I. Introduction

Asparagus officinalis L. is widely consumed as a vegetable crop, and increasing attention has also been directed toward its underground parts as potential sources of functional materials. Recent studies have shown that asparagus roots contain diverse bioactive constituents, including flavonoids, saponins, polyphenols, organic acids, and minerals, and that these compounds are associated with antioxidant, anti-inflammatory, antimicrobial, antidiabetic, and other pharmacological activities (Guo et al., 2023;Olas 2024). In addition to their pharmacological relevance, asparagus roots have recently been recognized as a valuable agricultural by-product rather than a simple cultivation residue. In particular, the roots of A. officinalis have been reported to be rich in fructans and other functional components, suggesting their potential use as raw materials for food, nutraceutical, and industrial applications (Viera-Alcaide et al., 2022). This renewed interest is important because the effective utilization of root materials may expand the value of asparagus beyond the edible spear and support the development of high-value functional resources.

Nevertheless, the presence of bioactive compounds and potential health benefits does not by itself guarantee safety. Botanical materials intended for food or functional use should be evaluated individually because their safety depends on factors such as identity, composition, processing, intended use, and exposure level (Schilter et al., 2003;Roe et al., 2018). In particular, the chemical profile of a botanical extract may vary substantially according to the plant part used and the extraction procedure applied, and such differences can influence the interpretation of toxicological outcomes (Roe et al., 2018;Heinrich et al., 2022). This point is especially relevant to A. officinalis because different plant parts have been reported to contain distinct bioactive constituents, and asparagus roots have recently attracted attention as a phytochemically valuable material beyond the edible spear (Guo et al., 2023;Olas 2024). Therefore, systematic safety assessment is necessary before root-derived asparagus preparations are considered for food or functional applications.

Previous studies on herbal and root-derived extracts have demonstrated that genotoxicity cannot be assumed to be uniformly absent even when the materials are traditionally used. For example, Hwanglyeonhaedok-tang water extract showed positive findings in parts of the bacterial reverse mutation and chromosomal aberration assays, suggesting that some herbal preparations may require careful interpretation across multiple test systems (Jin et al., 2017). Likewise, root extracts of Paeonia lactiflora showed differential responses between in vitro and in vivo genotoxicity assays, further supporting the need for case-by-case safety evaluation of botanical resources (Bak et al., 2023).

Therefore, the present study was conducted to obtain basic safety data for a water extract prepared from the root of A. officinalis L. To evaluate its safety as a potential food or functional material, acute oral toxicity and a standard genotoxicity battery, including a bacterial reverse mutation assay, an in vitro chromosomal aberration test, and an in vivo micronucleus test, were performed in accordance with internationally accepted guidelines.

II. Materials and Methods

1. Preparation of asparagus root extract

The roots of 5-year-old Asparagus officinalis L. ‘Atlas’ cultivated in Yanggu-gun, Korea, were harvested in September and obtained from a local farm in Yanggu-gun, Gangwon state, Korea. The identity of the plant material was confirmed by Won-kyung Lee, Gangwon State Agricultural research and Extension Services and a voucher specimen (No. GW330001) was deposited at Agro-food research Institute, Chuncheon, Korea. After removal of the edible shoots, only the roots were used for extraction. The roots were washed, dried, ground, and passed through an 80-mesh sieve using a grinder (DA280-S, Daesung Artlon, Paju, Korea). A 100 g portion of the powdered roots was mixed with 50 volumes of distilled water and extracted at 98 ± 2°C for 6 hr using a heating mantle (MS-EAM, MTOPs, Yangju, Korea). The extract was centrifuged at 3,570 ×g for 15 min (SUPRA 22K, Hanil, Incheon, Korea), and the supernatant was filtered through Whatman filter paper (Maidstone, Kent, UK). The filtrate was freeze-dried (PVTF20R, Ilshinbiobase, Dongducheon, Korea) to obtain a lyophilized asparagus root extract (ARE). The extraction yield was 14.4% (w/w) based on the dried root powder. The final extract, designated as ARE, was a yellowish-brown powder and was stored at room temperature in the shade until use. For basic phytochemical characterization, caffeic acid was used as a marker compound and analyzed by HPLC with a photodiode array detector (SLC-40 system, Shimadzu, Japan) using a Unison UK-C18 column (100 × 2 mm, 3 µm; Imtakt) maintained at 30°C. The mobile phase consisted of 1% formic acid in water and acetonitrile/methanol/water (80/10/10, v/v/v) mixed at 80:20 (v/v), with a flow rate of 0.3 mL/min, an injection volume of 2 µL, and detection at 323 nm. Accordingly, ARE used in this study was a standardized material, with caffeic acid employed as a marker compound for quality control.

