Saponin accumulation in the seedling root of Panax notoginseng

Background Panax notoginseng is an important Chinese medicinal plant. Dammarene-type triterpenoid saponins are main pharmacologically effective compounds in P. notoginseng. This study aims to investigate the formation and accumulation of saponins in P. notoginseng roots during germination and juvenile stage. Methods P. notoginseng seeds were collected and stored in wet sand. After germination, the seedlings were transplanted into a soil nursery bed and cultivated for one year. During this period, samples were collected every month and the concentrations of ginsengnosides Rg1, Re, Rb1, Rd and notoginsengnoside R1 were determined by HPLC. Results There was little saponin in the P. notoginseng seed. The chemical composition of seed was different from that of root. After germination, Rb1, Rg1, Re, Rd and R1 appeared successively in the seedling root. And in the five-month-old root, all these five main saponins came into existence. The accumulation of saponins in P. notoginseng root was affected by seasons. Conclusion The accumulation of saponins showed a time-dependent increase after germination of P. notoginseng.

Although detailed chemical studies have been carried out on mature root and leaf of P. notoginseng, very little information is available on the chemical composition of its seed or on saponin accumulation during root development. The present study aims to investigate the formation and accumulation of saponin constituents in P. notoginseng during germination and juvenile stages, particularly the developmental changes of Rg 1 , Re, Rb 1 , Rd and R 1 in the seed and seedling root.

Solvents and chemicals
Methanol (MeOH) was purchased from Tianjing chemical Ltd (China). Acetonitrile (MeCN) was purchased from Merck (Germany). Standard Rg 1 , Rb 1 , Re, Rd and R 1 were isolated from the root of P. notoginseng and their chemical structures were determined with nuclear magnetic resonance (NMR) and mass spectrometry (MS).

Plant materials
Mature seeds were collected from a 3-year-old P. notoginseng in November 2005 from the farm of Miaoxiang Ltd (Wenshan County, Yunnan Province, China). After removal of the pericarp, the seeds were stored in wet sand; after germination, the seedlings were transplanted into a soil nursery bed under a sheltering net with 80% shadowiness. During this period, 50 g seed or 20 individual seedling roots were sampled at intervals of one month until

Seed extraction
After removal of the pericarp, 100 g of fresh seeds was crushed and extracted at room temperature with MeOH (300 ml) for three times. The concentrated MeOH extract was partitioned between water and petroleum ether. The aqueous part was concentrated under reduced pressure as a crude seed extract (0.3 g).

Method validation
The method was validated by measuring Rg 1 , Re, Rb 1 , Rd and R 1 . Instrument precision was obtained by analyzing the peak areas of six injections. The relative standard deviations (RSDs) were: R 1 : 1.09%, Rg 1 : 0.97%, Re: 1.39%, Rb 1 : 1.04% and Rd: 1.03%. Stability of sample solution was measured by injecting the same sample solution at time points of 0, 6, 12 and 24 hour. The RSDs were: R 1 : 0.96%, Rg 1 : 0.43%, Re: 1.35%, Rb 1 : 0.54% and Rd: 1.41%. Reproducibility was evaluated by measuring the concentrations of these five analytes in six replicate samples with external standards. The RSDs were: R 1 : 0.80%, Rg 1 : 0.36%, Re: 1.24%, Rb 1 : 0.58% and Rd: 0.50%. The recovery rates of these five saponins were determined by the method of standard addition with six replications. The average recovery rates and the RSDs were calculated (Table 1).

Standard curve
Standard solutions were prepared by combining Rg 1 (0.332 mg/ml), Rb 1 (0.344 mg/ml) as standard solutions A whereas Re (0.047 mg/ml), Rd (0.082 mg/ml), R 1 (0.084 mg/ml) as solution B; both were dissolved in methanol. The standard curves were generated by injecting standard solutions of 5 μl and 10 μl to 90 μl at the intervals of 10 μl respectively. The peak area for each analyte was determined. Standard curves were then constructed ( Table 2).

Limits of detection and limits of quantitation
The standard solutions were diluted with 70% aqueous methanol to provide appropriate concentrations. When the ratio of the testing peak signal-to-noise (S/N) was 4, the limit of detection (LOD) for each analyte was determined; when the S/N ratio was 10, the limit of quantitation (LOQ) was determined.

