids in R. rosea depends on many factors, including harvesting time, plant material storage, rhizomes or roots used, genotype, extraction procedure, solvent type e.t.c [6]. Depending on plant age and solvent used, the amounts of salidroside varied between 0.3-0.4% DW ROStot between 1.2-3.0% [29]. Salidroside and ROStot content was 9 and 5 times higher in the rhizomes than the roots of R. rosea [22], while the salidroside and rosavin content in the rhizomes was approximately 10 and 120 times higher than in the leaves [43]. The research of [26] reported a drop in the content of rosavin between August and September [26]. The amount of salidroside and ROStot in our study is comparable with R. rosea from other localities. For example, the content of salidroside of R. rosea cultivated in Western Serbia varied between 0.4 and 1.4%, while the content of rosavins was ca. 1.1% depending on the concentration of ethanol used for extraction [40]. A broad study of thirteen R. rosea samples from various ecological and geographical conditions in the Altai Mountains showed varying values of total salidroside 0.1-0.4% and ROStot between 1.2 to 4.5% [44]. However, the study of [38] revealed the opposite, higher salidroside content (1.13 to 2.71%) and low rosavin content (0.21-0.30%) [38]. The salidroside content of R. rosea cultivated in Poland was 0.2% when extracted with water, while the rosavin content was 1.3% when extracted with 96% ethanol [35].

The phytochemical characterization of R. rosea extract in our study showed similar results with the analysis of R. rosea rhizomes collected from Pirin Mountain in 2018 at the same time interval, revealing consistency in the phytochemical profile and compound quantification [23]. Along with that, the rhizomes of R. rosea investigated in this study, meets the minimum requirements of the United States Pharmacopeia, Russia State Pharmacopeia and the Australian standard regarding the content of salidroside and rosavin(s), suggesting its therapeutic potential.

Oxidative stress and impaired mitochondrial homeostasis are increasingly recognized as central, rather than secondary, drivers of AD pathogenesis. Excessive ROS production damages mitochondrial and nuclear DNA, proteins, and lipids, progressively overwhelming the endogenous antioxidant network, including superoxide dismutase (SOD), glutathione (GSH), catalase (CAT), thioredoxin (Trx)/thioredoxin reductase (TrxR), glutathione peroxidase (Gpx), heme oxygenase-1 (HO-1), and NAD(P)H:quinone oxidoreductase 1 (NQO1) [45,46]. This antioxidant network is coordinated primarily through Nrf2 signaling (SKN-1 in C. elegans), whose activation attenuates oxidative stress and inflammation and protects tissue from oxidative damage. Because oxidative stress, neuroinflammation, and impaired proteostasis are strongly implicated in AD pathogenesis, the KEAP1-Nrf2/SKN-1 axis has emerged as a promising therapeutic target for the disease [45,46].

High stress resistance protects from AD pathology and correlates with increased lifespan in variety of organisms [19,47]. The effect of R. rosea on lifespan and stress resistance was a topic of considerable interest investigated mainly in short-lived biological models, such as yeast [19], C. elegans [48,49,50], silkworm [51] and Drosophila [52]. Characteristic feature is the dose-dependent fashion in the enhanced lifespan after extract treatment. For example R. rosea extract in a dose of 480 µg/mL increased the lifespan of C. elegans with 37.1% [49] and 10-20% after treatment with 10-25 µg/mL SHR-5 standardized R. rosea extract [48], respectively, while the lifespan of D. melanogaster increased 14 to 17% after feeding with 5 to 10 mg/mL R. roesea extract [52] and the lifespan of the silkworm increased 1.49 to 2.75% [51].

For our study we have selected doses for R. rosea extract based on the study of [48], where the range of doses was varying from 2.5 to 250 µg/mL. Considering that the most prominent effects were observed in the range of 10 to 25 µg/mL and beyond 100 µg/mL was observed reverse effects (e.g. a lifespan shortening effect of 25% was reported at concentration of 250 µg/mL), we also kept the tendency to have doses over a 100-fold concentrations, selecting 1, 10 and 100 µg/mL for the extract. Based on our HPLC results, the calculated molar concentration of salidroside in the highest dose of R. rosea extract was approximately 10 µM. Therefore, the preferred treatment doses for salidroside and rosavin have been selected to be 1, 10 and 100 µM.