2. Acute oral toxicity study

Acute oral toxicity was evaluated according to Organisation for Economic Co-operation and Development (OECD) Test Guideline 423 under Good Laboratory Practice (GLP) conditions at the Korea Testing and Research Institute. Specific pathogen-free (SPF) female Sprague-Dawley (SD) rats [Cr:CD(SD)] were used. Female rats were selected in accordance with OECD Test Guideline 423, which recommends the use of females as the default sex because females are generally considered slightly more sensitive when sex-related differences are observed. The animal experiment was approved by the Institutional Animal Care and Use Committee (IACUC No. IAC2021-2791). Animals were acclimated under controlled environmental conditions and were given free access to rodent diet and water except during the fasting period before dosing.

A total of 12 female rats were assigned to 4 sequential steps, with 3 animals in each step. ARE was suspended in sterile water for injection and administered once by oral gavage at a dose volume of 10 mL/kg body weight after approximately 18 hr of fasting. The dose levels were 300 mg/kg body weight in the first and second steps and 2,000 mg/kg body weight in the third and fourth steps. Mortality and clinical signs were monitored for 14 days after administration. On the day of dosing, animals were additionally observed at 0.5, 1, 2, 3, and 4 hr after treatment. Body weight was measured before administration and on Days 1, 3, 7, and 14 after dosing. At the end of the observation period, all surviving animals were necropsied after exsanguination under isoflurane anesthesia, and gross findings were recorded.

3. Bacterial reverse mutation assay

The mutagenic potential of the ARE was evaluated using a bacterial reverse mutation assay in accordance with OECD Test Guideline 471 under GLP conditions. Four histidine-dependent strains of Salmonella typhimurium (TA100, TA1535, TA98, and TA1537) and one tryptophan-dependent strain of Escherichia coli (WP2uvrA) were used. The assay was conducted by the pre-incubation method both in the presence and absence of metabolic activation with S9 mix. Sterile distilled water was used as the negative control. Furylfuramide (AF-2), sodium azide (NaN₃), 9-aminoacridine (9-AA), 2-aminoanthracene (2-AA), and benzo(a)pyrene (B(a)P) were used as positive controls depending on the tester strain and metabolic activation condition.

A concentration range-finding test was performed at 50, 150, 500, 1,500, and 5,000 µg/plate. Since no precipitation or microbial toxicity was observed, the main test was conducted at 313, 625, 1,250, 2,500, and 5,000 µg/plate. For each treatment condition, ARE, bacterial suspension, and S9 mix or buffer were pre-incubated at 37°C for 20 min before plating. Three plates were used for each concentration. A result was considered positive when a reproducible and concentration-dependent increase in revertant colonies reached at least 2-fold that of the negative control in 1 or more tester strains.

4. In vitro mammalian chromosomal aberration test

An in vitro mammalian chromosomal aberration test was performed according to OECD Test Guideline 473 under GLP conditions using Chinese hamster lung (CHL/IU) cells. Sterile distilled water was used as the negative control. Mitomycin C (MMC) was used as the positive control in the absence of metabolic activation, whereas cyclophosphamide monohydrate (CPA) was used as the positive control in the presence of S9 mix. In the concentration range-finding test, 5,000 µg/mL was set as the top concentration in accordance with the applied guideline. Based on the results of the preliminary test and the relative increase in cell counts, the concentrations for the main test were selected separately for each treatment condition.

