KXS is composed of ginseng, polygala, poria and acorus gramineus, with a proportion of 1:1:2:1, according to the Notice by the National Administration of Traditional Chinese Medicine of China.

Ginseng is the dried root and rhizome of Panax ginseng C. A. Mey. of the Araliaceae family. Polygala is the dried root of Polygala tenuifolia Willd. of the Polygalaceae family. Poria is the dried sclerotium of Poria cocos (Schw.) Wolf, a fungus of the Polyporaceae family. Acorus is the dried rhizome of Acorus tatarinowii Schott of the Araceae family. The above-mentioned medicinal materials were all provided by Shineway Pharmaceutical Group Co., Ltd., with a total of 15 batches (S1-S15).

Accurately weigh 1 g of KXS, add 5 mL of 95% methanol, and record the weight. Heat the mixture under reflux for 30 min, then allow it to cool. Re-weigh the solution and adjust to the original weight using 95% methanol. Mix thoroughly, filter, and collect the filtrate for use.

Polygalaxanthone III (DST230425-089), 3-6-disinapoylsucrose (DST210510-022), ginsenoside Rg1 (DSTDR000902), ginsenoside Rb1 (DSTDR000603), β-asarone (DSTDA001901), α-asarone (DST201025-044), dehydrotumulosic acid (DSTDQ010202), dehydropachymic acid (DSTDQ010101) were purchased from Dest Biotechnology Co., Ltd. (Sichuan, China). Acetonitrile (20240701, TEDIA) and methanol (20240628, TEDIA) were obtained from DiMa Technology Co., Ltd. (Beijing, China).

Dissolve an appropriate amount of the relevant components in 95% methanol to prepare the test solution. The concentrations of polygalaxanthone III, 3-6-disinapoylsucrose, ginsenoside Rg1, ginsenoside Rb1, β-asarone, α-asarone, dehydrotumulosic acid and dehydropachymic acid are 0.74 mg. mL⁻¹, 0.76 mg. mL⁻¹, 0.59 mg. mL⁻¹, 0.35 mg. mL⁻¹, 0.85 mg. mL⁻¹, 1.42 mg. mL⁻¹, 0.61 mg. mL⁻¹ and 0.58 mg. mL⁻¹, respectively.

Wild type W¹¹¹⁸, Elav-Gal4, and UAS-Aβ₄₂ drosophila were all provided by the Core Facility of Drosophila Resource and Technology, CEMCS, CAS.

Main groups: I Control group with the same genetic background (CA): Select male drosophila from the first generation of Elav-Gal4 virgin drosophila and W¹¹¹⁸ male drosophila hybrid offspring. II Aβ₄₂ group (MA): Red-eyed straight-winged males were selected from the progeny generation of Elav-Gal4 virgin drosophila and UAS-Aβ₄₂ male drosophila (Elav/Y; UAS-Aβ/+, Aβ₄₂ drosophila) (Tue et al., 2020).

Standard drosophila Medium: The flies culture medium consists of cornmeal, soybean flour, sucrose, maltose, agar, yeast, and water. Specifically, it contains 6.7 g of cornmeal, 0.8 g of soybean flour, 4.0 g of sucrose, 4.2 g of maltose, 1.0 g of agar, 2.5 g of yeast, and 100 mL of water. Besides, sodium benzoate, with a dosage of 0.1 g, is selected as the preservative. The flies were maintained at a temperature of 25°C, with 60% humidity, under a 12-h light and 12-h dark cycle. The medium was refreshed every 3–4 days for passaged flies.

KXS-Containing Medium: Following the method for preparing Chinese medicine-containing medium described in previous research (Zhang et al., 2016), the KXS-containing medium was prepared with concentrations ranging from 0.25% to 4%. Before each administration, the drosophila were fasted for 3 h.

The HPLC chromatographic analysis was performed using a Waters e2695 HPLC-DAD system (Waters, United States) equipped with a Waters Symmetry C₁₈ chromatographic column (4.6 × 250 mm, 5 μm) (Waters, United States). For gradient elution, the mobile phases consisted of acetonitrile solution (organic phase C) and 0.2% phosphoric acid aqueous solution (organic phase D), with a flow rate of 1 mL min⁻¹. The temperature of the chromatographic column was maintained at 30°C. The injection volume was set to 20 μL, and the chromatograms were detected at a wavelength of 203 nm. The gradient elution program was as follows: 0–2 min, 10–12%C; 2–10 min, 12–14%C; 10–14 min, 14–15%C; 14–21 min, 15–16%C; 21–23 min, 16–18%C; 23–33 min, 18–26%C; 33–48 min, 26–30%C; 48–58 min, 30–35%C; 58–68 min, 35–37%C; 68–78 min, 37%C; 78–93 min, 37–50%C; 93–103 min, 50–74%C; 103–115 min, 74–82%C; 115–120 min, 82–10%C.

The KXS sample test solution (S1) was prepared following the procedure described in Section 2.1. Six consecutive injections were performed under the conditions outlined in Section 2.4 to assess the precision of the instrument. The S1 test solution was injected at intervals of 0, 2, 4, 8, 12, and 24 h to evaluate sample stability. Additionally, six separate test solutions of S1 samples were prepared and injected to assess sample reproducibility.

The reference solution, prepared as described in Section 2.2, was analyzed using the chromatographic system. The retention time of the reference peak was recorded, and the retention times of the reference components in the sample solution were determined by comparison with the recorded reference peak retention times.