Alzheimer’s’ disease is a complex process, therefore locomotor assay is used to assess behavioral abnormalities. Therefore locomotion behavior is considered not only as a marker of aging [49], but also as marker of AD symptoms [53]. In our study the transgenic C. elegans CL2006 that express human AD-associated amyloid-beta Aβ-proteins showed a clear, progressive decrease in locomotion speed and motility compared to healthy wild-type worms. Rhodiola rosea extract, salidroside and rosavin expressed the most robust recovery at concentrations of 1 µg/mL and 1 µM, respectively. By increasing the concentrations of the extract and both compounds a decline in the locomotor activity improvement was observed. Our finding are in agreement with other studies, which were further overviewed by [54], concluding that the R. rosea extracts and salidroside commonly induce hormetic dose responses within a broad range of biological models, cell types, with particular emphasis on longevity and neuroprotective. Hormesis is defined as a biphasic dose/concentration response, whereby low doses/concentrations induce stimulation, and high doses/concentrations cause inhibition [54]. This effect was observed in the aging/longevity studies, where 10 µg/mL extract of R. rosea increased the lifespan of C. elegans up to 115.4%, while increasing treatment doses shortened the lifespan up to 86.02% at 250 µg/mL [48]. Decline in C. elegans motility on day 9, 14, and 18 was significantly delayed by treatment with R. rosea extract. The maximum effect on lifespan (103.8% of control) has been achieved at concentration of 240 mg/mL R. rosea extract and shortening of the lifespan (101.6 and 100.9% of the control) was reported with increasing the concentrations to 320 and 480 μg/mL extract [49]. Another confirmation about the hormetic-biphasic dose response effect of R. rosea was the increased median lifespan of D. melanogaster 14 and 17% when 5 and 10 mg/mL R. rosea extract was used as food supplement. Feeding with 30 mg/mL extract decreased the lifespan with 9 to 12%, demonstrating possible toxic effects [52]. Based on our observations, the next experiments were further continued with 10 µg/mL extract and 10 µM of each compound.

Egg-laying defects in C. elegans serve as a vital behavioral readout for studying AD mechanisms, particularly involving presenilin function and Aβ toxicity [55]. As an Aβ-expressing strain, CL2006 in our study exhibited a severe egg-laying deficiency compared to the N2 wild-type control. Other plant-derived products, e.g. rose essential oil (from 2 to 20 µg/mL) significantly suppressed Aβ deposits and reduced the Aβ oligomers to alleviate the toxicity induced by Aβ over-expression in C. elegans. The essential oil markedly activated the expression of gst-4 gene, which supported that SKN-1 signaling pathway was involved in the therapeutic effect of the oil on AD C. elegans [56]. The neuroprotective activity of caffeic acid (100 µM) was found to be due to the restored skn-1 levels in C. elegans [57].

The chemotaxis assay is used to evaluate cognitive decline, olfactory dysfunction, and memory-like learning deficits in C. elegans. In our study, the transgenic CL2355 strain had marked lower CI than the control CL2112 strain. Other studies also confirm the beneficial effect of natural products. The chemotaxis deficit in CL2355 was alleviated by feeding with black rice anthocyanin extract. The extract also downregulated the mRNA expression of age-1 and daf-2, while upregulated the daf-16 mRNA level and SOD-3, CTL-1, and GST-4 protein expression [58]. Resveratrol at final concentration of 50 µM significantly enhanced chemotaxis performance in GMC101, demonstrated 214% improvement in behavioral responses and suggesting partial restoration of learning-associated cognitive functions [59].

