The impact of thermal pretreatment on phytochemical profiles and bioactivity of freeze-dried lily bulbs (Lilium lancifolium)

Abstract

Lily bulbs are susceptible to deterioration during storage if improperly handled. To resolve this problem, it is necessary to investigate suitable processing techniques. The aim of this study is to evaluate the effects of steaming, blanching and microwave pretreatment on freeze-dried lily bulbs in terms of color, phenolic content and bioactivity. Results showed that appropriate steaming and blanching pretreatment could contribute to product characteristics similar to those of freeze-dried lily bulbs, with the maximum L* value reduced by only 7.57% and 0.55% respectively. Thermal pretreatment affected the retention, degradation and transformation of polyphenol, especially for regalosides. The polyphenol was closely associated with the browning of lily bulbs. Thermal processing caused the decline of regaloside A and the increase of regaloside B, which were the major phenolic monomers that can effectively inhibit the browning of lily bulbs. The antioxidant activity of freeze-dried lily pretreated with blanching for 6 min was the highest (4.39 ± 0.32 μmol TE/g DW), with an improvement of nearly 25.39% compared to that of untreated freeze-dried lily. Thus, the combination of freeze-dried with steaming or blanching pretreatment could be proposed as a sustainable strategy to improve the quality of lily bulbs for industrial application.

Highlights

• •Appropriate thermal pretreatments improved the appearance quality of lily bulbs.
• •Regaloside A and B are the major phenolic monomers in lily bulbs.
• •Changes of regaloside components were associated with the browning of lily bulbs.
• •Six minutes blanching gained the highest antioxidant activity in dried lily bulbs.

1. Introduction

Lilies are perennial herbs of the species Lilium in the family Liliaceae. Lilium lancifolium, known for its medicinal and culinary value, originates from Asia and in China is mainly found in Jiangsu, Zhejiang, Hunan and Anhui. The lily bulb, which appears bright white, has a nutritional value derived mainly from starch, protein, fibre, etc. and was an excellent source of active compounds such as polyphenols, saponins, alkaloids and polysaccharides (Tang et al., 2022). The natural products found in lily bulbs provide a range of benefits, including reducing cholesterol, anti-inflammatory, preventing oxidation, inhibiting tumour growth and delaying aging (Zhang et al., 2022).

Fresh lily bulbs often face problems such as browning, nutrient loss and microbial growth during storage, which can seriously affect their quality and commercial value (Yang et al., 2022). Therefore, it is necessary to process the lily bulbs in a proper way to minimize the negative effects on their physical and chemical properties and to prolong their shelf life. Drying is one of the most traditional methods of preserving food and significantly reduces the moisture activity of food, allowing it to be stored for long periods without deterioration. As a mild dehydration method, freeze-dried removes water mainly by sublimation. The food products that are subjected to freeze-dried retain a high level of vitamins and active ingredients, thus ensuring an excellent product quality (Bhatta, Stevanovic Janezic, & Ratti, 2020). Zhang et al. reported that freeze-dried Angelica sinensis exhibit superior quality compared to those prepared by hot air dried or microwave dried, effectively maximizing the retention of Z-ligustilide (13.91 mg/g), ferulic acid (1.82 mg/g) and other compositions (Zhang et al., 2021). Research has shown that freeze-dried has a compensatory effect on Gastrodia elata treated by steaming, increasing the content of gastrodin and p-hydroxybenzyl alcohol and restoring to a certain extent the antioxidant activity (Wu et al., 2022). However, drying, especially non-thermal drying techniques such as freeze-drying, only inhibits microorganisms growth by reducing the water activity of the food (Wang, Liu, Zhang, Li, & Ding, 2023). Some microorganisms that can survive in low water activity environments, such as Salmonella and Staphylococcus aureus, pose a serious threat to food safety (Chitrakar, Zhang, & Adhikari, 2019). Therefore, freeze-drying does not have a highly effective sterilization function and it is necessary to combine it with other heat treatment methods for sterilization to ensure the safety of dried foods.

The sensory characteristics and nutritional value of dehydrated foods are often impacted by pretreatment or drying parameters. Blanching at 95 °C or 100 °C can inactivate polyphenol oxidase and peroxidase, delaying the browning of fresh lily bulbs (Wang, Zhang, Ju, Mujumdar, & Yu, 2023). Liu et al. found that the brightness of lily bulbs blanched with steam in combination with air drying at 50 °C was higher than that of other samples, and that steaming for 1 min could effectively inhibit enzymatic browning (Liu et al., 2019). Compared to hot air drying, hybrid drying can significantly improve the total phenol content retention of Kedondong (p < 0.05) and show significantly higher antioxidant capacity (Ee, Hii, Ong, Law, & Tan, 2022).