Cells were treated under 3 conditions: short-term treatment for 6 hr followed by 18 hr recovery in the absence of S9 mix, short-term treatment for 6 hr followed by 18 hr recovery in the presence of S9 mix, and continuous treatment for 24 hr without S9 mix. Based on the preliminary range-finding test and relative increase in cell counts, the main treatment concentrations were 812.5, 1,625, 3,250, 3,500, and 3,750 µg/mL for short-term treatment without S9; 875, 1,750, 3,500, 4,000, and 4,500 µg/mL for short-term treatment with S9; and 700, 1,400, 2,800, 3,000, and 3,200 µg/mL for continuous treatment without S9. For short-term treatment without S9, the 3,750 µg/mL group was treated but excluded from chromosomal aberration scoring because of cytotoxicity. Two hr before harvest, 0.2 µg/mL Colcemid was added to arrest cells at metaphase. The harvested cells were stained with 5% Giemsa solution. A total of 300 metaphases per concentration, including 150 metaphases from each of 2 culture flasks, were scored for structural and numerical chromosomal aberrations. Frequencies of aberrant cells below 5% were considered negative, frequencies of 5-10% were considered inconclusive, and frequencies above 10% with dose dependency were considered positive.

5. In vivo micronucleus test

An in vivo mammalian erythrocyte micronucleus test was conducted according to OECD Test Guideline 474 under GLP conditions using SPF male Institute of Cancer Research (ICR) mice. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC No. IAC2021-2897). In the main study, the animals were 7 wk old and weighed 32.07-35.92 g at administration. Five animals were assigned to each group. A preliminary dose range-finding test was conducted at 0, 500, 1,000, 1,500, and 2,000 mg/kg body weight/day. As no mortality or treatment-related clinical signs were observed, 500, 1,000, and 2,000 mg/kg body weight/day were selected for the main study.

The ARE was administered orally twice at 24 hr intervals. The vehicle control group received the vehicle alone, and the positive control group received cyclophosphamide monohydrate at 70 mg/kg by intraperitoneal injection. Bone marrow cells were collected from the femurs approximately 24 hr after the final administration. Smears were prepared and stained with acridine orange. A total of 4,000 polychromatic erythrocytes (PCEs) per animal were screened to identify and enumerate micronucleated polychromatic erythrocytes (MNPCEs). In addition, the ratio of PCE to total erythrocytes [PCE/(PCE+NCE)] was calculated from 500 erythrocytes per animal as an indicator of bone marrow toxicity.

6. Statistical analysis

In the acute oral toxicity study, body weight data were analyzed by one-way analysis of variance (ANOVA) followed by Duncan's multiple range test. Mortality, clinical signs, and gross pathological findings were evaluated descriptively in accordance with OECD Test Guideline 423 and the Globally Harmonized System classification criteria. In the bacterial reverse mutation assay, the numbers of revertant colonies were evaluated according to OECD guideline criteria based on reproducibility, dose dependency, and comparison with the concurrent negative control; a result was considered positive when a reproducible and concentration-dependent increase reached at least 2-fold that of the negative control in one or more tester strains. In the in vitro chromosomal aberration test, the frequencies of cells with structural or numerical chromosomal aberrations were evaluated according to OECD guideline criteria based on comparison with the concurrent negative control, reproducibility, and dose dependency. Frequencies of aberrant cells below 5% were considered negative, frequencies of 5–10% were considered inconclusive, and frequencies above 10% with dose dependency were considered positive. In the in vivo micronucleus assay, differences in micronucleated polychromatic erythrocyte (MNPCE) frequency between treated groups and the vehicle control group were analyzed using the Kruskal–Wallis H-test, and the positive control group was compared with the vehicle control group using the Mann–Whitney U-test. The PCE/(PCE+NCE) ratio was analyzed by one-way ANOVA followed by Dunnett's test. Statistical significance was set at p<0.05.

III. Results and Discussion

1. Acute oral toxicity

In the acute oral toxicity study, no treatment-related mortality or abnormal clinical signs were observed throughout the observation period, including at 0.5, 1, 2, 3, and 4 hr after dosing, and body weight increased progressively in all treated animals during the 14-day observation period <Table 1>. No remarkable differences in body weight were observed between dose groups. Because body weight is one of the most sensitive general indicators of systemic toxicity, the absence of treatment-related suppression of body weight gain suggests that ARE did not induce overt acute toxic effects under the present experimental conditions. No treatment-related gross findings were observed at necropsy. This interpretation is consistent with previous safety studies of plant-derived materials. Tang et al. (2017) reported that Cajanus cajan leaf extracts did not induce mortality or meaningful toxic signs even at high oral doses, and Ferreira et al. (2022) also showed that aqueous extract of Eugenia uniflora leaves did not produce relevant acute or subacute in vivo toxicity. Taken together, these findings suggest that ARE, like these botanical extracts, may have a relatively wide safety margin in acute oral exposure. However, the absence of overt acute toxicity alone is not sufficient to exclude mutagenic or clastogenic risk. For that reason, the acute toxicity result in the present study should be interpreted as an initial indication of safety that requires confirmation through a standard genotoxicity battery.