Reducing daf-2 insulin/IGF-1 signaling (IIS) in C. elegans extends lifespan and significantly mitigates the toxicity and aggregation of Aβ plaques linked to AD [60]. Since R. rosea extract and its constituents regulate daf-2 it may mimic the stress conditions and reduce the IIS. In our study, rosavin showed the strongest decrease on daf-2/IGFR1. When daf-2 is reduced, daf-16 moves to the nucleus to turn on protective genes that clear toxic protein aggregates, such as sod-1 [61]. A central finding is the modulation of the conserved SKN-1/Nrf2-related antioxidant pathway. In C. elegans, SKN-1 is a key transcription factor and master protective regulator that mitigates Aβ toxicity, reduces oxidative stress, and preserves proteostasis involved in oxidative stress resistance and longevity-associated defense programs [50,62]. The increased expression of skn-1/nrf2 and sod-1, together with reduced ROS accumulation, supports the interpretation that these compounds strengthen endogenous redox homeostasis. Other plant-derived extracts, e.g. Selaginella spp. have been found to increase the stress resistance and key antioxidant genes, such as sod-3 through promotion of the nuclear translocation of DAF-16 in C. elegans [63]. Curcumin (in a dose of 100 µM) and its hemi-analogues significantly prolonged the survival of the nematodes (GMC101) protecting against Aβ toxicity, indicating SKN-1/Nrf activation as a possible mechanism of activity [64]. Berberine (80 µg/mL) extended the mean lifespan of wild-type C. elegans by 27%. By activating transcription factors such as DAF-16/FOXO, HSF-1, and SKN-1/NRF2, berberine upregulated antioxidant enzyme expression, reduced lipofuscin accumulation, and improved stress resistance [65]. The observed improved locomotion ability of Chinese olive (Canarium album L) and prolonged the survival time of C. elegans was due to affecting the Insulin/IGF-1 signaling pathway, DAF-16/FOXO, and SKN-1/Nrf2 transcription factors [66]. Lycopene was reported to paralysis in Aβ_1-42 transgenic C. elegans strain GMC101 in 1 µM concentration [20].

The observed stronger and more consistent effects of the whole R. rosea extract, especially in egg-laying rescue, ROS reduction, and Aβ aggregate decrease, indicate a phytocomplex effect. This means other bioactive molecules in the extract, even beyond the tested salidroside and rosavin, may work together to enhance neuroprotection.

4. Materials and Methods

4.1. Plant material

Rhodiola rosea rhizomes were collected in October 2023 with permission from the Ministry of Environment and Water (No 180/27.07.2023) from Pirin National Park near Fish Banderitsa Lake. The plant was recognized by Mrs. Gergana Chakarova, chief expert at Pirin National Park.

4.2. Plant material extraction

Rhodiola rosea rhizomes were extracted in triplicates with 70% methanol (1:20, w/v), at 35 ºC, in an ultrasonic bath for 15 min each. The extract was filtrated, vacuum concentrated at 40 ºC, lyophilized and stored at -20 ºC before use.

4.3. NMR Metabolite Profiling

The NMR metabolite profiling, including sample preparation, spectra recording and data processing were performed as previously described by [23]. Approximately 10 mg of the rhizome extract has been thoroughly dissolved with equal amounts of 0.4 mL CD₃OD and D₂O, containing KH₂PO₄ buffer with pH 6.0 and TSPA-d4 as internal standard with final concentration of 0.005% (w/v). After ca. 1 min of homogenization, the extract was placed in an ultrasonic bath for 20 min at ambient temperature. The procedure continued with 20 min of centrifugation at 12 000 prm and 4 ºC, after which 0.8 mL of the supernatant was transferred in a 5 mm NMR tube. The proton (¹H) and two-dimensional (2D) spectra, e.g. ¹H-¹³C heteronuclear single quantum coherence (HSQC), ¹H-¹H correlation spectroscopy (COSY) and ¹H-¹H total correlation spectroscopy (TOCSY) were recorded at 25 ºC on Bruker Avance III 500 MHz spectrometer. The obtained NMR spectra were manually phase- and base-line corrected, the peak of the internal standard TSPA-d4 was referenced to 0.0 ppm and scaled to 1.0 being used to normalize all signals in the spectra using MestReNova software, version 12.0.0 from Mestrelab Research (Santiago de Compostela, Spain).