However, current research on dried lily bulbs has mostly focused on the degree of browning and physicochemical properties, with little research on changes in biological activity. Furthermore, the quality change of freeze-dried lily bulbs in combination with pretreatment has not yet been reported. Therefore, the aim of this study was to investigate the effects of different pretreatments combined with freeze-dried on the appearance, phenolic compounds and their biological activities in lily bulbs. The research results can be used as a reference for optimizing the drying of lily bulbs in industrial application.

2. Materials and methods

2.1. Lily bulb sample preparation

Lily bulbs (Lilium lancifolium Thunb.) were cultivated at Tibet Agricultural and Animal Husbandry University. The bulbs were peeled and divided into ten equal parts, each weighing 500 g. After pretreatment by steaming, blanching (100 °C) and microwaving (500 W power) for a certain period of time, respectively, the lily bulbs were subjected to rapid freezing with liquid nitrogen and then were placed in a freeze dryer (Alpha 2–4 LD plus, Marin Christ Co., Ltd., Osterode am Harz, Germany). The freeze-dried process was maintained at 0.07 mbar vacuum and − 85 °C ice condenser temperature for 24 h. Freeze-dried lily bulbs without pretreatment were used as a control. The sample were coded as follows: steamed (5 min)-freeze-dried (S5—F), steamed (10 min)-freeze-dried (S10—F), steamed (15 min)-freeze-dried (S15—F), blanched (2 min)-freeze-dried (B2—F), blanched (4 min)-freeze-dried (B4—F), blanched (6 min)-freeze-dried (B6—F), microwaved (2 min)-freeze-dried (M2-F), microwaved (5 min)-freeze-dried (M5-F), microwaved (10 min)-freeze-dried (M10-F), freeze-dried (CK).

2.2. Color determination

The color of each sample was measured using the software Image J (National Institutes of Health, Bethesda, Maryland, USA). The color values of the lily bulbs were expressed in terms of L* (brightness/darkness), a* (redness/greenness) and b* (yellowness/blueness). The formula for calculating the total color difference, which was used to evaluate the color change of the lily bulbs, was as follows:

$$\mathbf{\Delta E}=\sqrt{{\left(\mathit{L}-{\mathit{L}}_{\mathbf{0}}\right)}^{\mathbf{2}}+{\left(\mathit{a}-{\mathit{a}}_{\mathbf{0}}\right)}^{\mathbf{2}}+{\left(\mathit{b}-{\mathit{b}}_{\mathbf{0}}\right)}^{\mathbf{2}}}$$

where L₀, a₀, b₀ are the color parameters of the freeze-dried lily bulbs(CK); L, a and b are the color parameters of the bulbs with pretreatment.

2.3. High-performance liquid chromatography (HPLC) analysis

Phenolic extracts were analyzed by HPLC (Waters Corporation, Milford, MA, USA) on a Waters Sunfire C18 column (250 mm × 4.6 mm, 5 μm). The chromatographic conditions were as follows: mobile phase A, water (including 0.1% TFA); mobile phase B, acetonitrile; injection volume, 20 μL; column temperature, 35 °C; flow rate, 1.0 mL/min and detection wavelength, 320 nm. The gradient elution program was as follows: 0 min 90% A; 5 min 83% A; 8 min 80% A; 12 min at 76% A; 15 min 72% A; 20 min at 5% A; 21–23 min 90% A. The individual components were quantified using the external calibration curves of standards. A set of dilutions were prepared from a standard stock solution to create a calibration curve (10–200 μg/mL). The coefficient of determination (R²) was consistently >0.99. The Limit of Detection (LOD) and Limit of Quantitation (LOQ) were determined based on signal-to-noise ratios (S/N) of 3 and 10, respectively.

2.4. Antioxidant activity determination

The 2,2′-Azinobis-(3-ethylbenzthiazoline-6-sulphonate) Radical Scavenging Capacity (ABTS) method to determine the antioxidant capacity of the extracts was performed with a T-AOC (Total antioxidant capacity) assay kit (Beyotime Biotechnology, Shanghai, China). The absorbance of the sample was measured and recorded at a wavelength of 734 nm. The Oxygen Radical Absorbance Capacity (ORAC) method was used for the determination of the antioxidant capacity of the lily bulb extract, with slight modifications as previously reported (Wang et al., 2016). 20 μL of Trolox standard or samples diluted with 75 mM phosphate buffer and 200 μL of 0.96 μM fluorescein were added to a 96-well plate, incubated for 20 min at 37 °C and then 20 μL of 119 mM 2,2-Azobis(2-amidinopropane) (ABAP) was added. Fluorescence decay was determined using a FilterMax F5 Multi-Mode Microplate Reader for 35 cycles every 4.5 min at 485 nm excitation and 535 nm emission. The measurements mentioned above are based on Trolox as the standard and the data were reported as micromoles of TE per gram of dry weight (μmol TE/g DW) with three replicates.