Fifteen batches of KXS samples were analyzed using the standardized HPLC method. The resulting chromatographic data were exported in CDF format and processed using the Similarity Evaluation System for Chromatographic Fingerprints of Traditional Chinese Medicines (version 2012).

Systematic cluster analysis was performed using SPSS statistical software (IBM, United States). The between-groups linkage method was employed to analyze the chromatographic peak area data from the fingerprint spectra of 15 batches of KXS. The similarity between samples was measured using the Euclidean distance.

The spectrum-effect relationship, which refers to the connection between the peak areas of the common peaks in KXS and its effect on enhancing cognitive ability, was established using grey relational analysis, bivariate correlation analysis, and partial least squares regression analysis.

Data are presented as mean ± SD. SPSS was used for statistical analysis, and GraphPad (United States) was employed for graphing. One-way ANOVA was conducted for data analysis. Statistical significance was indicated by P < 0.05.

Using the α-asarone peak (peak 0) as the reference, the relative standard deviation (RSD) for precision ranged from 0% to 0.6%, for reproducibility from 0.1% to 0.5%, and for stability from 0% to 1.3%. The results showed that the RSD of the relative peak area of each chromatographic peak in the fingerprint chromatogram was less than 2%. The HPLC method demonstrated excellent accuracy, and the KXS sample solution remained stable for up to 24 h. These findings confirm that the developed HPLC fingerprint method has superior reproducibility.

As illustrated in Figure 1, chromatograms were obtained for the mixed reference solution (Figure 1A), the KXS test samples (Figure 1B), and each batch of KXS (Figure 1C) under the specified chromatographic conditions. By comparing the ultraviolet spectra and the retention times of the reference substance peaks, 17 common peaks were identified within the 0–120 min range. Eight components were identified: polygalaxanthone III (b), 3-6-disinapoylsucrose (d), ginsenoside Rg1 (f), ginsenoside Rb1 (m), β-asarone (n), α-asarone (o), dehydrotumulosic acid (p), and dehydropachymic acid (q). Peak o, with the largest area in the fingerprint spectrum, was designated as the characteristic peak following chromatographic system analysis. The relative peak areas of each common peak are listed in Table 1, relative retention times in Table 2, and the proportion of non-common peak areas in Table 3. Variations in the RSD of relative peak areas reflect differences in the compound content among batches.

To assess sample similarity, a comparative analysis was performed between the fingerprint chromatograms of the 15 KXS batches and a reference chromatogram. The similarity range for the batches compared to the reference fingerprint was between 0.887 and 1, as detailed in Table 4. Most samples exhibited high similarity, with values close to or exceeding 0.99. For example, batches S2, S3, S6, and S7 showed similarities above 0.99, while S5 reached 1, indicating consistent chemical composition and relative content among these batches. This suggests stable quality and a well-controlled production process. However, batch S1, with a similarity of 0.887, displayed relatively lower similarity. This may point to minor differences in raw materials, processing, or storage, causing slight variations in chemical components. Despite this, the similarity remained high overall, indicating that the deviations were within an acceptable range.

The study employed the HCA function in SPSS statistical software (IBM, USA) and used the inter-group linkage method to analyze the fingerprints of 15 batches of KXS. HCA was applied to assess the correlation of the KXS HPLC fingerprints, and testing was performed using Heatmapper (http://www.heatmapper.ca/) (Xu et al., 2022). The HCA dendrogram (Figure 2A) and heatmap (Figure 2B) were generated. Each color unit in the heatmap corresponds to a matrix representing a numerical value. As seen in Figure 2, four clusters were identified. S1, S2, S3, S5, S6, S10, S11, S13, S14, and S15 form one group; S9 and S8 form another group; S4 forms a separate group; and S12 and S7 form a final group. The HCA results indicate that the different batches of KXS samples share similar fingerprints, but the peak areas of the chemical components differ, suggesting variations in the content of the chemical components. HCA can effectively differentiate samples at the chemical level.

The experiments demonstrated that, except for the 0.25% and 0.5% KXS groups, which were ineffective, all other concentrations of KXS significantly enhanced the PI of Aβ₄₂ transgenic drosophila (Figure 3A). Similarly, except for the groups treated for 1, 2, and 3 days, all other treatment duration groups also showed improvements in PI (Figure 3B). Based on these results, the optimal administration concentration was determined to be 1% KXS, with a treatment duration of 4 days.

As shown in Figure 4, the 15 batches of KXS samples all demonstrated an enhancement effect on cognitive impairment in Aβ₄₂ transgenic drosophila. However, there were variations among the different samples in the extent of cognitive improvement. This discrepancy may be attributed to the differences in the chemical composition of the samples. Therefore, the spectrum-effect relationship was further investigated.

Based on the results of the above spectrum-effect relationship analysis, b, c, d, e, f, g, h, i, j, k, l, m, n, o, and p may be potential active ingredients of KXS. Among these, seven identified ingredients were selected for subsequent verification to further analyze the impact of a single ingredient on the PI of Aβ₄₂ transgenic drosophila. As described in Figure 7, the PI values of the 15 batches of KXS increased by an average of 35% compared to the MA. However, the PI of polygalaxanthone III, ginsenoside Rg1, ginsenoside Rb1, β-asarone, and α-asarone increased by 30.26%, 31.60%, 27.36%, 28.55%, and 30.74%, respectively. Compared to KXS, the PI growth rate of these active ingredients decreased to varying degrees.