4.4. LC/MS Analysis

The analysis of the chemical extracts was performed by ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UHPLC-QTOFMS) using an Acquity UPLC™ I-Class coupled to a Synapt XS (Waters, Milford, USA). The system was controlled by MassLynx v.4.2. The separation was performed on a Waters Acquity HSS T3 column (100 × 2.1 mm i.d., 1.8 μm particle size) thermostated at 27 °C. The separation was performed in gradient mode using water containing 0.05% formic acid as mobile phase A and acetonitrile containing 0.05% formic acid as mobile phase B. The gradient started at 0% B and increased to 30% B in 6.0 min, then to 100% B in 4.0 min, followed by a hold at 100% B for 2 min and re-equilibration at 0% B for 4 min. The injection volume was 0.25 μl. MS detection was performed using data-dependant acquisition (DDA). The Synapt XS was operated in negaive electrospray using alternate scans at low (6 eV) and high (10-20 eV ramp) collision energies. An m/z range of 50-1200 Da was covered for a total cycle time of 0.60 s. Source parameters were: temperature 140°C, capillary voltage 1.0 kV, cone voltage 25V, desolvation gas flow and temperature 1000 L/h and 500°C, respectively, cone gas flow 250 L/h. A 50 ng/ml solution of the synthetic peptide leucine-enkephaline in water:acetonitrile:formic acid (50:50:0.1) was infused constantly into the mass spectrometer as internal reference to ensure accurate mass measurements (<2ppm).

4.5. HPLC-UV ANALYSIS

The HPLC-UV analysis followed procedure described previously by [23]. Waters 1525 Binary pumps with Waters 2487 Dual λ Absorbance Detector (Waters, Milford, MA, USA), equipped with a reverse-phase Kinetex^® C18, 100 Å (150 mm × 4.6 mm, 5 μm) core-shell column (Phenomenex, Torrance, CA, USA) operating at 26 ºC was used. The mobile phase consisted of acetonitrile (A) and water (B) with a flow rate 1 mL/min and gradient elution as follows: 0-20 min from 5% A to 20% A; 20-30 min the flow rate was held at 20% A; 30-40 min the flow decreased again to 5% A. Salidroside and p-tyrosol were detected at 220 nm, while rosarin, rosavin and rosin at 254 nm.

4.6. Worm Strains, Maintenance, and General Methods

Standard procedures were followed to culture C. elegans, unless otherwise noted. Strains used in this work and abbreviations are listed in Table 3. N2 (Bristol) and CL2122 wild type strains was used as control in the experiment, and CL2006 and CL2355 as AD transgenic strains. All the strains used in this experiment were grown in a temperature-controlled incubator at 16 °C on solid nematode growth medium (NGM) seeded with Escherichia coli (E. coli) OP50 as a food source. Bleaching technique was used to obtain the age synchronized eggs population by treating gravid adults with a bleach-containing alkaline hypochlorite solution (0.5 M NaOH, ∼2.6% NaClO) for 5 min. Fertilized eggs were spun down and resuspended in S-medium for 12 h. This way, the first larval stage (L1) was able to grow in the absence of nutrients overnight.

4.7. Compounds and Reagents Used for the Biological Assay

Rhodiola rosea extract, salidroside (Purity≥98%, Cat.# 43866, Sigma Aldrich, St. Louis, MO, USA) and rosavin (Purity≥98%, Cat.# SML0336, Sigma Aldrich, St. Louis, MO, USA) were serially diluted in a range between 10 mg/mL to 0.01 mg/mL and 10 mM to 0.01 mM in 100% dimethyl sulfoxide (DMSO, Sigma Aldrich, St. Louis, MO, USA), respectively. Each concentration was then diluted in MilliQ purified water to obtain a final concentration ranging between 100 µg/mL to 1 µg/mL and 100 µM to 1 µM in 1% DMSO in well. For liquid-culture assays, treatments were performed for 4 days at 20 °C. Each well contained a final volume of 60 µL and comprised 25-35 synchronized L1-stage animals, the test compound at the appropriate dose, and E. coli OP50 bacteria inactivated by freeze-thaw cycles and suspended in S-medium to a final optical density at 595 nm (OD₅₉₅) of 0.8-0.9 (Benchmark™ Plus Microplate Reader, Bio-Rad, Hercules, CA, USA). For solid-media assays (chemotaxis and other plate-based experiments), synchronized CL2355 and control CL2122 worms were exposed to salidroside, rosavin, or R. rosea extract on fresh NGM plates seeded with inactivated E. coli starting at the L1 stage; plates were incubated at 16 °C for 36 h and then at 23 °C for an additional 36 h.

4.8. Locomotor Activity Assay

CL2006 worms were treated as described above in liquid culture for four days, starting at the L1 stage. In day 4, the 96-well microplate was read by Wmicrotracker^® MINI and SMART (PhylumTech S. A., Santa Fe, Argentina). Activity counts measured for 20 min were obtained by using WMicrotracker MINI Software V1.4 API. Travelled distances measure for 5 min were obtained by using WMicrotracker SMART Software.