2.5. Cell culture

The Human Hepatocellular Carcinoma Cells (HepG2) were cultured in William's Medium E (WME) (500 mL) supplemented with 25 mL of fetal bovine serum (FBS), 5 mL of l-glutamine, 5 mL of HEPES buffer, antibiotics (50 units/mL penicillin and 50 μg/mL streptomycin), 5 μg/mL insulin, and 0.05 μg/mL hydrocortisone. The cells were incubated at 37 °C in 5% CO₂ to a confluence of 90–95%.

2.6. Cell cytotoxicity

The cytotoxicity of HepG2 cells was measured by the methylene blue assay as previously described (Cheng, Xiang, Cheng, Chen, & Guo, 2022). Briefly, the cells were cultivated at 2.5 × 10⁵ cells/mL on a 96-well plate in 100 μL of complete medium. After incubation for 24 h at 37 °C, the medium was removed by 100 μL of phosphate buffered saline (PBS) and 100 μL of complete medium containing ethanol extract of lily bulbs at different concentrations was added. Then discarded the medium after 24 h and added 50 μL methylene blue solution to each well. The plates were incubated for 1 h, then removed the dye and added 100 μL of elution solution. The absorbance was recorded at 595 nm by a microplate reader (Molecular Devices, Sunnyvale, CA, USA). The cytotoxicity was calculated from the half-maximal cytotoxic concentration (CC₅₀), which were measured in triplicate for accuracy and stated as mg per millilitre dry weight (mg/mL DW).

2.7. Cell Antiproliferation

The determination of antiproliferative activity was in accordance with the method previously reported (Cheng et al., 2022). The cells were cultured at a density of 1.5 × 10⁵ cells/mL in a plate for 4 h, after which the medium was replaced with samples at different concentrations. HepG2 cells were then cultured for 72 h. After that, the absorbance of the sample was recorded at 595 nm. The antiproliferative activity was calculated from the half-inhibitory concentration (IC₅₀), which are stated as mg per mL of dry weight (mg/mL DW) in triplicate.

2.8. Statistical analysis

All data were represented as the mean ± SD of three analyses for each sample. Statistical analysis was accomplished by Origin 2022 (Origin Lab Corporation, Northampton, MA, USA). Significance tests were analyzed using ANOVA followed by Duncan's test using SPSS 26 software (SPSS Inc., Chicago, 1 L, USA). The significant differences have been compared at a p-value of <0.05.

3. Results

3.1. Color changes

In comparison with the control group (freeze-dried lily bulbs), each of the treatment groups shows varying degrees of browning (Fig. A1). The degree of color change in the combination of blanching and freeze-dried treatment group (B—F) decreased with increasing blanching time compared to the CK group. The combination of microwave and freeze-dried treatment group (M-F) showed the most significant browning, while the combination of steaming and freeze-dried treatment group (S—F) showed relatively small color changes.

Supplementary Fig. A1.

Color changes in lily bulbs with different pretreatments.

Table 1 shows the variation of the color parameters of lily bulbs. The data show that the different processing methods reduce the L* values to different extents compared to CK, with the L* values increasing as the pretreatment time decreases. B6—F (0.55%) shows the smallest decrease in L* values, followed by S15—F (7.57%) and M2-F (39.38%) the largest. With the exception of B6—F and S10—F, the a* values of all samples were higher than those of the control group. M5-F has the highest a* values, which is 8.7 times that of CK, while S10—F has the lowest, which is only 0.37 times that of CK. In the B—F group, the a* values gradually decreased with increasing blanching time, while the minimum value showed no significant difference from CK (p < 0.05). The b* values in the experimental group ranged from 12 to 31, all higher than CK. In the S—F group the b* values decreased with increasing steaming time, with S15—F, the lowest value, showing an increase of almost 50.23% over CK. The group with the lowest b* value of all groups was M10-F, followed by B4—F, which increased by 2.51% and 4.62% respectively. The total color difference (ΔE) of the M-F group is the largest among all samples, which means the most significant color difference from CK. Interestingly, the microwave time difference within the group had no significant effect on the overall color difference. For the samples treated with blanching and steaming, the ΔE value decreased with increasing time, with B6—F having the lowest ΔE value, followed by S—15F.

Table 1.

Effect of pretreatment conditions on color parameters.