Sample preparation for HPLC analysis
For seed samples of germination test, 1.0 g of powder was weighed accurately and extracted ultrasonically for 30 minutes in 70% methanol in a 10 ml volumetric flask. After cooled down and made up the lost volume with methanol, the sample solution was obtained by filtering the supernate with a nylon filter membrane (0.45 μm) prior to the HPLC analysis. For seedling root samples, 50 mg of powder was weighed accurately and extracted in 70% methanol ultrasonically in a 5 ml volumetric flask. The other steps were similar to those of  the seed sample. Injection volumes of seed and seedling sample solutions for HPLC were 100 μl. As to HPLC analysis of seed extract, the raw extract was dissolved in MeOH (10 mg/ml) and filtered with 0.45 μm nylon filter membrane and 10 μl of solution was injected for HPLC analysis.

Statistical analysis
Linear regression was performed with Excel 2003 (Microsoft, USA). RSDs were also calculated with Excel 2003 (Microsoft, USA).

Results
Under the HPLC conditions used in this study, all five saponins were baseline separated and their calibration curves exhibited good linear regressions. The method validation analysis demonstrated that the analytical method developed in this study for all five saponins was accurate and precise. The P. notoginseng seeds were collected in November in our experiments; peeled seeds were stored in wet sand which started to germinate in the following January. Seedlings were then transplanted into a soil nursery bed in February and they would grow till the third February. Samples were collected every month, the contents of ginsenosides Rg 1 , Re, Rb 1 , Rd, notoginsenoside R 1 were analysed with HPLC. The RSDs of the concentrations of five analytes of each sample in triplicate analyses were all within 3% (Table 3).

Saponins in the seed
Results of HPLC analyses indicated that saponins were undetectable in P. notoginseng seed, even though the amount of seed sample was 10 times as much as that of seedling root. We obtained just 0.3 g MeOH extract from 100 g of fresh seed with the yield being 0.3%; yield was about 0.83% in dried seed while it may reach 9.25% and 11.74% of total saponin in dried two-year-old and three-year-old roots respectively [11]. Furthermore, HPLC analysis indicated that Rg 1 , Rb 1 , Rd, Re and R 1 (the main constituents in the root) were not detected in the crude seed extract whereas some other peaks appeared in the HPLC chromatogram, suggesting that the chemical composition of the seed was different from that of root (Figure 1). Further comprehensive chemical studies are required to determine the constituents of seed extract.

Saponin accumulation in the root
After germination, Rb 1 first appeared in the seedling root in February. Rg 1 , Re, and Rd began to appear in the fourth month of germination (the following May). After one month, all these five saponins were detected in the root. Then, with the growth of seedling, the saponin contents increased rapidly in the root. The accumulation showed a time-dependent increment of saponin contents. Rg 1 concentration reached its maximum in September (eight months after germination). Rb 1 , Rd and R 1 reached maximum in October whereas Re did in November. Later, as winter came, concentrations of all  these five saponins started to decline significantly ( Figure 2, Table 3).

Discussion
Dammarene-type triterpenoid saponins are main secondary metabolites of Panax notoginseng. The present study demonstrates a temporal and spatial distribution of saponins during the germination process and the growth of young plants. Our results show that ginsenosides Rg 1 , Re, Rb 1 , Rd and notoginsenoside R 1 were not detected in P. notoginseng seed. The formation of saponins in root is a gradual process. The synthesis and accumulation of saponins began after germination and continued with the growth of seedling. Saponin synthetases were activated after seed began to germinate. In young roots, saponin constituents formed and accumulated mainly between July and October, the most vigorous period of growth. This periodic change is, as in adult plant, closely related to the growth pattern of Panax notoginseng; the formation and accumulation of saponins were affected by seasons [12]. As a plant grows up, more and more saponins accumulate in the root.
Our previous work revealed that, in a 3-year-old root, the concentrations of Rg 1 , Rb 1 , Rd, Re and R 1 reached 4.11%, 4.12%, 0.82%, 0.83% and 1.14% respectively [11]. All these findings suggest that saponins may not serve as the nutrient storage in the seed. The protective functions of saponins in other plants are reported [13,14]. The role of this kind of secondary metabolites in P. notoginseng requires further investigation.

Conclusion
The accumulation of saponins showed a time-dependent increase after germination of P. notoginseng.  Figure 2 Time courses of saponin accumulation in Panax notoginseng seed and seedling root. The seeds were collected in November and stored in wet sand. They began to germinate the following January. The seedlings were then allowed to grow for one year in soil nursery bed.