4.9. Reproductive Capacity

CL2006 and N2 adult worms were treated as described above in liquid culture for four days, starting at the L1 stage. To assess egg laying, 10 adult worms were transferred onto fresh NGM plates at 20 °C for 6 h. Unhatched eggs were counted after 6 h. Adults at day 5 of age (day 4 after the start of treatment) corresponded to the stage known as “young adults.”

4.10. Chemotaxis Assay

Nematodes were collected after their respective treatments and washed with M9. Briefly, the assay was performed in 100 mm NGM plates, 10 µL of odorant (0.5% benzaldehyde in 96% ethanol) (CAS# 100-52-7, Sigma Aldrich, St. Louis, MO, USA), along with 1 M of sodium azide (CAS# 26628-22-8, Sigma Aldrich, St. Louis, MO, USA) added to the “attractant” spot. On the opposite side, 10 µL of control odorant (96% ethanol) along with 1 M of sodium azide were added. Immediately after, 50-60 worms were placed towards the center of the plate. Assay plates were incubated at 23 °C for 1 h and were scored according to the chemotaxis index (CI) as follows: CI = (number of worms at attractant−number of worms at control)/total number of worms. In each experiment, at least 60 worms from each group were analyzed.

4.11. RNA Extraction and Gene Determination

Total RNA was isolated from treated C. elegans samples using TRIzol^® reagent following the manufacturer’s protocol (Bioline Reagents, London, UK). RNA yield, purity, and quality were assessed using a NanoDrop™ ND-1000 spectrophotometer (Thermo Fisher, Waltham, MA, USA).

For reverse transcription, 2 μg of mRNA was converted to cDNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA). Real-time PCR was performed using SYBR^® Green chemistry on a StepOnePlus Detection System (Applied Biosystems). The reaction mixture contained 6.75 μL of cDNA (2 μg), 0.75 μL of each primer (100 nM), and 6.75 μL of SYBR Green PCR Master Mix (2×) (Applied Biosystems). Data analysis was conducted using the comparative cycle threshold (ΔΔCt) method, with a housekeeping gene used to normalize variations in sample loading and preparation.

Table 4. List of primers used in this work.

Gene	Forward (5’→3’)	Reverse (5’→3’)
skn-1	CCCAACATCCAACTACGCCTC	GTCTGCTGTTGACGTCCTGA
sek-6	CATTCTCCTTCTACAAAATGTACGC	AGCCCAAGAATCCTCTTCGC
daf-2	CCTCATCATTACTCAAACCAATAT	GTTACACTCGGTGCTCAGT
sod-1	TGTCGTTGGCCGATCTATGG	GTGGAGATTCAGAGACGCGA
act-1	ATCACCGCTCTTGCCCCATC	GGCCGGACTCGTCGTATTCTTG

Table 4. List of primers used in this work.

Gene	Forward (5’→3’)	Reverse (5’→3’)
skn-1	CCCAACATCCAACTACGCCTC	GTCTGCTGTTGACGTCCTGA
sek-6	CATTCTCCTTCTACAAAATGTACGC	AGCCCAAGAATCCTCTTCGC
daf-2	CCTCATCATTACTCAAACCAATAT	GTTACACTCGGTGCTCAGT
sod-1	TGTCGTTGGCCGATCTATGG	GTGGAGATTCAGAGACGCGA
act-1	ATCACCGCTCTTGCCCCATC	GGCCGGACTCGTCGTATTCTTG

4.12. Determination of ROS LEVEL

The ROS levels in C. elegans models were determined using 2’,7’-dichlorofluorescein diacetate (DCFH-DA), essentially as described in [67]. Briefly, CL2006 nematodes (80 per group) treated with salidroside, rosavin, R. rosea extract, or vehicle were collected 96 hours after the start of treatment at 20 °C and homogenized on ice in 100 µL of PBS containing 0.1% Tween-20. After adding 100 µL of 0.05% DCFH-DA, which is oxidized by ROS to the fluorescent compound 2’,7’-dichlorofluorescein (DCF), the samples were transferred into black 96-well plates and analyzed using a Varioskan microplate reader. Fluorescence was measured over 160 minutes at 10-minute intervals (excitation: 485 nm; emission: 538 nm). Relative ROS levels were expressed as DCF fluorescence intensity.