Treatments	L* value	a* value	b* value	ΔE value
S5-F	60.15 ± 0.36 d	1.66 ± 0.24 gh	30.30 ± 0.59 a	20.71 ± 1.47 c
S10-F	63.27 ± 0.59 c	0.78 ± 0.89 h	26.68 ± 1.12 b	16.06 ± 1.37 d
S15-F	65.73 ± 0.82 b	3.29 ± 0.59 f	25.66 ± 1.35 bc	14.04 ± 2.73 de
B2-F	48.93 ± 0.61 f	10.08 ± 0.34 d	18.49 ± 0.53 c	24.28 ± 0.62 b
B4-F	58.42 ± 0.31 e	7.98 ± 0.34 e	13.36 ± 0.42 f	14.03 ± 0.31 de
B6-F	70.72 ± 0.17 a	1.68 ± 0.33 g	23.88 ± 1.39 b	11.13 ± 2.53 f
M2-F	43.11 ± 0.71 h	15.79 ± 0.38 b	16.33 ± 0.89 e	31.41 ± 0.96 a
M5-F	45.41 ± 0.43 g	18.28 ± 0.18 a	17.49 ± 0.75 de	30.75 ± 0.30 a
M10-F	45.28 ± 0.87 g	14.74 ± 0.09 c	13.09 ± 0.28 f	28.79 ± 0.70 a
CK	71.11 ± 0.30 a	2.10 ± 0.40 g	12.77 ± 1.33 f	0.00

L* represents the degree of brightness. a* represents the color difference between red (positive values) and green (negative values). b* represents the color difference between yellow (positive values) and blue (negative values). Different letters (a-h) within each column indicate a significant difference at p < 0.05.

3.2. Phenolic acid content changes

Six phenolic acids from lily bulbs were identified by HPLC system, namely, regaloside A, regaloside B, regaloside C, regaloside E, regaloside K, and p-coumaric acid, of which the first five are phenylpropanoid glycerol glucosides (Table 2). And in the study, the cumulative sum of detected phenolic acid content was designated as the total phenolic acid content. Freeze-dried lily bulbs subjected to microwave pretreatment showed a significant reduction in total phenolic acid content, which was only 0.29–0.34 times that of the control (CK). Conversely, the samples pretreated with steam showed a varying degree of increase in total phenolic acid content, with sample S10—F showing the most significant increase, reaching an increase of 25.84%.

Table 2.

Effect of pretreatment conditions on phytochemical contents (μg/g DW).

Treatments	Regaloside A	Regaloside B	Regaloside C	Regaloside E	Regaloside K	p-Coumaric Acid	Total phenolic acid
S5-F	858.47 ± 11.61 d	1332.45 ± 11.95 d	438.72 ± 11.57 c	997.55 ± 33.74 c	624.47 ± 3.00 d	34.77 ± 1.55 c	4286.43 ± 58.46 d
S10-F	1145.22 ± 17.17 b	1563.47 ± 10.21 a	459.59 ± 3.82 ab	1533.57 ± 17.01 a	696.68 ± 2.57 c	60.19 ± 9.27 b	5458.72 ± 51.47 a
S15-F	979.47 ± 16.50 c	1486.08 ± 15.30 b	478.82 ± 9.75 a	1178.31 ± 10.68 b	805.72 ± 13.56 b	40.71 ± 1.64 c	4969.11 ± 20.78 c
B2-F	348.30 ± 10.30 e	771.08 ± 4.85 g	327.54 ± 2.65 f	571.47 ± 17.95 e	277.51 ± 8.52 f	19.21 ± 0.99 e	2315.11 ± 35.24 e
B4-F	846.51 ± 18.11 d	1261.27 ± 11.56 e	403.00 ± 4.65 d	1207.70 ± 68.89 b	523.55 ± 5.38 e	168.09 ± 5.44 a	4410.12 ± 128.06 d
B6-F	1153.29 ± 3.54 b	1446.05 ± 3.24 c	444.46 ± 11.07 bc	1238.46 ± 36.17 b	846.59 ± 3.13 a	34.85 ± 1.09 c	5163.7 ± 59.37 b
M2-F	269.63 ± 6.06 f	460.25 ± 11.50 g	366.90 ± 9.04 e	174.73 ± 5.12 g	125.21 ± 3.77 h	9.54 ± 2.98 f	1406.25 ± 41.53 f
M5-F	187.89 ± 2.30 h	463.30 ± 4.92 g	291.78 ± 3.17 g	199.52 ± 14.90 g	114.01 ± 0.55 h	7.00 ± 0.15 f	1263.49 ± 19.19 g
M10-F	227.38 ± 9.62 g	534.12 ± 13.41 h	287.08 ± 8.70 g	280.13 ± 8.11 f	146.99 ± 2.91 g	12.52 ± 0.58 ef	1488.22 ± 47.61 f
CK	1306.86 ± 8.05 a	929.44 ± 24.71 f	363.28 ± 9.59 e	843.44 ± 33.10 d	834.63 ± 0.50 a	59.75 ± 1.58 b	4337.4 ± 76.47 d

Different letters (a-g) within each column indicate a significant difference at p < 0.05.