2. Bacterial reverse mutation assay

In the bacterial reverse mutation assay, ARE did not increase the number of revertant colonies in Salmonella typhimurium TA98, TA100, TA1535, and TA1537 or in Escherichia coli WP2uvrA at any tested concentration up to 5,000 µg/plate, regardless of the presence or absence of metabolic activation <Table 2>. In contrast, the positive controls showed the expected increases in revertant colonies in the relevant strains, confirming that the assay system functioned appropriately. These findings indicate that ARE did not show mutagenic activity in this bacterial test system. No precipitation or bacterial growth inhibition was observed at any tested concentration up to 5,000 µg/plate. The present findings are comparable to previous reports for herbal formulas that were also negative in Ames tests. Guibi-Tang and So-ochim-tang-gamibang did not show mutagenic responses in standard bacterial reverse mutation assays, and these studies support the interpretation that the present negative result is toxicologically meaningful rather than incidental (Lee et al. 2014;Lee et al. 2018b). In addition, Oryeong-san extract and Dioscorea Rhizome water extract were likewise reported to be negative in Ames tests under standard conditions (Lee et al. 2015b;Park et al. 2021). These comparative data further support the view that ARE did not induce point mutations in the bacterial system used in this study. Nevertheless, a negative Ames result alone does not fully exclude genotoxic potential, because some herbal preparations have shown discordant findings in mammalian cell assays. Therefore, the present Ames test result should be interpreted together with the in vitro chromosomal aberration and in vivo micronucleus data.

3. In vitro chromosomal aberration assay

In the in vitro chromosomal aberration assay using CHL/IU cells, ARE did not induce any biologically relevant increase in either structural or numerical chromosomal aberrations under short-term treatment conditions with or without S9 mix or under continuous treatment conditions <Table 3>. In the short-term treatment without S9 mix, the 3,750 µg/mL group was treated but excluded from chromosomal aberration scoring because of cytotoxicity. At the other evaluated concentrations, although some reduction in cell growth was observed at higher concentrations, the frequencies of cells with structural and numerical chromosomal aberrations remained below 5% and did not show any dose-related increase. These findings indicate that ARE did not induce chromosomal aberrations in cultured mammalian cells under the present experimental conditions. This negative result is consistent with previous reports on Guibi-Tang and So-ochim-tang-gamibang, both of which were negative in chromosomal aberration assays as well as in complementary genotoxicity tests (Lee et al. 2014;Lee et al. 2018b). Oryeong-san extract and Dioscorea Rhizome water extract were also reported to be negative in mammalian chromosomal aberration assays, which further supports the present result (Lee et al. 2015b;Park et al. 2021). By contrast, So-Cheong-Ryong-Tang and Gyejibokryeong-hwan showed negative results in Ames and in vivo micronucleus assays but positive findings in in vitro chromosomal aberration assays at high concentrations (Lee et al. 2015a;Lee et al. 2018a). This contrast is important because it shows that a negative bacterial reverse mutation result does not necessarily exclude chromosomal effects in mammalian cells. This comparison contributes to the interpretation of the present findings in two ways. First, the agreement of the ARE results with those of other negative botanical preparations supports that the absence of chromosomal aberration was not incidental, but toxicologically meaningful. Second, the contrast with herbal preparations that were negative in bacterial or in vivo assays but positive in mammalian cell assays highlights the importance of evaluating ARE across multiple complementary test systems. Therefore, the present negative result in the chromosomal aberration assay strengthens the overall interpretation that ARE did not show genotoxic potential under the tested conditions.