4.13. Thioflavin-S Staining Aβ Aggregation

Adult CL2006 C. elegans were treated with the selected dose for 4 days, then fixed in 4% paraformaldehyde (PBS, pH 7.5) for 24 h at 4 °C. Worms were permeabilized in 5% β-mercaptoethanol, 1% Triton X-100, and 125 mM Tris (pH 7.5) at 37 °C for 24 h, then stained with 0.125% Thioflavin S (CAS# 1326121, Sigma Aldrich, St. Louis, MO, USA) in 50% ethanol for 2 min, destained for 2 min, and washed with PBS. Samples (~10 μL) were mounted in Fluoromount G (Electron Microscopy Sciences, Cat.#17984 25) and imaged using a Leica Thunder fluorescence microscope (40× objective). Fluorescence images were processed in ImageJ/Fiji. Each worm head region was defined, converted to greyscale, and segmented using a Triangle threshold. Background was corrected (rolling ball radius = 10 px), smoothed (Median filter, 1 px), and re-thresholded by the Otsu method. Binary images were refined with Open and Watershed functions to isolate Aβ plaques. Plaques were quantified and normalized to head area (plaque density). Approximately 25 worms per condition were analyzed across three independent experiments, providing sufficient statistical power to detect changes in Aβ aggregation.

4.14. Statistics

All statistical analysis related to C. elegans studies were conducted using GraphPad Prism GraphPad Prism Version 10.6.1 statistical software. Groups were compared with one-way analysis of variance (ANOVA), followed by the Dunnet post hoc test. In some cases, comparison between groups was also performed by t-test student. Data are expressed as the mean ± standard error of the mean (SEM). Statistical significance was considered when p values were <0.05. The statistical outliers were determined with Grubs’ test and removed from the analysis. The Kruskal-Wallis test followed by Dunn’s post hoc analysis was used for non-normally distributed data.

For the LC/MS data was applied MZmine analysis. Detection was performed for the natural population using noise levels for MS1 and MS2 of 1500 and 100, respectively. The parameters for chromatogram builder were: minimum consecutive scans 5, minimum intensity for consecutive scans 1500, minimum absolute height 2000, m/z tolerance 0.015 Da or 15 ppm. Chromatogram deconvolution parameters were: threshold 90%, minimum search range (retention time) 0.025, minimum relative height 0.00%, minimum absolute height 2000, minimum ratio top/edge 2.6, peak duration range 0.04-1, minimum scan 5, MS2 scan pairing 0.03 Da or 30 ppm, minimum relative feature height 25%, minimum required signal 1. The isotope peak grouper parameters were: m/z tolerance 0.015 Da or 15 ppm, retention time tolerance 0.07, maximum number of charges 2, most intense isotope. The joint aligner parameters were: m/z tolerance 0.015 Da or 15 ppm, retention time tolerance 0.1, weight for m/z 50 and weight for RT 50. Settings for feature list filtering were: minimum isotopes 2.

MZmine data were loaded into Global Natural Product Social Molecular Networking (GNPS) website with a minimum cosine set to 0.7 for compound identification from publicly available databases [68].

5. Conclusions

The present study provides evidence that R. rosea extract and its major constituents exert neuroprotective effects in C. elegans models of AD. The observed improvements in locomotion, chemotaxis, egg-laying, ROS burden, stress-gene expression, and amyloid-like aggregation suggest that these compounds act on several interconnected pathological processes rather than a single endpoint. This multi-parameter profile is consistent with the pharmacological nature of botanical extracts, which often contain multiple active molecules with overlapping or complementary mechanisms. Overall, the multi-parameter dataset, including behavioral, biochemical, molecular, and histological data converges on the finding that R. rosea and its marker compounds activate the SKN-1/Nrf2 antioxidant pathway to reduce AD-related traits in C. elegans.

Concerning the huge consumption of not only R. rosea products, but plant-derived products in general our finding could have also a practical implementation for reliable estimating the quantitative features of the dose response. The hormetic framework of R. rosea extract, salidroside and rosavin reveal their modest stimulatory response, which seems to be between 30-60% over the control group in doses at least 50% below the toxicological/pharmacological threshold. However, future research should focus on systematic dose-response studies, and validation in additional AD models, such as tau-expressing C. elegans strains or mammalian systems to enhance translational relevance.