The highest concentration of regaloside A in freeze-dried lily bulbs (CK) was found among the identified phenolic acids, representing approximately 30% of the total phenolic acid, followed by regaloside B with 21.42%. Interestingly, we observed that regaloside B was the highest (28–36.67%) in phenolic extracts of lily bulbs after pretreatment, followed by regaloside E at 12.43–28.09%. The content of regaloside A in all samples showed varying degrees of reduction compared to CK, ranging from 11.75% to 85.62%, with the lowest content in the M-F groups being only 0.14–0.21 times that of the freeze-dried sample. In the B—F groups, the concentration of regaloside A increased with blanching time, whereas there were no significant differences between B6—F and S10—F, which were the highest in both the B—F and S—F groups. Apart from the B2—F and M-F groups, the concentrations of regaloside B, regaloside C and regaloside E all increased in the samples, with the highest increases of 68.22%, 31.80% and 81.82% respectively. The concentration of regaloside K ranged from 114.01 ± 0.55 to 846.59 ± 3.13 μg/g dry weight (DW), accounting for 8.90–19.24% of the total phenolic acid. The B6—F group showed no significant difference compared to CK, whereas the other samples exhibited a decreasing trend in pretreatment time. The highest content of p-coumaric acid in the sample was relatively low, amounting to only 3.81% of the total phenolic extracts. In particular, an increase was observed only in the content of p-coumaric acid for samples S10—F and B4—F, with the content in B4—F being up to 2.81 times higher compared to the other freeze-dried samples.

3.3. Antioxidant activity

The ABTS and ORAC values, which indicate the antioxidant capacity of phenolic extracts from lily bulbs, are shown in Fig. 1. The ABTS values varied from 4.39 ± 0.32 to 7.31 ± 0.12 μmol TE/g DW, with a maximum of 1.67 times higher than a minimum. The samples of the S5—F, S10—F, S15—F and B4—F groups were not significantly different from the CK group in terms of ABTS value. However, it was noteworthy that only the B6—F group was higher than CK by almost 25.39%. In the B—F and M-F groups, the ABTS value was influenced by the pretreatment time and was positively proportional thereto. However, for steaming pretreatment, there was a slight increase in ABTS with the duration of pretreatment, but no significant difference.

Fig. 1.

Effect of pretreatment conditions on antioxidant activity. (A) ABTS. (B) ORAC. Different letters (a-d) indicate a significant difference at p < 0.05.

In the ORAC assay, the highest ORAC value (95.75 ± 5.78 μmol TE/g DW) was found in S5—F compared to CK, while the lowest ORAC value was found in M5-F with 12.58 ± 0.73 μmol TE/g DW. Of all the pretreatments, steaming resulted in the least reduction in ORAC values, ranging from 12.2% to 48.9%. Meanwhile, the antioxidant activity of the microwave group was found to be the lowest, only about 0.12–0.19 times that of CK. It was also observed that there was no correlation between pretreatment time and ORAC value in the M-F groups.

3.4. Cell cytotoxicity and Antiproliferation

We have investigated the effect of lily bulb extract on the cytotoxicity and proliferation of HepG2 cells. The methylene blue assay was used to evaluate the effect of different concentrations of the sample on the growth of HepG2 cells and the results were presented in Table 3. The cytotoxicity and antiproliferative activity of lily bulb extract were expressed as CC₅₀ and IC₅₀, respectively. And the lower the value of CC₅₀ and IC₅₀, the higher the value of cytotoxicity and anti-proliferative activity. It was found that all samples had a dose-dependent toxic effect on HepG2 cells in the concentration range of 50–200 mg/mL. Moreover, the CC₅₀ values for lily bulb extract were all higher than the IC₅₀ values. For lily bulb extract pretreated with blanching, the CC₅₀ value is positively correlated with treatment time, and only the B6—F group is less than CK, which is also the minimum of all samples. The data show that the sample with minimal cytotoxicity is S10—F, with no discernible difference to B2—F. Among the samples, S15—F, B6—F and M2-F had the lowest CC₅₀ values in steaming, blanching and microwave pretreatment, respectively. The half-inhibitory concentration range of the freeze-dried lily bulb extract is from 20.59 ± 0.55 to 53.36 ± 4.98 mg/mL, with a mean value of 39.78 mg/mL. The inhibitory effects of freeze-dried lily bulbs on cell proliferation were attenuated to varying degrees after pretreatment. B6—F displayed the greatest inhibition of HepG2 cell growth in relative terms, with an IC₅₀ that was only 1.3 times that of CK.