4. In vivo micronucleus assay

In the in vivo micronucleus assay, the frequencies of micronucleated polychromatic erythrocytes counted among 4,000 polychromatic erythrocytes per animal were comparable to those of the negative control group, and no dose-related increase was observed <Table 4>. In addition, the PCE/(PCE+NCE) ratios in the treated groups were similar to that of the negative control, suggesting that ARE did not suppress erythropoiesis or induce bone marrow toxicity. The use of a standardized ARE preparation, quality-controlled with caffeic acid as a marker compound, supports the analytical reproducibility of the test material and strengthens the interpretation of the present findings. The validated HPLC-PDA method showed acceptable linearity over 0.1–10 µg/mL, and the caffeic acid content of the test substance was 1.21 mg/g. In addition, the method showed acceptable intra-day accuracy and precision, and the dosing formulations were confirmed to be homogeneous and stable for 4 hr at room temperature and for 7 days under refrigerated conditions (2–8℃). Nevertheless, because the present study employed a single marker compound for basic characterization, broader chromatographic fingerprinting and additional marker-based standardization would further strengthen the chemical characterization and batch-to-batch reproducibility of ARE in future studies. The positive control produced the expected increase in micronucleus frequency, confirming that the assay system was valid. These findings indicate that ARE did not induce micronucleus formation in vivo under the present experimental conditions. The present findings are in agreement with previous in vivo studies of botanical extracts and herbal formulas. Ferreira et al. (2022) reported that aqueous extract of Eugenia uniflora leaves did not increase micronucleus formation in vivo, and So-ochim-tang-gamibang also showed a negative micronucleus result in mice (Ferreira et al. 2022;Lee et al. 2018b). These reports support the interpretation that ARE did not exert clastogenic or aneugenic effects in vivo. Similarly, Oryeong-san extract, Gyejibokryeong-hwan, and Dioscorea Rhizome water extract were all reported to be negative in in vivo micronucleus assays (Lee et al. 2015b;Lee et al. 2018a;Park et al. 2021). Therefore, the present result is consistent with a broader pattern of negative in vivo genotoxic findings reported for several herbal preparations. More importantly, because the current study showed concordant negative findings in the bacterial reverse mutation assay, the chromosomal aberration assay, and the micronucleus assay, the overall evidence for genotoxic safety is stronger than that provided by any single test alone. In this context, comparisons with previous studies help place the present findings within a broader toxicological pattern observed for botanical extracts. Rather than serving as simple examples, these comparisons show that the interpretation of ARE is strengthened by the consistency of negative findings across the standard genotoxicity battery and by the absence of the assay-specific discordance reported for some other herbal preparations.

From a practical perspective, the present findings provide initial hazard information supporting the potential use of ARE as a food or functional material, but they do not by themselves define safe human intake levels. In actual applications, ARE may be consumed in different forms, such as powders, beverages, capsules, or as an ingredient in functional food formulations, and the resulting exposure levels may vary according to formulation, frequency of consumption, and target population. Therefore, further studies addressing repeated-dose oral toxicity and realistic intake scenarios will be necessary to establish use levels that are relevant to practical human exposure.

IV. Summary and Conclusion

This study was conducted to evaluate the safety of a water extract prepared from the roots of Asparagus officinalis L. through an acute oral toxicity study and a standard genotoxicity battery. In the acute oral toxicity test, no treatment-related mortality, abnormal clinical signs, suppression of body weight gain, or gross pathological findings were observed in female Sprague-Dawley rats at dose levels up to 2,000 mg/kg body weight. In the bacterial reverse mutation assay, ARE did not induce any biologically relevant increase in revertant colonies in any tester strain, with or without metabolic activation, at concentrations up to 5,000 µg/plate. In the in vitro chromosomal aberration test using CHL/IU cells, no biologically relevant increase in structural or numerical chromosomal aberrations was detected under any scorable treatment condition. In the in vivo micronucleus assay, no dose-related increase in micronucleated polychromatic erythrocytes or evidence of bone marrow toxicity was observed in male ICR mice. Taken together, these findings indicate that ARE was not associated with acute oral toxicity or genotoxicity in the test systems used in this study under the present experimental conditions. Academically, the present study provides foundational toxicological data for the root of Asparagus officinalis L., a relatively underexplored non-edible plant part compared with the edible spear, and supports the usefulness of a standard genotoxicity battery in the safety assessment of botanical extracts. Industrially, these findings suggest that asparagus roots, which are generally regarded as an agricultural by-product, may have potential as a source material for food or functional applications when supported by appropriate quality control and safety evaluation. In addition, the use of a standardized ARE preparation, quality-controlled with caffeic acid as a marker compound, supported the analytical reproducibility of the test material and strengthens its potential applicability as a food or functional material. However, further studies on repeated-dose toxicity and realistic human intake scenarios will be needed to support practical application of ARE as a food or functional material.