Table 3.

Effect of pretreatment conditions on cell antiproliferative activities and cytotoxicity.

Treatments	Cytotoxicity CC50 (mg/mL)	Antiproliferation IC50 (mg/mL)
S5-F	117.37 ± 1.25 b	39.87 ± 1.21 de
S10-F	122.49 ± 1.27 a	53.36 ± 4.98 a
S15-F	101.81 ± 1.12 d	41.19 ± 0.36 cd
B2-F	121.85 ± 1.62 a	43.83 ± 0.65 bc
B4-F	92.66 ± 0.23 e	39.27 ± 0.73 de
B6-F	51.06 ± 0.25 h	26.85 ± 0.46 f
M2-F	85.35 ± 2.47 f	36.76 ± 0.22 e
M5-F	107.32 ± 0.97 c	51.02 ± 1.27 a
M10-F	99.64 ± 0.50 d	45.07 ± 0.40 b
CK	66.84 ± 0.07 g	20.59 ± 0.55 g

Different letters (a-h) within each column indicate a significant difference at p < 0.05.

The effect of different concentrations of lily bulb extracts on cell proliferation activity after different pretreatments is illustrated in Fig. 2. All of the lily bulb extracts were found to exert antiproliferative effects on HepG2 cells in a dose-dependent manner within the concentration of from 10 to 50 mg/mL. It can be observed in Fig. 2 that the effect of blanching time on cell antiproliferative activity was pronounced among the three pretreatment methods. With a concentration of 40 mg/mL of sample in B—F groups, the inhibition of cell proliferation could be up to 43.1%–65.23%, while the maximum inhibition rate in the M-F group was only 45.01%.

Fig. 2.

Effect of pretreatment conditions on antiproliferative activities. (A) Steaming pretreatment. (B) Blanching pretreatment. (C) Microwave pretreatment.

3.5. Correlation analysis

The correlation analysis was used to explore the relationships between treatment, color, total phenolic content, phytochemical compounds and antioxidant activity in lily bulb extracts (Fig. A2, Fig. A3, Table 4). Among the polyphenolic compounds, the content of p-coumaric acid and regaloside C showed no significant correlation with the various pretreatment methods used. However, regaloside B and regaloside E in lily bulbs subjected to various pretreatments show greater sensitivity compared to other phenolic acids. The ΔE value was found to be significantly related to the content of regaloside A (p<0.01), regaloside B (p<0.05), regaloside E (p<0.05) and regaloside K (p<0.01). The b* value is only significantly and positively correlated with the concentrations of regaloside B and regaloside C (p<0.05). ORAC was significantly correlated with regaloside A and regaloside K at 0.05 level. However, no significant correlation was found between regaloside C and ABTS. CC₅₀ and IC₅₀ are negatively correlated with phenolic acid, ORAC, and ABTS, but there is no significant difference.

Supplementary Fig. A2.

Correlation heatmap of color, phytochemical compositions and bioactivity.

Supplementary Fig. A3.

Clustering heatmap of pretreatment and phenolic acids.

Table 4.

Pearson correlation coefficient among color, phytochemical compositions, and bioactivity.

	L*	a*	b*	ΔE	Reg A	Reg B	Reg C	Reg E	Reg K	p-cou	ABTS	ORAC	CC50	IC50
L*	1	−0.922**	0.388	−0.935**	0.977**	0.819**	0.732*	0.829**	0.991**	0.390	0.836**	0.784**	−0.431	−0.551
a*		1	−0.592	0.814**	−0.936**	−0.900**	−0.822**	−0.893**	−0.950**	−0.346	−0.773**	−0.812**	0.152	0.374
b*			1	−0.071	0.377	0.684*	0.734*	0.556	0.469	−0.186	0.415	0.318	0.324	0.230
ΔE				1	−0.929**	−0.644*	−0.531	−0.708*	−0.904**	−0.514	−0.693*	−0.798**	0.484	0.655*
Reg A					1	0.831**	0.762*	0.865**	0.974**	0.464	0.763*	0.811**	−0.371	−0.501
Reg B						1	0.922**	0.977**	0.867**	0.476	0.798**	0.575	−0.022	−0.099
Reg C							1	0.865**	0.804**	0.348	0.623	0.540	−0.077	−0.153
Reg E								1	0.862**	0.583	0.784**	0.575	−0.044	−0.108
Reg K									1	0.382	0.822**	0.789**	−0.357	−0.499
p-cou										1	0.317	0.340	−0.106	−0.143
ABTS											1	0.430	−0.424	−0.375
ORAC												1	−0.175	−0.544
CC50													1	0.846**
IC50														1

Notes: Reg A: regaloside A; Reg B: regaloside B; Reg C: regaloside C; Reg E: regaloside E; Reg K: regaloside K; p-cou: p-coumaric acid. * and ** mean significant correlation at the 0.05 and 0.01 levels, respectively (2-tailed).

4. Discussion

4.1. Effect of pretreatment for color of freeze-dried lily bulbs

The color change of lily bulbs when stored is an important appearance factor affecting quality. Fresh lily bulbs appeared bright white and the color variation during browning was mainly violet-red and brown (Zhang et al., 2023). Color alteration in this study is mainly represented by L*, a*, b* and ΔE value. Less browning of the sample was observed with higher L* and lower a* values (Kan et al., 2022). As can be seen from Fig. A1 and Table 1, the M-F group had poorer effects in delaying the browning of lily bulbs compared to the other treatment groups. The processing time in the B—F group was found to be positively correlated with the browning inhibitory effect, which may be related to the enzymatic browning reaction. As reported by a previous study, increasing blanching time reduces the activities of polyphenol oxidase and peroxidase, which prevented the endogenous enzymes in the lily bulbs from reacting with the phenolic substances to form colored substances and inhibited the occurrence of browning (Wang et al., 2021). Furthermore, based on the correlation analysis, the content of regaloside A and regaloside K have a highly significant effect on the color change (ΔE) of lily bulbs (p < 0.01), which showed a positive correlation with brightness, but a negative correlation with a* value. Furthermore, the contents of regaloside B, regaloside C, and regaloside E were also significantly correlated with L* and a* values (p < 0.05). Based on the above data, we found that the B6—F group had the largest L* value, a smaller a* value and the smallest color difference with CK, which means that it had the best inhibitory effect on browning, followed by the S15—F and S10—F group.

4.2. Effect of pretreatment on the phenolic content of freeze-dried lily bulbs

Phenolics are among the most abundant bioactive components in lilies, with phenolic acids and flavonoids being the most predominant polyphenols, accounting for over half of the secondary metabolites detected (Tang et al., 2021). The results showed that the total phenolic acid content of M-F groups was the lowest, indicating that based on the experimental conditions, microwave pretreatment had the poorest retention effect on phenolic acid in lily bulbs. Owing to the high thermal efficiency induced by the instantaneous high temperature generated by the microwave, leading to more polyphenol degradation compared to blanching and steaming (Martínez, Armesto, Gómez-Limia, & Carballo, 2020). However, a considerable increase in the total content of phenolic acids was observed in certain pretreated samples, specifically S10—F and B6—F, which had been subjected to steaming and blanching. Similar results were reported by Ncube, who observed a significant increase in the content of phenolic acids such as chlorogenic acid and crypto-chlorogenic acid in blanched Corchorus olitorius (Ncube, Dlamini, & Beswa, 2022). According to Feumba Dibanda, phenolic acids with conjugated structures, such as ferulic acid and p-coumaric acid, which are found in fruits and vegetables, may be present in esterified forms (Feumba Dibanda, Panyoo Akdowa, Rani, Metsatedem Tongwa, & Mbofung, 2020). Appropriate thermal processing could potentially cleave the esters linking these compounds to cell wall components such as hemicelluloses, resulting in increased content. Regaloside and p-coumaric acid, both of which belong to the phenylpropanoid class of compounds, have similar chemical structures and functionalities. The increase in total phenolic acid content observed in some samples could be attributed to the disruption of cellular structures by steaming or blanching pretreatments, which increased the extractability of phenolic acids and released the bound forms of phenolic acids.

Regaloside A is present in all lily populations and was reported to be one of the most plentiful phenolic pigment monomers detected in fresh lily bulbs (Liang et al., 2022). However, regaloside C, regaloside E, and regaloside K are not ubiquitously present across all medicinal Lilium species. It has been reported that the existence of these three compounds in L. lancifolium, but they remain undetected in L. brownii, a finding closely linked to the key enzyme genes involved in the synthesis of phenylpropanoid glycerol glucosides (Qin et al., 2022). As shown in Table 2, regaloside A and regaloside B were the most abundant phenolic acids in CK. However, after processing, the content of regaloside A decreased, while some samples, represented by S10—F and B6—F, had increased levels of regaloside B, regaloside E and regaloside C. This implies that different pretreatment methods and processing times may affect the retention, degradation and transformation of regaloside. Similarly, Nayak et al. have reported that phenolic acids in foods are susceptible to degradation and changes in their structural forms during thermal and non-thermal processing, both in the free and bound states, occasionally leading to the formation of new compounds with potential antioxidant activities (Nayak, Liu, & Tang, 2015). A range of pretreatment conditions may influence the metabolism and transformation of regaloside A, resulting in a decrease in its content, while concurrently promoting the conversion of precursor substances into regalosides B and E. However, this hypothesis awaits empirical validation through subsequent research endeavors. According to Pearson correlation analysis, the phenolic monomers, except p-coumaric acid, were significantly positively correlated (p < 0.05) with the L* value and negatively correlated with the a* value, indicating that the five phenolic monomers were closely related to the brightness and redness of dried lily bulbs. Regaloside A and regaloside K showed a highly significant positive correlation (p < 0.01) with the ΔE value, which means that they are major phenolic monomers that can effectively inhibit the browning of lily bulbs.

4.3. Effect of pretreatment for bioactivity of freeze-dried lily bulbs

Numerous phenolics have been reported in members of the Lilium bulbs, including L. lancifolium, most of which are antioxidants and correlate strongly with antioxidant activity (Jin, Zhang, Yan, Guo, & Niu, 2012). However, Biological systems contain a wide variety of antioxidant systems, which may be involved in complex interactions, including synergistic or antagonistic reactions in the matrix of the food (Lutz, Jorquera, Cancino, Ruby, & Henriquez, 2011). We need to evaluate the antioxidant capacity of samples by detection methods, but currently different antioxidant capacity analysis methods have different mechanisms and reaction characteristics, and cannot rely solely on a single method for detection. Therefore, in this study, two detection methods, the ABTS and the ORAC assay, were used to investigate the antioxidant activity of phenolic extracts from freeze-dried lily bulbs after pretreatment. On the basis of the ABTS and ORAC results, it can be concluded that the antioxidant activity of the freeze-dried lily bulbs was slightly reduced after the steam pretreatment, but there was no significant difference, especially in the S5—F group. It is worth noting that B6—F is the only sample with an ABTS value higher than CK, which may be related to its high content of phenolic compounds. There is a significant positive correlation between the ORAC value and total phenolics, regaloside A and regaloside K, while regaloside B, regaloside E and regaloside K are significantly positively proportional to ABTS (Table 4). This may be due to the fact that the determination of antioxidant activity is affected to varying degrees by the different chemical structures, substituent types and steric hindrance effects of phenolic compounds.

Apart from the antioxidant activity, the biological activities of lily bulb extract, such as anti-inflammatory, antibacterial and anti-tumorigenic have been reported. The half-inhibitory concentration of all samples was below the CC₅₀ value, indicating that the inhibition of HepG2 cell proliferation by freeze-dried lily bulbs was not mediated by toxicity to the cells. Freeze-dried lily bulbs showed varying degrees of reduction in antiproliferative activity after different pretreatments, with B6—F showing the least reduction. This may be associated with the high antioxidant activity of B6—F. Recent research declared that polyphenols serve as potential anti-proliferative agents and may exert various health benefits by acting as ROS (reactive oxygen species) accumulators and causing oxidative damage to membrane lipids and other cellular components, leading to cancer cell death (do Carmo, Pressete, Marques, Granato, & Azevedo, 2018). Furthermore, the results show a negative correlation between regaloside and p-coumaric acid with CC₅₀ and IC₅₀ values, suggesting that an increase in the content of the detected phenolic acids may enhance the cell cytotoxicity and antiproliferative activity to some extent. However, the lack of a significant correlation also suggests that the complex composition of lily bulb extracts, including polyphenols and other constituents, cooperatively contribute to their bioactivity.

5. Conclusion

To summarize, different pretreatment methods and processing times could affect the appearance, phytochemical content, antioxidant activity and cell antiproliferative activity of freeze-dried lily. In this study, microwave pretreatment resulted in poor quality freeze-dried lily, whereas S—F and B6—F resulted in similar or better product quality compared to control. It is noteworthy that the antioxidant activity of B6—F increased by 25.39% compared to CK, while the L* value decreased by only 0.55%. The results showed that the combination of blanching pretreatment will provide the new idea and way to improve the quality characteristics of freeze-dried lily bulbs for food, medicine and cosmetic application.

The following are the supplementary data related to this article.

Funding

This research was funded by Tibet Autonomous Region Major Special Science and Technology, grant number: XZ202101ZD0023G; and the Forth National Survey of Traditional Chinese Medicine Resources, Chinese or Tibet Medicinal Resources Investigation in Tibet Autonomous Region, grant number: State Administration of Chinese Traditional Medicine 20200501–542301.

CRediT authorship contribution statement

Yixi Cai: Writing – original draft, Visualization, Software, Methodology, Formal analysis, Data curation. Hong Quan: Methodology, Investigation, Data curation. Ying Liu: Methodology, Formal analysis. Yazhou Lu: Investigation. Xiaozhong Lan: Writing – review & editing, Resources, Project administration, Funding acquisition, Conceptualization. Xinbo Guo: Writing – review & editing, Validation, Supervision, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data is contained within the article.